Source: RADAR HANDBOOK Electronic Counter-Countermeasures A. Farina Analysis of Integrated Systems SELEX Sistemi Integrati Chapter 24 24.

1 INTRODUCTION Since World War II, both radar and electronic warfare (EW)* have achieved a very high state of performance.1,2 Modern military forces depend heavily on electrom agnetic (EM) systems for surveillance, weapon control, communication, and naviga tion; thus access to, and control of, the EM spectrum is vital. Electronic count ermeasures (ECM) are likely to be taken by hostile forces to degrade the effecti veness of EM systems.37 As a direct consequence, EM systems are more and more fre quently equipped with socalled electronic counter-countermeasures (ECCM) to ensu re effective use of the EM spectrum despite an enemys use of EW actions. This cha pter is devoted to the description of the ECCM techniques and design principles to be used in radar systems when they are subject to an ECM threat. Section 24.2 starts with a recall of the definitions pertaining to EW and ECCM. The topic of radar signals interception by EW devices is introduced in Section 24.3; the fir st strategy to be adopted by radar designers is to try to avoid interception by the opponent electronic devices. Section 24.4 is dedicated entirely to the analy sis of the major ECM techniques and strategies. It is important to understand th e ECM threat to a radar system in order to be able to efficiently react to it. T o facilitate the description of the crowded family of ECCM techniques (Sections 24.6 through 24.10), a classification is attempted in Section 24.5. Then, the te chniques are introduced according to their use in the various sections of radar, namely, antenna, transmitter, receiver, and signal processing. A key role is al so played by those ECCM techniques that cannot be classified as electronic, such as human factors, methods of radar operation, and radar deployment tactics (Sec tion 24.10). The ensuing Section 24.11 shows the application of the aforemention ed techniques to the most common radar families, namely, surveillance, tracking, multifunctional, phased-array, imaging, and over-the-horizon radars. The main d esign principles (e.g., selection of transmitter power, frequency, waveform, and antenna gain) as dictated by the ECM threat are also discussed in some detail. * A list of acronyms is at the end of the chapter before the list of references. 24.1 Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengin eeringlibrary.com) Copyright 2008 The McGraw-Hill Companies. All rights reserved . Any use is subject to the Terms of Use as given at the website.

Electronic Counter-Countermeasures 24.2 RADAR HANDBOOK The chapter ends with an approach to the problem of evaluating the efficacy of E CCM and ECM techniques (Section 24.12). There is a lack of theory to properly qu antify the endless battle between ECCM and ECM techniques. Nevertheless, a commo nly adopted approach to determine the ECM effect on a radar system is based on e valuation of the radar range under jamming conditions. The advantage of using sp ecific ECCM techniques can be taken into account by calculating the radar range recovery. A list of acronyms used in the chapter and the references appear at th e end of the chapter. 24.2 TERMINOLOGY EW is defined as a military action involving the use of EM energy to determine, exploit, reduce, or prevent radar use of the EM spectrum.811 The operational empl oyment of EW relies upon the capture of radar EM emissions using electronic inte lligence (ELINT) devices, collating the information in support databases that ar e then used to interpret EM emission data, to understand the radar system functi ons, and to program reactions against the radar. EW is organized into two major categories: electronic warfare support measures (ESM) and ECM. Basically, the EW community takes as its job the degradation of radar capability. The radar commu nity takes as its job the successful application of radar in spite of what the E W community does; the goal is pursued by means of ECCM techniques. The definitio ns of ESM, ECM, and ECCM are listed below.8,11,12 ESM is that division of EW invo lving actions taken to search for, intercept, locate, record, and analyze radiat ed EM energy for the purpose of exploiting such radiations in the support of mil itary operations. Thus, ESM provides a source of EW information required to cond uct ECM, threat detection, warning, and avoidance. ECM is that division of EW in volving actions taken to prevent or reduce a radars effective use of the EM spect rum. ECCM comprises those radar actions taken to ensure effective use of the EM spectrum despite the enemys use of EW. The topic of EW is extremely rich in terms , some of which are also in general use in other electronic fields. A complete g lossary of terms in use in the ECM and ECCM fields is found in the literature.8, 11,13 24.3 ELECTRONIC WARFARE SUPPORT MEASURES ESM usually consists of several detection and measurement receivers and real tim e processor boards dedicated to the interception of radar emissions. The identif ication of specific emitters is based on comparison with tactical or strategic E LINT.9,1417 Emitter location can be additionally provided through several methods such as triangulation Since the publication of the second edition of this Handbook, the U.S. Air Force changed some of the EW terms we have gotten used to over these many years. ECM is now Electronic Attack (EA), ECCM is Electronic Protection (EP), and ESM is El ectronic Support (ES). These terms are not used in this chapter because they are seldom used by the radar community who seem to prefer retaining the more famili ar expressions ECM, ECCM, and ESM. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw-Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter-Countermeasures ELECTRONIC COUNTER-COUNTERMEASURES 24.3 from remote systems or single platform sequential bearing measurements, differen ce time of arrival (DToA) or hyperbolation, and Phase Difference Rate (PDR). Mod ern digital receiver technology, coupled with state-of-the-art deinterleaving si gnal processing to cleanly isolate and identify individual EM emitters, will enh ance situation awareness. Using techniques such as time and frequency difference of arrival will improve single and multiple platform spatial location; this wil l allow EW to be used to cue targeting systems. Radar interception, which is of particular interest in this section, is based on the reception and measurements of the signals transmitted by radar systems whether pulse or Continuous Wave (CW ). The operational scenario in which ESM should operate is generally crowded wit h pulsed radar signals: figures of 500,000 to 1 million pulses per second are fr equently quoted in the literature.9 The center frequency, amplitude, pulse width , time of arrival (ToA), and bearing of each detected pulse are measured, conver ted in digital format, and packed into a pulse descriptor word (PDW). The train of PDWs are sent to a pulse-sort processor that deinterleaves the sequences belo nging to different emitters and identifies Pulse Repetition Interval (PRI) value s and modulation laws (random jitter, stagger, switching). Further comparison ag ainst an emitter database, which contains the range of characteristic parameters (frequency, pulse width, PRI), the related pattern of agility (random, stagger, etc.) for each emitter, the type of antenna scanning pattern and periods permit s the generation of an emitter list with an identification score. The ESM receiv er is used to control the deployment and operation of ECM; the link between ESM and ECM is often automatic. A single received radar pulse is characterized by a number of measurable parameters. The availability, resolution, and accuracy of t hese measurements must all be taken into account when designing the deinterleavi ng system because the approach used depends on the parameter data set available. Obviously, the better the resolution and accuracy of any parameter measurement, the more efficiently the pulse-sort processor can carry out its task. However, there are limitations on the measurement process from outside the ESM system (e. g., multipath), from inside the system (e.g., timing constraints, dead time duri ng reception), and from cost-effectiveness considerations. Angle of arrival is t he most important sorting parameter available to the deinterleaving process sinc e the target bearing does not vary from pulse to pulse. Therefore, amplitude com parison monopulse antennas or multiple base interferometric (phase comparison) s ystems are often used in order to warrant both 360 spatial coverage and pulse-bas ed angle of arrival measurement. Monopulse rotating antennas can also be used wh en the time to intercept is not critical (this is the ELINT case), and it is pos sible to scan sequentially the operational scenario. The carrier frequency is th e next most important pulse parameter for deinterleaving. A common method of fre quency measurement is to use a scanning superheterodyne receiver that has the ad vantage of high sensitivity, good frequency resolution, and immunity with respec t to the interference of nearby emitters.9 Unfortunately, this type of receiver has a poor probability of intercept for the same reasons as the rotating bearing measurement system. The situation is much worse if the emitter is also frequenc y-agile (random variation) or frequency-hopping (systematic variation). A common method to allow for wideband frequency measurements is based on interferometric devices that provide instantaneous frequency measurement with good accuracy and are able to reject signal interference with lower intensity. The higher sensiti vity and probability of interception are provided by wide instantaneous band sup erheterodyne receivers followed by banks of contiguous receiver channels. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw-Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter-Countermeasures 24.4 RADAR HANDBOOK Several technologies have been proposed in the past such as surface acoustic wav e (SAW) filters and Bragg cells.9 The preferred approach is based on digital rec eivers that integrate wideband spectral analysis and several post-detection func tions, such as intrapulse modulation measurement and waveform code reconnaissanc e. Pulse width is an unreliable sorting parameter because of the high degree of corruption resulting from multipath transmission. Multipath effects can severely distort the pulse envelope, for example, by creating a long tail to the pulse a nd even displacing the position of the peak. The ToA of the pulse can be taken a s the instant that a threshold is crossed, but in the presence of noise and dist ortion, this becomes a variable measurement. Nevertheless, the ToA is used for d eriving the PRI of the radar. The amplitude of the pulse is taken as the peak va lue. Dynamic-range considerations must take into account at least some three ord ers of magnitude for range variation and three orders of magnitude for scan patt ern variations. In practice, 60 dB instantaneous dynamic range sounds like a min imum value; in many applications, it should be larger. The amplitude measurement is used (along with ToA) for deriving the scan pattern of the emitter.9 The cla ssification of radar interception systems is based on the type of representation they provide of the electronic environment. A radar warning receiver (RWR) in a n airborne installation provides alerts of the presence of threats such as radar on a missile, supplying the relative bearing on a cockpit-based display. Search radars are not the primary target for these systems, though range advantage due to one-way propagation with respect to two-way propagation allows radar interce ption at farther range than own platform detection. Required sensitivity values range from 38 dBm (dB milliWatt with respect to the isotropic) to 60 dBm. ESM is t he most complex system and usually comprises the capability to produce a picture of the complete electronic order of battle in its deployment area and alert fun ction. This kind of system is able to detect and analyze emitter waveforms and s canning patterns. The reaction time for the reconnaissance of the operational en vironment may be less than 10 s, though dangerous emitters and alert functions c all for tighter constraints. Required sensitivity ranges from 55 dBm to better th an 80 dBm. ELINT systems are similar to ESM, but may not require 100% probability of intercept. The reaction time may be minutes or hours. The purpose is not to detect emitters as soon as they switch on in the operational environment, but to provide detailed characteristics of emitters to allow the generation of an iden tification database for RWR and ESM systems. ELINT system sensitivity may reach 9 0 dBm, but they dont need to provide 360 surveillance, and they can reach such per formance by means of several directive antennas. The range at which a radar emis sion is detected by an RWR depends primarily on the sensitivity of the receiver and the radiated power of the victim radar. The calculation of the warning range can be obtained by the basic one-way beacon equation, which provides the signal -to-noise ratio (SNR) at the RWR: S P Gr 2 1 1 = G N at RWR where P is the radar radiated ower; R is the range from the RWR to the radar; G t is the transmitting-antenna gain of the radar; Gr is the receiving-antenna gai n of the RWR; is the radar wave ength; the quantity kTsB is the tota system n oise ower of the RWR; and L is the osses. Down oaded from Digita Engineering Library @ McGraw-Hi (www.digita engineerin g ibrary.com) Co yright 2008 The McGraw-Hi Com anies. A rights reserved. Any use is subject to the Terms of Use as given at the website.

4 R 2

E ectronic Counter-Countermeasures ELECTRONIC COUNTER-COUNTERMEASURES 24.5 Equation 24.1 is the basis of erformance ca cu ation for an RWR. It is noted th at the RWR detection erformance is inverse y ro ortiona to R2 rather than to R4 of the radar target detection equation. For this reason, the RWR can detect a radiating radar at distances far beyond those of a radars own target detection c a abi ity. The radarversus-interce tor rob em is a batt e in which the radars ad vantage ies in the use of matched fi tering, which cannot be du icated by the interce tor (it does not know the exact radar waveform), whi e the interce tors a dvantage ies in the fundamenta R2 advantage of one-way versus two-way radar r o agation.1518 Low robabi ity of interce t (LPI) techniques are a ied to radar to win the batt e of to see and not to be seen: see Sch eher19 and references the rein. 24.4 ELECTRONIC COUNTERMEASURES The objectives of an ECM system are to deny information (detection, osition, tr ack initiation, track u date, and c assification of one or more targets) that th e radar seeks or to surround desired radar echoes with so many fa se targets tha t the true information cannot be extracted.37 ECM tactics and techniques may be c assified in a number of ways, i.e., by main ur ose, whether active or assive, by de oyment, by atform, by victim radar, or by a combination of these.1316,2 0 An encyc o edia of ECM tactics and techniques can be found in the iterature.3 ,13 Here descri tion is imited to the most common ty es of ECM. ECM inc udes bo th jamming and dece tion. Jamming is the intentiona and de iberate transmission or retransmission of am itude, frequency, hase, or otherwise modu ated interm ittent, CW, or noise- ike signa s for the ur ose of interfering with, disturbin g, ex oiting, deceiving, masking, or otherwise degrading the rece tion of other signa s that are used by radar systems.313 A jammer is any ECM device that trans mits a signa of any duty cyc e for the so e or artia ur ose of jamming a rad ar system.313 Radio signa s by s ecia transmitters intended for interfering with or rec uding the norma o eration of a victim radar system are ca ed active j amming. They roduce at the in ut of a victim system a background that im edes t he detection and recognition of usefu signa s and determination of their arame ters. The most common forms of active noise jamming are s ot, swe t, and barrage noises. S ot noise is used when the center frequency and bandwidth of the victi m system to be jammed are known and confined to a narrow band. However, many rad ars are frequency-agi e over a wideband as an ECCM against s ot jamming. If the rate of frequency agi ity is s ow enough, the jammer can fo ow the frequency ch anges and maintain the effect of s ot jamming. Barrage or broadband jamming is s imu taneous y radiated across the entire band of the radar s ectrum of interest. This method is used against frequency-agi e systems whose rates are too fast to fo ow or when the victims frequency arameters are im recise y known. Jammer si ze is characterized by the effective radiated ower; ERP = GjPj, where Gj is the transmit antenna gain of the jammer and Pj is the jammer ower. Passive ECM is synonymous with chaff, decoys, and other ref ectors that require no rime ower. The chaff is made of e ementa assive ref ectors that can be f oated or otherw ise sus ended in the atmos here or exoatmos here for the ur ose of confusing, s creening, or otherwise adverse y affecting the victim e ectronic system. Exam e s are meta foi s, meta -coated die ectrics (a uminum, si ver, or zinc over fibe rg ass or ny on Down oaded from Digita Engineering Library @ McGraw-Hi (www.digita engineerin g ibrary.com) Co yright 2008 The McGraw-Hi Com anies. A rights reserved. Any use is subject to the Terms of Use as given at the website.

E ectronic Counter-Countermeasures 24.6 RADAR HANDBOOK being the most common), string ba s, ro e, and semiconductors.3,13 Chaff consis ts of di o es cut to a roximate y a ha f wave ength of the radar frequency. It is suitab y ackaged to contain a broad range of di o e engths designed to be e ffective over a wide frequency band. The basic ro erties of chaff are effective scatter area, the character and time of deve o ment of a chaff c oud, the s ect ra of the signa s ref ected by the c oud, and the width of the band that concea s the target.3,9,21,22 From a radar view oint, the ro erties of chaff are very simi ar to those of weather c utter, exce t that its broadband in frequency can extend down to VHF. The mean do er frequency of the chaff s ectrum is determin ed by the mean wind ve ocity, whi e the s ectrum s read is determined by wind tu rbu ence and a shearing effect due to different wind ve ocities as a function of a titude.3 Decoys, which are another ty e of assive ECM, are a c ass of hysic a y sma radar targets whose radar cross sections (RCS) are genera y enhanced by using ref ectors or a Luneburg ens to simu ate fighter or bomber aircraft. The objective of decoys is to cause a di ution of the assets of the defensive sy stem, thereby increasing the survivabi ity of the enetrating aircraft. However, when the decoys grow too arge, they have to be engaged if they are thought ar ge enough to carry a wea on. Penetration Aid (Penaids) cou d be used by incoming ba istic missi es (BMs).23 Penaid decoys are on y one of severa ossib e ena ids. A decoy rovides another target that the defense has to hand e if the defen se cannot distinguish a decoy from a re-entry vehic e. The other major ty e of a ctive jammer is dece tive ECM (DECM). Dece tion is the intentiona and de iberat e transmission or retransmission of am itude, frequency, hase, or otherwise mo du ated intermittent or CW signa s for the ur ose of mis eading in the inter re tation or use of information by e ectronic systems.3,13 The categories of dece t ion are mani u ative and imitative. Mani u ative im ies the a teration of frien d y EM signa s to accom ish dece tion, whereas imitative consists of introducin g radiation into radar channe s that imitates a hosti e emission. DECM is a so d ivided into trans onders and re eaters.3 Trans onders generate noncoherent signa s that emu ate the tem ora characteristics of the actua radar return. Re eate rs generate coherent returns that attem t to emu ate the am itude, frequency, a nd tem ora characteristics of the actua radar return. Re eaters usua y requir e some form of memory for microwave signa s to a ow antici atory returns to be generated; this is usua y im emented by using a microwave acoustic memory or a digita RF memory (DRFM).3 In a DRFM system, the in ut RF signa is genera y f irst down-shifted in frequency and then sam ed with a high-s eed ana ogue-to-di gita converter (ADC). The sam es, stored in memory, can be mani u ated in am itude, frequency, and hase to generate a wide range of jamming signa s. The sto red sam es are ater reca ed, rocessed by the digita -to-ana ogue converter ( DAC), u converted, and transmitted back to the victim radar.24 The information content of an interce ted radar signa is main y carried in the hase of the sig na , and then the am itude is usua y discarded and on y the hase contribution is quantized and rocessed.25 The hase quantization is erformed by the DRFM b y means of M bits into N = 2M eve s. After the hase quantization, introduced o n the signa by the DRFM, the jamming signa is transmitted back to the victim r adar with an increasing de ay with res ect to the received radar signa . This de ay is quantized by a range gate u off (RGPO) device. A range gate stea er sy stem inear y de ays the quantized signa in order to generate a constant rangerate fa se target. The joint effect of hase and de ay quantization in DRFM can be ana yzed as re orted in Greco, Gini, and Farina.26 Other artifacts in the dec e tion signa s can be introduced by im erfections in the down-u conversion and demodu ation/modu ation of the signa erformed in the DRFM device. A detai ed a na ysis of this kind of errors is in Berger.27

Down oaded from Digita Engineering Library @ McGraw-Hi (www.digita engineerin g ibrary.com) Co yright 2008 The McGraw-Hi Com anies. A rights reserved. Any use is subject to the Terms of Use as given at the website.

E ectronic Counter-Countermeasures ELECTRONIC COUNTER-COUNTERMEASURES 24.7 The DRFM is the rinci a means to im ement a dece tion jammer; the range-gate stea er u s the radar-tracking gate from the target osition through the intro duction of a fa se target into the radars range-tracking circuits. A re eater jam mer sends back an am ified version of the signa received from the radar. The d ece tion signa , being stronger than the radars return signa , ca tures the range -tracking circuits. The dece tion signa is then rogressive y de ayed by using the DRFM, thereby wa king the range gate off the actua target (RGPO technique). W hen the range gate is sufficient y removed from the actua target, the dece tion jammer is turned off, forcing the tracking radar into a target reacquisition mo de.3 Another form of dece tion is the ve ocity gate u off (VGPO); a combinati on of RGPO and VGPO is a so ossib e. Another DECM technique is ca ed inverse-g ain jamming; it is used to ca ture the ang e-tracking circuits of a conica -scan tracking radar.3,13 This technique re eats a re ica of the received signa wit h an induced am itude modu ation that is the inverse of the victim radars combin ed transmitting and receiving antenna scan atterns. Against a conica y scannin g tracking radar, an inverse-gain re eater jammer has the effect of causing osi tive feedback, which ushes the tracking-radar antenna away from the target rath er than toward the target. Inverse-gain jamming and RGPO are combined in many ca ses to counter conica -scan tracking radars.3 The vu nerabi ity of conica scan to such countermeasures motivates the use of mono u se trackers that are a most a ways used in mi itary tracking radars. A different form of DECM used against t he main beam of survei ance radar attem ts to cover the targets skin return with a wide u se in order to confuse the radars signa - rocessing circuitry into su ressing the actua target return. What the radar can do against DECM is discuss ed ater; see Section 24.11. In the de oyment of ECM, severa c asses can be si ng ed out.3 In the stand-off jammer (SOJ) case, the jamming atform remains c o se to but outside the etha range of enemy wea on systems and jams these system s to rotect the attacking vehic es. Stand-off ECM systems em oy high- ower noi se jamming that must enetrate through the radar antenna receiving side obes at ong ranges. Escort jamming is another ECM tactic in which the jamming atform accom anies the strike vehic es and jams radars to rotect the strike vehic es. Mutua -su ort, or coo erative, ECM invo ves the coordinated conduct of ECM by c ombat e ements against acquisition and wea on contro radars. One advantage of m utua -su ort jamming is the greater ERP avai ab e from a co ection of atform s in contrast with a sing e atform. However, the rea va ue of mutua -su ort jamming is in the coordinated tactics that can be em oyed. A favorite tactic em oyed against tracking radars, for exam e, is to switch between jammers ocate d on se arate aircraft within the radars beamwidth. This b inking has the effect of introducing artificia g int into the radar tracking circuits, which, if intr oduced at the ro er rate (ty ica y 0.1 to 10 Hz), can cause the radar to break ang e track. In addition, b inking has the desirab e effect of confusing radiat ion homing missi es that might be directed against the jammer radiations.3 Stand -forward jamming is an ECM tactic in which the jamming atform is ocated betwe en the wea on systems and the strike vehic es and jams the radars to rotect the strike vehic es. The stand-forward jammer is usua y within the etha range of defensive wea on systems for a considerab e time. Therefore, on y the use of re ative y ow-cost remote y i oted vehic es might be ractica ; they can assist strike aircraft or missi es in enetrating radar-defended areas by jamming, ejec ting chaff, dro ing ex endab e jammers or decoys, acting as decoys themse ves, and erforming other re ated ECM tasks. Down oaded from Digita Engineering Library @ McGraw-Hi (www.digita engineerin g ibrary.com) Co yright 2008 The McGraw-Hi Com anies. A rights reserved. Any use is subject to the Terms of Use as given at the website.

E ectronic Counter-Countermeasures 24.8 RADAR HANDBOOK A se f-screening jammer (SSJ) is used to rotect the carrying vehic e. This situ ation stresses the ca abi ity of an ECM system re ating to its ower, signa - ro cessing, and ESM ca abi ities. Se f- rotection (SP) decoy jamming is an off-boar d technique intended to create ang e dece tion by causing a missi e seeker to tr ansfer ang e track from the target to a decoy. Consequent y, the missi e guides toward the decoy and away from the target. A se f- rotection decoy is most ike y to be used by arge fighter/attack and bomber aircraft. The SP decoys are ex e ndab e or towed. Ex endab e decoys are ejected (or dro ed) from the aircraft wh ereas towed decoys are tethered behind the aircraft. Ex endab e decoys contain m iniature jamming systems that are sma enough to fit into a standard chaff/f ar e dis enser. The decoy orients itse f to the air stream by de oying ow-drag ae rodynamic fins sufficient to maintain stab e f ight. The decoy diverges from the ve ocity vector of the aunch aircraft by natura dece eration as a resu t of a ir stream and fa ing due to gravity. The decoy ty ica y starts radiating jammi ng signa s toward the missi e seeker immediate y after ejection from the aircraf t and continues radiation throughout its f ight. Decoy ejection is ty ica y com menced when the RWR detects incoming radar-guided missi es. Mu ti e decoys are sometimes dis ensed at redetermined rates in order to im rove the cumu ative r obabi ity of aircraft surviva . A towed decoy is a sma aerodynamica y stab e body that houses a miniature jammer. The decoy is de oyed by ree ing it out on a cab e behind the aircraft to a fixed distance or offset. This offset is chosen such that even if a missi e hits the decoy, the aircraft wi not be damaged. T he decoy can either be owered by the aircraft via the cab e or be se f owered. Besides roviding ower to the decoy, the cab e can a so be used as a data ink to contro jammer o eration. Once de oyed, the towed decoy can begin radiating jamming signa s toward the missi e seeker. When the towed decoy is no onger ne eded, it is either ree ed in or jettisoned. The major drawback with towed decoys is that they might severe y degrade aircraft maneuverabi ity. According to the atform, the jammer can be c assified as airborne, missi e-borne, nava -based, or ground-based. A s ecia c ass of missi e-borne threat is the anti-radiation m issi e (ARM), having the objective of homing on and destroying the victim radar. The sorting and acquisition of radar signa s is re iminari y made by an ESM sy stem; afterward it cues the ARM, which continues homing on the victim radar by m eans of its own antenna, receiver, and signa rocessor. Acquisition de ends on the direction of arriva (DoA), o erating band, carrier frequency, u se width, PRI, scan rate, and other arameters of the victim radar. An ARM homes on the co ntinuous radiation from the radar side obes or on the f ash of energy from the m ain beam. ARM benefits from the one-way-on y radar signa attenuation. However, ARM receiver sensitivity is affected by mismatching osses; accuracy in ocating the victim radar is affected by the imited dimension of the ARM antenna.

24.5 OBJECTIVES AND TAXONOMY OF ECCM TECHNIQUES The rimary objective of ECCM techniques when a ied to a radar system is to a ow the accom ishment of the radar intended mission whi e countering the effect s of the enemys ECM. In greater detai , the benefits of using ECCM techniques may be summarized as fo ows: (1) revention of radar saturation, (2) enhancement o f the signa -to-jamming ratio, (3) discrimination of directiona interference, ( 4) rejection of Down oaded from Digita Engineering Library @ McGraw-Hi (www.digita engineerin g ibrary.com) Co yright 2008 The McGraw-Hi Com anies. A rights reserved. Any use is subject to the Terms of Use as given at the website.

