CHAPTER 3

SYSTEM DESIGN

3.1 SPECTRUM SENSING

Spectrum Sensing is defined by “as the task of finding spectrum holes by sensing the
radio spectrum in the local neighborhood of the cognitive radio receiver in an
unsupervised manner”. A number of schemes have been developed for detecting whether
the primary user (PU) is present in a particular frequency band. The various spectrum
sensing techniques were proposed to identify the presence of primary user signal and if
the primary user is absent then the spectrum will be allocated to the secondary user.
Spectrum sensing is the ability to measure, sense and be aware of the parameters related
to the radio channel characteristics, availability of spectrum and transmit power,
interference and noise, radio’s operating environment, user requirements and
applications, available networks (infrastructures) and nodes, local policies and other
operating restrictions. It is done across Frequency, Time, Geographical Space, Code and
Phase. Spectrum sensing [5] is the most important task among others for the
establishment of CRs because they need to sense the spectrum band for a spectrum hole
[5-9], decide to use the spectrum band or not. A number of different techniques have been
proposed for identifying the presence of the PU signal transmission. The existing
spectrum sensing techniques can be broadly divided into three categories [6]: energy
detection, matched filter detection, and cyclostationary detection. Matched filter [6][7],
energy detection [11][12] and cyclostationary detection [8] are widely used techniques as
detection techniques. Since interference actually takes place at the receiver location,
active awareness should be focused on detecting the receiving activities of PUs but,
without a cooperation between primary and secondary users, SUs hardly detect PU’s
communication. Due to the dynamism of the spectrum utilization, it is not possible to
adopt solely the approaches provided by passive awareness and, in general, those based
on active awareness and spectrum sensing are preferred. As shown in Figure 10, there are
several aspects (i.e. algorithms, challenges, etc.) to be taken into account for the spectrum
sensing operation deployed on a CR device. In the following sections, a description of the
main approaches and challenges of the Spectrum Sensing operation will be discussed
[67].
Spectrum Sensing Model
Spectrum sensing is one of the most important processes performed by cognitive radio
systems. It allows the SUs to learn about the radio environment by detecting the presence
of the PU signals using one or multiple techniques and decide to transmit or not in its
frequency band [1-5]. The spectrum sensing model can be formulated as:

…………….. (3.1)

where n=1….N, N is the number of samples, y(n) is the SU received signal, s(n) is the PU
signal,
w(n) is the additive white Gaussian noise (AWGN) with zero mean and variance 𝛿<= ,
and h is the
complex channel gain of the sensing channel. H0 and H1 denote respectively the absence
and the presence of the PU signal. The PU signal detection is performed using one of the
spectrum sensing techniques to decide between the two hypotheses H0 and H1. The
detector output, also called the test statistic, is then compared to a threshold in order to
make the sensing decision about the PU signal presence. The sensing decision is
performed as:

where T denotes the test statistic of the detector and 𝛾 denotes the sensing threshold. If
the PU signal is absent, SU can access to the PU channel. Otherwise, it cannot access to
that channel at that time.
 Y(n)
Average sum of Test statistic
Xp(n)
Convolution N

Threshold Threshold
comparison

Sensing
decision

Figure 4: General model of spectrum sensing [26]

3.1 SPECTRUM SENSING METHODS

A number of different methods are proposed for identifying the presence of signal
transmission

all of which are in early development stage. They are:

 Energy – Detection Based

 Cyclostationary Based
 Matched filtering Based
1 ENERGY DETECTION
Energy detector is also known as radiometry and it is most common method of spectrum
sensing because of its low computational and implementation complexities. Moreover,
the cognitive user’s receivers do not need any prior knowledge of the primary user’s
signal. The signal is detected by comparing the output of energy detector with threshold
which depends on noise floor. The most advantages of using energy detection, low
computational cost, easy implementation, less complexity and does not need any prior
knowledge of primary user. In contrast, in this technique the signal detection depends on
comparing the power of the received signal to the threshold level, whereas threshold level
rely on the noise floor which can be estimated but the signal power is difficult to estimate
as it changes relying on two factors, distance between primary user and cognitive radio,
another factor is ongoing transmission characteristics. Energy detection is popular till
now, but the major drawback with energy detection method is that the poor performance
under low SNR conditions and also no proper distinction between primary users and
noise. In signal processing, for a given signal s(t), the autocorrelation function is defined
as [9]The sensing decision is based on the knowledge of the statistical distribution of the
autocorrelation function. For random noise, the first lag of the autocorrelation is very
small or negative, but when there is a signal, the autocorrelation at the first lag represents
a significant value. Thus the sensing method consists of comparing lag0 and lag1 of the
autocorrelation if the received signal and sensing decision is made as following:

