asas

1. Analysis on Decision Fusion Strategies on Spectrum Sensing in
Cognitive Radio Networks
Priya Geete
Ph. D, Research Scholar,
Suresh Gyan Vihar University, Jaipur
Megha Motta
Department of Electronics and Communication
Acropolis Technical Campus,Indore
Abstract—Cognitive radio is a innovative technology which provides a new way to extend
utilization of available spectrum. Spectrum sensing is a essential problem for cognitive radio.
Cooperative spectrum sensing is an efficient way to detect spectrum holes in cognitive radio
network. In this paper, we analysis that in cooperative sensing for decision fusion we
perform some hypothesis test in which we study different methods of hypothesis testing
based on various fusion rules, Likelihood ratio test (LRT) and Neymon Pearson Criteria. Some
serial and parallel topologies of distributed network in which secondary users are
connected to each other for performing their operation are also shown. Soft combination
scheme exceeds hard combination scheme at the cost of complexity. Therefore quantized
combination scheme provides a better compromise between detection performance and
complexity.
Index-Cognitive radio (CR), Dynamic Spectrum Access (DSA), Cooperative spectrum sensing,
Energy detection, Likelihood ratio test (LRT), Fusion rules, Decision fusion, Data fusion.
I. INTRODUCTION
The radio spectrum which is very essential for
wireless communication is a nature limited resource.
Fixed Spectrum Access (FSA) policy has
traditionally been adopted by spectrum regulators
to support various wireless applications. According to
FSA each part of spectrum with definite bandwidth
will be hand over to one or more dedicated users also
known as licensed user’s. Only these users have right
to use the allocated spectrum and other users are not
allowed to use it. On the other hand, recent studies of
spectrum utilization measurements shows that a large
segments of licensed spectrum experiences less
utilization ,i.e, most of the time spectrum is in ideal
condition and is not used by ts licensed users[1]-[3].
To overcome this situation Dynamic Spectrum Access
(DSA), was introduced. It allows radio spectrum to be
used in a more effective manner. According to DSA a
small part of spectrum can be allocated to one or more
users, which are called primary users (PUs); however
the use of that spectrum is not fully granted to these
users, although they have higher priority in using it.
Other users, which are referred to as secondary
users (SUs),can also access the allocated spectrum
as long as the PUs are not temporally using it.
This opportunistic access should be in a manner that it
does not interrupt any primary user in band.
Secondary users must be aware of the activities done
by the primary user so that they could spot the
spectrum holes and the ideal state of the primary users
in order to utilize the free band and also rapidly
evacuate the band as soon as the primary users

2. becomes active. Very low utilization of spectrum from
0-6 GHz is shown in Fig. 1
Fig: 1. Spectrum Utilization Measurements [4]
The rest of the paper is organized as follows. In
section II, we revealed cognitive functionalities which
includes cognitive cycle too. Formulation methods for
hypothesis are being examined in Section III. Section
IV, illustrates different types of spectrum sensing
techniques. In Section V, we formulate the system
model in CR networks. Then we investigate different
fusion rules and propose a new quantized foue-bit
hard combination scheme in Section VI. Comparison
between one to four bit hard combination scheme is
shown in Section VII, respectively. Conclusions are
given in Section VIII
II. COGNITIVE RADIO FUNCTIONALITIES
According to S.Hykin “Cognitive Radio is an
intelligent wireless communication system that is
aware of its surround- ing environment (i.e. outside
world), and uses the methodology of understanding by
building to learn from the environment and adapt its
internal states to statistical variations in the incoming
RF stimuli by making corresponding changes in
certain operating parameters in real-time, with two
primary objectives in mind:
• Highly reliable communications whenever and
wherever needed.
• Efficient utilization of the radio.”[5]
From the above mention definition two characteristics
of cognitive radio can be summarized as cognitive
radio can be summarized as cognitive and
recofigurability. The first one enables the cognitive
radio to interact with its environment in a real-time
manner, and intelligently determine based on quality
of service (QoS) requirements. Thus these tasks can be
implemented by a basic cognitive cycle: Spectrum
sensing, spectrum analysis and spectrum decision as
shown in Fig. 2
Fig. 2. The Cognitive Capability of cognitive radio
enabled by a basic cognitive cycle
1. Spectrum Sensing: It is done by either
cooperative or non cooperative technique in
which cognitive radio nodes continuously
monitor the RF environment
2. Spectrum Analysis: It estimates the
characteristics of spectral bands that are
sensed through spectrum sensing.
3. Spectrum Decision: An appropriate spectral
band will be chosen according to the spectrum
characteristics analyzed for a particular
cognitive radio node. Then the cognitive radio
determines new configuration parameters.
The other feature of cognitive radio is
reconfigurability. Therefore in order to get adapted to
RF environment, cognitive radio should change its