E ectronic Counter-Countermeasures ELECTRONIC COUNTER-COUNTERMEASURES 24.9 fa se targets, (5) maintenance of target tracks, (6) counteraction of ESM, and ( 7) radar system survivabi ity.3 There are two broad c asses of ECCM: e ectronic techniques (Section 24.624.9) and o erationa doctrines (Section 24.10). S ecific e ectronic techniques take ace in the main radar subsystems, name y, the ante nna, transmitter, receiver, and signa rocessor. Tab e 24.1 shows a categorizat ion of some ECCM techniques a ong with the ECM techniques that are used to count er.5,28 Suitab e b ending of these ECCM techniques can be im emented in the var ious ty es of radars, as discussed in Section 24.11. The ensuing descri tion is imited to the major ECCM techniques; the reader shou d be aware that an a habe tica y isted co ection of 150 ECCM techniques and an encyc o edia of ECCM tac tics and techniques can be found in the iterature.8,29 Many other references de scribe the ECCM rob em, among which S ocumb and West,5 Maksimov et a .,21 Gros et a .,30 and Johnson and Stoner31 are worth noting. ECCM Techniques Versus ECM Technique Countered (Re roduced with ermission from S ocumb and West5 Artech House 2000 and G. V. Morris28) TABLE 24.1

ECM Technique Category Countered Radar Subsystem Fa se Target e Gate Pu Off Pu Off

Noise Ang e Antenna re ated Transmitter re ated Receiver re ated

Down oaded from Digita Engineering Library @ McGraw-Hi (www.digita engineerin g ibrary.com) Co yright 2008 The McGraw-Hi Com anies. A rights reserved. Any use is subject to the Terms of Use as given at the website.

Signa

rocessing re ated

ECCM Technique Low or u tra- ow side obes Mono u se ang e tracking Low cross- o arized res onse SLB (side obe b anking) SLC (side obe cance er) E ectronic scan Ada tive receive o arization Cross o arization cance ation Low cross- o arize d antenna High ower Pu se com ression Frequency diversity Frequency agi ity PRF jitter RGPO memory nu ing Bandwidth ex ansion Beat frequency detector Cover u se channe rocessing Home-on-jam Leading/trai ing edge track Narrowband do e r noise detector Ve ocity guard gates VGPO reset Signa rea ism Acce eration im iting Censored or ordered statistic CFAR Do er/range rate com arison Time aver age CFAR Tota energy test

E ectronic Counter-Countermeasures 24.10 RADAR HANDBOOK 24.6 ANTENNA-RELATED ECCM Because the antenna re resents the transducer between the radar and the environm ent, it is the first ine of defense against jamming. The directivity of the ant enna in the transmission and rece tion hases a ows s ace discrimination to be used as an ECCM strategy. Techniques for s ace discrimination inc ude antenna co verage and scan contro , reduction of main-beam width, ow side obes, side obe b anking, side obe cance ers, and ada tive array systems. Some of these technique s are usefu during transmission, whereas others o erate in the rece tion hase. Additiona y, some are active against main-beam jammers, and others rovide ben efits against side obe jammers. B anking or turning off the receiver whi e the r adar is scanning across the azimuth sector containing the jammer or reducing the scan sector covered are means to revent the radar from ooking at the jammer. Certain dece tion jammers de end on antici ation of the beam scan or on know edg e or measurement of the antenna scan rate. Random e ectronic scanning effective y revents these dece tion jammers from synchronizing to the antenna scan rate, thus defeating this ty e of jammer. A highgain antenna can be em oyed to s ot i ght a target and burn through the jammers. An antenna having mu ti e beams can a so be used to a ow de etion of the beam containing the jammer and sti maint ain detection ca abi ities with the remaining beams. A though they add com exit y, cost, and ossib y weight to the antenna, reduction of main-beam width and co ntro of coverage and scan are va uab e and worthwhi e ECCM features of a rada rs. If an air defense radar o erates in a severe ECM environment, the detection range can be degraded because of jamming entering the side obes. On transmit, th e energy radiated into s atia regions outside of the main beam is subject to be ing received by enemy RWRs or ARMs. For these reasons, ow side obes are desirab e on both receive and transmit (see Schrank,32 Patton,33 and Cha ter 2 in Farin a34). Sometimes the increase in main-beam width that resu ts from ow side obes worsens the rob em of main-beam jamming; this consequence shou d be carefu y c onsidered in s ecifying the antenna radiation attern. Usua y, s ecification of the side obes as a sing e number (e.g., 30 dB) means that the peak of the highes t sidelobe is 30 dB below the peak of the main beam. The average, or root mean s quare (rms), sidelobe level is often more important. For example, if 10% of the radiated power is in the sidelobes, the average sidelobe level is 10 dB, where dB refers to the number of decibels by which the average sidelobe level is below t he gain of an isotropic (ideal) radiator. In theory, extremely low sidelobes can be achieved with aperture illumination functions that are appropriately tapered . This leads to the well known tradeoffs among gain, beamwidth, and sidelobe lev el.35 In order to keep the beamwidth small with low sidelobes, a larger and cost ly (the cost could not be that large unless the radar uses an active aperture) a ntenna is needed. The chief problem with the low sidelobe antenna in its early d ays was that it had more mechanical problems because it was a waveguide array an d not a reflector. Other design principles involved in low antenna sidelobes are the use of radar absorbent material about the antenna structure, the use of a f ence on ground installations, and the use of polarization screens and reflectors . This means that very low sidelobe antennas are costly in terms of size and com plexity when compared with conventional antennas of similar gain and beamwidth c haracteristics. Second, as the design sidelobes are pushed lower and lower, a po int is reached where Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEASURES 24.11 minor error contributions to scattered energy (random errors) or misdirected rad iation (systematic errors) become significant. In practice, peak sidelobe levels as low as 30 to 35 dB (average level, 5 to 20 dB) can be readily realized with phas edarray antennas that electronically scan. To obtain sidelobes at levels 45 dB do wn from the main beam (average level, below 20 dB), the total phase error budget is required to be in the order of 5 rms or less. This is difficult in arrays that electronically scan: the errors induced by phase shifters, active components, a nd feed elements must be included in this budget. Arrays have been realized in p ractice that have peak sidelobes in the vicinity of the 45 dB level; however, the se are generally mechanically scanned, and the low error budgets are achieved by using all passive feed components. Phased array developments, which do scan ele ctronically, also foresee fairly good sidelobe performance; see the following re ferences3640 for a view of relevant developments. Two additional techniques to pr event jamming from entering through the radars sidelobes are the so called sidelo be blanking (SLB) and sidelobe canceler (SLC). An example of the practical effec tiveness of the SLB and SLC devices is presented in the literature, where the pl an position indicator (PPI) display is shown for a radar, subject to an ECM, equ ipped with and without the SLB and SLC systems.31 Other discrimination means are based on polarization. The polarization characteristics of a radar can be explo ited as ECCM techniques in two ways. First, the crosspolarized pattern (i.e., th e orthogonal polarization to the main plane of polarization) of a radar antenna should be kept as low as possible consistent with radar system cost. Ratios of c opolarized main beam peak gain to cross polarized gain anywhere in the antenna p attern should be greater than 25 dB to provide protection against common cross p olarized jamming. This is thought of as an ECCM technique, but it is really no m ore than good antenna design. The cross polarized jamming in this case attacks a design deficiency in the radar. The requirement for good cross polarization des ign practice in a radar antenna system extends to any auxiliary ECCM antennas as well. If their cross polarized gains are high, ECCM techniques such as SLC and SLB may not be effective against cross polarized noise or repeater jammers.29 In the second use of polarization, the radar antenna system purposely receives the cross polarization component of the radar wave in addition to the copolarized c omponent. The two orthogonally polarized components can be used to discriminate the useful target from chaff and jammer on the basis of their different polariza tions.41 However, limited benefits (a few decibels of cancellation ratio) can be obtained at the expense of a more complex antenna system (consider, for example , a phasedarray with radiating elements able to separately receive and possibly transmit the two orthogonal components of a radar wave) and of a duplication of the receiver and signal processing. Sidelobe Blanking (SLB) System. The purpose of an SLB system is to prevent the detection of strong targets and interference pulses (as they might appear after pulse compression) entering the radar receive r via the antenna sidelobes. Thus, SLB is mainly used to eliminate interference from other pulse transmissions and deliberate pulse like jamming. Also, SLB is e ffective against coherent repeater interference (CRI); here coherent means that th e interference tries to mimic the coded waveform radiated by the radar appearing as a spike signal after pulse compression.4245,34 A method of achieving this is to employ an auxiliary antenna coupled to a parallel receiving channel so that t wo signals from a single source are available for comparison. Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter Countermeasures 24.12 RADAR HANDBOOK FIGURE 24.1a Main and auxiliary antenna patterns for the SLB (after L. Maisel42 IEEE 1968) By suitable choice of the antenna gains, one may distinguish signals entering th e sidelobes from those entering the main beam, and the former may be suppressed. Figure 24.1a illustrates the radiation pattern of the main antenna together wit h a low gain auxiliary antenna. An implementation of the SLB processor is shown in Figure 24.1b, where the square law detected outputs of the two channels, idea lly identical except for the antenna patterns, are compared. The comparison is m ade at each range bin for each pulse received and processed by the two parallel channels. Thus, the SLB decides whether or not to blank the main channel on a si ngle sweep basis and for each range bin. A target A in the main beam will result in a large signal in the main receiving channel and a small signal in the auxil iary receiving channel. A proper blanking logic allows this signal to pass. Targ ets and/or jammers J situated in the sidelobes give small main but large auxilia ry signals so that these targets are suppressed by the blanking logic. It is ass umed that the gain GA of the auxiliary antenna is higher than the maximum gain G sl of the sidelobes of the radar antenna. The performance of the SLB may be anal yzed by looking at the different outcomes obtained as a consequence of the pair (u, v) of the processed signals (see Figure 24.1b). FIGURE 24.1b Scheme of sidelobe blanking system (after L. Maisel42 IEEE 1968) Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEASURES 24.13 Three hypotheses have to be tested: (1) the null hypothesis H0 corresponding to the presence of noise in the two channels, (2) the H1 hypothesis pertaining to t he target in the main beam, and (3) the H2 hypothesis corresponding to target or interference signal in the sidelobe region. The null and H1 hypotheses correspo nd to the usual decisions of no detection and target detection, respectively. The bl anking command is delivered when H2 is detected. SLB performance can be expresse d in terms of the following probabilities: (i) The probability PB of blanking a jammer in the radar sidelobes, which is the probability of associating the recei ved signals (u, v) with H2 when the same hypothesis is true; PB is a function of the jammer to noise ratio (JNR) value, the blanking threshold F, and the gain m argin b = GA/Gsl of the auxiliary antenna with respect to the radar antenna side lobes. (ii) The probability PFA of false alarm, which is the probability of asso ciating the received signals (u, v) with the hypothesis H1 when the true hypothe sis is H0; PFA is a function of the detection threshold a normalized to the nois e power level and of the blanking threshold F. (iii) The probability PD of detec ting a target in the main beam, which is the probability of associating the rece ived signal (u, v) with H1 when the same hypothesis is true; PD depends, among o ther things, on the signal to noise power ratio SNR, PFA, and the blanking thres hold F. (iv) The probability PFT of detecting a false target produced by a jamme r entering through the radar sidelobes. PFT is the probability of associating (u , v) with H1 when H2 is true; it is a function of JNR, the thresholds a and F, a nd the gain margin b. (v) The probability PTB of blanking a target received in t he main beam. This is the probability of associating (u, v) with H2 when H1 is t he true hypothesis. PTB is related to SNR, F, and the auxiliary gain w = GA/Gt n ormalized to the gain Gt of the main beam. To complete the list of parameters ne eded to describe the SLB performance, the last figure to consider is the detecti on loss L on the main beam target. This can be found by comparing the SNR values required to achieve a specified PD value for the radar system with and without the SLB. L is a function of many parameters such as PD, PFA, F, GA, JNR, and b. A numerical evaluation of these performance parameters can be found in the liter ature (specifically Chapter 3 of Farina,34among others4250). The SLB design requi res the selection of suitable values for the following parameters (Chapter 3 of Farina,34): (i) the gain margin b and then the gain GA of the auxiliary antenna, (ii) the blanking threshold F, and the normalized detection threshold a. The a priori known parameters are hypothesized to be the radar sidelobe level Gsl and the values of SNR and JNR. The design parameters can be selected by trying to ma ximize the detection probability PD while keeping at prescribed values the proba bilities PB and PFA and trying to minimize PFT, PTB, and L. The choice of the po sition of the auxiliary antenna has an impact on SLB performance in presence, fo r instance, of multipath; to avoid its effect, the phase centers of main and aux iliary antennas should be positioned at the same height with respect to the terr ain surface. In modern radar, the blanking of sidelobe impulsive interference ma y be achieved by the comparison of signals, pertaining to the same cells of the range filter map (RFM) of main beam and SLB channels. The RFM is a two dimension al map collecting the radar echoes of all range cells (after pulse compression) and all doppler filters of a radar burst. The two RFMs are independently generat ed for the main and auxiliary signals and the testing of the main and auxiliary received power values is performed for all range cells and all doppler filters. This is different from a conventional SLB approach (such as the one illustrated in Figure 24.1) operating so that if interfering/repeater jammer power is detect ed at a particular range cell, then that range cell has to be effectively blanke d. The RFM based SLB logic greatly reduces Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companies. All rights reserved. Any

use is subject to the Terms of Use as given at the website.

Electronic Counter Countermeasures 24.14 RADAR HANDBOOK the risk of successfully emulating a useful target because a repeater has to app ear in the same target range cell and has to emulate the same target doppler. Si delobe Canceler (SLC) System. The objective of the SLC is to suppress high duty cycle or even continuous noise like interferences (NLI) (e.g., SOJ) received thr ough the sidelobes of the radar. This is accomplished by equipping the radar wit h an array of auxiliary antennas used to adaptively estimate the DoA and the pow er of the jammers and, subsequently, to modify the receiving pattern of the rada r antenna to place nulls in the jammers directions. The SLC was invented by P. Ho wells and S. Applebaum.5152 A sample of subsequent references on SLC are also in the literature.34,5355 The conceptual scheme of an SLC system is shown in Figure 24.2. The auxiliary antennas provide replicas of the jamming signals in the rada r antenna sidelobes. To this end, the auxiliary patterns approximate the average sidelobe level of the radar receiving pattern. In addition, the auxiliaries are placed sufficiently close to the phase center of the radar antenna to ensure th at the samples of the interference that they obtain are statistically correlated with the radar jamming signal. It is also noted that as many auxiliary antennas are needed as there are jamming signals to be suppressed. In fact, at least N a uxiliary patterns properly controlled in amplitude TARGET SIGNAL JAMMERS SIDELOBES A1 RX & ADC RX & ADC AUXILIARY ARRAY A2 AN RX & ADC RX & ADC W1 ADAPTIVE SYSTEM VM a + V1 V2 VN W2 WN + Z OUTPUT FIGURE 24.2 Principle of SLC operation (connection a only in the closed loop imp lementation techniques); RX: receiver Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEASURES 24.15 and phase are needed to force to zero the main antenna receiving pattern in N gi ven directions. The auxiliaries may be individual antennas or groups of receivin g elements of a phased array antenna. The amplitude and phase of the signals del ivered by the N auxiliaries are controlled by a set of suitable weights: denote the set with the N dimensional vector W = (W1, W2, , WN). The jamming signal is c anceled by a linear combination of the signals from the auxiliaries and the main antenna. The problem is to find a suitable means of controlling the weights W o f the linear combination so that the maximum possible cancellation is achieved. Owing to the stochastic nature of the jamming signals in the radar and in the au xiliary channels and to the hypothesized linear combination of signals, it is ad visable to resort to the techniques of linear prediction theory for stochastic p rocesses. Denote with VM, the radar signal at a certain range bin and with V = ( V1, V2, , VN) the N dimensional vector containing the set of signals, at the same range bin, from the N auxiliary antennas. It is assumed that all the signals ha ve bandpass frequency spectra; therefore, the signals can be represented by thei r complex envelopes, which modulate a common carrier frequency that does not app ear explicitly. The jamming signals in the channels may be regarded as samples o f a stochastic process having zero mean value and a certain time autocorrelation function. For linear prediction problems, the set of samples V is completely de scribed by its N dimensional covariance matrix M = E(V*VT), where E() denotes the statistical expectation, the asterisk ()* indicates the complex conjugate, and V T is the transpose vector of V. The statistical relationship between VM and V is mathematically represented by the N dimensional covariance vector R = E(VMV*). The optimum weight vector W is determined by minimizing the mean square predicti on error, which equals the output residual power: PZ = E{| Z |2} = E{| VM WT V | 2} (24.2) where Z is the system output. It is found that the following fundamental equatio n applies: W = mM 1R (24.3) where m is an arbitrary ured by introducing the the output noise power V |2} = E{| VM |2} RT

constant value. The benefit of using the SLC can be meas jammer cancellation ratio (JCR), defined as the ratio of without and with the SLC: JCR = E{| VM |2} E{| VM |2} T M 1R * E{| VM W (24.4)

By applying Eqs. 24.3 and 24.4 to the simple case of one auxiliary antenna and o ne jammer, the following results are found: E{VM VA *} W= = E{| VA |2 } JCR = 1 1 | |2 (24.5) It is noted that the optimum weight is elated to the co elation coefficient between the main signal, VM, and the auxilia y signal, VA; high values of the co elation coefficient p ovide high values of JCR. The p oblem of implementing th e optimum weight set (Eq. 24.3) is essentially elated to the eal time estimati on of M and R and to the inve sion of M. Seve al p ocessing

ownloaded f om igital Enginee ing Lib a y @ McG aw Hill (www.digitalenginee in glib a y.com) Copy ight 2008 The McG aw Hill Companies. All ights ese ved. Any use is subject to the Te ms of Use as given at the website.

Elect onic Counte Counte measu es 24.16 RA AR HAN BOOK schemes have been conceived that may be classified in two main catego ies: (1) c losedloop techniques, in which the output esidue (connection a in Figu e 24.2) is fed back into the adaptive system; and (2) di ect solution methods, often ef e ed to as open loop, which ope ate just on the incoming signals VM and V. B oa dly speaking, closed loop methods a e cheape and simple to implement than di e ct solution methods; one of seve al p actical implementations is desc ibed in G iffiths.56 By vi tue of thei self co ecting natu e, they do not equi e compon ents that have a wide dynamic ange o a high deg ee of linea ity, and so they a e well suited to analogue implementation. Howeve , closed loop methods suffe f om the fundamental limitation that thei speed of esponse must be est icted i n o de to achieve a stable and not noisy steady state. i ect solution methods, on the othe hand, do not suffe f om p oblems of slow conve gence but, in gene al, equi e components of such high accu acy and wide dynamic ange that they c an only be ealized by digital means. Of cou se, closed loop methods can also be implemented by using digital ci cuit y; in which case, the const aints on nume ical accu acy a e g eatly elaxed, and the total numbe of a ithmetic ope ations is much educed by compa ison with di ect solution methods. The majo ity of imp lementations has become open loop with digital technology. P actical conside ati ons (see Chapte 4 of Fa ina34 fo a detailed analysis) often limit the SLC null ing capabilities to a JCR of about 30 to 40 dB, but thei theo etical pe fo manc e is potentially much highe . Adequate cancellation of the di ectional inte fe e nce is obtained if the eceiving channels a e p ope ly matched in amplitude and phase ac oss the ada eceiving bandwidth. This condition is necessa y to att i bute the amplitude and phase diffe ences measu ed ac oss the channels only to th e natu e (powe and oA) of the impinging inte fe ence. The e a e seve al sou ce s of mismatching; the impe fect matching of the analogue eceiving channels is o ne of the main limitations to the inte fe ence cancellation. The effect of this mismatch on the JCR has been studied in the lite atu e; see Fa ina57 and efe en ces the ein. Fo contempo a y p esence of amplitude and phase mismatches, the JC R has an exp ession that is de ived in Appendix 2.57 A nume ical application of this equation is shown in Figu e 24.3; the pa amete values of the study case a e quoted in Fa ina.57 Figu e 24.3 shows the JCR contou cu ves ve sus the no mal ized amplitude an and the phase b (deg ees) mismatches of the analogue eceiving channels (see Fa ina57 fo the p ecise definition of these pa amete s). It is s een that to have 40 dB of JCR, one needs to specify tight equi ements fo both amplitude (below 1%) and phase (below 0.7) mismatches. This figu e motivates the need to eso t to equalization digital filte s to compensate fo the mismatches of the auxilia y channels (in thei analogue pa t) with espect to the main chan nel. This subject is cove ed in Fa ina57 and efe ences the ein. Examples of oth e possible limitations to cancellation a e listed below34,53,58,59: 1. Mismatch between the main and auxilia y signals including the p opagation paths, the pat te ns of the main and auxilia y antennas, the paths inte nal to the system up to the cancellation point, and the c osstalk between the channels6062 2. The limite d numbe of auxilia y channels adopted in a p actical system as compa ed with th e numbe of jamming signals 3. Ape tu e f equency dispe sion, often exp essed in te ms of ape tu e bandwidth p oduct37,59,63 4. The limited bandwidth of the maj o ity of the schemes implementing Eq. 24.3, as compa ed with the wideband of a b a age jamme that can be ega ded as a cluste , sp ead in angle, of na owband jamme s 5. Quad atu e e o s in synch onous (i.e., I, Q) detecto s6466 ownloaded f om igital Enginee ing Lib a y @ McG aw Hill (www.digitalenginee in glib a y.com) Copy ight 2008 The McG aw Hill Companies. All ights ese ved. Any use is subject to the Te ms of Use as given at the website.

Elect onic Counte Counte measu es ELECTRONIC COUNTER COUNTERMEASURES 24.17 2 30 1.5 25 b () 1 35 30 25 0.5 40 0 0 35 0.02 30 0.04 an 0.06 25 0.08 FIGURE 24.3 Contou cu ve of JCR (dB) ve sus the amplitude (in natu al numbe , a long the ho izontal axis) and phase (in deg ees, along the ve tical axis) mismat ches of the analogue eceiving channels 6. igital eceive channel e o s such as A C quantization, sample/hold jitte , and digital conve te offset67,68 7. The pulse width that limits the eaction t ime of the adaptive system, in o de to avoid the cancellation of ta get signal 8. The ta get signal in the auxilia y a ay that may esult in nonnegligible ste e ing of the auxilia ies towa d the main beam di ection 9. Multipath delay, ofte n exp essed in te ms of delay bandwidth p oduct6970 10. The p esence of clutte t hat, if not p ope ly emoved, may captu e the adaptive system, giving ise to nu lls along di ections diffe ent f om those of the jamme s37,71 11. The t adeoff t hat has to be sought between the accu acy of weights estimation and the eaction time of the adaptive system 12. The limited numbe of time samples available to estimate the jamme cova iance mat ix; usually 3N sample should be available if N is the numbe of adaptive channels101 13. The antenna otation ate that migh t p oduce a fast time va ying powe and jamme oA101 Joint SLB and SLC. SLB is effective against spiky signal afte pulse comp ession (like CRI), whe eas SLC c ombats the continuous NLI. As p eviously stated, both techniques combat the inte fe ences impinging on the main antenna sidelobes. The two techniques can be joi ntly used against the simultaneous p esence of CRI and NLI. An app oach is to ca scade the SLC and SLB techniques as shown in Figu e 24.4. The scheme depicts th ee eceiving channels, each one having an antenna, a eceive , and an A C; they p ovide th ee signals labeled, espectively, as SLC, MAIN, and SLB. The left han d side antenna is a low gain auxilia y pe fo ming the SLC p ocessing in ownloaded f om igital Enginee ing Lib a y @ McG aw Hill (www.digitalenginee in glib a y.com) Copy ight 2008 The McG aw Hill Companies. All ights ese ved. Any use is subject to the Te ms of Use as given at the website.

Elect onic Counte Counte measu es 24.18 Auxilia y channel RX & A C SLC W1 x Mai n channel RX & A C MAIN 1 x + MAIN ' SLB ' W2 x 1 RA AR HAN BOOK SLB channel RX & A C SLB x + MAIN' Gate To ta get detection SLB' Compa ison F blanking th eshold SLB' >F MAIN' Blanking of adapted main signal ange cell FIGURE 24.4 A p ocessing scheme inco po ating SLC and SLB devices the main and sidelobe blanking channels. The cente antenna is the high gain ad a antenna to detect ta gets notwithstanding impulsive and noise like inte fe en ces. The ight hand side antenna is a low gain auxilia y that is used fo SLB p ocessing in the main channel. The adaptive cancellation of NLI eceived by the m ain antenna is achieved by the linea combination of the SLC and MAIN signals wi th the adaptive weights W1 and 1, espectively; the esulting adapted signal MAI N' doesnt contain the NLI. Similarly, the adaptive cancellation of NLI received b y the rhs auxiliary antenna is reached by the linear combination of the SLC and SLB signals with the adaptive weights W2 and 1, respectively; the adapted signal SLB doesnt contain the NLI. Once the NLI is removed from the two channels, then the classic SLB logic can be applied against the CRI by comparing the amplitude |MAIN | of the main channel with that |SLB | of the blanking channel, which are both NLI free.72 Because the phase centers of the three antennas (the main and the two auxiliaries) are spaced, in general, more than 0.5 l (where l is the len gth of the radiated EM wave), the adapted patterns of the main and SLB channels fluctuate around average curves due to the presence of grating lobes.72 Neverthe less, a reasonable gain margin is present between the pattern of the adapted SLB and the sidelobes of the adapted main antenna; thus, an adequate probability of blanking the CRI in the presence of adaptively nulled NLI should be expected. I n order to improve the above gain margin and, consequently, the blanking probabi lity of CRI, the following processing strategies are suggested72,73: spatial and frequency diversity. Spatial Diversity. The rationale is to use two low gain au xiliaries (instead of one, as shown in Figure 24.4) for the SLB; because their p hase centers will be different, the grating lobes affecting the adapted patterns of the two SLB antennas will be different too. Taking the greater of the two ad apted SLB signals, the gain margin between the SLB and the main antenna sidelobe s will increase with a consequent improvement of the performance of the blanking logic. Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEASURES 24.19 Frequency Diversity. Another technique to improve the blanking performance is to resort to the diversity of the radar carrier frequency; in this case, we need j ust one low gain antenna (as shown in Figure 24.4) for the SLB. The radar operat es in frequency diversity mode, i.e., it radiates a burst of L pulses (T seconds apart) with slightly different carrier frequencies.73 The grating lobes in the adapted main and the L SLB patterns will change as a function of the carrier fre quency. Taking the max of the output of the L SLB signals is equivalent to a smo othing of the grating lobes. In a specific example presented in Farina and Timmo neri,73 two carrier frequencies are used and the values of d/l (where d is the int er element distance) for the array of receiving elements are respectively 0.5 an d 0.55. The blanking is separately applied on the received data at the two carri er frequencies; subsequently, the separate blanking bits are processed by a logi c OR (the global blanking logic). The ensuing Figure 24.5 displays the blanking curves for the two separate carrier frequencies and for the logic OR. It is note d that the frequency diversity and the logic OR provide an improvement of the bl anking probability; this is due to the different shapes of the antenna patterns at the two slightly different carrier frequencies. Figure 24.5 also presents the probability of blanking a useful target (PTB) received by the main antenna beam . The probabilities are estimated via 200 independent Monte Carlo simulations. T he target SNR is 20 dB; the JNR is 20 dB; the target DoA is assumed to be evenly distributed in the main beam angular interval [4, 4]; details on the numerical par ameters used in the study case are in the reference.73 It is noted that PTB is n egligible for F = 0 dB, while PB 0.9. After a careful performance evaluation of the system depicted in Figure 24.4, it might be necessary to always resort to ei ther spatial or frequency diversity to improve the SLB performance. The selectio n of one of the two diversity techniques depends on overall system consideration s related to the impact of adding more auxiliaries and/ or radiating, with the r adar, proper carrier frequencies. Furthermore, if compact and high speed process ing are requested, spatial and frequency diversity techniques can be fruitfully implemented resorting to systolic schemes. 1 0.9 0.8 0.7 PB, PTB 0.6 0.5 0.4 0.3 d/ = 0.55 0.2 0.1 10 d/ = 0.5 5 0 F (dB) 5 PTB OR d/ = 0.55 d/ = 0.5 PB OR 10 FIGURE 24.5 B anking robabi ity (PB) and target b anking robabi ity (PTB) vers us the b anking thresho d F (in dB) for the frequency diversity scheme Down oaded from Digita Engineering Library @ McGraw-Hi (www.digita engineerin g ibrary.com) Co yright 2008 The McGraw-Hi Com anies. A rights reserved. Any use is subject to the Terms of Use as given at the website.