If lag0 >> lag1 PU transmission is absent

If lag0 ≈ lag1 PU transmission is present

Correlation threshold is the margin between the two lag values, for example, lag0 is
superior to
(λ %) of lag1, where λ is a percentage value that represents a decision margin [3].

2. CYCLOSTATIONARY BASED
Cyclostationary feature detection in a cognitive radio network refers to the process of
identifying and exploiting cyclostationary properties present in the received signals to
perform tasks such as spectrum sensing, signal classification, interference detection, and
modulation recognition. Cyclostationarity refers to the periodic or quasi-periodic
statistical properties of a signal, which can arise from modulation schemes, time-varying
channels, or other factors. In a cognitive radio network, where secondary users
opportunistically access the spectrum, detecting cyclostationary features can provide
valuable information for efficient spectrum utilization and interference mitigation. By
exploiting the cyclostationary properties of primary user signals and the surrounding
environment, cognitive radio devices can make informed decisions about channel
occupancy, interference levels, and optimal channel selection.

Figure Block diagram of Cyclostationary feature detection

MATCHED FILTER BASED

In cognitive radio networks, spectrum sensing aims to detect the unused spectrum
channels in order to use the radio spectrum more efficiently. Various methods have been
proposed in the past, such as energy, feature detection, and matched filter. These methods
are characterized by a sensing threshold, which plays an important role in the sensing
performance. Most of the existing techniques used a static threshold. However, the noise
is random, and, thus the threshold should be dynamic. In this paper, we suggest an
approach with an estimated and dynamic sensing threshold to increase the efficiency of
the sensing detection. The matched filter method with dynamic threshold is simulated and
its results are compared to those of other existing techniques.
One of the important functions in a cognitive radio is spectrum sensing, which allows an
SU to detect the presence or absence of PU transmission in licensed frequency bands, and
allow the SU to borrow unused spectrum from the PU. A number of sensing techniques
have been proposed over the last decade [4]. These techniques are classified into two
categories: cooperative sensing and non-cooperative sensing [5, 6]. The non-cooperative
sensing category is sub-divided into three subcategories: energy, feature, and matched
filter based sensing. Energy detection is the simplest technique which does not require
any information about the PU signal to operate. It works by comparing the received
signal energy with an estimated threshold which depends on the noise power [7].
Correlation based sensing is a method based on the value of the autocorrelation
coefficient of the received signal and exploits the autocorrelation features that exist in the
transmitted signal and that are not present in the noise [8, 5]. Matched filter based sensing
is a coherent pilot sensor which detects known PU signals. It maximizes the signal to
noise ratio (SNR) at the output of the detector, but it requires a prior knowledge about the
PU signal. Thus, it is assumed that the PU transmitter sends a pilot stream simultaneously
with the data and the SU receiver has a perfect knowledge of the pilot stream to verify its
transmission at the frequency band [1].

Sensing threshold is an important parameter in spectrum sensing. When a detector does

not properly adjust its threshold, the sensing performance is degraded. A number of
approaches, based on energy detection, were proposed [11-16]. As the sensing
performance is highly affected by the estimation error of noise power, a dynamic
estimation style of noise power is recommended in [11]. Adaptive threshold control is
implemented in [12] with linear adaption of the threshold based on signal to interference
plus noise ratio (SINR). This approach attains a considerably higher SU throughput than
the fixed threshold approach, but maintains unacceptable false alarms. Adaptive threshold
in unknown white Gaussian noise is presented in [13] with noise power estimation,
keeping the false alarm rate at a preferred point under any noise level. It is based on a
concept of a dedicated noise estimation channel in which only noise is received by SU.
An improved energy detection method is proposed in [14], where misdetection of PU
transmission due to sudden drop in PU transmission power is addressed by keeping an
additional updated list of the latest fixed number of sensing events that are used to
calculate an average test statistic value. A double-threshold technique is proposed in [15]
with the aim of finding and localizing narrowband signals. Another technique is
presented in [16] based on wideband spectrum sensing, which senses the signal strength
levels within several frequency ranges to improve the opportunistic throughput of SU and
decrease the interference to the PU.