3. operational parameters[5]:
1. Operating Frequency: Cognitive radio is
capable of varying its operating frequency in
order to avoid the PU to share spectrum with
other users.
2. Modulation Scheme: According to the user
requirements of the user and channel condition
cognitive radio should adaptively reconfigure
the modulation scheme.
3. Transmission Power: In order to improve
spectral efficiency or diminish interference
transmission power can be reconfigured.
4. Communication Technology: By changing
modulation scheme interoperability among
different communication systems can also be
provided by cognitive radio.
Decision Fusion versus Data Fusion
The cooperative spectrum sensing approach can be
seen where each cooperative partner makes a binary
decision based on the local observation and then
forwards one bit of the decision to the common
receiver. At the common receiver, all 1-bit decisions
are fused together according to an OR logic. We refer
to this approach as decision fusion. An alternative
form of cooperative spectrum sensing can be
performed as follows. Instead of transmitting the 1-bit
decision to the common receiver each CR can just
send its observation value directly to the common
receiver. This alternative approach is referred to this
approach as data fusion. Obviously, the 1-bit decision
needs a low bandwidth control channel.
III. FORMULATION METHODS
FOR HYPOTHESIS
A. Neyman Pearson Decision
Criterion:
It is considered for the estimation of minimum
error probability when information of a priori
probabilities is not available [6]. Thus in this type of
situation two different types of probabilities are of
importance. One is the probability of False Alarm
and the other is the probability of Miss Alarm.
Therefore both probabilities are defined on the basis
of two hypothesis H1 and H0 . H1 hypothesis is
considered when signal and noise both are present
where as H0 hypothesis is considered when only
noise is present. Errors take place in either of two
situations. First type of error arises when choice is
made in favor of H1 but H0 is true. It is denoted
by P (D1 /H0 ) and is known as probability of
false Alarm Pf . And the other error occurs when
choice is made in favor of H0 although H1 is true.
This is denoted by P (D0 /H1 ) and is known as
Miss Alarm Pm [6]. Probability of correct decision
is denoted in equation A
PD = 1 − P (Do /H1 ) = 1 − Pm (A)
In Neyman Pearson criterion an approach is made to
maximize probability of of detection for an consigned
probability of false alarm. Effectively, a function
defined by QN P = Pm + µPf is minimized for
an assigned Pf and a given constant. Thus the plot
between PD versus Pf is known as Receiver
Operating Characteristics (ROC) as shown in Fig. 3
Fig. 3. ROC
B. Likelihood Ratio Test:

4. H 1
For establishing the receiver decision rule for the
case of two signal classes a practical starting point
is
P (S0 |Z ) ≷H 0
P (S1 |Z ) (B)
his equation states that we should choose hypothesis
H0 if the posterior probability P (S0 |Z ) is
greater than the posterior probability P (S1 |Z ).
Otherwise we should choose hypothesis H1 . Above
equation can also be written as:
P (Z |S0 )P (S0 ) ≷H 0 P (Z |S1 )P (S1 ) (C)
IV. SPECTRUM SENSING
Spectrum sensing is a key element in cognitive
radio net- work. In fact it is a major challenge in
cognitive radio for secondary users to detect the
presence of primary users in a licensed spectrum
and quit the frequency band immediately if the
corresponding primary user emerges in order to
avoid interference to primary users.[7]
Spectrum sensing technique can be further
categorized as Non- cooperative and Cooperative as
shown in Fig. 4
Fig. 4. Spectrum Sensing
Techniques
Spectrum sensing can be simply reduced to an
identification problem, modelled as a hypothesis test .
The sensing equipment has to just decide between for
one of the two hypotheses:-
H1: y (n) = r(n) + x(n),n=1,2,3....N (1)
H 0: x (n) = x(n), n=1,2,3....N (2)
where ‘y(n)’ is the signal transmitted by the primary
users.
‘r(n)’ being the signal received by the
secondary users.
‘x (n)’ is the additive white Gaussian noise
with Variances σ2n
x.
Hypothesis ‘H0’ indicates absence of primary user
Hypothesis ‘H1’ points towards presence of primary
user
Thus
for
the three state hypotheses numbers of important cases
are:-- H1 turns out to be TRUE in case of presence of
primary user i.e. P(H1 / H1) is
known as Probability of Detection (Pd).
- H0 turns out to be TRUE in case of presence
of primary user i.e. P(H0 / H1) is known
as Probability of Miss-Detection (Pm).
- H1 turns out to be TRUE in case of absence
of primary user i.e. P(H1 / H0) is known
as Probability of False Alarm (Pf).
Probability of detection is of main concern as it gives
the probability of correctly sensing for the presence
of primary users in the frequency band. Probability of
miss-detection is just the complement of detection
probability. The goal of the sensing schemes is to
maximize the detection probability for a low
probability of false alarm. But there is always a trade-
off between these two probabilities. Receiver
Operating Characteristics (ROC) presents very