E ectronic Counter-Countermeasures 24.20 RADAR HANDBOOK Systo ic Schemes for SLB and SLC. In the quest for efficient ara e rocessing , the systo ic schemes come into the scene; their use has been described for the im ementation of SLC and more genera ada tive array rob ems in the iteratur e34,74,75 The rationa e and the use of a systo ic array that rocesses the signa s received by SLC and main channe is re orted on ages 146156 of Farina34 and i n Farina and Timmoneri.73 Figures 1 to 473 de ict the use of a systo ic rocessi ng scheme that incor orates the SLB and SLC. The advantage of these schemes resi des in the decom osition of the com ex rocessing for ada tive cance ation of the NLI into a network of sim e rocessing e ements that can be convenient y ma ed onto a ara e rocessing architecture based either on Commercia off the She f (COTS) techno ogy or custom Very Large Sca e Integration (VLSI) devices. I n the iterature7680 it has been shown that a wide s ectrum of techno ogies can b e used ike Fie d Programmab e Gate Arrays (FPGA), Coordinate Rotation Digita C om uter (CORDIC) im emented with VLSI and o tica com uters. Pioneering work on the use of CORDIC for ada tive nu ing dates back to C. Rader at MIT-Linco n La boratory.81,82 The advantage of systo ic im ementation is high rocessing s eed and com act, ow weight, ow ower consum tion hardware. Ada tive Arrays. An ad a tive array (Figure 24.6) is a co ection of N antennas with their own receiver s (RX) and ADC, feeding a weighting and summing network, with automatic signa -d e endent weight adjustment to reduce the effect of unwanted JAMMER TARGET qS qJ d V1 V2 ... VN RX & ADC RX & ADC RX & ADC w1 x w2 x wN x + Z ||2 COMPARISON WITH THRESHOLD a

H1 H0 FIGURE 24.6 The ada tive array scheme Down oaded from Digita Engineering Library @ McGraw-Hi (www.digita engineerin g ibrary.com) Co yright 2008 The McGraw-Hi Com anies. A rights reserved. Any use is subject to the Terms of Use as given at the website.

E ectronic Counter-Countermeasures ELECTRONIC COUNTER-COUNTERMEASURES 24.21 signa s and/or to em hasize the desired signa or signa s in the summing network out ut. Out ut signa Z is enve o e-detected and com ared with a suitab e thres ho d a to detect the resence of a usefu target (see Cha ter 5 in Farina34 and other sources.5357,83,84) The ada tive array is a genera ization of the SLC descr ibed in the receding subsection. The basic theory of jammer cance ation and ta rget enhancement is considered first; attention is then focused on the fo owing to ics: main-beam jammer cance ation, target DoA estimation in resence of jam mer, two-dimensiona ada tive rocessing for joint c utter and jammer cance ati on, ada tivity at the subarray eve , and su erreso ution. The im ementation of the ada tive array conce t is more and more re ated to digita beamforming8587 a nd to digita array radar (DAR) techno ogies.88,89 Jammer Cance ation and Targe t Signa Enhancement. Ada tive array rinci es have found a thorough mathematic a treatment since the ate 1960s83,84; for a brief history of ada tive arrays, see Reed90; for an overview of east squares ada tive rocessing in mi itary a ications with ce ebration of B. Frank in meda to B. Widrow for ioneering work on ada tive signa rocessing, see Etter et a .91 The theory and a ication of ada tive array rinci es to radar is we estab ished; for a ook to o u ar ub ications see, for instance, Haykin and Steinhardt,92 Smith,93 and Farina et a .94 The basic resu t is given by the ex ression of the o timum set of weights: W = M 1S* (24.6) whe e M = E(V*VT) is the N dimensional cova iance mat ix of the ove all distu ba nce (noise and jamme ) V eceived by the a ay, and S is the N dimensional vecto containing the expected signal samples in the a ay f om a ta get along a ce t ain di ection of a ival. The simila ity of Eq. 24.6 to Eq. 24.3 gove ning the S LC is immediately ecognized. With espect to SLC, adaptive a ay techniques off e the capability of enhancing the ta get signal while canceling the distu bance . The adaptive system allocates in an optimum fashion its deg ees of f eedom to the enhancement of the ta get signal and to the cancellation of jamme . Seve al gene alizations of the basic theo y have been conside ed, including: (i) the ta get model S is not known a p io i, as it is assumed in de iving Eq. 24.6; (ii) i n addition to spatial filte ing, dopple filte ing is pe fo med to cancel clutte and chaff; and (iii) the ada platfo m is moving as in shipbo ne, ai bo ne, o even spacebo ne applications. A elevant advancement of the adaptive a ay con cept is space time adaptive p ocessing (STAP).9598 STAP may be thought of as a tw o dimensional adaptive filte that combines eceive beamfo ming and dopple filt e ing. A basic illust ation of STAP is given in Figu e 1 of Wa d95 whe e a picto ial view of the inte fe ence envi onment seen by an ai bo ne ada and the co esponding adapted two dimensional filte esponse a e shown. The powe spect al density esulting f om jamme and clutte is depicted as a function of the spati al (i.e., the sin(angle)) and the tempo al (i.e., dopple ) f equencies. Ba age noise jamming appea s as a wall localized in angle and dist ibuted all ove dopp le f equencies. The clutte echo f om a single g ound patch has a dopple f equ ency that depends on the angle between the clutte patch and the platfo m flight di ection; clutte f om all angles lies on a diagonal idge ac oss the space ti me f equency plane. A mainbeam ta get competes with both main beam and sidelobe clutte as well as jamming. ownloaded f om igital Enginee ing Lib a y @ McG aw Hill (www.digitalenginee in glib a y.com) Copy ight 2008 The McG aw Hill Companies. All ights ese ved. Any use is subject to the Te ms of Use as given at the website.

Elect onic Counte Counte measu es 24.22 RA AR HAN BOOK The STAP gene ates a space time filte ing esponse with a main beam along the ex pected dopple f equency and angle of a ival of ta get and deep nulls along the jamme wall and the clutte idge. To pe fo m STAP, the ada should have an a ay of N antennas, each with thei own eceiving channel and A C. Each channel eceives M echoes f om a t ansmitted t ain of M cohe ent pulses. Adaptivity invol ves the NM echoes. The detection p obability P fo the optimum filte of Eq. 24 .6 is fo a constant c oss section ta get model84: P = Q ( 2iS M T 1 * S, 2 ln(1 / PFA ) ) (24.7) whe e Q (. , .) is the Ma cum Q function and PFA is the p esc ibed p obability of false ala m. It is also shown that the set of weights of Eq. 24.6 p ovides the maximum value of the imp ovement facto If, which is defined as follows: If = si gnaltointe fe ence plus noise powe atio at the output signaltointe fe ence plus no ise powe atio at the input (24.8) The If value co esponding to the optimum set of weights of Eq. 24.6 is84 If = S T M 1S* ( SINR) I (24.9) The signal to inte fe ence plus noise powe atio (SINR)I is measu ed at the inp ut of a eceiving element of the a ay and efe s to one echo pulse. The If ep esents the pe fo mance of the adaptive a ay: it accounts fo the ta get signal integ ation and the inte fe ence cancellation. P actical applications of the equ ation above a e, fo instance, in Chapte 5 of Fa ina.34 C ucial fo the unde st anding of the adapted a ay patte n is the concept of eigenvalue eigenvecto dec omposition of the inte fe ence cova iance mat ix M: see again Chapte 5 of Fa in a34 and Testa and Vannicola.99 An impo tant technique that mitigates the delete ious effects of the noise eigenvecto s, thus continuing to maintain a p esc ibed level of low sidelobes in the adapted a ay patte n is the so called diagonal l oading.100,101 Adaptive a ays came about afte the successful application of SL C, the application of Eq. 24.6, and of mo e gene al and powe ful adaptive a ay concepts (e.g., GSLC: gene alized SLC34). Clea ly the efficiency of the adaptive a ay depends on the numbe of deg ees of f eedom (dof) and the accu acy of ec eiving channels (e.g., deg ees of matching). The e is some t ade off between acc u acy and numbe of channels; a system with one dof is less efficient (and equi es maximum accu acy) than a system with, say, fou dof. An adaptive system with N dof can theo etically supp ess (N 1) jamme s, ealisticallyas a ule of thumbN/ 2 o N/3. If the numbe of jamme s is highe , the adaptive a ay is still useful because some jamme supp ession is achieved with an acco dingly educed detecti on ange. The Ma cum Q function is defined as Q( a, b ) =

 b x exp whe e Io (.) is the modified Bessel function of o de 0. x2 + a2 I ( ax )dx 2 o

ownloaded f om igital Enginee ing Lib a y @ McG aw Hill (www.digitalenginee in glib a y.com) Copy ight 2008 The McG aw Hill Companies. All ights ese ved. Any use is subject to the Te ms of Use as given at the website.

Elect onic Counte Counte measu es ELECTRONIC COUNTER COUNTERMEASURES 24.23 Conce ning the compa ison between adaptive beamfo ming and a ve y low sidelobe a ntenna, it is in the impo tant case of close to main beam jamming whe e adaptive beamfo ming is supe io . On the othe hand, the adaptive a ay allows one to ob tain ce tain lowe ed sidelobes simultaneously to jamming nulling. Conce ning the p actical applicability of adaptive a ays some conside ations follow. A numbe of ope ational ada systems a e adaptive; they a e desc ibed in the technical lite atu e.3840,102 A mode n ada with digital p ocessing al eady has at least f ou digital channels (sum, diffe ence in azimuth, diffe ence in elevation, and g ua d). In gene al, the numbe of implemented eceiving channels is mainly a matt e of cost. It has been a gued that ada systems with a numbe of adaptive dof of a few tens a e al eady in ope ation in the mic owave band; the numbe of adap tive dof may be mo e in the ove the ho izon (OTH) ada . Fo the fo eseeable fu tu e, the fully adaptive a ay (i.e., with adaptivity at eceiving element level ) has only theo etical value fo antennas with a thousand elements. The e a e a da s that a e fully adaptive, but they have only a limited numbe of elements th at can be economically handled in an adaptive a ay. A ays with a la ge numbe of eceiving elements need some fo m of p ocessing eduction. One method of pa t ial adaptivity is to a ange the a ay elements in subg oups that fo m the input s of the adaptive p ocesso . Ca eful selection of the subg oup elements is neces sa y to avoid g ating lobes; this topic is discussed in a following section. Ano the simplification of the fully adaptive a ay is the dete ministic spatial fil te ing, whe e a fixed eduction of the sidelobes is ope ated in those di ections o solid angles f om which the inte fe ences a e expected to come. As an exampl e, a p obable egion with inte fe ences is the ho izon o pa t of it because jam me s a e mostly g ound based o at long ange. The weights a e computed offline, by assuming an a p io i known cova iance mat ix M, and sto ed in a memo y whe e a menu of weights is available to an ope ato o an automatic decision system (pp . 277283 of Fa ina34). Main Beam Cancellation (MBC) Systems. The objective of the MBC is to supp ess high duty cycle and NLI eceived th ough the main beam of th e ada . The conceptual scheme of MBC is analogous to the scheme of SLC; howeve , high gain beams a e employed in lieu of low gain auxilia y antennas. Jamming i s cancelled by a linea combination of the signals f om the high gain beams and the main antenna. The weights to be applied can be computed by Eq. 24.3. The cap ability to cancel a ce tain numbe of main beam inte fe ences depends on the ava ilable numbe of high gain beams. A so called fou lobed patte n can be used fo main beam inte fe ence cancellation.103,104 The use of low gain auxilia y anten nas joined to high gain beams allows the contempo aneous cancellation of sidelob e and main beam inte fe ences. Ta get oA Estimation in P esence of Sidelobe and Main Beam Inte fe ences. Phased a ay ada s a e equi ed to detect, locate, an d t ack ta gets in the p esence of natu al inte fe ence and jamming. Monopulse i s the technique of choice to dete mine the ta get angula coo dinates when ECM i s encounte ed since it is much ha de to deceive than a conical scan. Howeve , t he application of adaptive beamfo ming (to bette mitigate the p esence of an in tense jamme ) with the elated disto tion of sum and diffe ence beam shapes may int oduce e o s in the conventional monopulse technique, in pa ticula , if the jamme is close to the main beam105; thus, the conventional monopulse technique cannot be applied.106 A Maximum Likelihood (ML) app oach fo ta get oA estimati on is conside ed, which gene alizes the monopulse concept.107114,104 ownloaded f om igital Enginee ing Lib a y @ McG aw Hill (www.digitalenginee in glib a y.com) Copy ight 2008 The McG aw Hill Companies. All ights ese ved. Any use is subject to the Te ms of Use as given at the website.

Elect onic Counte Counte measu es 24.24 RA AR HAN BOOK The ta get angula coo dinatesazimuth and elevation (q, f)can be estimated by ML, also in the p esence of main beam and sidelobe jamming, by p ocessing the data eceived by a set of low and high gain beams. The set of eceived ada echoes, V bS(qT, fT) + d, depends on the angular coordinates of the target, (qT, fT), the complex target amplitude b, and white gaussian zero mean noise plus jamming dis turbance d. S is a vector containing the values of the patterns of high and low gain antennas in a certain direction (q, f). The data V are characterized by a g aussian probability density function conditioned to the target unknown parameter s, i.e., pv(V/b, qT, fT) . The ML estimation of the targets unknown parameters is obtained as follows: ( b,T , T ) = arg min b, , [V bS( , )] M 1 [V bS( , b, , {F ( b, , )} d H { }

(24.10) where Md is the disturbance covariance matrix, Md = [I + JNR S( J, J) S ( J, J)H], depending on the angular coordinates o the jamming ( J, J) and on t he 2 jamming to noise power ratio, JNR = PJ / n ; in Eq. 24.10, ()H tand for th e complex conjugate tran po e operation. The amplitude b can be eparately e tim ated by nulling the fir t derivative of the function to be minimized. By replaci ng the ampli tude e timation b into the function to be minimized, the following D oA e timator i obtained: 2 1 H , ) = argmax {U ( , )} = argmax S ( , ) i M d i V (T T , , ) 2 n (24.11) It can be noted that the numerator of the functional U(q, f) i the quared adap ted output (| S H ( , ) i M 1 i V |2 ) o a generalized array o high and low gain antenna d patterns; the denominator [S H ( , ) i M 1 i S ( , )] is a normalizing t erm that, as we d will see in a moment, plays a key role. The U unction or a c ertain pair o angles ( , ) determines, a ter comparison with a suitable thresh old, i a target is detected. The same U unctional when scanned across a suitab le set o ( , ) angle values provides, by means o E . 24.11, the target DoA es timate. We re er to E . 24.11 and its practical implementation as the generalize d monopulse techni ue. The algorithm needs the estimation o the disturbance cov ariance matrix Md, which is obtained by the radar echoes corresponding to range cells adjacent to the cell under test where a potential target is sought. The ma ximum o the U unctional can be estimated by an exhaustive search in the range o values o interest o ( , ) or by using a ast recursive algorithm.114 The r ecursion can be initialized with the angular coordinates o the main beam pointi ng. By replacing the estimated disturbance covariance matrix into the U unction al, a Constant False Alarm Rate (CFAR) detector is obtained.115 Thus, the compar ison o the U unctional with a suitable threshold permits the target detection maintaining the prescribed CFAR. Only or the range cell in which the detection occurred, the radar signals are taken and urther processed by the ML algorithm to produce the target DoA estimate. The per ormance o the ML estimation algorit hm o target DoA can be studied by resorting to the Cramer Rao Lower Bound (CRLB ) analysis and Monte Carlo simulations.104,107109,113,114 In these studies, it is shown that the shape o the U unctional Here, just one jammer is considered; but the mathematical approach is easily ext ended to more than one jammer. Downloaded rom Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companies. All rights reserved. Any

use is subject to the Terms o Use as given at the website.

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEASURES 24.25 depicts the presence o the target as well as o the jamming. It has been demons trated that Monte Carlo simulation is in close agreement with CRLB analysis. It has been ound that the use o the our lobed antenna pattern in addition to the conventional monopulse beams (sum, di erence in azimuth, and elevation) can im prove the estimation o the target DoA in presence o a jammer. Joint Adaptive J ammer and Clutter Cancellation. Clutter, always present in a radar, negatively a ects the per ormance o the adaptive jammer cancellation; there ore, means hav e to be adopted to e ectively contrast the contemporaneous presence o clutter and jammer. When heavy clutter is present, the SLC and adaptive array will attem pt to minimize the power in the adapted output without di erentiating between c lutter and other orms o inter erence. In other words, the adapted pattern will contain nulls steered in the direction o the main beam antenna. A number o te chni ues may be used to avoid the problems raised by the presence o clutter. A techni ue, particularly suitable or low PRF (Pulse Repetition Fre uency) radar, avoids the in luence o close in clutter returns on adaptive weights by simply selecting or adaptation the clutter ree ranges at the end o each PRI. This te chni ue does not apply to radars operating in high PRF range ambiguous modes wit h signi icant clutter in all range cells. I the clutter and jammer cannot be se parated, either in range or doppler domains, then a two dimensional (in doppler re uency and angle) adaptive ilter might be re uired; this is particularly tru e when the statistical eatures o clutter and jammer are not known a priori. In act, when either the jamming or clutter statistics cannot be estimated indepen dently o one another, it becomes di icult to design an e ective spatial adapt ive ilter or jamming rejection or a temporal adaptive ilter or clutter mitig ation since the presence o one contaminates the estimation process or the othe r.116 This problem is most accentuated when the clutter to jamming ratio approac hes unity; in which case, the cascade o a spatial and a temporal adaptive proce ssors may per orm poorly. In such situations, a joint two dimensional adaptive iltering in doppler and angle domains represents a means to cancel the composite disturbance (i.e., the superposition o jamming and clutter) jointly rather tha n se uentially.117 The per ormance advantages o twodimensional adaptivity shall be traded o with the computational cost. To reduce the computational load, di erent computational strategies may be devised, or example, by calculating the adapted two dimensional weights at a rate lower than the input data and applyin g them to the radar snapshots at their natural rate. An e icient algorithmic pr ocedure to extract the weights, named Inverse QR, is detailed in Bollini et al.7 5 Other possibilities are to use modern computing technologies like the FPGA, Po wer PC, or high speed optical processor to support the two dimensional adaptive processing.80 Adaptivity at the Sub Array Level. For an operational phased array radar (PAR) with thousands o elements, it is not possible to adapt directly th e signals rom each radiating element. It is necessary to reduce the system comp lexity by using sub arrays. A sub array is an aggregation o antenna elementary radiators; the whole antenna can be considered as an array o these super elemen ts. Adaptive processing can be applied at the output signals o each sub array, thus reducing the system complexity. Provided that the sub arrays are con igured reasonably, the number o sub arrays and the receiving channel errors (e.g, cha nnel mismatching) determine the cancellation per ormance. Thus, the number o su b arrays is a trade o between hardware complexity, cost, and achievable per or mance. Downloaded rom Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companies. All rights reserved. Any use is subject to the Terms o Use as given at the website.

Electronic Counter Countermeasures 24.26 RADAR HANDBOOK It is highly desirable in PAR to have low sidelobes; this is obtained by (i) ix ed weighting layer with analogue technology (i.e., at the microwave element stag e) to reduce the sidelobe level everywhere; (ii) ixed weights at the digital su b array level to reach a prescribed peak to sidelobe ratio (PSLR); and (iii) an adaptive weighting layer with digital technology to put nulls along the jammer D oA o high directional beams (sum, di erence, cluster o high gain patterns) an d low gain, possibly omnidirectional, beams (e.g., guard channel: ). Figure 24.7 presents a simpli ied scheme o a modern PAR. Formation o Sum and Di erence Pa tterns. Consider the problem o how to orm sum and di erence beams with prescr ibed low sidelobes in a PAR with sub arrays. A strategy is to apply a tapering a t element level (i.e., in the analogue receiving section where one attenuator is generally available per element; thus one taper unction is available to achiev e reasonable low sidelobes or both sum and di erence beams). Subse uently a i xed digital taper, a ter the ormation o sub arrays, is applied with a set o w eights or the sum and a di erent set o weights or each di erence channel. T his is schematically illustrated in Figure 24.8 or a uni orm linear array (ULA) that generates a sum and a di erence channel. The igure depicts a ULA with 24 receiving elements clustered into our not overlapping and not regular sub arra ys.118 The calculation o analogue taper is made by resorting to the nulling o ictitious wide angle jamming, which occupies the whole angular sector where sid elobes o sum and di erence beams have to be kept low. In Farina et al.,118 it was ound that the analogue tapering is a compromise between the Taylor (which i s the best taper or the sum beam) and the Bayliss (which is the best taper or the di erence beam), the degree o compromise being regulated by amount o ict itious JNR selected or the sum and di erence beams. A numerical example, repor ted in Farina et al.118, or a ULA o N = 24 elements and a uni orm distribution o the ictitious jammer out o the main beams o the sum and di erence beams, gives a PSLR o 17.5 dB and 16.5 dB or the sum and di erence beams, respectiv ely. The next step is to derive the ixed tapers at a digital level or sum and di erence beams; a suitable techni ue is described in Nickel.119,120 The ration ale o the approach Thousands o radiating elements Tapering weight wi j j ... j j Analogue microwave layer ... j j RX & ADC T/R modules j Phase shifters Data rate reduction j j RX & ADC RX & ADC Hundreds of sub-arrays

... Digital layer with fixed weights ... ... Digital layer with adaptive weights To further digital processing and display everal tens of beams

Tens of beams , , , , ... FIGURE 24.7 Scheme o a PAR Downloaded rom Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companies. All rights reserved. Any use is subject to the Terms o Use as given at the website.

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEASURES 24.27 24 Analogue section 1 ... j 8 9 ... 12 13 ... 16 17 ... Tapering wi at element level j ... j j ... j j ... j j ... Phase shi ters Sub array 1 ormation Sub array 2 ormation Sub array 3 ormation Sub array 4 ormation RX & ADC RX & ADC RX & ADC RX & ADC

Digital tapering for difference beam formation FIGURE 24.8 Example of a ULA with sub a ays that gene ate sum and diffe ence ch annels is to obtain the sum beam by compensating, at the sub a ay level, the analogue tape at the element level to achieve an ove all tape mo e simila to the Taylo one; this is obtained by inc easing the cont ibution of cent al sub a ay weig hts (i.e., the suba ays 2 and 3 shown in Figu e 24.8) with espect to the weigh ts of the side sub a ays 1 and 4. To obtain the diffe ence beam, the analogue t ape ing at the element level is compensated, at the sub a ay level, to achieve an ove all tape function mo e simila to the Bayliss one; this is obtained by d ec easing the cont ibution of the cent al suba ays 2 and 3. A nume ical example is epo ted in Fa ina et al.118 with an ULA of N = 24 elements and M = 4 sub a ays. The chosen weight is a Taylo tape ing with 30 dB of PSLR. The e a e only 4 digital deg ees of f eedom; this means that ma ginal imp ovement of pe fo manc e in te ms of PSLR can be achieved. Neve theless, a PSLR of 25 dB was obtained f o the combination of 24 analogue weights and 4 digital weights. Fo the same UL A about 20 dB of PSLR was obtained fo the diffe ence channel. Conside ations Re lated to Sub a ay Adaptivity. Tape ing at the a ay element level p oduces uneq ual noise powe at sub a ay outputs because of the diffe ent numbe of elements in each sub a ay. Adaptivity would t y to equalize the noise between channels, thus negating the tape ing effect.119 The t ansfo mation T that encodes the sub a ay a chitectu e* should be such that THT = I. In this way, the noise powe a t sub a ay outputs is equal; subsequently the missing tape weights a e applied digitally at sub a ay outputs (weight escaling).119 As an example conside a linea a ay of 12 elements and aised cosine tape ing. Figu e 24.9 depicts the following. Continuous cu ve: patte n of a ay * The sub a ays a chitectu e can be ep esented by a mat ix T; having a numbe M of columns equal to the numbe of sub a ays and a numbe N of ows equal to t he numbe of elementa y adiato s. The element tij of the mat ix is defined eith e as wi if the i th elementa y adiato belongs to the j th sub a ay o as 0 i f the i th elementa y adiato does not belong to the j th sub a ay, whe e wi i s the tape ing weight in the analogue laye of Figu e 24.8. ownloaded f om igital Enginee ing Lib a y @ McG aw Hill (www.digitalenginee in glib a y.com) Copy ight 2008 The McG aw Hill Companies. All ights ese ved. Any use is subject to the Te ms of Use as given at the website.

