Camellia Institute Of Technology

Department Of Electronics And Communications Engineering

Name Of The Topic : - EQUALIZER

University Roll Number : - 23000322020

Registration Number : - 222300120230

Semester : - 5th Sem

Subject : - Digital communication & Stochastic Process

Exam : - Continuous Assessment 2 (CA2).

Of
 Equalization

ABSTRACT the optimal demodulator is the one that employs the matched
The purpose of this work is the design and implementation of filter, i.e., we can pass the received signal through the
an signal in Labview. This work deals with the study of the matched filter v(t) = v(−t) and then sample the
various kinds of interferences in a communication channel like matched filter output at time t=0 to obtain the decision
Inter Symbol Interference (ISI), and Additive White Gaussian statistic. When a sequence of symbols is transmitted, the
noise(AWGN) and counteracts their effects atthe receiver. matched filter is used to perform demodulation. One way is to
sample the matched filter output at time t = mT to obtain the
Keywords decision statistic for the symbol b. At t = mT, the output of the
ISI, AWGN, Equalizer, LMS, EVM, DSSS matched filter is

1. INTRODUCTION 𝑁
Equalization is a process in which active or passive electronic (𝑚) = ∑ b * v * v(mT − nT) + n(t)
elements are used for fulfilling the purpose of altering the 𝑛=0
frequency response characteristics for any system. It is the
process of adjusting the balance between frequency Where n(t) is a zero-mean Gaussian random variable. The first
components within an electronic signal. Equalization term is the desired signal contribution due to the symbol b and
compensates for the differences in signal attenuation and the second term contains contributions from the other
delay associated with different frequency components. There symbols. These extra symbols are called intersymbol
are various methods of signal equalization[1].Equalized signal interference (ISI). The causes of ISI are multipath propagation
can be analysed at the output using EVM [2].The received and band limited channels.
signal in this case is IEEE standardized by IEEE 802.11b
standards [3].The type of modulation used is CCK [4]. The 2.1 Multipath Propagation
design involves finding the channel response H(f) of the One of the causes of inter-symbol interference is what is
medium through which signal travelled. The received signal is known as multipath propagation in which a wireless signal
afterwards passed through an adaptive filter or an inverse from a transmitter reaches the receiver via many different
filter (1/H(f)) which nullifies the effect of the channel. H(f) paths. The causes of this include reflection , refraction (such
can be estimated at the receiver by using methods like training as through the foliage of a tree) and atmospheric effects such
sequence, pilot carrier etc. The main problems for inaccurate as atmospheric ducting and ionosphere reflection. Since all of
measurement of the signal by the equalizer are Inter symbol these paths are different lengths - plus some of these effects
Interference (ISI) [5]-[11] and Adaptive white Gaussian noise will also slow the signal down - this results in the different
[12] (AWGN). versions of the signal arriving at different times. This delay
results in the spreading of a symbol to all the other
2. INTER SYMBOL INTERFERENCE subsequent symbols, interfering with the detection of these
(ISI) symbols .Multipath propagation often distorts the amplitude
In telecommunication, inter-symbol interference (ISI) is a or phase of the signal causing interference with the received
form of distortion of a signal in which one symbol interferes signal.
with another symbols. This is an unwanted phenomenon as
the previous symbols have similar effect as noise, making the 2.2 Band Limited Channel
process less efficient.Mathematically the ISI can be Another cause of intersymbol interference is the transmission
represented by, of a signal through a band limited channel, i.e., one where the
Let us consider the transmission of a sequence of symbols frequency response is zero above a certain frequency (the
with the basic Waveform (𝑡). To send the nth symbol b, we cutoff frequency). Passing the signal through a band limited
send b * u(t − nT),where T is the interval and transmitted channel results in the removal of frequency .The amplitude of
signal given by the frequency components below the cutoff frequency may
also be attenuated .
(𝑛) = ∑𝑁 (b * u(t − nT) The ISI is caused at the transmitter due to the presence of
𝑛=0
Based on the dispersive channel model, the received signal pulse shaping filters at the transmitter end. Three types of
can be represented by, pulse shaping filters have been used
Root cosine, Root raised cosine and Gaussian.
𝑁
(𝑛) = ∑(b * u(t − nT) + 𝑛(𝑛)
𝑛=0
Where r(t) = x(t) * h(t) is the received waveform for a
symbol. If a single symbol, say the symbol 𝑏0, is transmitted,