5. valuable information as regards the behaviour of
detection probability with changing false alarm
probability (Pd v/s Pf) or miss-detection probability
(Pd v/s Pm).
4.1 Transmission Detection (Non-Cooperative)
A. Matched-Filtering Technique:
Matched-filtering is known as the optimum method
for detection of primary users when the transmitted
signal is known. The main advantage of matched
filtering is the short time to achieve a certain
probability of false alarm or probability of miss
detection as compared to other methods. Matched-
filtering requires cognitive radio to demodulate
received signals. Hence, it requires perfect
knowledge of the primary users signalling features
such as bandwidth, operating frequency, modulation
type and order, pulse shaping, and frame format.
Fig 5: Matched-Filtering Technique [5]
The output of the matched filter, given that ‘x[n]’ is
the received signal and ‘s[n]’ is the filter response, is
given as
(7)
B. Energy Detector Based Sensing:
Energy detector based approach which is also
known as radiometry or periodogram, is the most
common way of spectrum sensing because of
its low computational and implementation
complexities. It is more generic method as
receivers do not need any knowledge on the
primary users signal.
Fig 6: Energy Detector Based Sensing [5]
It has the following components:-
- Band-pass filter -- Limits the bandwidth of the
received Signal to the frequency band of
interest.
- Square Law Device – Squares each term of
the received Signal.
- Summation Device – Add all the squared
values to Compute the energy.
A threshold value is required for comparison of the
energy found by the detector. Energy greater than the
threshold values indicates the presence of the primary
user. The principle of energy detection is shown in
figure 6. The energy is calculated as
(8)
the Energy is now compared to a threshold for
checking which hypothesis turns out to be true.
E > λ => H1 (9)
E < λ => H0 (10)
C. Cyclostationary-Based Sensing:
Nature has its way in such a manner that many of its
processes arise due to periodic phenomenon.
Examples include fields like radio astronomy wherein
the periodicity is due to the rotation and revolution of
the planets, weather of the earth due to periodic
variation of seasons [13]. In telecommunication, radar
and sonar fields it arises due to modulation, coding
etc. It might be that all the processes are not periodic
function of time but their statistical features indicate
periodicities and such processes are called cyclo-
stationary process. For a process that is wide sense
stationary and exhibits cyclostationary has an auto-
correlation function which is periodic in time domain.
Now when the auto-correlation function is expanded
in term of the Fourier series co-efficient it comes out
that the function is only dependent on the lag
parameter which is nothing but frequency. The
spectral components of a wide sense cyclostationary
process are completely uncorrelated from each other.
The Fourier series expansion is known as cyclic auto-
correlation function (CAF) and the lag parameter i.e.
the frequencies is given the name of cyclic
frequencies. The cyclic frequencies are multiples of
the reciprocal of period of cyclostationary. The cyclic
spectrum density (CSD) which is obtained by taking
the Fourier transform of the cyclic auto-correlation
function (CAF) represents the density of the
correlation between two spectral components that are
separated by a quantity equal to the cyclic frequency.