Digital tapering

or sum beam ormation

Elect onic Counte Counte measu es 24.28 RA AR HAN BOOK 0 Unequal noise at the suba ay outputs Antenna gain (dB) 20 A ay of suba ays A ay of elements; aised cosine tape ing A ay of elements ; unifo m tape ing 50 0 Angle () 50 100 40 100 FIGURE 24.9 Examples of the antenna patte ns achieved in seve al cases of sub a ays without noise no malization at the output of sub a ays; it app ox imately follows the unifo m tape ing (dashed line). otted line: patte n of a a y of elements and of a ay of sub a ays afte noise no malization and weight e scaling. The nume ical example follows with Figu e 24.10, which po t ays the can cellation of a jamme with the oA = 50 and a JNR of 30 dB. The continuous line e fe s to an unadapted patte n, tape ed at the element level, whe eas the dotted l ine pe tains to the adapted patte n at the sub a ay level. Sub a ays a e, in g ene al, chosen to be i egula in thei shape and position to avoid g ating lobe s. If a jamme impinges on a g ating lobe, the jamme will be nulled by disto ti ng the g ating lobe and, as a consequence, the a ay main beam (g ating notch). Fo example, conside a ULA with N = 12 elements; then, fo m two types of not ov e lapping sub a ay configu ations both having M = 6 sub a ays. 0 Gain of adapted patte n (dB) 20

40 Unadapted, tape ed at element level Adapted at suba ay level 50 0 50 100 Angle () J oA = 50 60 100 FIGURE 24.10 Jamme cancellation at sub a ay level ownloaded f om igital Enginee ing Lib a y @ McG aw Hill (www.digitalenginee in glib a y.com) Copy ight 2008 The McG aw Hill Companies. All ights ese ved. Any use is subject to the Te ms of Use as given at the website.

Jamme

Elect onic Counte Counte measu es ELECTRONIC COUNTER COUNTERMEASURES 24.29 20 SINRout (dB) 0

20 40

The fi st configu ation is egula with two elements fo each sub a ay. The sec ond configu ation is i egula with 2, 1, 4, 2, 1, and 2 elements, espectively. Figu e 24.11 shows the SINR at the output of the a ay ve sus the jamme oA. T he ta get oA is at 0, the SNR is 0 dB, the JNR is 30 dB. Th ee cu ves summa ize the system pe fo mance. The dashed line is the a ay patte n with aised cosine tape : this is shown fo the sake of compa ison with the othe two cu ves of the SINR. The dotted line is the SINR fo the quiescent (absence of adaptivity) pat te n: it mimics the ecip ocal of the sidelobe and main beam patte n. The contin uous line is the SINR fo the adaptive i egula sub a ay a chitectu e; the max imum value of the SINR is (10log1012 tape ing losses). Figu e 24.12 depicts the SINR fo the egula a ay configu ation and absence of tape ing. It is noted th at when the jamme oA is a ound 80, the SINR dec eases; this is because of the g ating lobe. The maximum value of SINR is 10.79 dB = 10log1012, because the e a e no tape ing losses. 20 10 0 10 20 30 40 SINRout (dB) 0 20 40 60 80 100

ownloaded f om igital Enginee ing Lib a y @ McG aw Hill (www.digitalenginee in glib a y.com) Copy ight 2008 The McG aw Hill Companies. All ights ese ved. Any use is subject to the Te ms of Use as given at the website.

J oA () FIGURE 24.12 SINR ve sus the J oA fo a pe ing

egula

sub a ay a chitectu e and no ta

FIGURE 24.11 SINR ve sus the jamme

oA (J oA)

Adaptivity at suba ay level No adaptivity, aised cosine tape sed cosine 0 20 40 60 J oA () 80 100

Patte n with

ai

Elect onic Counte Counte measu es 24.30 RA AR HAN BOOK Supe esolution. The esolution of a conventional antenna is limited by the well known Rayleigh c ite ion, which states that two equal amplitude noise sou ces ca n be esolved if they a e sepa ated in angle by 0.8 l/L, in adians, whe e l is the wavelength and L is the ape tu e length. When the incident wave is eceived with a high JNR, an adaptive a ay antenna mayin p incipleachieve a na owe adapt ive beamwidth, giving a sha pe bea ing estimation of the incident wave. If accu ate st obes of the jamme s can be obtained, these can be exploited to fo m beam s in the jamme di ections, which a e used as auxilia y channels fo adaptive in te fe ence supp ession.121 The inte fe ence di ections can also be used fo dete ministic nulling, which is of inte est fo main beam nulling.122 In addition to the inte fe ence sou ce di ections and sou ce st engths, this technique can p o vide othe info mation as to the numbe of sou ces and any c oss co elations (c ohe ence) between the sou ces. Such info mation can be used to t ack and catalog ue the inte fe ence sou ces in o de to p ope ly eact to them; the jamme mappi nga function unning in the backg oundis useful to select the modes (e.g., admissi ble pointing di ections and wavefo ms) of multifunction ada and fo gene al si tuation awa eness. Supe esolution might be able to esolve multiple independent sou ces; due to sidelobe supe position and masking p oblems, supe esolution mi ght be vital fo jamme mapping in case of multiple jamme s. Supe esolution is also of inte est as an ECCM to counte c oss eye jamming in seeke head applicat ions: see Section 12.1 of Wi th.102 The supe esolution concept was mainly devel oped and analyzed by W. F. Gab iel123 at the Naval Resea ch Labo ato y (U.S.). iffe ent methods fo bea ing estimation we e desc ibed by Gab iel and, subsequen tly, by othe autho s.34,124130 One is the maximum ent opy method (MEM) invented by J. P. Bu g. It wo ks well with a HowellsApplebaum adaptive beamfo me , which has an omnidi ectional eceiving patte n except whe e jamme s a e p esent. The p esence of jamme s is indicated by nulls in the eceiving patte n. Because nulls a e always sha pe than antenna lobes, jamme bea ings can be obtained mo e acc u ately f om the adaptive beam patte n, and supe esolution is the esult. The d esi ed spatial spect um patte n is obtained as simply the inve se of the adapted patte n. As Gab iel pointed out, the e is not a t ue antenna patte n because th e e is no linea combination of the signals f om an a ay that could p oduce suc h a peaked spatial patte n. It is simply a function computed f om the ecip ocal of a t ue adapted antenna patte n. Supe esolution and adaptive antennas fo ja mme cancellation a e intimately elated. Roughly speaking, the diffe ence is th at one p oduces a patte n with the nulls down (adaptive antenna fo jamme cance llation) and the othe with the nulls up, i.e., peaks (supe esolution of jamme s). One limiting facto of supe esolution techniques is that they often equi e the eceived signals to obey accu ate models of the a ay manifold. This can be violated due to p opagation effects (e.g., spatial sp eading and nonstationa it y) as well as inst umental effects (e.g., channel mismatch). These facto s also affect the pe fo mance of adaptive antennas fo jamme cancellation, but model m ismatches can deg ade the pe fo mance of supe esolution techniques mo e seve el y. The highe pe fo mance of the supe esolution techniques is often obtained at the expense of equi ing a st icte adhe ence to the assumed model; if the mode l is inaccu ate, these techniqueswhich ely on its heavy exploitationsubsequently become the most sensitive and mo e p one to pe fo m badly. Fo efficient supe e solution, an a ay with a easonable numbe of sub a ays is equi ed; this may be the eason fo the lack of application of this technique to p actical ada s ystems except fo expe imental pu poses. Supe esolution based on a small ownloaded f om igital Enginee ing Lib a y @ McG aw Hill (www.digitalenginee in glib a y.com) Copy ight 2008 The McG aw Hill Companies. All ights ese ved. Any use is subject to the Te ms of Use as given at the website.

Elect onic Counte Counte measu es ELECTRONIC COUNTER COUNTERMEASURES 24.31 SLC configu ation is not efficient, because this leads to the maximum ent opy o auto eg essive methods that, being nonlinea p ocessing, have a high p obabilit y of spu ious peaks. P actical expe iences indicate that the esolution limit is dete mined athe mo e by implementation and envi onmental facto s than by the pu e JNR conside ations. 24.7 TRANSMITTER RELATE ECCM The diffe ent types of ECCM a e elated to the p ope use and cont ol of the pow e , f equency, and wavefo m of the adiated signal. One b ute fo ce app oach to defeat noise jamming is to inc ease the ada s transmitter power. This techni ue, when coupled with spotlighting the radar antenna on the target, results in an inc rease of the radars detection range. potlighting or burnthrough modes might be e ffective, but a price must be paid. As the radar dwells in a particular directio n, it is not looking elsewherewhere it is supposed to look. In addition, the burn through mode is not effective against chaff, decoys, repeaters, spoofers, and so on. More effective is the use of complex, variable, and dissimilar transmitted signals that place a maximum burden on E M and ECM. Different ways of operation refer to the change of the transmitted fre uency in fre uency-agility or fre uen cy-diversity modes or to the use of wide instantaneous bandwidth.131133 Fre uency agility usually refers to the radars ability to change the transmitter fre uency on a pulse-to-pulse or a batchto-batch basis. The batch-to-batch approach allow s doppler processing, which is not compatible with fre uency agility on a pulseto-pulse basis. In a waveform with pulseto-pulse fre uency agility, the center f re uency of each transmitted pulse is moved, in either a random or a programmed schedule, between a large number of center fre uencies. The fre uency of the nex t pulse cannot generally be predicted from the fre uency of the current pulse.13 4 Fre uency diversity refers to the use of several complementary radar transmiss ions at different fre uencies, either from a single radar (e.g., a radar having stacked beams in elevation by employing different fre uencies on each elevation beam38) or from several radars. The objective of fre uency agility and diversity is to force the jammer to spread its energy over the entire agile bandwidth of the radar; this corresponds to a reduction of the jammer density and resulting E CM effectiveness.29 A good example of the use of the fre uency domain for purpos e of ECCM is the enrad, an experimental long-range air-surveillance radar built and tested at the Naval Research Laboratory (U. .)135 enrad was an example of how to build a radar so as to force the jammer to dilute its radiated energy per unit bandwidth; it includes both fre uency agility and fre uency diversity. Thi s radar shows that its unusually wide bandwidth allows a reduction of the effect iveness of the noise jammer that can seriously degrade more narrowband radars. F re uency agility, diversity, and instantaneous wideband techni ues represent a f orm of ECCM in which the information-carrying signal is spread over as wide a fr e uency (or space or time) region as possible to reduce detectability by E M and /or ARM and make jamming more difficult. This ECCM techni ue pertains to the rea lm of waveform coding.3,136137 The ambiguity function (AF) is the tool to charact erize waveform coding in terms of resolution, sidelobe level, and ambiguity.137 In selecting a waveform for a given radar application, the AF should be tested a gainst the environment in which the radar Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw-Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter-Countermeasures 24.32 RADAR HANDBOOK will be expected to operate. The so-called environmental diagram depicts spectra l, spatial, and amplitude characteristics of the radar environment (clutter, ECM such as chaff, intentional interferences, andperhapsinterferences from neighborin g EM apparatuses) and is used to assist the radar waveform design. An example of an environmental diagram is on p. 15 of Levanon and Mozeson137: on the range-do ppler plane are shown the regions in which several types of clutter and high alt itude chaff are expected. On the same diagram are superimposed the expected targ et trajectories and the AF contour of a, say, pulse-burst waveform. As the targe t follows a particular trajectory, the AF will move accordingly and the spurious AF peaks will slide across the clutter and chaff regions determining the intens ity and features of the radar echoes. Waveform coding includes PRF jitter and PR F stagger, which are helpful for some deception jammer but dont help against nois e jammer. Waveform coding makes deception jamming or spoofing of the radar diffi cult, since the enemy should not know or anticipate the fine structure of the tr ansmitted waveform; as a conse uence, it gives assurance of maximum range perfor mance against such types of jamming. Intrapulse coding to achieve pulse compress ion may be particularly effective in improving target detection capability by ra diation of enough average radar power without exceeding peak power limitations w ithin the radar and by improving range resolution (larger bandwidth), which, in turn, reduces chaff returns and resolves targets to a higher degree. ome advant age can be gained by including the capability to examine the jammer signals, fin d holes in their transmitted spectra, and select the radar fre uency with the lo west level of jamming. This approach is particularly useful against pulsed ECM, spot noise, and nonuniform barrage noise; its effectiveness depends primarily on the extent of the radar agile bandwidth and the ac uisition speed and fre uency tracking of an intelligent jammer. A techni ue suited to this purpose is referred to as automatic fre uency selection (AF ).133,138 Another method to reduce the effect of main-beam noise jamming is to increase the transmitter fre uency (as a n alternative means to the use of a larger antenna) in order to narrow the anten nas beamwidth. This restricts the sector that is blanked by mainbeam jamming and also provides a strobe in the direction of the jammer. trobes from a few spatia lly separated radars allow the jammer to be located. The availability of solid-s tate transmitter technology3640 allows the generation of high duty cycle waveform s, which may be of some help to realize LPI radar. As a general remark, one of t he factors preventing good ECCM is the reduction of EM spectrum allocated to rad ar. As discussed, operating over a wide spectral range has important advantages for ECCM, but the civilian and commercial telecommunication systems erode more a nd more portions of the spectrum at the expense of military ECCM capability. 24.8 RECEIVER-RELATED ECCM Jamming signals that survive the antenna ECCM expedients can, if large enough, s aturate the radar processing chain. Wide dynamic range receivers need to be used to avoid saturation. A logarithmic (log) receiver might help against noise jamm ing, but it has detrimental effects against clutter when doppler processing is u sed. A log receiver is a device whose video output is proportional to the logari thm of the envelope of the RF input signal over a specified range. It might be u seful in preventing receiver saturation in Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw-Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter-Countermeasures ELECTRONIC COUNTER-COUNTERMEA URE 24.33 the presence of variable intensities of jamming noise. By comparison with a line ar receiver of low dynamic range, moderate jamming noise levels will normally ca use the computer to saturate so that the target signal will not be detected. The main disadvantage is that a log characteristic causes spectral spreading of the received echoes. It would not be possible to maintain clutter rejection in an M TI or pulse doppler radar if the spectrum of clutter echoes were to spread into the spectral region in which target returns were expected.21, 29 The main messag e is that the dynamic range problem is important for the rejection of jammer as well as clutter, the latter always present in a radar. Thus, the recommendation is to implement, in a modern radar, a receiver with a wide linear dynamic range (e.g., 100 dB). This wide dynamic range needs to be maintained also in the ADC d evices with a suitable number of bits; as a rule of thumb, each bit adds 6 dB to the dynamic range count. Hard or soft limiters can also be used to counter jamm ing signals. They are nonlinear memoryless devices that cut jamming signals havi ng wide amplitudes. The Dicke-Fix receiver counters high rates of swept-fre uenc y CW jamming and swept spot noise jammers.29,139 In a radar receiver, the DickeFix uses a wideband intermediate-fre uency (IF) amplifier and a limiter ahead of the narrow-bandwidth IF amplifier. The wideband amplifier allows a rapid recove ry time from the effects of the swept jammer, and the limiter cuts the jamming s ignal. The narrowband target signal, after transit through the wideband amplifie r and the limiter without remarkable degradation, is integrated by the narrowban d filter matched to the signal. The word Fix in the Dicke-Fix was put there many years ago to indicate it was a fix for a problem that occurred at the time and wa s to be replaced by something better. It was usually installed with a switch to turn it off, if necessary. Today Dicke-Fix is not used in a modern radar, especi ally one that employs doppler processing; thus, it is no longer of interest in m any radar applications. Other special processing circuits can be used in the rad ar to avoid saturation, i.e., fast-time-constant (FTC) devices (perhaps of littl e use in modern radar), automatic gain control (AGC), and CFAR.8,29,31 However, they cannot be said to be ECCM techni ues. For example, FTC allows the detection of signals that are greater than clutter by preventing the clutter from saturat ing the computer. FTC does not provide subclutter visibility. AGC keeps the rada r receiver operating within its dynamic range, preventing system overload and pr oviding proper normalization so as to furnish signals of standardized amplitude to radar range, velocity, and angle processing-tracking circuits. In conclusion, these devices have a place in the radar but not as means for fighting the ECM b attle. In summary, there isnt much that has been done in the receiver to combat E CM other than to insure there is a good receiver that does its job. Today, moder n phasedarray multi-channel radar are going to adopt fully digital, software con trolled receivers, as in the DAR case; here, the expected advantages are the wid er linear dynamic range and the within-band calibration of the receivers that wi ll support the adaptivity on several tens of channels: a distinctive advantage a gainst directional noise jammers. 24.9 IGNAL-PROCE ING-RELATED ECCM Digital coherent signal processing greatly alleviates the effects of clutter and chaff.3,140 This is motivated by the use of coherent doppler processing techni ues such as fixed, adaptive MTI, or optimum pulse-doppler processing. Noncoheren t devices are also

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw-Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter-Countermeasures 24.34 RADAR HANDBOOK re uired because of the limited degree of clutter, chaff, and jammer suppression practically achieved by coherent devices, so that the cancellation residue may still be a significant source of false alarm. Among the noncoherent devices, it is worth mentioning the CFAR detector141145 and the pulse-width discriminator, th is latter being effective against pulsed jammers. The pulse-width discrimination circuit measures the width of each received pulse. If the received pulse is not of approximately the same width as the transmitted pulse, it is rejected. A pul se-width discrimination techni ue can help in rejecting chaff; in fact, echo ret urns from chaff corridors are much wider than the transmitted pulse. However, if a target is within the chaff corridor the pulse width discriminator might also eliminate the target. Coherent Processing. The most effective anti-chaff techni ue available to radar is the use of doppler filtering, which exploits the differ ent motion characteristics of the target and the chaff.3 The characteristics of chaff are similar to those of weather clutter, except that the chaff-scattering elements are cut to respond to a broad spectrum of radar fre uencies. Weather cl utter and chaff differ from ground clutter in that both the mean doppler shift a nd the spread are determined by wind velocity and wind shear, the latter arising from the variation of wind velocity with height. Chaff moves with the local win d, and there are ways (adaptive MTI and optimum doppler processing) to make an MT I null out both moving and stationary unwanted echoes.55,136,146 There are two b asic doppler filtering techni ues that are used. The first is the MTI, which emp loys a PRF that provides unambiguous range coverage while using a comb doppler f ilter whose nulls are tuned to the average radial speed of chaff.3 The second is the pulse doppler, which can use a high PRF to provide unambiguous doppler cove rage in conjunction with a doppler filter bank, allowing separation of target fr om the chaff.3 A problem with chaff might be when there is a significant wind sh ear in the atmosphere. With wind shear, the doppler spectrum from chaff can have a wide spectral width (unless the elevation beamwidth is very narrow as might o ccur with tri-dimensional radar with stacked beams in elevation38) so that it is difficult to cancel moving chaff echoes. A pulse doppler radar has a better cha nce, but it has problems of its own because of the foldover of the clutter that might occupy a large range extent. Coherent doppler processors might re uire rel atively large amounts of pulses (e.g., more than 10), which must be transmitted at a stable fre uency and PRF. A responsive jammer could measure the fre uency o f the first transmitted pulse and then center the jammer to spot-jam the followi ng pulses. Also, the re uirement for a stable PRF precludes the use of pulse-topulse jitter, which is one of the most effective techni ues against deception ja mmers that rely on anticipating the radar transmitter pulse. Coherent doppler pr ocessors are also generally vulnerable to impulsive radio fre uency interference , especially in radars with a limited number of coherent bursts on target.147 An other ECCM techni ue to be considered is pulse compression by matched filtering; it is intimately related to the waveform coding discussed in ection 24.7. Puls e compression136,137,141 is a pulse radar techni ue in which long pulses are tra nsmitted to increase the energy on the target while still retaining the target r ange resolution of a short pulse transmission. It is almost always used in radar for achieving The adaptive MTI estimates the mean doppler fre uency of a moving clutter source and places the null of, say, a binomial MTI. Optimum doppler processing estimat es the whole spectrum of clutter and shapes, by means of the inverse of the clut ter covariance matrix, the cancellation filter accordingly; furthermore, with a doppler filter bank, it integrates the echo signal from moving targets. The opti mum filter weights are calculated by an e uation similar to E . 24.6 applied to the radar echoes received by a coherent pulse train transmitted by the radar. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw-Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter-Countermeasures ELECTRONIC COUNTER-COUNTERMEA URE 24.35 high range resolution or reducing the peak power. Pulse compression offers some ECCM advantage that is discussed hereafter.3,148 When the pulse compression sear ch radar is compared, from an E M standpoint, with a conventional search radar w ith the same wide pulse, the enemy receiver on a jamming platform will not know (in the general case) the pulse compression reference code and will be at a disa dvantage. Compared with a radar that uses an uncompressed wide pulse, the pulse compression techni ue increases the radars capability against extended signal ret urns like chaff and clutter. In addition, noise from a jammer does not pulse-com press. Extended clutter tends to be noise like and will not pulse-compress, whic h keeps down interference displayed to operators.29 The disadvantages of pulse c ompression are related to the long duration of the coded pulse, which gives more time for the ECM e uipment to process the pulse. In many cases, pulse compressi on can provide the means for easy radar jamming for the enemy ECM operator. Puls e compression is also vulnerable to cover-pulse jamming, in which the ECM pulse is returned to the radar with a high JNR such that the normal target return is c overed by the jamming pulse. The width of the ECM pulse is normally wider than t he radar skin return.13 This type of deception can be counteracted by an ECCM te chni ue such as the cover-pulse channel, where the tracking is on the ECM transm ission rather than on the skin return from the target.29 The digital coherent im plementation of the Dicke-Fix receiver concept re uires the use of a coherent ha rd limiter that preserves the phase of the signal while keeping the amplitude at a constant value. The coherent limiter is inserted upstream of the pulse compres sion filter in a radar that uses phase-coded signals. In reception, the jammer a nd target signals are cut in amplitude. The preservation of the target signal ph ase coding allows the integration of target energy by means of the pulse compres sion filter matched to the phase code. The Dicke-Fix processing scheme suffers f rom three limitations. The first is related to the detection loss experienced wh en the target does not compete with the jammer. The second disadvantage refers t o the masking effect of a weak target signal sufficiently close in range (compar ed with the spatial extension of the code) to a strong target. Furthermore, it c annot be used in conjunction with doppler processing. CFAR. CFAR is a techni ue made necessary to prevent the computer from being overloaded by false alarms, wh ich reduce the capability of the radar to detect desired targets.141 This proces sing also plays a role as ECCM; there are three motivations for this. First, the scope of ECM techni ues, in a broad sense, is to impair the target detection an d tracking performance of a radar system. Detection performance is measured by t he probability of detection; tracking performance is determined by the probabili ty of detection and the probability of false alarm as well. Conventional (cell a veraging) CFAR raises the threshold in the presence of noise jamming and reduces the number of targets detected. However, the targets that survive can be effect ively tracked because the probability of false alarm has been maintained at suff iciently low levels. Without CFAR and appropriate threshold adjustments, perhaps no targets will be tracked due to the very large number of false peaks (detecti ons on the jammer), making it through to saturate the tracker. Conventional CFAR is not really removing the interference: it is just hiding it from the radar oper ator. However, it is allowing the tracker to operate This is used with Barker codes, for instance, where the amplitude limitation doe snt impair the phase code. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw-Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter-Countermeasures 24.36 RADAR HANDBOOK effectively for the targets that survive, and so in this way, it can prevent the overall failure of the radar. In the limit of no targets being detected (i.e., a very powerful jammer), then it could still be argued that no tracks are better than many false tracks. econd, not all jammers are noise jammers. ome indeed have a structure in range-doppler space, and CFAR techni ues can potentially be used to lower these unwanted signals beneath the detection threshold, once again preventing the detection of false tracks, whichfrom a tactical perspectivecan cau se serious dilemmas for a radar operator. Third, there are the adaptive CFAR det ectors (AMF, or adaptive matched filter, for example115) that really are ECCM te chni ues in the sense that they enhance the probability of detection against str uctured interferences (in space and/or time) while maintaining the constant fals e alarm rate property that allows these detected targets to be effectively track ed, rather than being seduced by a high number of false detections. This type of processingor similarly derived from the generalized likelihood ratio test (GLRT)h as been used in some practical radar. Furthermore, any self-respecting radar sys tem should make the operator aware of higher noise levels due to jamming even th ough they may not be visible on the CFAR display; the act of performing CFAR sho uld not exclude the operator from knowing that jamming is present and that the d etection threshold has been raised. 24.10 OPERATIONAL-DEPLOYMENT TECHNIQUE To this point in the chapter, only electronic ECCM techni ues have been consider ed. However, radar operational philosophy and deployment tactics may also have a significant effect on the radars resistance to ECM, E M, and ARM. This group of techni ues can be subdivided into those involving the operator, the methods of o peration, the radar deployment tactics, and the friend E M in support to ECCM.8 An operational techni ue against ECM is to use missiles with home on jam (HOJ) g uidance to intercept noise jammers. HOJ is a means whereby a missile guidance re ceiver utilizes the self-screening target jamming signal to develop angular stee ring information so the missile can home on that target. The role of the operato r in the ECCM chain pertains to the more general topic of human-factor ECCM.29 T his is a generic ECCM techni ue that covers the ability of an air defense office r, a radar operator, a commanding officer, and/or any other air defense associat ed personnel to recognize the various kinds of ECM, to analyze the effect of the ECM, to decide what the appropriate ECCM should be, and/or to take the necessar y ECCM action within the framework of the persons command structure. However, the human operator is less effective against a simultaneous attack of many enemy ve hicles supported by a strong ECM force. An operator confronted with a large mix of ECM types and a large number of ECCM techni ues is likely to do the wrong thi ng and/or react too slowly. In this situation, it might be proper to resort to a utomatically applied ECCM techni ues; this is the tendency today. However, a pos sible concern is that this could, sometimes, hurt the ECCM capability of the rad ar since a well-trained operator can often figure out what is happening; but the automatic processor can only make decisions based on preprogrammed logics insta lled in its computer and might not recognize when something unusual (as in jammi ng) occurs. This is a possible adverse effect of the absence of a decision-makin g operator. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw-Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter-Countermeasures ELECTRONIC COUNTER-COUNTERMEA URE 24.37 The operational methods include emission control (EMCON), the appropriate assign ment of operating fre uencies to various radars, the use of combined ECCMs to me et combined ECMs, the use of dummy transmitters to draw ECM to other fre uencies , and so on. EMCON is a techni ue for the management of all EM radiations of a f riendly system, force, or complex to obtain maximum advantages in the areas of i ntelligence data reception, detection, identification, navigation, missile guida nce, etc., over the enemy in a given situation. EMCON permits essential operatio ns while minimizing the disclosure of location, identification, force level, or operational intentions to enemy intelligence receptors. It includes the authoriz ation to radiate, the control of radiation parameters such as amplitude, fre uen cy, phase, direction, and time, the prohibition of radiation, and the scheduling of such actions for all units and e uipment of a complex.29 The on-off scheduli ng of the radars operation, to include only those time intervals when surveillanc e is re uired, can reduce the probability of the radar location being found by d irection finding (DF) e uipment or radar homing and warning receivers. Radar bli nking (using multiple radars with coordinated on-off times) can confuse an ARM s eeker and guidance or a DF receiver. Decoy transmitters, radiating from antennas not located at the radar, may also be employed to confuse DF receivers and ARMs ; these decoys can also operate in conjunction with the radars in a blinking mod e. Proper site selection for ground-based radars in fixed installations can prov ide a degree of natural signal masking to prevent, for example, detection by gro und-based E M e uipment. A high degree of mobility for tactical systems allows ra diate and run operations, which are designed to prevent the radar from being enga ged by DF location techni ues and associated weapons.3 The deployment of a radar network with overlapping coverage could provide some ECCM benefits. In the nett ed monostatic case, the radars have different fre uencies for interference reduc tion purposes; conse uently, the ECM has to consider jamming all radars in the o verlapping zone, thus reducing its efficacy. This is the kind of fre uency diver sity discussed in ection 24.7. Finally, it is worth noting that friendly E M ca n support ECCM action by warning of possible hostile activity, providing angular locations of hostile jammers and information characteristics of jammers. This i nformation is helpful in the selection of a suitable ECCM action. 24.11 APPLICATION OF ECCM TECHNIQUE This section shows the application of the previously described ECCM techni ues t o surveillance, tracking, phased-array, imaging, and over-the-horizon radars. Th e use of ECCM techni ues in other types of radars such as mortar location radars , missile guidance radars, and navigation radars is considered in the literature .3 urveillance Radars. The function of a surveillance radar is to search a larg e volume of space and locate the position of targets within the search coverage. The radar range and the azimuth-elevation coverage depend on the specific radar applications. The target reports generated by a surveillance radar are processe d to form target tracks. The key features of a surveillance radar are the detect ion range in clear, clutter, and jamming environments, the accuracy and rate of the extracted data, and the falsealarm rate. In the ensuing discussion, the desi gn principles, driven by the re uirements forced by the threat, are mainly addre ssed.3 Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw-Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter-Countermeasures 24.38 RADAR HANDBOOK Detection in a clear environment is a feature of early-warning radars, which look primarily for high-altitude targets at long ranges beyond the surface horizon, where the effects of clutter can be ignored. Under these conditions, a simplifie d analysis states that radar performance is relatively insensitive to transmitte r fre uency and waveform shape; in practice, the lower microwave fre uencies are preferred because it is easier to obtain large antenna and high average power a t lower fre uencies, and rain clutter is not important. The maximum detection ra nge on a target with a certain RC (s) in free space, for a surveillance radar t hat must uniformly search a specified volume in a given time period, depends on the product of the average transmitter power ( P ) and the effective antenna ape rture (Ar). It also depends on the inverse of the system noise temperature, but this is of little conse uence since the noise temperature is not a major design issue anymore. The situation is more complex when the target to be detected is o f the stealth type.149 Waveform design and operating fre uency are relevant para meters in tactical and volume surveillance radars, which must be able to detect low-flying penetrating targets that attempt to use terrain-shielding effects to escape radar detection. In this case, the selection of waveform and fre uency is made to tackle the problems of masking, multipath, chaff, clutter, and ECM.134, 138,150 The major EW threats to a surveillance radar are (i) noise jamming, (ii) chaff, (iii) deception jamming, (iv) decoys and expendables, and (v) ARM. Commo n types of jamming are main-beam noise jamming and sidelobe noise jamming. Again st this threat, good radar ECCM performance is achieved by increasing the produc t ( PAr ) of average transmitter power by the effective antenna aperture. A mili tary radar should always have 20 dB more power-aperture product than given by st andard designs, yet this is seldom allowed. The re uest for a low sidelobe level has to be traded off with the corresponding degradation of the main-beam width; the widening of the main-beam width may make the radar more vulnerable to mainb eam jamming. The noise jammer situation is basically an energy battle between th e radar and the jammer. In the main-beam noise-jamming situation, the advantage is with the jammer because the radar experiences a two-way propagation loss of i ts energy as contrasted with the one-way propagation loss between the jammer and the radar. With sidelobe jamming, the radar designer can reduce the jammer adva ntage by low sidelobe design coupled with the use of sidelobe cancellation techn i ues. With main-beam noise jamming, the radar can maximize the received target energy by transmitting more average power, dwelling longer on the target, or inc reasing the antenna gain. If the radars data rate is fixed and a uniform angular search rate is dictated by mechanical or search strategy, then the only option f or the radar is to increase its average transmitter power. The next option is to manage the data rate, thereby allowing a longer dwell time on the target (burnt hrough mode) along specific spatial sectors where needed. The ability to vary th e data rate in an optimal manner is one of the principal advantages of phasedarr ay radars.3 Another principle of ECCM design against main-beam noise jamming is to minimize the amount of jamming energy accepted by the radar. This is accompli shed by spreading the transmitted fre uency range of the radar over as wide a ba nd as possible, thus forcing the jammer into a barrage-jamming mode. This can be obtained by Of course, such radars have to see also targets at shorter ranges where clutter echoes can mask the targets echo; for this reason, all long range civil air traff ic control radar employ doppler processing. Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw-Hill Companies. All rights reserved. Any use is subject to the Terms of Use as given at the website.