In [7], each pair of (Pd, Pf) was associated with a particular threshold to make sensing
decision. In [17], the sensing threshold is determined dynamically by multiplying the
theoretical threshold by a positive factor. Several papers do not mention how the
threshold was selected. However, with a static threshold, the sensing decision is not
reliable because of the uncertainty of the noise. In this work, we suggest an estimated and
dynamic sensing threshold for energy detection, matched filter detection and correlation
based detection to increase the probability of detection and the decision reliability.

To evaluate the performance of spectrum sensing techniques, several metrics were used,
including the probability of detection, Pd, and the probability of false alarm, Pf. Pd is the
probability that the SU declares the presence of a PU signal. The higher the Pd, the better
the PU protection is. Pf is the probability that SU declares the presence of PU when the
spectrum is idle. The lower the Pf, the more the sensing method is effective in detecting
the signal [4].
Figure Simulation model for matched filter detection

MATHEMATICAL EXPRESSION FOR MATCHED FILTER

WORKFLOW OF MATCHED FILTER DETECTION

The workflow of matched filter detection in a cognitive radio network involves several
steps, from signal acquisition to decision making. Here is a detailed explanation of the
workflow:

Signal Acquisition: The cognitive radio device acquires the received waveform, which
includes the desired signal along with noise and potential interference from other sources.
This waveform is typically obtained through an antenna or other receiving devices.

Preprocessing: The received waveform undergoes preprocessing to enhance the desired

signal and mitigate noise or interference. This stage may include operations such as
filtering, amplification, and analog-to-digital conversion. The purpose is to improve the
quality of the received signal before further processing.

Matched Filter Design: The matched filter is designed based on the characteristics of the
known signal that the cognitive radio device aims to detect. The matched filter is
typically a replica of the known signal, but it is time-reversed and time-scaled. Time-
reversal compensates for the reversal caused by the transmission and reception process,
while time-scaling aligns the symbol rate of the known signal with the received signal.

Matched Filter Operation: The received waveform is correlated with the matched filter
using convolution or cross-correlation. This operation measures the similarity between
the received signal and the known signal at different time offsets. The output of the
correlation operation is a sequence of correlation values, representing the level of
resemblance between the received signal and the known signal at each time offset.

Detection Threshold Determination: A detection threshold is determined based on the

noise characteristics and the desired detection performance. The threshold is typically set
such that the probability of false alarm (detecting the signal when it is not present) is
below a certain threshold. The threshold can be determined through mathematical
analysis, simulation, or empirical measurements.

Detection Decision: The correlated output from the matched filter operation is compared
to the detection threshold. If the correlation value exceeds the threshold, it is considered a
detection of the known signal. If the correlation value is below the threshold, it is
considered a non-detection. The decision is made based on a binary hypothesis test,
where the null hypothesis is no signal presence, and the alternative hypothesis is signal
presence.

Decision Making: Based on the detection decision, the cognitive radio device can make
decisions related to spectrum access, channel selection, or interference mitigation. For
example, if the known signal is a primary user signal, and it is detected, the cognitive
radio device may decide to vacate the channel or adjust its transmission parameters to
avoid interference. On the other hand, if the known signal is not detected, the cognitive
radio device may continue to sense the channel or explore other available channels.
 Figure: Process Flow Diagram for Matched Filter Detection.

CHANNEL ALLOCATION

In the last decade, studies in the area of cognitive radio (CR) has shown good results
regarding the use of the RF spectrum opportunistically. The RF spectrum opportunities
may vary from small frequency ranges to whole bands with low utilization. In some cases
this resource may be available during a specific time of the day, when the license holder
or primary user (PU) is not transmitting any information. The dynamic spectrum access
(DSA) is an important part of CR studies. Its main role is to grant secondary users (SUs)
access to spectrum holes or white spaces [1], in such a way that SUs can transmit
information in a licensed band without causing any interference with primary PUs.