6. The following conditions are essential to be filled by a
process for it to be wide sense cyclostationary:-
Fig 7: Cyclostationary Sensing Technique [5]
E{x(t+T0) = E{x(t)} (11)
Rx (t+T0,τ) = Rx(t, τ) (12)
Where Rx = E { x(t + τ) x(t)} (13)
Thus both the mean and auto-correlation function for
such a process needs to be periodic with some period
say T The cyclic auto-correlation function (CAF) is
represented in terms of Fourier co-efficient as:-
(14)
‘n/T0’ represent the cyclic frequencies and can be
written as ‘α’. A wide sense stationary process is a
special case of a wide sense cyclostationary process
for ‘n/T0 = α=0’. The cyclic spectral density (CSD)
representing the time averaged correlation between
two spectral components of a process which are
separated in frequencies by ‘α’ is given as
(15)
The power spectral density (PSD) is a special case of
cyclic spectral density (CSD) for ‘α=0’. It is
equivalent to taking the Fourier transform of special
case of wide sense cyclostationary for ‘n/T0 = α=0’.
V. SYSTEM MODEL
Let there be a cognitive network with K cognitive
users (such that K = 1,2,3,.....K) to sense the spectrum
in order to detect the presence of PU. Assume that
each CR performs local spectrum sensing
independently by using N samples of the received
signal. By taking two possible hypothesis H0 and H1
in binary hypothesis testing problem the problem of
spectrum sensing can be formulated as:
H0 : xk (n) = wk (n) (16)
H1 : xk (n) = hk s(n) + wk(n) (17)
here s(n) are samples of transmitted signal also known
as primary signal, wk(n) is the receiver noise for the
kth CR user, which is assumed to be an i.i.d. random
process with zero mean and variance and hk is the
complex gain of the channel between the PU and the
kth
CR user. H0 and H1 represents whether the signal
is present or absent correspondingly. Using energy
detector, the kth
CR user will calculate the received
energy as:
Ek = (18)
If we consider the case of soft decision, each CR
user forwards the entire result Ek to the FC where
as in case of hard decision, the CR user makes one-
bit decision given by ∆k by comparing the received
energy Ek with the predefined threshold λk .
∆k={1,Ek>λk} (4)
∆k={0,otherwise} (5)
Detection probability Pd,k and false alarm
probability Pf,k of the CR user K are defined as:
Pd,k = P r{∆k=1|H1} =Pr {Ek >λk |H1 } (21)
Pf,k = P r {∆k = 1|H0}=P r {Ek >λk |H0} (22)
Let λk = λ for all CR users, the detection
probability, false alarm probability and miss
detection Pm,k over AWGN channels can be
expressed respectively.
Pd,k = Qm( ) (23)
Pf,k =Γ(m, λ/2)/Γ(m) (24)
Pm,k = 1-Pd,k (25)
where γ is the signal to noise ratio (SNR),
m=TW is the time bandwidth product, QN (., .) is