Electronic Counter-Countermeasures ELECTRONIC COUNTER-COUNTERMEA URE 24.39 resorting to fre uency agility and/or fre uency diversity. ome radars incorpora te an AF device that allows the radar fre uency to be tuned to that part of the spectrum containing the minimum jamming energy.133,138 In accordance with the s earch-radar e uation (see ection 24.12), ECCM performance appears (explicitly) to be insensitive to fre uency.* Increasing the radar fre uency does not affect the signal-to-interference energy ratio within a radar resolution cell when the antenna aperture and the radar data rate are held constant. The increased fre ue ncy increases both the antenna gain and the number of radar resolution cells tha t must be searched by e uivalent amounts; the net effect is that the target retu rn power is increased by the same amount by which the target dwell time is decre ased, thereby holding the target-to-jamming energy ratio constant. Nevertheless, in practice, the effect of main-beam noise jamming can be reduced with high rad ar fre uency. Higher fre uency radars tend to have narrower antenna beamwidths a nd larger operating fre uency bandwidths (5 to 10 percent of radar center fre ue ncy) than lower fre uency radars. Thus, main-beam jammers will blank smaller sec tors of high fre uency radars than of low fre uency radars. In addition, main-be am jamming of a narrow beam radar tends to provide a strobe in the direction of the jammer, which can be used to triangulate and reveal the jammer location. Wid er radar bandwidth, with appropriate coding, forces the jammer to spread its ene rgy over a wider band, thereby diluting the effective jamming energy.3 ECCM desi gn principles for main-beam noise jamming also apply to sidelobe noise jamming, with the addition that the sidelobe response in the direction of the jammer must be minimized. Ultralow sidelobes in the order of, say, 45 dB below the antennas main-beam peak response are feasible by using advanced technology. ometimes the control of sidelobe noise jamming by using ultralow-sidelobe antennas is not pr oper; this is true because the main-beam width might be increased two to three t imes. In addition, most operational radars do not use ultralow (less than 40 dB) or low (30 to 40 dB) idelobe antenna and have antenna idelobe in the 20 to 30 dB region with average idelobe of 0 to 5 dB below i otropic. SLC ha the potenti al of reducing noi e jamming through the antenna idelobe , and it i u ed for t hi purpo e in operational radar .3 A explained in Section 24.9, ECCM technique again t chaff are mainly tho e ba ed on coherent doppler proce ing.3,152 In p articular, the reference152 de cribe a compari on of fixed and adaptive doppler canceler applied to chaff data recorded by a multifunctional pha ed array rada r operating at S band. Both canceler proce an 8 pul e coherent bur t. The fix ed (i.e., nonadaptive) proce ing i a Dolph Cheby hev filter with 60 dB of ide lobe attenuation with re pect to peak. The adaptive filter, ba ed on the optimum doppler filtering ( ee Section 24.9 and the literature55,136,146) ha the weigh t built around the e timation and inver ion of the di turbance (chaff and noi e ) covariance matrix. The target detectability i evaluated again t a den e chaff cloud. For the particular et of recorded mea urement , it ha been hown a ub tantially enhanced performance of the adaptive filter over the nonadaptive filt er. * A mentioned previou ly, the lower frequencie might be preferred for long ran ge urveillance becau e the u ual radar equation doe not include all the pertin ent factor . In the jamming ca e, one hould take account that the jamming anten na on an aircraft ha a lower gain at lower frequencie o the jamming power den ity might be le at the lower frequencie . Al o, when multipath i important, by electing the radar frequency properly, one might reduce the jamming power re ceived by being in a null of the jammer tran mitting antenna. Chaff might not be a ea y to deploy at the lower frequencie .151 In conclu ion, the lower radar f requencie might not be a vulnerable a one might think by examining the tradit ional radar equation. Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin

glibrary.com) Copyright 2008 The McGraw Hill Companie . All right u e i ubject to the Term of U e a given at the web ite.

re erved. Any

Another cla of ECCM technique i aimed to contra t the deceptive ECM. Decepti on jammer have a number of pecific characteri tic that can be u ed by radar to identify their pre ence. The mo t prominent i that fal e target return mu t u ually follow the return from the jammer carrying target and mu t all lie in t he ame direction within a radar PRI. If the deception jammer u e a delay that i greater than a PRI period to generate an anticipatory fal e target return, th en pul e to pul e PRI jitter identifie the fal e target return . The generation of fal e target in direction different from that of the jammer carrying aircr aft require injecting pul e jamming ignal into the radar idelobe . Many rada r employ the SLB ( ee Section 24.6) to defeat thi type of ECM. True target ret urn tend to fluctuate from can to can with fixed frequency radar and from pu l e to pul e with frequency agility radar . Tran ponder jammer generally end t he ame amplitude reply to all ignal they receive above a thre hold and, there fore, do not imulate actual target fluctuation re pon e . In addition, they u u ally appear wider in azimuth than real target owing to the modulation effect of the radar canning antenna re pon e on the real target. Repeater jammer can be made to imulate the actual amplitude re pon e of real target and, hence, are more effective over tran ponder type jammer from an ECM viewpoint. An operating mode to be included in a radar to di tingui h u eful target from tran ponder a nd repeater jammer i ba ed on a doppler pectrum analy i provided enough time on target i available. Additional expen ive technique again t deceptive jammi ng can be ba ed on the mea urement and analy i of the angular and polarization ignature of the echo ignal . The ame ECCM con ideration apply with decoy ta rget that have the general attribute of real target and are very difficult to identify a fal e target . A method ometime employed i to te t the cintilla tion characteri tic of the detected target to determine whether or not they fo llow tho e of real target . Expendable that tend to be de igned under tringent economic con traint often return only a teady ignal to the radar. With doppl er pectrum analy i , it i po ible to look for return from rotating component of the target that any form of powered target mu t po e . Example are jet e ngine or propeller modulation return a ociated with aircraft target . ARM po e a eriou threat to a urveillance radar. The urvivability of a urveillance radar to an ARM attack relie upon waveform coding (to dilute the energy in the frequency range), the management of radiated energy in time and along the angula r ector , and the adoption of low idelobe in tran mi ion. The e action make it more difficult for an ARM to home on radar. When an ARM attack i detected, it may be u eful to turn on patially remote decoy tran mitter to draw the ARM away from the radar ite. Blinking with a network of radar achieve better re u lt . The ARM trajectory i u ually elected to attack the radar through the zeni th hole region above the radar, where it detection capability i minimal. Thu , a upplemental radar that provide a high probability of detection in the zenit h hole region i required. There are certain advantage in choo ing a low tran m itting frequency (UHF or VHF) for the upplemental radar. The RCS of the ARM bec ome greater a the wavelength of the radar approache the mi ile dimen ion , c au ing a re onance effect.3 A low frequency radar i omewhat le vulnerable to an ARM attack owing to the difficulty of implementing a low frequency antenna w ith the limited aperture available in the mi ile.151 However, low frequency rad ar ha poor angular re olution. Tracking Radar . Tracking radar provide good re olution and preci e mea urement of the kinematic parameter (po ition, velocity , and acceleration) of target . The e timation, updated with mea urement , and p rediction of the kinematic parameter Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companie . All right re erved. Any u e i ubject to the Term of U e a given at the web ite.

Electronic Counter Countermea ure RADAR HANDBOOK

24.40

Electronic Counter Countermea ure ELECTRONIC COUNTER COUNTERMEASURES 24.41 a the time run are the proce ing tep u ed to build up the track of target . Track allow guidance and control of friendly force , threat a e ment, and e nemy target engagement by weapon . Tracking can be accompli hed in four way : (i ) The dedicated radar tracker ( ometime called a ingle target tracker and deno ted STT) continuou ly point it antenna at a ingle target by en ing error fr om the true target po ition and correcting the e error by a ervo control y te m. Then there are two different type of radar called, in the pa t, track while can. (ii) One i a limited angle can a in ome air defen e radar and aircra ft landing radar , which earch a limited angular ector at a rapid rate (e.g., 10 or 20 time a econd). (iii) The other type of track while can (TWS) wa wha t i now called automatic detection and tracking (ADT). The ADT y tem generate track of more than one target by u ing a erie of can to can target mea ure ment taken a the antenna ample the target path . (iv) The multifunctional ph a ed array radar track multiple target by multiple independent beam , formed b y the ame array aperture, that are allotted to different target . Thi ub ecti on i limited to the de ign principle , driven by the threat requirement , of th e dedicated radar tracker.3,153 The en uing ub ection will refer to multifuncti onal pha ed array radar. Good ECCM performance i achieved by radiating a large an average tran mitter power at the highe t tran mitter frequency practicable, coupled with a low a idelobe level a achievable. Increa ing the tran mitter f requency, for a fixed antenna ize, increa e the antenna gain Gt, which, in tur n, increa e the received target power a Gt2. For main beam noi e jamming, the received jamming power increa e directly a Gt, re ulting in a net increa e in ignal to jamming power by a factor proportional to the antenna gain Gt. Here, i t i noted a ba ic difference between urveillance and tracking radar: the detec tion range of a tracking radar improve a the frequency i increa ed for a fixe d ize antenna. The rea on for thi improvement i that the antenna gain i dire ctly increa ed with frequency, thereby focu ing more power on the target. Thi i ncrea ed power i integrated for a time, which i inver ely proportional to the bandwidth of the ervo control loop. For a urveillance radar, thi increa ed po wer i collected for a proportionally horter time, ince the radar mu t earch more cell in the ame time becau e of the narrower antenna beamwidth. With ide lobe jamming, the received jamming power i proportional to the idelobe antenna gain, (G l) re ulting in a net increa e in ignal to jamming power ratio by the factor GtG l1. A with urveillance radar , idelobe noi e jamming and deception can be further attenuated by the u e of SLC in conjunction with SLB, a de crib ed in Section 24.6. The u e of higher tran mi ion frequencie for tracking rada r generally make them le u ceptible to noi e jamming than urveillance radar . In addition, tactical tracking radar may track the noi e jammer in angle. Tr acking a noi e jammer in angle from two patially eparated radar provide enou gh information to locate a jammer with ufficient accuracy. A more threatening E CM again t tracking radar i DECM. The e threat require con iderably le ener gy than noi e jamming (a feature particularly important on tactical aircraft, wh ere available pace i limited). Neverthele , they are very effective in captur ing and deceiving the range gate (with the RGPO technique), the velocity gate (w ith the VGPO technique), and the angle tracking circuit . A primary ECCM defen e again t RGPO i the u e of a leading edge range tracker. The a umption i that the deception jammer need time to react and that the leading edge of the retur n pul e will not be covered by the jammer. PRI jitter and frequency agility both help to en ure that the jammer will not be able to anticipate the radar pul e a nd lead the actual kin interval. Alternatively, the tracking radar might employ a multigate range tracking Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companie . All right re erved. Any

u e i

ubject to the Term of U e a

given at the web ite.

y tem to imultaneou ly track both the kin and fal e target return . Thi appr oach utilize the fact that both the jamming ignal and the target return come from the ame angular direction, o that the radar angle tracking circuit are a lway locked onto the real target.3 The methodology of introducing VGPO into the radar tracking circuit i analogou to the method u ed with RGPO. The frequenc y hift i initially programmed o that the repeated ignal i within the pa ba nd of the doppler filter containing the target return. Thi i needed to capture the doppler filter containing the target, through the radar AGC action. The rep eater jammer ignal i then further hifted in frequency to the maximum expected doppler frequency of the radar. The repeated ignal i then witched off, forci ng the victim radar to reacquire the target.3 Coherent tracking radar can check the radial velocity derived from doppler mea urement with that derived from di fferentiated range data. Anomalou difference provide a warning of the probable pre ence of a deception jammer. When RGPO and VGPO operate imultaneou ly, the be t defen e i the contemporary tracking of true and fal e target in both rang e and doppler dimen ion . The u e of multimode (high, low, and medium PRF) radar can al o be an effective ECCM mea ure helping to counter range gate and veloci ty gate tealer by witching radar mode . Angle gate tealing i particularly e ffective again t conical canning or equential lobing tracking radar . It i fo r thi rea on that uch tracker cannot be u ed in military application . The fu ndamental problem with the e radar i that angle tracking i accompli hed by de modulating the amplitude modulation impo ed on the target return pul e over a c omplete canning or lobing cycle. To jam thi type of radar effectively, the rad ar angle tracking error en ing circuit mu t be captured with a fal e amplitude modulated ignal, at the canning or lobing rate, which i ignificantly out of pha e with that from the target return. When the conical can or lobing modulat ion i impo ed on both the tran mitter and the receiver beam , it i relatively imple for a jammer to ynthe ize the appropriate jamming ignal by inverting an d repeating the tran mitter modulation (inver e gain repeater).154 Thi can be p artially overcome by a conical can on receiver only (COSRO) y tem, where the t racking radar radiate a non canning tran mitting beam but receive with a conic al can beam. The jammer then ha no knowledge of the pha e of the conically ca nned receiving beam and mu t adopt a trial and error method of canning the jamm ing modulation until a noticeable reaction occur in the tracking radar beam. (T hi jamming technique i called jog detection.13) A equential lobingon receive only (LORO) y tem conceal the lobing rate from a potential jammer.3 Conical c an and equential lobing are going to be replaced by the monopul e technique; th u , COSRO and LORO are becoming ob olete. Monopul e tracking i inherently in en itive to angle deceptive jamming from a ingle point ource. Thi i a re ult o f the monopul e angle error en ing mechani m that form an error proportional t o the angle between the target and the antenna bore ight on each return pul e. T hi i accompli hed by comparing ignal received imultaneou ly in two or more antenna beam , a di tingui hed from technique uch a lobe witching or conica l canning, in which angle information require multiple pul e . Effective monop ul e jamming technique generally attempt to exploit the monopul e radar u cept ibility to target glint or multipath ignal .13 One jamming approach, known a c ro eye, u ed again t monopul e radar generate artificial glint into the mono pul e tracking loop.13 The inventor of the cro eye technique are B. Lewi (NR L, USA) and D. Howard; ee their patent155 originally filed the 1958. Cro eye i ba ically a two ource interferometer who e antenna u ually are Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companie . All right re erved. Any u e i ubject to the Term of U e a given at the web ite.

Electronic Counter Countermea ure RADAR HANDBOOK

24.42

Electronic Counter Countermea ure ELECTRONIC COUNTER COUNTERMEASURES 24.43 mounted on the aircraft wingtip a far apart a po ible. The ignal received i n each wingtip antenna are repeated in the oppo ite wingtip antenna, except for a 180 pha e hift, which i in erted in one line to direct an interferometric nul l toward the victim radar. In effect, thi create an apparent change of target direction a viewed from the radar. A large repeater gain i required to generat e a high jammer to ignal ratio; otherwi e, the kin echo will overwhelm the jam ming ignal in the interferometer pattern null . The maximum effectivene of t he technique implie a con iderable delay (on the order of 100 n ) in the repeat ed ignal, owing to the tran mi ion line and amplifier between the receiver and tran mitter antenna . Thu , leading edge or multigate range tracking hould be an effective ECCM technique again t cro eye jamming.3,13 Terrain bounce jammin g or terrain cattered interference (TSI) or hot clutter i another monopul e ja mming technique that i u ed again t emiactive mi ile eeker and airborne tra cking radar . With thi technique, the jammer aircraft illuminate the Earth ur face in front of and below it, o that the emiactive mi ile home on the illum inated ground pot and not on the jammer aircraft. The uncertainty of the terrai n cattering parameter and the po ible depolarizing effect of urface reflect ion are ome of the problem a ociated with thi technique.3 The TSI again t ai rborne radar and the corre ponding mitigation technique are de cribed in detail in the literature156158 TSI i a ignificant problem to military airborne radar; in fact, an often weak target ignal in the main beam ha to compete with jamme r that propagate not only via direct path but al o via multipath from the under lying terrain. Mitigation technique have been focu ed on e timating the direct jammer ignal, e timating the linear y tem created by the multipath, and removi ng an e timate of the reflected jammer ignal from the main received radar igna l158; thi i al o allowed by u ing reference beam pointed at hot clutter.157 A daptive cancellation technique have to be able to account for the doppler induc ed by relative motion between airborne radar and jammer platform and the jammer ignal non tationarity that i produced from uch a bi tatic geometry. TSI miti gation for over the horizon (OTH) radar i de cribed in Abramovich et al.159 Mon opul e radar that u e parabolic reflector antenna are u ceptible to jamming t hrough cro polarization lobe generated by the reflector urface.3,13 Thi occ ur becau e the angle error en ing di criminator ha an inver e lope for a cro polarized ignal, which cau e the angle tracking ervo to have po itive feed back in tead of the negative feedback required for tracking. Monopul e e timate with planar array antenna u ually have a high re i tance to cro polarization jamming ( ee Section 11.5 of Wirth102). With array antenna in contra t to reflec tor antenna all the ingle antenna element have the ame polarization dependent pattern. Thi i multiplied with the array factor and al o applie for the um a nd difference pattern . The re ultant form of the beam pattern will thu be inde pendent of polarization. Therefore, the monopul e operation will al o not be di turbed.102 Pha ed Array Radar . In thi ub ection, we illu trate, by a numerica l example, the role played by the cheduler in a multifunctional PAR to combat E CM. To thi end, we re ort to a benchmark tudy de cribed in the literature, whi ch define typical ECM threat , operational cenario , and pha ed array performa nce mainly in term of target tracking under ECM. The imulation benchmark160 in clude two type of ECM, namely SOJ and RGPO. The SOJ, mounted on an aircraft, t ran mit broadband noi e toward the radar. The SOJ flie an oval (race cour e) h olding pattern in a clockwi e direction at an altitude of 3050 m and a peed of 168 m/ ; it i approximately Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companie . All right re erved. Any u e i ubject to the Term of U e a given at the web ite.

150 km from the radar. The two circular turn are performed at an acceleration o f 1.5 g. The tran mitted SOJ noi e impact the radar with power g 0 not exceedin g eight time the receiver noi e power. Thu , a SOJ will not completely hide a t arget, and it can be defeated with a higher energy waveform. In RGPO, the target under track repeat with delay and amplification the radar pul e o a to pull the radar range gate off the target. The time delay i controlled o the fal e t arget i eparated from the true one with either linear or quadratic motion. For the linear ca e, the range of the fal e target Rkft i related to the range of the true target Rkt via t ft Rk = Rk + vpo (tk t0 ) (24.12) where vpo i the pull off rate, tk i the time at which the target i being ob e rved, and t0 i the initial reference time of the RGPO fal e target. Alternative ly, for the quadratic ca e t ft Rk = Rk + 1 a (t t ) 2 2 po k 0 (24.13) where apo i the pull off acceleration. Radar Scheduling. The cheduling and the tracking function clo ely cooperate; both interact to update, with current mea urement , the target tate vector, and make the prediction nece ary to point the radar beam at the target the next time it i ob erved, elect the type of wa veform to radiate, and elect the thre hold to apply for target detection. A con ceptual cheme howing the interaction of cheduling and tracking i hown in Fi gure 24.13, where rk, bk, ek are the range, bearing, and elevation mea urement at tk; SNRk i the ob erved SNR at tk; tk+1 i the commanded time for the next t arget ob ervation; rk+1|k, bk+1|k, ek+1|k are the predicted range, bearing, and elevation for beam pointing control at tk+1; Wk+1 i the waveform election at t k+1; bk+1 i the detection thre hold for the dwell et at tk+1; and Xk|k, Pk|k a re the target filtered tate e timate and covariance matrix at tk given all the radar mea urement up to tk. SURVEILLANCE SCENARIO RADAR MODEL tk, rk, bk, ek, SNRk tk 1, Xk 1/k 1, Pk 1/k 1 SCHEDULER rk+1/k, bk+1/k ek+1/k TRACKING FILTER TIME DELAY tk+1, Wk+1, bk+1 FIGURE 24.13 Interaction of radar cheduler and tracking filter tk, Xk/k, Pk/k Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companie . All right re erved. Any u e i ubject to the Term of U e a given at the web ite.