Envisioning the next generation of wireless networks and the rational use of RF
spectrum, the IEEE 802.22 Wireless Regional Area Networks (WRAN) standard was
defined in 2011 [2]. This standard allows the use of VHF and UHF bands by SUs in
sparsely populated areas, which are allocated in the licensed bands through the CR
technology. Among the definitions of the IEEE 802.22, the Spectrum Manager (SM) and
the Database Service (DS) have great importance for the channel allocation context. The
SM is responsible for the user and channel allocation under the coverage area of the CR
base station (BS). The DS contains the list of available channels for CR use, which can be
provided by the operator or by the regulation authority. The SM accesses the DS
periodically in order to update the channel list, but the BS must be able to operate
independently in case of DS unavailability. The IEEE 802.22 specifies that the BS must
be able to obtain the available channels from spectrum observation, when the DS is not
available.

The use of spectrum data prediction is not detailed in IEEE 802.22, but it is a commonly
used resource in the channel allocation problem. In prediction rules are applied to
estimate channel usage for different types of traffic. This method senses the channels of
PUs in order to identify the traffic, and then apply a specific prediction model that
represents the channel occupation. Since the measured signal is not known, a longer
period of sensing must be considered. In works presented in the channel database
classification depends essentially from spectrum observation in the BS initialization.
Although in the prediction data is considered in the database creation, both approaches do
not consider the use of allocation result information to reclassify the channels. This work
presents a novel approach to channel allocation in CR networks, proposing the use of
spectrum holes in licensed bands Terrestrial Trunked Radio (TETRA), Global System
Mobile for communications (GSM), and Digital Cellular System (DCS). Although IEEE
802.22 provides for the use of WRANs in areas of low population, this paper proposes
the use of bands typically provided to urban areas, but with substantial availability of
spectrum holes [7]. The proposed method is based on a channel database, which is built
from prediction data. The channels are sorted in such a way that those with the longest
idle time are preferred to SU allocation. This ensures the predictive characteristic of the
system in that it uses the spectrum occupancy knowledge to rank the channels. The SM
uses the best-classified channels to SUs allocation and the channel sensing results are
used to calculate the channel success rate. The main improvement of the proposed
method is that, once the database is built, it is dynamically updated by channel sensing
results. Hence, the result of each channel sensing operation is used to feed back the
database. The channels with higher allocation success rate are kept on the top of the DB
channel list. Consequently, channels with lower rates are moved to the end of the list,
thus ensuring the adaptive characteristic of the system. Through this procedure, the SM
increases the use of those channels with best performance and decreases the utilization of
channels with lower channel sensing success rate. This allocation scheme improves the
overall network performance and reduces the number of channel sensing operations.
Another important improvement regarding the adaptive characteristic of the proposed
system is that, it is able to build the channel database through its self-learning capability.
Hence, when the system operates in self-learning mode the spectrum holes are tracked
exclusively from channel sensing results. Thus, this operation mode allows the system to
work independently of prediction data. In order to evaluate the performance of the
proposed approach, the simulation results are compared in four different channel
allocation methods. For each method, the following criteria are evaluated (1) the channel
allocation success rate, (2) the user allocation success rate, (3) the channel sensing
success rate, and (4) number of channels sensed per user. Furthermore, the statistics of
the spectrum allocation are analyzed in order to evaluate the efficiency of the proposed
method.

SYSTEM DESCRIPTION
The proposed CR channel allocation method is composed by a spectrum manager (SM)
entity and a predictive and adaptive database (DB). The database is built from predicted
channel occupation statistical data of the considered bands (TETRA, GSM and DCS).
The spectrum-modeling algorithm presented in [7] is used to simulate spectrum
occupation, preserving the statistical properties of the channel. Based on channel
prediction data, the SM ranks the channels in the DB according to their probability of
been idle. This procedure is performed at the DB initialization and it allows the SUs to be
opportunistically allocated, by the SM, at those bands with a higher number of spectrum
holes, since the system startup.