Of
3. DESIGNING 3.2 Equalizer Design
3.1 Channel Estimation Two types of equalizers are designed for the purpose of
The small-scale variations of a mobile radio signal can be equalization
directly related to the impulse response of the mobile radio 1. Adaptive equalizer
channel. The impulse response can be considered as a
wideband channel characterization and can be used to 2. Linear equalizer
represent all information necessary to simulate or analyze any
type of radio transmission through the channel. The filtering 3.3 Adaptive Equalizer
nature of the channel is caused by the summation of An adaptive equalizer is an equalization filter that
amplitudes and delays of the multiple arriving waves. The automatically adapts to time-varying properties of the
impulse response is used to predict and compare the communication channel. It is implemented for performing tap-
performance of many different mobile communication weight adjustments periodically or continually. In general, for
systems and transmission bandwidths for a particular mobile the case of first call or reset adaptive equalizer uses a
channel. The channel impulse response relates the transmitted training sequence but for this case as we have the channel
and received signals as estimate in Fig 1. so we will be using these co-efficient as
the training sequence. Periodic adjustments are accomplished
(𝑛) = X(𝑛) * ℎ(𝑛) by periodically transmitting a preamble or short training
From, the above mentioned equation the channel impulse sequence of digital data known by the receiver. Continual
response can be calculated as adjustment are accomplished by replacing the known training
ℎ(𝑛) = 𝑦(𝑛) * 𝑥(−𝑛)/𝑝(𝑛) sequence with a sequence of data symbols estimated from the
Where y(n) is the received signal, x(-n) is the transmitted equalizer output and treated as known symbols. When the
signal and p(n) is a constant. adaptive filter functions continuously and automatically for a
period of time it is called decision directed. Adaptive filters
For estimating the channel impulse response,
depend on recursive algorithms to update their coefficients
A computationally efficient method is used [5],in accordance
and train them to near the optimum solution. In this case The
with the IEEE 802.11b WLAN standards that utilizes DSSS.
LMS algorithm is used for continuously updating the filter
This method comprises of the auto-correlation and cross- coefficients as done in [14]-[17]. The equation is given by
correlation of the 11 chip barker sequence with the received (𝑛 + 1) = 𝐶(𝑛) + 𝑈 * 𝑒(𝑛) * 𝑥(𝑛)
signal.In WLAN the type of modulation used is CCK
modulation [4]. In CCK modulation ,the sampling rate is First the received signals is compensated for its gain. This is
fixed at M times the bandwidth of baseband signal..The done by comparing its power level with the ideal case and
channel estimation is done by using the technique of training compensating for the power loss. As CCK modulation have
bits. In this technique first the received signal is demodulated been used, so the symbol rate is 8 samples per symbol. The
using the barker sequence .The demodulated signal is then symbol rate is changed to sample rate by passing the signal
used to find the strongest signal. The strongest signal is through a multi-rate filter. The oversampling is removed
spreaded using the barker sequence and samples is then before passing the signal through the equalizer. The result of
correlated with the received samples. The correlated output is the adaptive filter in terms of its co-efficient is shown below .
divided by a constant which is, the autocorrelation of the
barker sequence. A separate unit for obtaining the symbol RESULTS
information may be used and the symbol information can be
fed to the channel estimator. But in this implementation, we
have merged the unit for obtaining the symbol information
into the channel estimator. This would provide us the
flexibility of performing channel estimation anywhere in the
preamble without bothering much about synchronizing the
unit for obtaining the symbol information and the channel
estimator.

RESULTS

FIG 2. INPUT COEFFICIENTS

FIG 1.Channel Estimate

Of
 stream of I-Q points which can be used as a reasonably
reliable estimate for the ideal transmitted signal in EVM
calculation.
An error vector is a vector in the I-Q plane between the ideal
constellation point and the point received by the receiver. In
other words, it is the difference between actual received
symbols and ideal symbols.
The average power of the error vector, normalized to signal
power, is the EVM. For the percentage format, root mean
square (RMS) average is used. The EVM can be measured
mathematically as
𝑁

𝐸𝑉𝑀 = ∑ X(𝑛) − X^(𝑛)