Electronic Counter Countermea ure RADAR HANDBOOK

24.44

Electronic Counter Countermea ure ELECTRONIC COUNTER COUNTERMEASURES 24.45 The cheme i built around two clo ed loop : (i) the loop encompa ing the radar model, the tracking filter, and the cheduler, and (ii) the tracking filter loo p. The cenario i ob erved by the radar at time tk; the radar model provide th e mea urement rk, bk, ek and SNRk. The tracking filter update the previou tar get tate e timate Xk 1|k 1, and it covariance matrix Pk 1|k 1 at time tk provi ding the new e timate Xk|k and Pk|k and the value rk+1|k, bk+1|k and ek+1|k at the next time in tant tk+1. The cheduler provide the waveform to radiate Wk+1 and the thre hold bk+1 to apply for target detection at tk+1. Selection of the Sampling Period. The ampling period i cho en among a finite number of po ible different value ba ed on kinematic con ideration on the target (e timated pe ed) a well a on whether mi ed detection have occurred. If there i no mea ur ement to be a ociated to the target, the ampling period i et equal to T = 0 .1 and the waveform of highe t energy i elected, o a to po ibly avoid a econd mi ed detection due to the po ibly low target RCS. Conver ely, the ampl ing period i elected a follow : T = 0.5 for target with e timated peed greater than 400 m/ T = 2 for target with e timated peed between 100 an d 400 m/ T = 3 for target with e timated peed le than 100 m/ Even tho ugh the target may accelerate or maneuver, for the ake of implicity, the ampl ing period i elected only on the ba i of the target e timated peed. Selectio n of the Detection Thre hold. The pre ence of a jamming ignal can increa e the number of fal e alarm and wrong plot track a ociation up to an unacceptable l evel, thu increa ing ignificantly, the probability of loo ing a target under t rack. It i , therefore, important that the radar receiver be equipped with a CFA R. Since the fal e alarm probability i related to the detection thre hold, the latter hould be adapted online ba ed on the inten ity of di turbance . Selectio n of the Waveform. The benchmark161 include 8 waveform , indexed by i and chara cterized by a different pul e width te(i), o that the waveform can be elected in order to provide a SNR greater than the detection thre hold and thu maintain an a igned probability of target detection. Thi can be accompli hed by fir t e timating the average target RCSk at time tk, and then computing for each wavefo rm i, the predicted SNRk(i), and finally electing the waveform index i uch tha t the corre ponding SNRk(i) i ju t greater than the de ired detection thre hold plu a given tolerance.161 ECCM: A SOJ and A RGPO. Hereafter, pecific anti SOJ (A SOJ) and antiRGPO (A RGPO) technique will be de cribed. A SOJ i ba ed on e timating the jammer po ition and power level and then u ing uch e timate to a dapt the radar detection thre hold online. Jammer tate e timation. Whenever the radar operate in the pa ive mode, i.e., without emitting pul e , bearing bkj and elevation ekj of the jammer a well a the relative tandard deviation kjb and kje, and the jammer to noi e ratio rkj (in the following The RCS i certainly a fluctuating quantity ver u time; it al o depend upon th e target a pect angle. However, if enough time on target i available, the RCS e timate can be ufficiently accurate. Downloaded from Digital Engineering Library @ McGraw Hill (www.digitalengineerin glibrary.com) Copyright 2008 The McGraw Hill Companie . All right re erved. Any u e i ubject to the Term of U e a given at the web ite.

expre ed in dB) are mea ured. Thi allow the tracking filter to e timate the j ammer tate made up of four tate component : the two angular po ition (bearing and elevation) and the relative angular peed . The jammer track i initialized by u ing the fir t two mea urement provided by the radar. Jammer power level e timation. An e timate of the power level can be obtained by the fir t order lin ear filter initialized from g0(t1j) = 1 for a uitable filter coefficient aj (0, 1): o (tkj ) = j o (tkj1) + (1 j ) 10 k j 10 (24.14) Ad pt tion of the detection th eshold. Fo iven detection th eshold b (in dB) , the p ob bility of f lse l m tu ns out to be: 10 /10 Pfa = exp j 2 ) ( k ) + 1 (24.15) Hence, the detection threshold can e selected at each time instant, tk, in the followin way: 2 k = max 9.64, 10 lo 10 0 (tkj1 ) Gstc (rk / k ) ( kj ) + 1 ln Pfa { } (24.16)

j where 0(t k 1) is the most recent availa le estimate of the jammer power leve l; the value k = 9.64 dB is the one which allows, in the a sence of jammers, th e desired false alarm pro a ility Pfa = 10 4; rk|k is the filtered estimate of t he tar et ran e; Gstc(.) is the sensitivity time control ain; and kj is the norm alized antenna ain for the received si nal computed in the radar active mode.16 1 Because the phased array radar considered here is a multifunctional one, it has also a trackin mode that mi ht e affected y the RGPO; for this reason, an A R GPO is considered an ECCM technique. Whenever RGPO is active, two hi h amplitude si nals are received from the radar: the true tar ets echo and an RGPO induced s i nal. ince the time at which the tar et under track activates RGPO is unknown to the trackin al orithm, the latter must first reco nize that RGPO is active a nd then implement an appropriate A RGPO technique. In order to esta lish whether RGPO is active, the followin test can e adopted. Let N e the num er of measu rements exceedin the detection threshold of more than 3 dB. Then if N < 2, it i s decided that RGPO is not active and no A RGPO action is undertaken; otherwise, if N = 2, the A RGPO ECCMs descri ed elow are applied. Notice that the test al so aims at discriminatin the type of ECM ein active, i.e., OJ or RGPO. In fa ct, whenever the noise jammer is in the antenna main eam many false measurement s with hi h jammer to noise ratio are induced; in this case, it turns out that N > 2 and RGPO is declared inactive. Whenever the jammer is no lon er in the tar ets line of si ht, it may happen that multiple measurements exceed the detection threshold, ut the condition that the excess is reater than 3 dB will e very u nlikely to e fulfilled in practice. Once it has een esta lished that RGPO is a ctive, several devices can e adopted in order to prevent the loss of the tar et under track. 1. A first approach consists of maintainin two tracks until the R

 

 

Electronic Counter Countermea ure RADAR HANDBOOK

24.46

 

 

  

  

0Gstc

GPO is deactivated. 2. A second approach consists of penalizin , in the data ass ociation, the measurements whose ran e is reater than the avera e ran e of meas urements with NR hi her than the detection threshold.161 Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEA URE 24.47 3. A third, more drastic approach, consists of discardin the measurement with h i her ran e amon the two measurements that have exceeded the detection threshol d of more than 3 dB. It is important to uarantee, under RGPO, a very hi h NR f or the tar et. In fact, it mi ht happen that the si nal produced y the false ta r et overcomes the detection threshold whereas the one from the true tar et does not, thus causin an association error with possi le serious consequences in th e tar ets trackin . Hence, whenever RGPO is active, a hi h ener y waveform must e selected. A further precaution is the followin : if there are missed detection s for at least two out of the last three scans, an immediate revisit with sampli n interval Ts = 0.1 s in search dwell mode is performed. In search mode, the ra n e ate is 10 km instead of 1.5 km, so in this way, it is possi le to o tain a new tar et measurement for updatin the trackin filter and thus avoid the tar e ts loss. imulation Results. Monte Carlo simulation experiments usin the enchma rk160 have een carried out in order to assess the enefits of the a ove descri ed ECCMs. More specifically, the adaptation of the detection threshold has een used as A OJ whereas the technique ased on discardin the measurements with hi her ran e has een adopted as A RGPO. Three types of tar ets (num ered 1, 5, an d 6) have een considered: tar et 1 represents a car o aircraft while tar ets 5 and 6 represent fi hter/ attack aircrafts with a much hi her de ree of maneuvera ility. For each experiment, the followin results are displayed: num er of lost tar ets (over 50 Monte Carlo trials), Ts (radar samplin time), Tave (avera e f raction of time required y the radar for tar et trackin ), PM (avera e power), position error, and velocity error. Ta le 24.2 shows simulation results for the Interactin Multiple Model (IMM) trackin 162 al orithm in the a sence of ECMs. T a les 24.3 and 24.4 report the results in the presence of OJ, without and respe ctively with A OJ. imilarly Ta les 24.5 and 24.6 report the results in the pre sence of RGPO, without and respectively with A RGPO. The examination of these ta les reveals that the presence of ECMs considera ly deteriorates the trackin pe rformance if no appropriate ECCMs are undertaken. Conversely, the adoption of th e a ove descri ed A OJ and A RGPO techniques allows the performance that would e attained in the a sence of the correspondin ECMs to e restored. TABLE 24.2 imulation Results Without ECM LO T TARGET 0 1 1 Ts (s) 1.958 0.6772 1.112 Tave (s) 0.5106 103 1.477 103 0.899 103 PM (W) 5.7985 68.898 10.774 PO ERR (m) 116.8 9 5.39 82.94 VEL ERR (m/s) 65.26 61.29 58.43 TARGET NO. 1 5 6 TABLE 24.3 imulation Results With OJ and Without A OJ LO T TARGET 34 15 50 Ts (s) 1.919 0.6923 Tave (s) 0.521 103 1.444 103 PM (W) 6.6179 68.411 PO ERR (m) 127.5 103 VE L ERR (m/s) 71.09 66.78 TARGET NO. 1 5 6 Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

 

  

 

  

 

Electronic Counter Countermeasures 24.48 TABLE 24.4 RADAR HANDBOOK imulation Results With OJ and A OJ LO T TARGET 1 1 4 Ts (s) 1.944 0.6888 1.1 18 Tave (s) 0.5144 103 1.452 103 0.8944 103 PM (W) 6.6179 68.411 15.11 PO ERR (m) 127.5 103 80.49 VEL ERR (m/s) 71.09 66.78 59.59 TARGET NO. 1 5 6 TABLE 24.5 imulation Results With RGPO and Without A RGPO LO T TARGET 48 50 50 Ts (s) 1.9 63 Tave (s) 0.5095 103 PM (W) 5.044 PO ERR (m) 120.5 VEL ERR (m/s) 66.6 TARGET NO. 1 5 6 TABLE 24.6 imulation Results With RGPO and A RGPO LO T TARGET 0 1 0 Ts (s) 1.889 0.7045 1 .156 Tave (s) 0.5295 103 1.419 103 0.8651 103 PM (W) 6.6179 68.411 15.586 PO ERR ( m) 127.5 103 124.9 VEL ERR (m/s) 71.09 66.78 80.26 TARGET NO. 1 5 6 Ima in Radar. There are two types of ima in radar that will e discussed: synt hetic aperture radar ( AR) and inverse AR (I AR). AR. AR allows us to have a hi h resolution mappin of the EM ackscatter from an o served scene. More preci sely, the radar data is o tained in polar coordinates, i.e., slant ran e and azi muth, while a two dimensional ima e in the rectan ular coordinates (x, y) is pro vided. Hi h resolution in slant ran e is o tained y transmittin a coded wavefo rm, with a lar e value of the time andwidth product, and coherently processin in a filter matched to the waveformthe echo si nals. Hi h resolution alon the tran sversal direction is achieved y formin a synthetic aperture. This requires (i) to put the radar on oard a movin platform, e. ., an aircraft or a satellite; ( ii) to record the EM si nals from each scatterer that is illuminated y the movi n antenna eam in successive instants of time, and (iii) to coherently com ine the si nalsvia a suita le azimuthal matched filterthus focusin the slidin antenn a pattern in a narrower synthetic eam. Radiometric resolution, another key para meter, is related to the capa ility of AR of distin uishin different o jects i n the scene on the asis of their EM reflectivity. Radiometric resolution determ ines how fine a sensor can distin uish etween o jects with similar EM reflectio n properties. It is a parameter of reat importance, especially for those applic ations oriented to extended tar et exploitation like polarimetry and classificat ion. Thus, the radiometric resolution should e optimized mainly for ood extend ed tar et interpretation, accountin for all kind of ack scatterers. Multilook processin is commonly used in AR ima e Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

  

 

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEA URE 24.49 formation in order to reduce the speckle noise. Traditional di ital multilook pr ocessin consists of an incoherent addition of independent ima es (looks) of the same scene. The looks can e o tained y partitionin the availa le si nal and width (ran e and/ or azimuth) and processin each look independently. The final ima e is produced y addin the looks incoherently, pixel y pixel. The direct t rade off etween eometric and radiometric resolution must e considered when ch oosin the num er of looks for processin . One look processin means a fully coh erent use of the andwidth ( est eometric resolution), and in this case, the sp eckle noise will o ey an exponential distri ution where the standard deviation i s equal to the mean value in the intensity ima e (multiplicative characteristic) . For multilook processin , the eometric resolution will de rade as the num er of looks increases and the speckle statistics of the intensity ima e o ey a amm a distri ution, where the standard deviation decreases with the square root of t he num er of independent looks.163 AR ima es are useful for surveillance and re connaissance applications. However, jammin could make AR ima es unusa le. The use of ECCM is, therefore, essential to reducin the vulnera ility of AR to jam mers. The suscepti ility to intercept si nals from AR and vulnera ility to jamm in are descri ed in Goj.164 A simulated noise jammin produces stripes on the AR ima es that demonstrate the effectiveness of jammin a ainst tar ets like str on point scatterers such as electric power line towers, as well as low reflecti vity a ricultural patterns and desert land. The references165,166 discuss the si nificant vulnera ility to ECM of space orne AR durin maritime reconnaissance missions. In 1978, typical ima ery from easat AR showed several features that made the AR a powerful maritime surveillance sensor. The ships and the wakes pr oduced y the ship motion were ima ed. The ship ima e (a lo ) appeared displace d from its wake due to the doppler shift caused y the ships motion relative to t he spacecraft. However, the ships position at the time of ima in and its course could e determined from the wake. The ship speed can e calculated too y the d isplacement of the ship from its wake. All this information is o taina le only i f enou h NR is availa le for the identification of these ima e features either y a human operator or an automatic processor. There exists, therefore, a potent ial vulnera ility to AR in the maritime surveillance application if a hi h leve l of ack round noise causes de radations of the AR ima e to an extent where sh ip and wake can no lon er e identified in the ima e. In the literature165,166 s ome critical aspects, in terms of jammer receiver sensitivity and transmitted po wer, for spot noise jammin are considered and the system requirements are deriv ed to determine the feasi ility and practicality of such jammer. Results of a co mputer simulation of an en a ement etween AR and representative jammin system are iven to ena le the effectiveness of ECM to e assessed. The threats to a AR are arra e jammin , spot jammin , random pulse jammin , and repeater jammin . Repeater/deception jammin is a major threat ecause it mi ht not e reco niza le, whereas the others are, at least in principle, reco niza le. The impact of each threat and the possi le countermeasures are descri ed in the remainin part of this section. Barra e jammer. The distur ance noise extends over the entire swath of the AR ima e, and it shows enerally a uniform intensity. The radar im a e of arra e jammer noise will exhi it speckle, i.e., a ri htness variation f rom one resolution cell to another. In addition, ecause a lar e num er of noise samples are added noncoherently, the multiple looks of jammer noise tend to smo oth out the intensity variation from pixel to pixel, just as in the case of ther mal noise. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

 

  

 

 

  

 

 

 

   

 

  

 

 

 

Electronic Counter Countermeasures 24.50 RADAR HANDBOOK pot jammer. It also covers the entire swath and with uniform intensity distur i n noise as for the arra e jammer; however, its ima e will differ from the arr a e noise ecause the Fourier transform of narrower and jammer noise will resul t in speckle size in the ran e dimension that is lar er than that of thermal noi se or clutter. The processed cross ran e dimension is a ain equal to that of clu tter or thermal noise. pot jammer noise will appear to e stretched in ran e. R andom pulse jammin . The jammer pulses may also e transmitted at random interva ls, so that such noise pulses can appear in any part of the ran e swath. When o served over a sufficient num er of samples, the noise pulses will occupy all par ts of the ran e swath in one sample or another. The azimuth processor forms the sum of the noise power from all samples within one synthetic aperture len th. Th at sum will e equal to the total noise power in the aperture, which is proporti onal to the avera e jammer noise power. Also, in this case, the speckle dimensio n will appear stretched in ran e, just as in the spot jammer case. However, the random pulse jammer speckle will exhi it more pronounced ri htness variations t han that from spot or arra e jammin , ecause fewer noise samples are added non coherently, there y reducin the smoothin effect of multiple looks. Repeater ja mmin .167 The enemy may utilize the transmittin radar to send out a si nal with in the and of the AR to confuse the AR system receiver. The jammin si nal ca uses the AR to receive and process erroneous information that results in severe de radations in the AR ima es and/or formation of the ima e of nonexistent tar ets. A deception jammin could e composed of manipulated replicas of the trans mitted radar si nals via DRFM. In Hy er 168 the possi ility of preventin AR ma ppin throu h coherent DRFM jammin has een investi ated. A software model has een developed and verified in several fli ht trials in the case of a round as ed DRFM jammer. ECCM techniques for AR can e divided into (i) antenna ased te chniques (low sidelo es, adaptive arrays) and (ii) transmitter/receiver/processi n ased techniques (frequency a ility, pulse codin ). Low sidelo es. AR antennas with low sidelo es reduce the level of jammin power received and, in addition, reduce the pro a ility of ein intercepted y ECM stations (in the sidelo e re ion).169 In relation to low sidelo es, the followin comment is in order. In a conventional radar, the effect of low sidelo es is clear, ut there is a differe nce in AR ecause the eamwidth is much wider than in other radar applications. In principle, the finer the resolution the smaller the AR physical antenna and the wider is its eamwidth. Thus, jammin in the main eam is more likely in a AR than in other radars ecause of the wide main eam. To et false tar ets int o the AR, it would have to come from the main eam, so low sidelo es mi ht not e much involved in deception jammer. Likewise, main eam jammin may e more of a threat to AR than sidelo e jammin . Adaptive arrays. The references170172 dea l with the rejection of a arra e noise jammin usin an adaptive spatial nullin . Equippin the AR system with an antenna partitioned into several su apertur es connected to parallel channels (i.e., multichannel AR) allows spatial adapti ve processin to suppress the interferin si nal. In Farina and Lom ardo,170 the performance of such a technique is evaluated in terms of AR impulse response, detection performance of point tar et, and radiometric resolution of an extended scene. In Ender171 a AR ima e, taken with an experimental four channel AR jam med y a small 1 watt Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

  

 

 

 

  

 

 

 

 

   

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEA URE 24.51 noise jammer leadin to a JNR in the raw data of a out 30 dB when the jammer pas ses the center of the main eam, is depicted. The de jammed ima e y adaptive sp atial suppression is also shown demonstratin the ood performance of adaptive s patial cancellation. The reference171 provides a comprehensive study of anti jam min spatial adaptive techniques includin also the space/slow time anti jammin filter with suita le ima e reconstruction al orithm. Results indicate that the slow time TAP provides superior interference cancellation than spatial only fil terin . AR usually involves wide and processin , requirin for adaptive nullin techniques, peculiar al orithms. Efficient road and jammer nullin has to e c ountered with space/fast time (i.e., ran e cell) processin .172 The expected num er of spatial dof is not hi her; we have only to add the dof in time. The adapt ive eamformin al orithms have to e implemented into the AR processin which is always space time processin (typically post doppler). Rosen er and Gray173 tackle the pro lem of miti atin the effects of an air orne road and jammer pre sent in the main eam of a AR. In addition to this, multipath reflections from the round, known as hot clutter, will add a nonstationary interference componen t to the ima e. The authors show the ima e de radation from hot clutter, the lim ited restoration that multi channel spatial ima in and slow time TAP can provi de, and how fast time TAP can improve the final ima e quality. Frequency a ilit y. AR processin needs phase coherence for o tainin the synthetic aperture, th us frequency a ility has to e used with care. Frequency chan in durin a synth etic aperture len th time results in a chan e of focal len th (different coeffic ient of the quadratic phase term) of the phase history of the illuminated tar et s that de rades the cross ran e resolution. AR operatin in urst mode can chan e its central frequency from one look to another, without any de radation in im a e quality. Given the efficiency of simple road and jammin and modern E M, we need to conclude that frequency a ility is not of reat help in AR ECCM. Pulse codin .167 An effective ECCM a ainst a DRFM repeat jammer is to chan e the rada r transmitted pulse code from one PRI to another. The radar maintains the same c arrier and andwidth; however, the pulses are coded to e approximately ortho on al to each other (i.e., their cross correlation is approximately equal to zero). uch a radar is less suscepti le to a DRFM repeater ecause (i) the jammer cann ot adapt easily since the radar si nal is varyin in the PRI domain, and (ii) th e si nal transmitted y a DRFM repeater jammer at a iven PRI (i.e., the radar s i nal that is used y the AR at the previous PRI) is approximately ortho onal t o the radar si nal that the AR is utilizin at the current PRI, and thus, a mat ched filterin with the current PRI radar si nal would weaken the DRFM repeater jammer si nal. In oumekh167 a novel method is outlined that com ines the a ove mentioned pulse diversity radar si nalin with a new coherent two dimensional pr ocessin of the measured data to effectively suppress a DRFM repeater jammer. I AR. The inverse AR is a method of reconstructin a hi h resolution twodimension al EM intensity ima e of movin tar ets (e. ., ships, aircraft) in the ran e and cross ran e (doppler) domains. I AR ima in is important in military applicatio ns such as tar et reco nition and classification (since it can usually reco nize the class of tar et) that can also e used to cue weapon systems. The need for coherent counterin of these ima in sensors is a hi h priority for EW. The refe rences174,175 present the Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

  

 

 

 

 

 

  

 

 

 

 

Electronic Counter Countermeasures 24.52 RADAR HANDBOOK desi n of a pipelined all di ital ima e synthesizer capa le of eneratin false tar et ima es from a series of intercepted I AR chirp pulses, thus providin RF ima in decoy capa ility. The ima e synthesizer modulates the phase samples from a phase samplin DRFM that stores intercepted I AR pulses. The ima e synthesize r must also synthesize the temporal len thenin and amplitude modulation caused y the many reflective surfaces of a tar et and must enerate a realistic dopple r profile for each surface. The position of a false tar et ima e in ran e can e controlled y delayin in time the read out samples oin to the ima e synthesi zer. The ran e doppler ima e of a ship with 32 ran e ins is synthesized as an e xample in Pace et al.175 ECCM techniques to defeat this type of jammin si nals are similar to those proposed for AR. Over the Horizon Radar. An important defe nse related role of hi h frequency (HF) over the horizon (OTH) radar is to provi de a capa ility for early warnin detection and trackin of air and ship tar ets . By usin the ionosphere as a propa ation medium, skywave OTH radars can operat e at very lon distances to achieve detection and trackin at ran es of 5003000 k m. On the other hand, surface wave OTH radars exploit vertically polarized HF si nals (330 MHz) and the conductive properties of sea water to detect tar ets at r an es limited to a out 250 km. This upper limit enerally applies to lar e ships and frequencies in the lower HF and.176,177 ECM to OTH Radar. For oth skywave and surface wave OTH radar systems, the ionosphere also propa ates unwanted int erference si nals to the radar site, particularly at ni ht when the ionosphere i s prone to propa atin radio frequency interference (RFI) sources from very lon distances. RFI can arise from unintentional and intentional anthropo enic emitt ers in the user con ested HF and, as well as jammin sources. Jammin sources m ay e located on the tar et platform itself (self screenin ) and received y the main antenna eam or radiate from a separate location (stand off) and e receiv ed mainly throu h the antenna eampattern sidelo es. The jammin si nal may e i ncoherent with the radar waveform and operate in a spot or arra e fashion to raise the noise floor in oth ran e and doppler search spaces to potentially impair de tection performance, or it can e coherent with the radar waveform, as in the ca se of deception jammin , which may enerate false tar ets and potentially impair the trackin system from followin the true tar et. Impact of Ionosphere. An im portant aspect that distin uishes OTH radar from line of si ht systems is the im pact of the ionospheric propa ation medium on the characteristics of the receive d interference. The ionosphere is stratified with different reflectin layers, s o a sin le interference source is often received as a num er of multipath compon ents with different DoAs, oth in elevation (due to the different hei hts of ref lection points) and in azimuth (due to layer dependent ionospheric tilts or rad ients). In addition to multipath, each interference component is su jected to te mporal and spatial distortions caused y the dynamic ehavior of electron densit y irre ularities present in the individual reflectin layers.178 This physical p henomenon is known not only to deform the interference wavefronts relative to th e anticipated plane wavefront, ut also to induce a si nificant level of spatial nonstationarity on the various interference components over time intervals comm ensurate with the coherent processin interval of OTH radars (in the order of a few to tens of seconds).179,180 Relevance to Interferin i nals. ources of int erference within the radar covera e (e. ., on an air orne platform) can potentia lly screen the platform in ran e and impair the detection of other tar ets with similar azimuth ut possi ly at different ran es. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

 

 

 

 

  

 

 

      

 

 

 

 

 

     

 

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEA URE 24.53 uch sources can expect ood propa ation to the radar receiver ecause the choic e of operatin frequency is usually optimized for the covera e area. In the case of standoff interference sources, which are located ar itrarily with respect to the surveillance re ion (e. ., a round ased emitter), propa ation conditions will enerally e su optimum. However, such sources may have reater power and a ntenna ain at their disposal, allowin the si nals to reach the radar receiver with apprecia le stren th, sometimes after propa ation via hi hly distur ed and nonstationary ionospheric paths (as commonly occurs in the equatorial and polar re ions). Under normal circumstances, OTH radars seek to find relatively clear f requency channels in the user con ested HF spectrum, so the presence of interfer ence from other manmade sources is effectively diminished y suita le frequency selection. When jammin is present, the radar may need to operate with much hi h er than usual levels of interference that can de rade performance. For this reas on, protection in the form of ECCM techniques ecomes necessary. ECCM Techniques . Electronic protection for OTH radar antenna arrays can e provided in the form of adaptive si nal processin in space and time. The stochastic constraints ada ptive eamformin and TAP methods181184 were developed specifically for the HF e nvironment to address the rejection of nonstationary interference while protecti n the clutter doppler spectrum properties. A method for time varyin spatial ad aptive processin (TV AP)185 that addresses the same pro lem was found to e mo re attractive for practical implementation due to the much lower computational c ost in real time applications, as well as reater ro ustness in protectin su c lutter visi ility after doppler processin . The pro lem of reducin false alarms caused y stron sidelo e tar ets and spatially structured (non aussian distri uted) RFI was treated in Fa rizio et al.,186 where the advanta es of adaptive s u space detectors relative to conventional approaches were shown. TAP technique s with temporal de rees of freedom spaced at the PRI (i.e., slow time) have een proposed in Farina et al.187 to jointly cancel RFI and clutter when oth are of similar stren th ut neither can e isolated for estimation, whereas an alterna tive low dimension TAP formulation with temporal taps spaced at the ran e cell interval (i.e., fast time) has een proposed in Fa rizio et al.,188 to jointly c ancel sidelo e and main eam RFI that exhi its correlation in the ran e dimensio n. The TAP methods used in OTH radar are very similar conceptually to those ado pted for air orne radar, especially the former tap architecture.187 The chief di fference is that in eni n conditions (free of si nificant co channel interferen ce), TAP is not indicated for OTH radar ecause the sidelo e clutter does not ty pically mask doppler shifted tar ets any more than the main eam clutter. A possi le exception to this is ship orne HF surface wave radar,* althou h such systems have een proposed, they have not yet demonstrated their practical utility. Co channel interference for OTH radar refers mainly to other transmissions in th e HF spectrum that either fully or partially overlap the radar andwidth. The ma in eam and sidelo e clutter received y air orne radars can have quite differen t doppler shifts due to the movement of the platform with respect to the round resultin in the an le doppler couplin of the clutter. However in OTH radar, ma in eam and sidelo e clutter from a sin le ionospheric mode typically have simil ar doppler spectrum characteristics ecause the radar is stationary. This means that sidelo e clutter appears at rou hly the same doppler shift as main eam clu tter and doppler filterin can e used effectively for detectin tar ets usually without special need to reject the sidelo e clutter spatially. * O viously the situation may chan e in ship orne HF surface wave radar ecause the platform is movin with respect to the sea surface and hence a conceptually similar situatio n arises for the clutter as encountered in air orne radar. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any

 

 

 

 

 

 

 

 

 

 

use is su ject to the Terms of Use as iven at the we site.