Once the DB is built, the channel sensing results are used to update it so that channels
with best performance in terms of channel allocation success are prioritized for SUs
allocation. This feature gives the adaptive characteristic of the system, which brings two
main advantages: (1) it reduces collision with other users and (2) it minimizes user
allocation delay due to search for available channels. Thus, the proposed solution
combines the predictive and adaptive features, which eliminates the needs of performing
the spectrum sensing in order to identify opportunities during the BS initialization, as
proposed in [4, 5, 6]. Hence, it makes the BS more efficient during the system startup.
With the aim of evaluating the proposed solution, the main CR network elements have
been developed in MATLAB. Five entity programs have been created in order to simulate
the cognitive plane functions, as follows [2]:

Spectrum Manager Entity (SME): It is the main program responsible for triggering
other entities and keeping the database updated. The SME also controls the channel
classification entity (CCE) and the channel allocation entity (CAE). The DB is part of the
SME and it is adapted according to the channel sensing results.

Spectrum Generation Entity (SGE): It is the entity responsible for emulating the
spectrum occupancy and for providing prediction data regarding spectrum occupation in
TETRA, GSM e DCS bands.

Channel Classification Entity (CCE): This entity is responsible for uploading the
predicted data from SGE and for the channels classification based on each channel idle
time.
Channel Allocation Entity (CAE): The main function of this entity is to allocate SUs
according to channel availability defined in the DB.

Spectrum Sensing Entity (SSE): It emulates the Spectrum Sensing Function (SSF) [2]
providing instant channel status (busy or idle).

Figure 1 shows the flowchart for channel allocation procedure. Note that the DB is
initially built through the predicted data. Once the system starts operating, the DB is
dynamically adapted according to the channel sensing results. The “allocation cycle”
variable specifies the number of SU allocations that is performed to generate the channel
statistics that are used for the DB update.

FAIR ADAPTIVE ALGORITHM

The Fair Adaptive allocation algorithm is based on the fairness in allocation bit rates for
each secondary user in the CR system. This algorithm is fair in the sense that it tries to
allocate bits to users who have not received their fair share of service as much as possible
[1]. The algorithm first allocates bits to users to ensure fairness, and then subcarrier and
power are decided in greedy manner.

SYSTEM MODEL

We have assumed a system consisting of base station which serves both primary and
secondary users. Let us consider M secondary users are operating in the CR system in the
vicinity of only one primary or licensed user. The primary and secondary users have
adjacent frequency bands. The bandwidth of the primary user band is Wp Hz and that of
secondary sub-band is Ws Hz. We assume the presence of K orthogonal subcarriers such
that K/2 subcarriers are present in either side of the primary band. Hence the total
bandwidth of the CR system is Ws*K/2 Hz. Since orthogonal subcarriers have no
interference, only interference due to primary and secondary users has been considered.

The power spectral density (PSD) of kth subcarrier signal is assumed to be:

…………………………

where,

Pk is the transmit power of kth subcarrier

Ts is the symbol duration

Let Ik be the interference power introduced by the secondary signal into the primary
band.

So,

……………

where,

gk is the channel gain from base station to primary user for kth subcarrier dk is the
spectral distance between kth subcarrier and primary band

IFk is the interference factor for kth subcarrier

Let Smk be the interference power introduced by primary signal into kth secondary band
at

mth user. So,

 …………..
where

Hmk is subcarrier k gain from base station to user m

ϕRR ( ⅇ jω ) is the PSD of primary user‟s signal

Now, maximum number of bits in a symbol transmitted in the kth subcarrier is given by:

……………………..

where

No is the one sided noise PSD,

Pmk is the transmit power allocated to kth subcarrier of mth user,

Γ is set to unity for simplicity, amk є (0,1) is a subcarrier allocation indicator.

amk = 1 if kth subcarrier is allocated to mth user. The main objective is to maximize the
total bit rate for secondary users constrained by total transmit power, fairness and
interference levels. So, the optimization problem can be expressed as:

……………………..

where
Ptotal is the total CRU power

Ith is primary user‟s maximum tolerable interference level

The nominal bit rate weight (NBRW) for mth user is denoted by λm so that (
) is the fraction of total secondary user bits loaded to be fairly
allocated to user m.

This chapter describes the research methodology followed in this work to answer the
research questions and test formulated hypothesis. Also, the methods used for data
collection are summarized here. The main objectives of this work are proposing and
virtual prototyping the subset of radio scene analysis algorithms, including 1. detection of
vacant frequency channels for spectrum sharing; 2. waveform estimation including
modulation type, symbol rate, and central frequency estimation for the target application:
real-time portable