𝑛=0

The error vector magnitude is equal to the ratio of the power

of the error vector to the root mean square (RMS) power of
FIG 3.OUTPUT COEFFICIENTS the reference. It is defined in dB as:
3.4 Linear Equalizer EVM(dB)=10log10(Perror/Preference)
A linear equalizer works on the concept that, if the transfer
function of the equalizer is made the inverse of the channel, Where Perror is the RMS power of the error vector. For single
then Inter Symbol Interference (ISI) can be completely carrier modulations, the power of the outermost (highest
eliminated at the receiver. Linear Equalizers are easy to power) point in the reference signal constellation.
implement and are highly eff ective where ithe ISI is not
severe (like the wire line telephone channel). The major issue EVM is defined as a percentage as :
in designing is the problem of stability. The equalizer co-
efficient are not stable as the region of convergence(ROC) is EVM(%)=(√Perror/Preference)*100%
beyond the unit circle as some of the poles are beyond the unit
circle.. This process of filter stability check be done in number EVM, as conventionally defined for single carrier
of ways [18]-[22] . To stabilize the filter the whole process modulations, is a ratio of a mean power to a peak power.
of taking Z-transform and Inverse Z-transform is taken on an Because the relationship between the peak and mean signal
unit circle. So, that the resultant poles are inside the unit circle power is dependent on constellation geometry, different
and the filter is both causal and stable. constellation types (e.g. 16-QAM and 64-QAM), the same
mean level of interference, will give different EVM values.
RESULTS
The equalized signals are represented by the constellation
graph by:
 Adaptive equalizer

AT SNR=30dB

4. RESULTS AND ANALYSIS

The output results can be measured by the help of the error
vector magnitude or EVM (sometimes also called receive
constellation error or RCE).It is a measure of the performance
of a digital transmitter or receiver. The transmitted signal is
represented by constellation points. Various imperfections in
the implementation like carrier leakage, phase noise cause
the actual constellation points to deviate from the ideal
locations. EVM can be defined as a measure of how far the
points are from the ideal locations [23]-[25].
Transmitter EVM can be measured by specialized equipment,
which demodulates the received signal in a similar way to
how a real radio demodulator does it. One of the stages in a
typical phase-shift keying demodulation process produces a

Of
 AT SNR=35dB while linear equalizer is designed on the concept of inverse
filter .The output is analysed with EVM and results are
matched with the industry standards. There are a number of
areas where future developments can be made .The channel
estimation can be improved by taking more number of
averages. The convergence factor of the adaptive filter can be
improved by continuously changing the convergence factor
with respect to the SNR of the signal.

ACKNOWLEDGMENT
This paper is published by permission of National instrument,
India .The opinions expressed are those of the author and do
not necessarily represent those of NI.

REFERENCE
[1] Alan v.oppenheim, ronal w.schafer “Discrete-time signal
processing”. Pearson education
[2] TAN xiao-heng , LI Teng-jiao “EVM simulation and
analysis in digital transmitter”. College of
 Linear Equalizer communication Engineering, Chongqing University,
AT SNR=35 dB Chongqing 400044,china
[3] “IEEE Standard for Wireless LAN Medium Access
Control and Physical Layer Specifications,”IEEE
802.11b,Nov, 1999.
[4] C.Andren and M.Webster,”CCK Modulation Delivers
11Mbps for High Rate IEEE 802.11
Extensions,”Wireless Symposium/Portable Design
Conference,Spring,1999
[5] Intersymbol interference and probability of error in
digital systems
[6] S. Benedetto, E. Biglieri, and R. Daffara, "Performance
of multilevel baseband digital systems in a nonlinear
environment", IEEE Trans. Commun., vol. COM-24,
pp.1166 -1175 1976
[7] R. W. Lucky "A functional analysis relating delay
variation and intersymbol interference in data
transmission", Bell Sys. Tech. J., vol. 42, pp.2427 -
2483 1963
AT SNR=30 dB [8] K. Metzger, "On the probability density of intersymbol
interference", IEEE Trans. Commun., vol. COM-35,
pp.396 -402 1987
[9] J. M. Aein and J. C. Hancock "Reducing the effects of
intersymbol interference with correlation receivers",
IEEE Trans. Information Theory, vol. IT-9, pp.167 -175
1963
[10] M. R. Aaron and D. W. Tufts "Intersymbol interference
and error probability", IEEE Trans. Information Theory,
vol. IT-12, pp.26 -34 1966
[11] R. W. Lucky "A functional analysis relating delay
variation and intersymbol interference in data
transmission", Bell Sys. Tech. J., vol. 42, pp.2427 -
2483 1963
[12] “Performance loss of dirty-paper codes in additive white
Gaussian noise and jitter channels “Licks, V. ; Dept. of
Electr. & Comput. Eng
5. CONCLUSION
In this paper, we have shown how it is possible to implement [13] “A computationally efficient method for estimating the
the equalization of a IEEE 802.11b signal in Labview. A channel impulse response for the IEEE 802.11b
channel estimation code is designed first and on the basis of (WLAN)” by kedia ,A dept ofElect.& comput.ENG. ,
that Adaptive and linear equalizer are designed. The Adaptive British Columbia Univ., Vancouver, BC,Canada.
equalizer is designed on the basis of a recursive algorithm,

Of
THANK YOU