Electronic Counter Countermeasures 24.54 RADAR HANDBOOK 24.12 ECCM AND ECM EFFICACY There is a need for a quantitative measurement of the efficacy of one or more EC CM electronic techniques when a radar equipped with these devices is su ject to an ECM threat. One performance measure enerally used for an unjammed search rad ar is the detection ran e of a certain tar et a ainst a system noise ack round; this situation is referred to as detection in clear environment. When the radar is jammed, it is of interest to calculate the de radation of the detection ran e with respect to self screenin , standoff, and escort jammers. These calculatio ns apply to oth search and trackin radars. For trackin radars, it is also wor thwhile to consider the de radation of measurement accuracy and resolution. The enefits of usin ECCM techniques such as frequency a ility, coherent doppler pr ocessin , very low sidelo e antennas, and LC can e easily assessed at a first approximation y properly modifyin the parameters involved in the radar equatio n. If, for instance, an LC is adopted a ainst an OJ, its net effect is to redu ce jammin power y the amount of jammer cancellation ratio that the LC can off er. The prediction of radar ran e is difficult ecause of the many factors that are hard to represent with models of the required accuracy. The factors involve the tar et to e detected (tar et returns of an unknown statistical nature), the natural environment in which the tar et is em edded (e. ., clutter returns, uni ntentional interference, uncontrolla le environmental refraction, and a sorption ), the random nature of the interference, and the radar itself (system noise tem perature, si nal distortions, etc.). Nevertheless, radar ran e prediction made u nder avera e conditions provides a preliminary and useful indication of performa nce under ECM threat and ECCM desi n effectiveness that produces aseline values prior to simulation and operational tests. A classical ook presents accurate d etection ran e equations in a variety of practical situations.189 In the second part of this section, a review of software tools availa le for the prediction of ran e equation in jammin and chaff conditions is iven. Of course, the radar e quation is a simplification in assessin ECM ECCM interactions; a measure of ECC M effectiveness should involve the whole weapon system in which the radar operat es. The measure of effectiveness should e expressed in terms of the num er of a ttackers destroyed or the pro a ility of radar survival. References in the liter ature attempt to assess the ECCM efficacy.190194 imulation is another means to a ssess the ECCM enefits in radar and weapon systems.193 An advanta e of this app roach resides in the capa ility to artificially enerate different types of thre ats and to look at the radar160,161 and weapon system reactions. However, the si mulation of such a complex system is a difficult, timeconsumin task that someti mes involves the use of ad hoc pro rammin lan ua es suita le for simulation. i mulation of a complex system on a di ital computer is a technique used for the a nalysis, desi n, and testin of a system whose ehavior cannot e easily evaluat ed y means of analysis or computation. The procedure essentially consists of re producin the al orithms of a suita le model of the examined system y means of computer pro rams. Proper inputs to the model, correspondin to the most relevan t operational conditions for the real system, can e enerated y the same compu ter pro rams. The outputs o tained are compared with some reference values (expe cted or theoretical) to assess system performance. When random inputs are provid ed, a num er of statistically independent trials are performed to achieve a si n ificant sample of the output values from which relia le statistics can e estima ted. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

 

 

 

  

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEA URE 24.55 The accuracy and detail of the model may vary from a coarse functional descripti on of the system to a very accurate one, accordin to the purpose of the simulat ion and the required accuracy of the results. However, it is desira le to limit the complexity of the simulation tools in order to have mana ea le pro rams, iv in results that are easily interpreted. The accuracy in representin each syste m function depends upon its relevance with respect to system performance. When a very complex system is to e simulated, it is enerally preferred to resort to several pro rams of limited complexity in lieu of a sin le ulky simulation. Thi s approach corresponds to partitionin the whole system into su systems separate ly modeled in detail. From each partial simulation, a limited num er of relevant features are extracted and employed to uild a simplified model of the overall system. imulation is particularly important to account for the adaptive nature (e. ., CFAR, adaptive eamformin , automatic radar mana ement, adaptive trackin , adaptive clutter cancellation) of modern radar systems.195 In this case, tradi tional static measures such as detection ran e a ainst a iven tar et will no lo n er adequately define the capa ilities of radar systems. Measures of radar dyna mic characteristics, such as the suscepti ility to processor overload or the tim e to adapt in chan in conditions, are more important. Modelin and simulations to evaluate the radar response to standardized chan in scenarios represent an a ttractive technical solution.195 imulation is always of value; however, the eff ectiveness of ECM and ECCM is ultimately done, when possi le, with tests of real EW capa ilities a ainst real radar systems under real world conditions. This is especially important for radar equipped with adaptive techniques since they mi ht not e always fully modeled in a simulation as they are in the real world env ironment in which they must operate. The Radar Equation in Jammin and Chaff Con ditions. An example of radar ran e performance under noise jammin is reported o n pp. 1419 of Farina,34 where the important role played y a radar with low sidel o e antennas is also noted. Today the use of computer pro rams for predictin ra dar performance under jammin , clutter, and chaff, and in the presence of variou s refined propa ation models is well esta lished: there are pro rams developed i n house y individual radar companies196 or availa le on the market.197 The Rada r Work tation (RW ) is an example of a developed in house pro ram196; RW ori i nates from the modelin and simulation activities carried out for prediction of radar performance in several scenarios. One main o jective of RW is to provide the radar analyst or system desi ner with a friendly ut comprehensive toolkit f or prediction of radar performance ased on well reco nized, flexi le, and docum ented mathematical models. A road ran e of radar types ( i dimensional, multi eam threedimensional, phased array), composite clutter, ECM and propa ation scen arios, and a tar ets kinematics and RC features are covered. Input and output da ta can e saved, loaded, and exported to other similar applications or for ener al use (i.e., M Office tools for data analysis). A second purpose is to provide a handy and relia le tool for technicians and en ineers performin system setup at the site or acceptance tests y means of field trials, y providin not only the software tools and models ut also, where required, a data ase of predictio n results, and allowin simple parametric excursions thereof, without the need t o consult a ulky reference documentation. In rief, the most valua le outcomes that can e o tained with the RW are radar ran e calculation, radar elevation c overa e dia rams in clear, ECM and multi propa ation oth for coherent and nonco herent radars; ran e and velocity responses in complex Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

 

 

  

 

 

 

Electronic Counter Countermeasures 24.56 RADAR HANDBOOK scenarios (multiple clutter sources, user defined trajectory) in terms of si nal to distur ance power ratio and detection pro a ility; and radar ran e and hei h t accuracy calculation, radar resolution evaluations employin suita le data ext ractin lo ics. The RW suite consists of the followin main modules: C\C++ and Fortran (to calculate special functions like Gamma and Bessel K) coded li raries ; a standard li rary of Windows APIs (Application Pro rammin Interfaces) to dra w the covera e dia rams; a template li rary to implement the matrix al e ra; an application ased on M Office Excel to code the Blake Chart; a set of Visual Bas ic tools to evaluate some aspects of radar performance (e. ., ADC jitter, atmosp heric loss, taperin loss, etc.); and an unformatted archive of radar data, as s imple A CII files, pertinent to performance, environment, trajectories, terrain hei ht, and waveform. User friendly interfaces run on low cost platforms (PC) an d popular environments (Win98, WinNT, Win2000, Windows XP, Vista) for users and developers. In the RW , the chaff volume clutter model is characterized in terms of volume extent, EM reflectivity, and doppler spectrum. The si nal to noise pl us chaff ratio is determined on the asis of the chaff location in the space, th e antenna receivin pattern, and the radiated radar waveform. The radar equation can e applied, and the si nal processin scheme can e emulated to determine t he amount of chaff miti ation. A arra e noise jammer is modeled in terms of eff ective radiated power (ERP) and frequency and of operation. The si nal to noise plus jammer ratio is determined on the asis of the JDoA and the antenna receiv in patterns; the radar equation is then applied; and suita le ECCM si nal proce ssin schemes can e emulated to determine to which extent the jammer is attenua ted. Computer Aided Radar Performance Evaluation Tool (CARPET) is an example of availa le software on the market. In the CARPET 1.0 manual,197 the equations for calculatin the contri utions from chaff (volume clutter) in the si nal tointer ference ratio are descri ed on pp. 59 and 60, and the equations for calculatin the contri ution from noise jammin ( arra e or responsive) are descri ed on p. 61. CARPET is pro rammed in C++ and has a Windows XP compati le user friendly r aphical interface. ACRONYM LI T ADC ADT AF AF AGC AMF API A RGPO ARM A OJ Analo ue to Di ital Converter Automa tic Detection and Trackin Am i uity Function Automatic Frequency election Auto matic Gain Control Adaptive Matched Filter Application Pro rammin Interface Ant i Ran e Gate Pull Off Anti Radiation Missile Anti tand Off Jammer Actually, it is a eneralized Blake chart (that improves on the ori inal Blake c hart), which includes details such as antenna patterns, processin , system losse s, etc., in a suita le electronic format. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

 

 

 

 

  

  

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEA URE 24.57 BM CARPET CFAR CORDIC CO RO COT CRI CRLB CUT CW DAC DAR DECM DF DoA Dof DRFM DT oA EM EA ECCM ECM ELINT EMCON EP ERP E E M EW FFT FPGA FTC GA G LC HOJ HF IF IM M I AR JCR JDoA JNR LORO LPI MBC MEM ML MTD MTI Ballistic Missile Computer Aided Radar Performance Evaluation Tool Constant Fals e Alarm Rate COordinate Rotation Di ital Computer Conical can on Receiver Only Commercial Off The helf Coherent Repeater Interference Cramer Rao Lower Bound C ell Under Test Continuous Wave Di ital to Analo ue Converter Di ital Array Radar Deceptive ECM Direction Findin Direction of Arrival De ree of freedom Di ital Radio Frequency Memory Difference Time of Arrival Electroma netic Electronic Att ack Electronic Counter Counter Measure Electronic Counter Measure ELectronic INT elli ence EMission CONtrol Electronic Protection Effective Radiated Power Electr onic upport Electronic warfare upport Measure Electronic Warfare Fast Fourier Transform Field Pro ramma le Gate Arrays Fast Time Constant Genetic Al orithm Ge neralized ide Lo e Canceler Home On Jam Hi h Frequency Intermediate Frequency I nteractin Multiple Model Inverse ynthetic Aperture Radar Jammer Cancellation R atio Jammer Direction of Arrival Jammer to Noise Ratio Lo in On Receive Only Lo w Pro a ility of Intercept Main Beam Canceler Maximum Entropy Method Maximum Lik elihood Movin Tar et Detector Movin Tar et Indicator Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

Electronic Counter Countermeasures 24.58 RADAR HANDBOOK NLI OTH PAR PDR PDW Penaids PPI PRF PRI P LR RC RF RFI RFM RGPO RWR Rms RW RX AR AW INR LB LC NR OJ P J TAP TT ToA T I TV AP TW UHF ULA VGPO VHF VL I Noise Like Interference Over The Horizon Phased Array Radar Phase Difference Rat e Pulse Description Word Penetration Aid Decoy Plan Position Indicator Pulse Rep etition Frequency Pulse Repetition Interval Peak to ide Lo e Ratio Radar Cross ection Radio Frequency Radio Frequency Interference Ran e Filter Map Ran e Gate Pull Off Radar Warnin Receiver Root Mean quare Radar Work tation Receiver y nthetic Aperture Radar urface Acoustic Wave i nal to Interference plus Noise R atio ideLo e Blankin ideLo e Canceler i nal to Noise Ratio tand Off Jammer elf Protection elf creenin Jammer pace Time Adaptive Processin in le Tar et Tracker Time of Arrival Terrain cattered Interference Time Varyin patial A daptive Processin Track While can Ultra Hi h Frequency Uniform Linear Array Ve locity Gate Pull Off Very Hi h Frequency Very Lar e cale Inte ration ACKNOWLEDGMENT The Author wishes to warmly thank his collea ues for their cooperation in this w ork: Dr. L. Timmoneri, Dr. L. Ortenzi, and Dr. E. Andreta ( ELEX istemi Inte ra ti, Italy), Dr. G. A. Fa rizio (D TO, Australia), Dr. U. Nickel (FGAN, Germany), Prof. L. Chisci, Dr. A. Benavoli, and Dr. . Roma noli (University of Florence, Italy), Dr. M. Grazzini (Elettronica pA, Italy), and Dr. . Ko on (MIT, Lincol n La oratory, U. .). Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

  

 

 

 

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEA URE 24.59 REFERENCE 1. . L. Johnston, World War II ECCM history, suppl. to IEEE Int. Radar Conf. Rec. , May 69, 1985, pp. 5.25.7. 2. A. E. Hoffmann Heiden, Anti jammin techniques at th e German AAA radars in World War II, suppl. to IEEE Int. Radar Conf. Rec., pp. 5. 225.29, May 69, 1985. 3. D. C. chleher, Introduction to Electronic Warfare, Norwo od, MA: Artech House, Inc., 1986. 4. D. C. chleher, Electronic Warfare in the I nformation A e, Norwood, MA: Artech House, Inc., 1999. 5. B. J. locum and P. D . West, ECM modelin for multitar et trackin and data association, in Multitar et Multisensor Trackin : Applications and Advances, vol. III, Y. Bar halom and W. D. Blair (eds.), Norwood, MA: Artech House, Inc., 2000, pp. 395458. 6. F. Neri, Introduction to Electronic Defense, 2nd Ed., Norwood, MA: Artech House, Inc., 20 01. 7. L. Nen jin and Z. Yi Tin , A survey of radar ECM ECCM, IEEE Trans., vol. A E 31, no. 3, pp. 11101120, July 1995. 8. . L. Johnston (ed.), Radar Electronic Co unter Countermeasures, Norwood, MA: Artech House, Inc., 1979. 9. pecial Issue o n electronic warfare, IEE Proc., vol. 129, pt. F, no. 3, pp. 113232, June 1982. 1 0. W. A. Davis, Principles of electronic warfare: Radar and EW, Microwave J., vol. 33, pp. 5254, 5659, Fe ruary 1980. 11. L. B. Van Brunt, The Glossary of Electroni c Warfare, Dunn Lorin , VA: EW En ineerin , Inc., 1984. 12. Department of Defens e, Joint Chiefs of taff, Dictionary of Military and Associated Terms, JC Pu 1 , eptem er 1974. 13. L. B. Van Brunt, Applied ECM, vol. 1, Dunn Lorin , VA: EW En ineerin , Inc., 1978. 14. R. G. Wiley, Electronic Intelli ence: The Analysis of Radar i nals, Norwood, MA: Artech House, Inc., 1985. 15. R. G. Wiley, Electr onic Intelli ence: The Interception of Radar i nals, Norwood, MA: Artech House, Inc., 1986. 16. R. G. Wiley, ELINT: The Interception and Analysis of Radar i n als, Norwood, MA: Artech House, Inc., 2006. 17. R. A. Poisel, Electronic Warfare Tar et Location Methods, Norwood, MA: Artech House, Inc., 2005. 18. E. P. Pace, Detectin and Classifyin Low Pro a ility of Intercept Radar, Norwood, MA: Arte ch House, Inc., 2003. 19. D. C. chleher, LPI radar: Fact or fiction, IEEE AE Ma azine, vol. 21, no. 5, pp. 36, May 2006. 20. . L. Johnston, Philosophy of ECCM ut ilization, Electron. Warfare, vol. 7, pp. 5961, MayJune, 1975. 21. M. V. Maksimov, et al., Radar Anti Jammin Techniques, Norwood, MA: Artech House, Inc., 1979. (T ranslated from Russian, Zaschita at Radiopomekh, oviet Radio, 1976.) 22. D. Cli fford Bell, Radar countermeasures and counter countermeasures, Mil. Technol., pp. 96111, May 1986. 23. J. A. Adam and M. A. Fischetti, tar Wars. DI: The rand exp eriment, IEEE pectrum, vol. 23, no. 9, pp. 3446, eptem er 1985. 24. . J. Roome, Di ital radio frequency memory, Electronic & Communication En ineerin Journal, p p. 147153, Au ust 1990. 25. J. W. Goodman and M. ilvestri, ome effects of Fourie r Domain Phase Quantization, IBM J. Res. Develop., pp. 478484, eptem er 1970. 26. M. Greco, F. Gini, and A. Farina, Com ined effect of phase and RGPO delay quanti zation on jammin si nal spectrum, Proc. of IEEE Int. Conf. on Radar, Radar 2005, Washin ton, DC (U A), May 1012, 2005, pp. 3742. 27. . D. Ber er, Di ital radio fr equency memory linear ate stealer spectrum, IEEE Trans., vol. AE 29, no. 2, pp. 7 25735, April 2003. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

  

Electronic Counter Countermeasures 24.60 RADAR HANDBOOK 28. G. V. Morris et al., Principles of electronic counter countermeasures, short l ecture notes, Geor ia Institute of Technolo y, 1999. 29. L. B. Van Brunt, Applie d ECM, vol. 2, Dunn Lorin , VA: EW En ineerin , Inc., 1982. 30. P. J. Gros, D. C . ammons, and A. C. Cruce, ECCM Advanced Radar Test Bed (E/ARTB) systems definit ion, IEEE Nat. Aerosp. Electron. Conf. NAECON 1986, May 1923, 1986, pp. 251257. 31. M. A. Johnson and D. C. toner, ECCM from the radar desi ners view point, Microwav e J., vol. 21, pp. 5963, March 1978. 32. H. E. chrank, Low sidelo es phased array and reflectors antennas, in Aspects of Modern Radar, E. Brookner (ed.), Norwood, MA: Artech House, Inc., 1988. 33. W. T. Patton, Low idelo e Antennas for Tactic al Radars, IEEE Int. Radar Conf. Rec., April 2830, 1980, pp. 243254. 34. A. Farina, Antenna Based i nal Processin Techniques for Radar ystems, Norwood, MA: Arte ch House, Inc., 1992. 35. F. J. Harrys, On the use of windows for harmonic analys is with the Discrete Fourier Transform, Proc. IEEE, vol. 66, pp. 5183, January 197 8. 36. E. Brookner, Trends in radar systems and technolo y to the year 2000 and eyond, in Aspects of Modern Radar, E. Brookner (ed.), Artech House, Inc., Norwood , MA, 1988. 37. E. Brookner, Phased array around the world. Pro ress and future t rends, IEEE Int. ymp. on Phased Array ystems and Technolo y 2003, Boston (U A), Octo er 1417, 2003, pp. 18. 38. M. Cicolani, A. Farina, E. Giaccari, F. Madia, R. Ronconi, and . a atini, ome phased array systems and technolo ies in AM , IEEE Int. ymp. on Phased Array ystems and Technolo y, Boston (U A), Octo er 1417, 2 003, pp. 2330. 39. W. Kuhn, W. ieprath, L. Timmoneri, and A. Farina, Phased array radar systems in support of the Medium Extended Air Defense ystem (MEAD ), IEEE Int. ymp. on Phased Array ystems and Technolo y, Boston (U A), Octo er 1417, 2 003, pp. 94100. 40. A. R. Moore, D. M. alter, and W. K. tafford, ME AR (Multi Fu nction, Electronically canned, Adaptive Radar), Proc. of Int. Conf. Radar 97, Ed in ur h, Octo er 1416, 1997, Pu lication no. 449, London, UK: IEE, pp. 5559. 41. D . Giuli, Polarization diversity in radars, Proc. IEEE, vol. 74, pp. 245269, Fe ruar y 1986. 42. L. Maisel, Performance of sidelo e lankin systems, IEEE Trans., vol. AE 4, no. 1, pp. 174180, March 1968. 43. P. O. Aranci ia, A sidelo e lankin syst em desi n and demonstration, Microwave J., vol. 21, pp. 6973, March 1978; reprinte d in Ref. 8, 1979. 44. D. H. Harvey and T. L. Wood, Desi ns for sidelo e lankin systems, IEEE Int. Radar Conf. Rec., April 1980, pp. 41416. 45. M. O ullivan, A com parison of sidelo e lankin systems, IEE Int. Conf. Radar87, Conf. Pu . 281, Lon don, UK, Octo er 1921, 1987, pp. 345349. 46. A. Farina and F. Gini, Calculation of lankin pro a ility for the sidelo e lankin ( LB) for two interference statis tical models, IEEE i nal Processin Letters, vol. 5, no. 4, pp. 98100, April 1998 . 47. A. Farina and F. Gini, Blankin pro a ilities for LB system in correlated clutter plus thermal noise, IEEE Trans., vol. P48, no. 5, pp. 14811485, May 2000. 48. A. Farina and F. Gini, Desi n of LB systems in presence of correlated round clutter, IEE Proc., vol. 147, pt. F, no. 4, pp. 199207, 2000. 49. A. De Maio, A. Farina, and F. Gini, Performance analysis of the sidelo e lankin system for two fluctuatin jammer models, IEEE Trans., vol. AE 41, no. 3, pp. 10821090, July 2005 . 50. D. A. hnidman and . . Toumod e, idelo e lankin with inte ration and t ar et fluctuation, IEEE Trans., vol. AE 38, no. 3, pp. 10231037, July 2002. 51. P. W. Howells, Intermediate Frequency idelo e Canceler, U. . Patent 3,202,990, Au us t 24, 1965. 52. . P. Apple aum, P. W. Howells, and C. Kovarik, Multiple Intermed iate Frequency ide Lo e Canceler, U. . Patent 4,044,359, Au ust 23, 1977. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

  

 

 

 

 

 

 

   

 

 

  

 

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEA URE 24.61 53. R. A. Monzin o and T.W. Miller, Introduction to Adaptive Arrays, New York: J ohn Wiley & ons, 1980. 54. J. Hudson, Adaptive Array Principles, London: Peter Pere rinus Ltd., 1981. 55. R. Nitz er , Adaptive i nal Processin for Radar, No rwood, MA: Artech House, Inc., 1992. 56. H. D. Griffiths, A four element VHF adap tive array processor, Proc. 2nd IEE Int. Conf. on Antennas and Propa ation, IEE C onf. Pu . no. 195, pt.1, York (UK), April 1316, 1981, pp. 185189. 57. A. Farina, Di ital equalisation in adaptive spatial filterin : a survey, i nal Processin , El sevier, vol. 83, no. 1, pp. 1129, January 2003. 58. B. D. Carlson, L. M. Goodman, J. Austin, M. W. Ganz, and L. O. Upton, An ultralow sidelo e adaptive array ante nna, The Lincoln La oratory Journal, vol. 3, no. 2, pp. 291310, 1990. 59. W. F. Ga riel, Adaptive di ital processin investi ation of DFT su andin vs. transvers al filter canceler, Naval Research La oratory, NRL Report 8981, July 28, 1986, Wa shin ton, DC (U A). 60. A. Farina and R. anzullo, Performance limitations in ada ptive spatial filterin , i nal Processin , Elsevier, vol. 81, no.10, pp. 21552170 , Octo er 2001. 61. K. Gerlach, The effects of IF andpass mismatch errors on ada ptive cancellation, IEEE Trans., vol. AE 26, no. 3, pp. 455468, May 1990. 62. A. Fa rina, G. Golino, L. Timmoneri, and G. Tonelli, Di ital equalisation in adaptive s patial filterin for radar systems: Application to live data acquired with a ro und ased phased array radar, Radar 2004, Toulouse, France, Octo er 1921, 2004. 63 . R. Fante, R. Davis, and T. Guella, Wide and cancellation of multiple main eam j ammers, IEEE Trans., vol. AP44, no. 10, pp. 14021413, Octo er 1996. 64. F. E. Churc hill, G. W. O ar, and B. J. Thompson, The correction of I and Q errors in a coher ent processor, IEEE Trans., vol. AE 17, no. 1, pp. 131137, January 1981. 65. K. Ger lach, The effect of I, Q mismatchin errors on adaptive cancellation, IEEE Trans., vol. AE 28, no. 7, pp. 729740, July 1992. 66. K. Gerlach and M. J. teiner, An ada ptive matched filter that compensates for I, Q mismatch errors, IEEE Trans., vol. P45, no. 12, pp.31043107, Decem er 1997. 67. A. Farina and L. Ortenzi, Effect of ADC and receiver saturation on adaptive spatial filterin of directional interfe rence, i nal Processin , Elsevier, vol. 83, no. 5, pp. 10651078, 2003. 68. A. Far ina, R. anzullo, and L. Timmoneri, Performance limitations and remedies in adapt ive spatial filterin with timin errors, i nal Processin , Elsevier, vol. 82, n o. 2, pp. 195204, Fe ruary 2002. 69. D. R. Mor an and A. Arid ides, Adaptive sidel o es cancellation of wide and multipath interference, IEEE Trans., vol. AP33, no. 8, pp. 908917, Au ust 1985. 70. R. L. Fante, Cancellation of specular and diffuse jammer multipath usin a hy rid adaptive array, IEEE Trans., vol. AE 27, no. 10, pp. 823837, eptem er 1991. 71. A. Farina and L. Timmoneri, Cancellation of clutte r and e.m. interference with TAP al orithms. Application to live data acquired with a round ased phased array radar demonstrator, Proc. of 2004 IEEE Radar Con f., Philadelphia (U A), April 2629, 2004, pp. 486491. 72. A. Farina, L. Timmoneri, and R. Tosini, Cascadin LB and LC devices, i nal Processin , Elsevier, vol. 4 5, no. 2, pp. 261266, 1995. 73. A. Farina and L. Timmoneri, ystolic schemes for J oint LB, LC and adaptive phased array, Proc. of Int. Conf. on Radar, Radar 2000 , Washin ton, DC, U A, May 712, 2000, pp. 602607. 74. L. Timmoneri, I. K. Proudler , A. Farina, and J. G. McWhirter, QRD Based MVDR al orithm for adaptive multipuls e antenna array si nal processin , IEE Proc., vol. 141, pt. F, no. 2, pp. 93102, A pril 1994. 75. P. Bollini, L. Chisci, A. Farina, M. Giannelli, L. Timmoneri, and G. Zappa, QR versus IQR al orithms for adaptive si nal processin : performance e valuation for radar applications, IEE Proc., vol. 143, pt. F, no. 5, pp. 328340, O cto er 1996. 76. A. Farina and L. Timmoneri, Real time TAP techniques, Electronic s & Communications En ineerin Journal, pecial Issue on TAP, vol. 11, no.1, pp . 1322, Fe ruary 1999. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 