7. the generalized Marcum Q function(.) and Γ(., .) are
complete and incomplete gamma function
respectively.
VI. CONCLUSION
In this paper, the effect of fusion rules for cooperative
spectrum sensing is shown. We have seen the data and
decision fusion in cooperative sensing using some
hypothesis test. These hypothesis testing was based on
various fusion rules, Likelihood ratio test (LRT) and
Neymon Pearson Criteria. Some serial and parallel
topologies of distributed network in which secondary
users are connected to each other for performing their
operation are also shown. The hypothesis testing and
all fusion rules are applied in centralized network of
secondary users. We have extended the combination
of bits till 4 bits. The proposed quantized four-bit
combination scheme wins advantage of the soft and
the hard decisions schemes with a tradeoff between
overhead and detection performance. Simulation
comparison will be done between various fusion rules.
With the help of simulation we will see that soft
combination scheme exceeds hard combination
scheme at the cost of complexity. Therefore quantized
combination scheme provides a better compromise
betweendetection performance and complexity.
VII. REFERENCES
1. Adel Gaafar A. Elrahim, and Nada Mohamed Elfatih, “A
Survey for Cognitive Radio Networks” International
Journal of Computer Science and Telecommunications"
,Volume 5, Issue 11, November 2014
2. Ekram Hossain, Dusit Niyato and Dong In Kim, ”Evolution
and future trends of research in cognitive radio: a
contemporary survey”, Wirel. Commun. Mob. Comput.John
Wiley & Sons, Ltd., December 2013.
3. Ayubi Preet, Amandeep Kaur, “Review paper on Cognitive
Radio Networking and Communications”, International
Journal of Computer Science and Information Technologies,
Vol. 5 (4) , 2014,.
4. Shankar, S.N.," Squeezing the Most Out of Cognitive
Radio: A Joint MAC/PHY Perspective", In the
proceedings of IEEE International Conference on
Acoustics, Speech and Signal Processing, 2007.
5. Md. Manjurul Hasan Khan, Dr. Paresh Chandra Barman
"Investigation of Cognitive Radio System Using MATLAB"
World Vision Research Journal Vol.8,No. 1 ,November
2014.
6. Rehan Ahmed & Yasir ArfatGhous, “Detection of vacant
frequency bands in Cognitive Radio,” Blekinge Institute of
Technology May 2010.
7. D.Manish “Spectrum Sensing in Cognitive Radio: Use of
Cyclo-Stationary Detector,” National Institute of
Technology Rourkela, Orissa-769008, India May 2012.
8. Zhaolong Ning, Yao Yu, Qingyang Song, Yuhuai
Peng, Bo Zhang “Interference-aware spectrum sensing
mechanisms in cognitive radio networks” Article in
Press Computers and Electrical
Engineering,Elsevier,2014
9. Li, X., Cao, J., Ji, Q., & Hei, Y. (2013, April). Energy
efficient techniques with sensing time optimization in
cognitive radio networks. In: Wireless Communications and
Networking Conference (WCNC),
10. S. Atapattu, C. Tellambura, and Hai Jiang, “Energy
Detection Based Cooperative Spectrum Sensing in
Cognitive Radio Networks,” IEEE Transactions on
Wireless Communications, vol. 10, no. 4, pp. 1232-1241,
2011.
11. Mahmood A. Abdulsattar and Zahir A Hussein “Energy
Detection Technique for Spectrum Sensing in Cognitive Radio:
A Survey” International Journal of Computer Networks &
Communication (IJCNC) Vol.4, No.5 September 2012
12. Hongjian S u n , “Wideband Spectrum Sensing For
Cognitive Radio Networks: A Survey” Volume-2,
Issue-5, June 2013.

8. 13. Tevfik Yucek and Huseyin Arsla, "A Survey of Spectrum
Sensing Algorithms for Cognitive Radio
Applications",IEEE Communication Surveys & Tutorial
,Vol.11,No.1,2009.
Table1: Summary of advantages and disadvantages of Transmission detection (non cooperative) spectrum
sensing techniques [5]
Spectrum
sensing
Techniques
Advantages Disadvantages
Matched
filtering
Optimal
performance
Low
computational
cost
Requires prior
information of
the primary user
Energy
detection
Does not
require prior
information
Low
computational
cost
Poor
performance for
low SNR
Cannot
differentiate
users
Cyclostationary
Feature
Valid in low
SNR region
Robust against
interference
Requires partial
prior information
High
computational
cost

9. Table2: Summary of advantages and disadvantages of Transmission detection (non cooperative) spectrum
sensing techniques [5]
Spectrum sensing
techniques
Advantages Disadvantages
Centralized
Sensing
Bandwidth efficient
for same number of
cooperating CRs as
compared to
distributed
cooperation
One CR i.e. FC becomes very
critical as well as
complex to carry the burden of all
cooperating CRs
Distributed
Sensing
No need of
backbone
infrastructure
resulting
in low
implementation
cost
Large control bandwidth
required for information
exchange among all
cooperating CRs
Finding neighbors in itself is a
challenging task for CRs
Large sensing duration
resulting from iterative nature of
distribute techniques

10. Table2: Summary of advantages and disadvantages of Transmission detection (non cooperative) spectrum
sensing techniques [5]
Spectrum sensing
techniques
Advantages Disadvantages
Centralized
Sensing
Bandwidth efficient
for same number of
cooperating CRs as
compared to
distributed
cooperation
One CR i.e. FC becomes very
critical as well as
complex to carry the burden of all
cooperating CRs
Distributed
Sensing
No need of
backbone
infrastructure
resulting
in low
implementation
cost
Large control bandwidth
required for information
exchange among all
cooperating CRs
Finding neighbors in itself is a
challenging task for CRs
Large sensing duration
resulting from iterative nature of
distribute techniques