 

 

 

Electronic Counter Countermeasures 24.62 RADAR HANDBOOK 77. P. Kapteijin, E. Deprettere, L. Timmoneri, and A. Farina, Implementation of t he recursive QR al orithm on a 2*2 CORDIC test oard: a case study for radar app lication, Proc. of the 25th European Microwave Conf., Bolo na (Italy), eptem er 47, 1995, pp. 500505. 78. A. DAcierno, M. Ceccarelli, A. Farina, A. Petrosino, and L. Timmoneri, Mappin QR decomposition on parallel computers: a study case for ra dar applications, IEICE Trans. on Communications, vol. E77B, no. 10, pp. 12641271, Octo er 1994. 79. A. Farina and L. Timmoneri, Parallel processin architectures f or TAP, in Applications of pace Time Adaptive Processin , R. Klemm (ed.), Londo n, UK, IEE Radar, onar and Navi ation eries 14, 2004, pp. 265302. 80. A. Farina , A. Aver ouch, D. Gi or, L. Lescarini, . Levit, . tefanini, and L. Timmoneri , Multichannel radar: Advanced implementation technolo y and experimental results , Proc. of Int. Radar ymp., IR 2005, Berlin (Germany), eptem er 68, 2005, pp. 31 7329. 81. C. M. Rader, Wafer scale inte ration of a lar e scale systolic array for adaptive nullin , The Lincoln La oratory Journal, vol. 4, no. 1, pp. 329, 1991. 8 2. C. M. Rader, VL I systolic array for adaptive nullin , IEEE i nal Processin M a azine, vol. 13, no. 4, pp. 2949, July 1996. 83. . P. Apple aum, Adaptive arrays , yracuse University Research Corporation Rept. PL TR 661, 1966. This report is reproduced in IEEE Trans., vol. AP24, pp. 585598, eptem er 1976. 84. L. E. Brenna n and I. . Reed, Theory of adaptive radar, IEEE Trans., vol. AE 9, no. 1, pp. 23 7252, March 1973. 85. B. Wardrop, The role of di ital processin in radar eamform in , GEC J. Res., vol. 3, no. 1, pp. 3445, 1985. 86. P. Valentino, Di ital eamform in : new technolo y for tomorrows radars, Def. Electron., pp. 102107, Octo er 1984. 87. H. teyskal, Di ital eamformin antennas: an introduction, Microwave J., pp. 107124, January 1987. 88. B. Cantrell, J. de Graaf, L. Lei owitz, E. Willwerth, G. Meurer, C. Parris, and R. tapleton, Development of a Di ital Array Radar (DAR ), Proc. of IEEE Radar Conf. 2001, Atlanta (Geor ia), May 13, 2001 pp. 157162. 89. M. Zatman, Di itization requirements for di ital radar arrays, IEEE Radar Conf. 20 01, Atlanta (Geor ia), May 13, 2001, pp. 163168. 90. I. . Reed, A rief history of adaptive arrays, u dury/Wayland Lecture eries, Raytheon Div. Education, notes 23, Octo er 1985. 91. D. Etter, A. teinhardt, and . toner, Least squares adapt ive processin in military applications, IEEE i nal Processin Ma azine, vol. 19 , no. 3, pp. 6673, May 2002. On occasion of the 2001 B. Franklin Medal awarded to B. Widrow for pioneerin work on adaptive si nal processin . 92. . Haykin and A. teinhardt, Adaptive Radar Detection and Estimation, New York: John Wiley & ons, Inc., 1992. 93. . T. mith, Adaptive Radar, in Wiley Encyclopedia of Electri cal and Electronic En ineerin , J. G. We ster (ed.), vol. 1, New York: Wiley, 19 99 (updated 13 July 2007), pp. 263289. 94. A. Farina, C. H. Gierull, F. Gini, and U. Nickel (eds.), pecial Issue New trends and findin s in antenna array process in , i nal Processin , Elsevier, vol. 84, no. 9, pp. 14771688, eptem er 2004. 95 . J. Ward, pace time adaptive processin for air orne radar, MIT Lincoln La orato ry Technical Report TR1015, Decem er 13, 1994. 96. R. Klemm, Principles of pace Time Adaptive Processin , 3rd Ed., London, UK: IET Radar, onar and Navi ation eries 21, 2006. 97. R. Klemm (ed.), Applications of pace Time Adaptive Processi n , London, UK: IEE Radar, onar and Navi ation, eries 14, 2004. 98. J. R. Guer ci, pace Time Adaptive Processin for Radar, Norwood, MA: Artech House, Inc., 2 003. 99. B. Testa and V. Vannicola, The physical si nificance of the ei envalues in adaptive arrays, Di ital i nal Processin , vol. 15, pp. 9196, 1995. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

 

 

 

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEA URE 24.63 100. B. D. Carlson, Covariance matrix estimation errors and dia onal loadin in a daptive arrays, IEEE Trans., vol. AE 24, no. 3, pp. 397401, July 1988. 101. A. Fari na, P. Lan sford, G. C. arno, L. Timmoneri, and R. Tosini, ECCM techniques for a rotatin , multifunction, phased array radar, Proc. of the 25th European Microwav e Conf, Bolo na (Italy), eptem er 47, 1995, pp. 490495. 102. W. D. Wirth, Radar T echniques Usin Array Antennas, London, UK: IEE Radar, onar, Navi ation and Avi onics, eries 10, 2001. 103. J. B. Hoffman and B. L. Ga elach, Four channel monop ulse for main eam nullin and trackin , Proc. of IEEE National Radar Conf. NATRA D 97, yracuse, New York, May 1315, 1997, pp. 9498. 104. A. Farina, P. Lom ardo, an d L. Ortenzi, A unified approach to adaptive radar processin with eneral antenn a array confi uration, pecial Issue on New trends and findin s in antenna array p rocessin for radar, i nal Processin , Elsevier, vol. 84, no. 9, pp. 15931623, e ptem er 2004. 105. R. C. Davis, L. E. Brennan, and I. . Reed, An le estimation w ith adaptive arrays in external noise field, IEEE Trans., vol. AE 12, no. 2 pp. 17 9186, March 1976. 106. P. Lan sford A. Farina, L. Timmoneri, and R. Tosini, Monopu lse direction findin in presence of adaptive nullin , presented at IEE Colloquiu m on Advances in Adaptive Beamformin , Romsey, UK, June 13, 1995. 107. F. C. Lin and F. F. Kretschmer, An le measurement in the presence of main eam interference , Proc. of IEEE 1990 Int. Radar Conf., Arlin ton (VA), U A, May 710, 1990, pp. 4444 50. 108. U. Nickel, Monopulse estimation with adaptive arrays. IEE Proc., vol. 130 , pt. F, no. 5, pp. 303308, Octo er 1993. 109. M. Valeri, . Bar arossa, A. Farin a, and L. Timmoneri, Monopulse estimation of tar et DoA in external fields with a daptive arrays, IEEE ymp. of Phased Array ystems and Technolo y, Boston (MA), U A, Octo er 1518, 1996, pp. 386390. 110. U. Nickel, Performance of corrected adapti ve monopulse estimation, IEE Proc., vol. 146, pt. F, no. 1, pp. 1724, Fe ruary 199 9. 111. J. Worms, Monopulse estimation and LC confi urations, Proc. of IEEE Radar Conf. 1998, Dallas, TX, May 1114, 1998, pp. 5661. 112. U. Nickel, Overview of ene ralized monopulse estimation, IEEE AE Ma azine, vol. 21, no. 6, part 2 of 2, pp. 2756, June 2006. 113. A. Farina, G. Golino, and L. Timmoneri, Maximum likelihood estimator approach for the estimation of tar et an ular coordinates in presence of main eam interference: Application to live data acquired with a round ased phased array radar, Proc. of IEEE 2005 Int. Radar Conf., Alexandria (VA), U A, M ay 912, 2005, pp. 6166. 114. A. Farina, G. Golino, and L. Timmoneri, Maximum likeli hood estimate of tar et an ular coordinates under main eam interference: Applic ation to recorded live data, in Advances in Direction of Arrival Estimation, . C handran (ed.), Norwood, MA: Artech House, Inc., 2006, pp. 285303. 115. J. Ro ey, D. Fuhrmann, E. Kelly, and R. Nitz er , A CFAR adaptive matched filter detector, I EEE Trans., vol. AE 28, no. 1, pp. 208216, January 1982. 116. A. Farina, G. Golino , and L. Timmoneri, Comparison etween L and TL in adaptive processin for rada r systems, Proc. of IEE, vol. 150, pt. F, no. 1, pp. 26, Fe ruary 2003. 117. A. Fa rina and L. Timmoneri, Cancellation of clutter and e.m. interference with TAP al orithm. Application to live data acquired with a round ased phased array rada r, Proc. of IEEE 2004 Radar Conf., Philadelphia (U A), April 2629, 2004, pp. 486491 . 118. A. Farina, G. Golino, . Immediata, L. Ortenzi, and L. Timmoneri, Techniqu es to desi n su arrays for radar phased array antennas, IEE Int. Conf. on Antenna s and Propa ation (ICAP) 2003, March 31 April 3, 2003, pp. 1723. 119. U. Nickel, u array confi urations for di ital eamformin with low sidelo es and adaptive i nterferences suppression, Proc. IEEE 1995 Int. Radar Conf., Alexandria (VA), U A, May 811, 1995, pp. 714719. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

 

 

Electronic Counter Countermeasures 24.64 RADAR HANDBOOK 120. U. Nickel, Monopulse estimation with su array output adaptive eam formin and low side lo e sum and difference eams, IEEE ymp. on Phased Array ystems an d Technolo y, Boston (MA), U A, Octo er 1518, 1996, pp. 283288. 121. E. Brookner a nd J. M. Howells, Adaptive Adaptive Array Processin , IEE Int. Conf. Radar87, Conf. Pu . 281, London, Octo er 1921, 1987, pp. 257263. 122. L. W. Dicken, The use of nu ll steerin in suppressin main eam interference, IEE Int. Conf. Radar77, Conf. P u . 155, London, Octo er 2528, 1977, pp. 226231. 123. W. F. Ga riel, pectral analy sis and adaptive array superresolution techniques, Proc. IEEE, vol. 68, pp. 654666 , June 1980. 124. U. Nickel, Fast su space methods for radar applications, in Adva nced i nal Processin : Al orithms, Architectures and Implementation VII, F. T. Luk (ed.), PIE Proc. eries vol. 3162 (Conf. Rec. PIE an Die o 1997), pp. 4384 48. 125. U. Nickel, Aspects of implementin superresolution methods into phased a rray radar, Int. Journal Electronics and Communications (AE), vol. 53, no. 6, pp. 315323, 1999. 126. U. Nickel, potli ht MU IC: uperresolution with su arrays wit h low cali ration effort, IEE Proc., vol. 149, pt. F, no. 4, pp. 166173, Au ust 20 02. 127. U. Nickel, uperresolution and jammer suppression with road and arrays for multi function radar, Chapter 16 in Applications of pace Time Adaptive Proce ssin , R. Klemm (ed.), London: IEE, 2004, pp. 543599. 128. H. Lee, Ei envalues and ei envectors of covariance matrices for si nal closely spaced in frequency, IEEE Trans., vol. P40, no. 10, pp. 25182535, Octo er 1992. 129. pecial Issue on upe rresolution, The Lincoln La oratory Journal, vol. 10, no. 2, pp. 83222. 130. . T . mith, tatistical resolution limits and complexified Cramer Rao ound, IEEE Tra ns., vol. P53, no. 5, pp. 15971609, May 2005. 131. D. K. Barton, Radar, vol. 6, F requency A ility and Diversity, Norwood, MA: Artech House, Inc., 1977. 132. B. B er kvist, Jammin frequency a ile radars, Def. Electron., vol. 12, pp. 75.7881.83, January 1980. 133. . trappaveccia, patial jammer suppression y means of an au tomatic frequency selection device, IEE Int. Conf. Radar87, Conf. Pu . 281, London , Octo er1921, 1987, pp. 582587. 134. C. H. Ga er, The impact of waveform andwidth upon tactical radar desi n, IEE Int. Conf. Radar 82, London, Octo er 1820, 1982, pp. 278282. 135. M. I. kolnik, G. Linde, and K. Meads, enrad: An advanced wide a nd air surveillance radar, IEEE Trans., vol. AE 37, no. 4, pp. 11631175, Octo er 20 01. 136. B. L. Lewis, F. F. Kretschmer, and W. W. helton, Aspects of Radar i n al Processin , Norwood, MA: Artech House, Inc., 1986. 137. N. Levanon and E. Moz eson, Radar i nals, New York: John Wiley & ons, Inc., 2004. 138. G. Petrocchi, . Rampazzo, and G. Rodri uez, Anticlutter and ECCM desi n criteria for a low co vera e radar, Proc. Int. Conf. Radar, Paris, France, Decem er 48, 1978, pp. 194200. 139. V. G. Hansen and A. J. Zottl, The detection performance of the ie ert and Dicke Fix CFAR detectors, IEEE Trans., vol. AE 7, pp. 706709, July 1971. 140. . L. Johnston, Radar electronic counter countermeasures a ainst chaff, Proc. Int. Conf . Radar, Paris, France, May 1984, pp. 517522. 141. M. I. kolnik, Introduction to Radar ystems, 3rd Ed., New York: McGraw Hill, 2001. 142. A. Farina and F. A. tuder, A review of CFAR detection techniques in radar systems, Microwave Journal, pp. 115128, eptem er 1986. 143. E. Conte and A. De Maio, Miti ation techniques fo r non aussian sea clutter, IEEE Journal of Oceanic En ineerin , vol. 29, no. 2, pp. 284302, April 2004. 144. E. Conte, A. De Maio, A Farina, and G. Fo lia, CFAR ehavior of adaptive detectors: an experimental analysis, IEEE Trans., vol. AE 41, no. 1, pp. 233251, January 2005. 145. M. C. Wicks, W. J. Baldy o, and R. D. Brown , Expert ystem Application to Constant False Alarm Rate (CFAR) Processor, U. . Pa tent 5, 499, 030, March 12, 1996. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

 

 

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEA URE 24.65 146. A. Farina (ed.), Optimised Radar Processors, London: Peter Pere rinus, Ltd. , 1987. 147. E. Fon , J. A. Walker, and W. G. Bath, Movin tar et indication in t he presence of radio frequency interference, Proc. IEEE 1985 Int. Radar Conf., Ar lin ton (VA), U A, May 69, 1985, pp. 292296. 148. L. B.Van Brunt, Pulse compression radar: ECM and ECCM, Def. Electron., vol. 16, pp. 170185, Octo er 1984. 149. H. K ushel, VHF/UHF. Part 1: characteristics, Electronics & Communications En ineerin Journal, vol. 14, no. 2, pp. 6172, April 2002. 150. R. J. Galejs, Volume surveilla nce radar frequency selection, Proc. of IEEE 2000 Int. Radar Conf., Alexandria (V A), U A, May 712, 2000, pp. 187192. 151. H. Kushel, VHF/UHF. Part 2: operational as pects and applications, Electronics & Communications En ineerin Journal, vol. 14 , no. 3, pp. 101111, June 2002. 152. W. N. Daw er and N. M. Harwood, Comparison of doppler clutter cancellation techniques for naval multi function radars, IEE Int . Conf. Radar 2002, Conf. Pu . No. 490, Edin ur h, UK, 15 17 Octo er 2002, pp. 4 24 428. 153. A. I. Leonov and K. J. Fomichev, Monopulse Radar, Norwood, MA: Arte ch House, Inc., 1987. 154. . L. Johnston, Trackin radar electronic counter coun termeasures a ainst inverse ain jammers, IEE Int. Conf. Radar82, Conf. Pu . 216, London, Octo er 1982, pp. 444447. 155. B. L. Lewis and D. H. Howard, ecurity Devi ce, U. . Patent, 4, 006, 478, Fe ruary 1, 1977, filed Au ust 15, 1958. 156. R. L. Fante and J. A. Torres, Cancellation of diffuse jammer multipath y an air orne adaptive radar, IEEE Trans., vol. AE 31, no. 2, pp. 805820, April 1995. 157. . Ko on, Al orithms for miti atin terrain scattered interference, Electronics & Commun ications En ineerin Journal, vol. 11, no. 1, pp. 4956, Fe ruary 1999. 158. . Bj orklund and A. Nelander, Theoretical aspects on a method for terrain scattered in terference miti ation in radar, Proc. of IEEE 2000 Int. Radar Conf., Alexandria ( VA), U A, May 912, 2005, pp. 663668. 159. Y. A ramovich, . J. Anderson, and A. Y. Gorokov, tochastically constrained spatial and spatio temporal adaptive process in for non stationary hot clutter cancellation, Chapter 17 in Applications of p ace Time Adaptive Processin , R. Klemm (ed.), London: IEE Radar, onar and Navi ation, eries 14, 2004, pp. 603697. 160. W. D. Blair, G. A. Watson, T. Kiru araja n, and Y. Bar halom, Benchmark for radar allocation and trackin in ECM, IEEE Tra ns., vol. AE 34, no.4, pp.10971114, 1998. 161. T. Kiru arajan, Y. Bar halom, W. D . Blair, and G. A. Watson, IMMPDAF for radar mana ement and trackin enchmark wi th ECM, IEEE Trans., vol. AE 34, no.4, pp.11151134, 1998. 162. H. Blom and Y. Bar halom, The interactin multiple model al orithm for systems with Markovian switch in coefficients, IEEE Trans., vol. AC33, no. 8, pp. 780783, Au ust 1988. 163. A. M oreira, Improved multilook techniques applied to AR and CAN AR ima ery, IEEE Tra ns. on Geoscience and Remote ensin , vol. 29, no. 4, pp. 529534, July 1991. 164. W. Goj, ynthetic Aperture Radar and Electronic Warfare, Dedham, MA: Artech Hou se, Inc., 1989. 165. C. J. Condley, The potential vulnera ility to increased ack round noise of synthetic aperture radar in the maritime environment, IEE Colloqu ium on ynthetic Aperture Radar, Novem er 29, 1989, pp. 10/110/5. 166. C. J. Cond ley, ome system considerations for electronic countermeasures to synthetic apert ure radar, IEE Colloquium on Electronic Warfare ystems, January 14, 1991, pp. 8/ 18/7. 167. M. oumekh, AR ECCM usin phased pertur ed LFM chirp si nals and DRFM repeat jammer penalizer, IEEE Trans., vol. AE 42, no. 1, pp. 191205, January 2006. 168. P. Hy er , Assessment of modern coherent jammin methods a ainst synthetic a perture radar ( AR), Proc. of EU AR 98, European Conf. on ynthetic Aperture Radar , Friedrichshafen, Germany, May 2527, 1998, pp. 391394. 169. C. Boesswetter, ECCM e ffectiveness of a low sidelo e antenna for AR round mappin , AGARD AVP ymp. Mul tifunction Radar for Air orne Applications, Toulouse, 1985. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

 

 

 

 

Electronic Counter Countermeasures 24.66 RADAR HANDBOOK 170. A. Farina and P. Lom ardo, AR ECCM usin adaptive antennas, Proc. of IEEE Lo n Island ection, Adaptive Antenna ystems ymp., Lon Island, U A, Novem er 19 94, pp. 7984. 171. J. H. Ender, Anti jammin adaptive filterin for AR ima in , Pr oc. of IR 98, Int. Radar ymp., Munich, Germany, eptem er 1517, 1998, pp. 1403141 3. 172. J. A. Torres, R. M. Davis, J. D. R. Kramer, and R. L. Fante, Efficient wi de and jammer nullin when usin stretch processin , IEEE Trans., vol. AE 36, no. 4, pp. 11671178, Octo er 2000. 173. L. Rosen er and D. Gray, Anti jammin techniq ues for multi channel AR ima in , IEE Proc., pt. F, vol. 133, no. 3, pp. 234242, June 2006. 174. P. E. Pace, D. J. Fouts, . Ekestrom, and C. Karow, Di ital false tar et ima e synthesizer for counterin I AR, IEE Proc., pt. F, vol. 149, no. 5, pp. 248257, Octo er 2002. 175. P. E. Pace, D. J. Fouts, and D. P. Zulaica, Di ita l ima e synthesizer: Are enemy sensors really seein whats there?, IEEE Aerospace and Electronic ystems Ma azine, vol. 24, no. 2, pp. 37, Fe ruary 2006. 176. L. ev i, A. Ponsford, and H. C. Chan, An inte rated maritime surveillance system as ed on hi h frequency surface wave radars, part 1: Theoretical ack round and num erical simulations, IEEE Antennas and Propa ation Ma azine, vol. 43, no. 5, pp. 2 843, Octo er 2001. 177. A. Ponsford, L. ev i, and H. C. Chan, An inte rated marit ime surveillance system ased on hi h frequency surface wave radars, part 2: Ope rational status and system performance, IEEE Antennas and Propa ation Ma azine, v ol. 43, no. 5, pp. 5263, Octo er 2001. 178. G. A. Fa rizio, pace time characteriz ation and adaptive processin of ionospherically propa ated HF si nals, Ph.D. dis sertation, Adelaide University, Australia, July 2000. 179. G. A. Fa rizio, D. A. Gray, and M. D. Turley, Experimental evaluation of adaptive eamformin methods and interference models for hi h frequency over the horizon radar, Multidimension al ystems and i nal Processin pecial Issue on Radar i nal Processin Techni ques, vol.14, no. 1/2/3, pp. 241263, JanuaryJuly 2003. 180. G. A. Fa rizio, Y. I. A ramovich, . J. Anderson, D. A. Gray, and M. D. Turley, Adaptive cancellation o f nonstationary interference in HF antenna arrays, IEE Proc., vol. 145, pt. F, no . 1, pp. 1924, Fe ruary 1998. 181. Y. I. A ramovich, A. Y. Gorokhov, V. N. Mikhay lyukov, and I. P. Malyavin, Exterior noise adaptive rejection for OTH radar imple mentations, IEEE Int. Conf. on Acoustics, peech, and i nal Processin 1994, ICA P94, Adelaide (Australia), 1994, pp. 105107. 182. . J. Anderson, Y. I. A ramovi ch, and G. A. Fa rizio, tochastic constraints in non stationary hot clutter canc ellation, IEEE Int. Conf. on Acoustics, peech, and i nal Processin 1997, ICA P97, Munich, Germany, vol. 5, pp. 2124, April 1997, vol. 5, pp. 3753 3756. 183. Y. I. A ramovich, N. pencer, and . J. Anderson, tochastic constraints method in non stationary hot clutter cancellationpart 1: Fundamentals and supervised traini n applications, IEEE Trans., AE 34, no. 4, pp. 12711292, 1998. 184. Y. I. A ramovi ch, N. pencer, and . J. Anderson, tochastic constraints method in non stationa ry hot clutter cancellationpart 2: Unsupervised trainin applications, IEEE Trans. , vol. AE 36, no. 1, pp. 132150, 2000. 185. G. A. Fa rizio, A. B. Gershman, and M. D. Turley, Ro ust adaptive eamformin for HF surface wave over the horizon, IEEE Trans., vol. AE 40, no. 2, pp. 510525, April 2004. 186. G. A. Fa rizio, A. Farina , and M. D. Turley, patial adaptive su space detection in OTH radar, IEEE Trans., vol. AE 39, no. 4, pp. 14071428, Octo er 2003. 187. A. Farina, G. A. Fa rizio, W. L. Melvin, and L. Timmoneri, Multichannel array processin in radar: tate of th e art, hot topics and way ahead, Proc. ensor Array and Multichannel i nal Proce ssin IEEE Workshop (invited paper), it es, pain, July 1821, 2004, pp. 1119. 188 . G. A. Fa rizio, G. J. Frazer, and M. D. Turley, TAP for Clutter and Interferen ce Cancellation in a HF Radar ystem, IEEE Int. Conf. on Acoustics, peech, and i nal Processin 2006, ICA P 2006, Toulouse, France, May 2006. 189. D. K. Barto n, Radar ystem Analysis and Modelin , Norwood, MA: Artech House, Inc., 2005. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

 

 

 

 

 

  

  

Electronic Counter Countermeasures ELECTRONIC COUNTER COUNTERMEA URE 24.67 190. . L. Johnston, The ECCM improvement factor (EIF): illustration examples, ap plications, and considerations in its utilization in radar ECCM performance asse ssment, Int. Conf. Radar, Nanjin (China), Novem er 47, 1986, pp. 149154. 191. J. C larke and A. R. u ramanian, A ame theory approach to radar ECCM evaluation, Proc . of IEEE 1985 Int. Radar Conf., Arlin ton (VA), U A, May 69, 1985, pp. 197203. 19 2. L. Nen jin , Formulas for measurin radar ECCM capa ility, IEE Proc., vol. 131, pt. F, pp. 417423, July 1984. 193. L. Nen jin , ECCM efficacy assessment in surve illance radar analysis and simulation, IR 98, Int. Radar ymp., Munich, Germany, eptem er 1517, 1998, pp. 14151419. 194. D. H. Cook, ECM/ECCM systems simulation pr o ram, electronic and aerospace systems record, IEEE Conv. Rec. EA CON 68, eptem er 911, 1968, pp. 181186. 195. . Watts, H. D. Griffiths, J. R. Hollaway, A. M. Ki n horn, D. G. Money, D. J. Price, A. M. Whitehead, A. R. Moore, M. A. Wood, and D. J. Bannister, The specification and measurement of radar performance, IEE Int. Conf. Radar 2002, Conf. Pu . no. 490, Edin ur h, UK, Octo er 1517, 2002, pp. 54254 6. 196. F. A. tuder, M. Toma, and F. Vinelli, Modern software tools for radar pe rformance assessment, Proc. of IR 98, Int. Radar ymp., Munich, Germany, eptem e r 1517, 1998, pp. 10791090. 197. A. G. Huizin and A. Theil, CARPET 2.11 oftware + User Manual, The Ha ue, The Netherlands: TNO Defense, ecurity and afety, 200 4. Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.

 

 

 

 

 

 

Electronic Counter Countermeasures Downloaded from Di ital En ineerin Li rary @ McGraw Hill (www.di italen ineerin li rary.com) Copyri ht 2008 The McGraw Hill Companies. All ri hts reserved. Any use is su ject to the Terms of Use as iven at the we site.