Energy detection Analysis of Cluster based Cooperative Spectrum Sensing in Cognitive Relay Network

— In this study, we analyzed an intra-cluster based cooperative energy detection between source (S) and destination (D), the network model consisting of one source, one destination and M number of clusters between the source and destination and each cluster consisting of N number of cognitive relays. We analyzed the performance of inter-cluster and intra-cluster cooperative energy detection. We derived the detection probability and false alarm for the improvement of energy detection of the proposed new scheme; we also provided the efficient number of relays in each cluster to improve the performance of the network. For relative comparison, we simulated similar situation like direct (non-cooperative), inter-cluster and intra-cluster communication between source and destination. Simulation results show that the proposed intra-cluster based network improves the network performance in comparison to direct and inter-cluster communication in terms of energy detection.

e) Simulation results are discussed to compare the performance of the proposed method with existing methods. The content of this paper is arranged in the following sections. Section II describes the System model for proposed cluster based cooperative communication. The Proposed methodology is discussed in section III. Analytical Result is presented in Section IV. Section V gives the simulation results and the conclusion of this paper is drawn in Section VI.

II.
SYSTEM MODEL In this study, we consider a clustered based wireless network shown in Fig. 1. It consist of one source and one destination node denoted by S and D, respectively in the XY plane. This XY plane is divided into areas of almost equal size. These areas are created by grouping of randomly generated cognitive relays (CR), having single transmitting and receiving antenna. This group of cognitive relays built a cluster. The distance between each cluster to destination is calculated and assuming that the distance between every cluster is almost equal and it is greater than the distance between the intra-cluster cognitive relays. Let us consider an M number of clusters are formed and each cluster consists of N number of cognitive relays (CR) cooperates for spectrum sensing. Let the channels between source (S) and cognitive relays (CR) are Rayleigh faded with AWGN. Basically our system model works in 2 steps as mentioned below: A. Inter-Cluster Communication) As the transmission distance is larger among the clusters, the source (S) can primarily broadcast the message signal to the relays of the first cluster. Then the CRs calculate the energy and the relay having the maximum energy is selected as best-relay for communication. Suppose that the elected best-relay in a cluster wants to send the message signal to the elected best-relay of the adjacent cluster, then the best-relay of first cluster will act as a source and the best-relay of second cluster will act as a destination.

B. Intra-Cluster Communication
In Intra cluster communication source (S) broadcasts the message signal or data bits with certain SNR to all CRs of the 1 st cluster. Then the relays calculate the energy and the relays having the maximum energy are selected as best-relays. Here we choose two cognitive relays which have received the maximum energy. In 1 st hop source (S) transmit the message signal to the 1 st best-relay and in 2 nd hop, communication 2 nd best-relay receives the message signal from the 1 st best-relay. Now the 2 nd best-relay acts as a source and broadcast to the next cluster.
In this study, we analyze the best-relay selection in terms of detection probability and false-alarm to improve the energy detection in cluster based network for cooperative relay transmission. At 1 st time slot, the message signal (x) from source reaches to all the relays of the 1 st cluster. The received signal at i th relay is represented by [8], The spectrum sensing for i th CR is represented as a binary hypothesis, i.e. [4] , 0 S CR h is represented as the channel coefficient between the source and cognitive relay. The energy detector at CR uses binary hypothesis determines the presence or absence of Source (S).

III.
PROPOSED METHODOLOGY In this section we discuss about the proposed methodology of intra-cluster communication shown in Fig. 2. The energy detection clustering consist of three phases-the broadcasting phase, relay selection phase and received phase. In the broadcasting phase, the source (S) broadcasts the primary message signal to the 1 st cluster with certain SNR. In the relay selection phase, the received energy at each relay of 1 st cluster is calculated and the relays having the maximum energy is selected as best-relay of a cluster which will send their sensing information to the next cognitive relay. The received energy is represented as, , , The received energy of the best-relay is compared with a preselected threshold  at the output of energy detector. If the energy is greater than the preselected threshold, then it broadcasts the message signal to the next cognitive relay (CR) else it sends an acknowledgement to the previous cognitive relay (CR).
; C R broadcast the next C R ; Send an acknoledgement to previous C R In the received phase, the message signal is received from the cognitive relay of the last cluster in the Network.

IV. ANALYTICAL MODEL OF DETECTION PROBABILITY
Substituting equation (8) and equation (9) in equation (7), the PDF of E bestrelay under hypothesis H 0 and H 1 can be represented as, [ It is noticed that P f is independent of fading mechanism. Thus the probability of false-alarm (P f ) under Rayleigh faded can be expressed as [7], The selected best-relay of a cluster sends their sensing information under the Rayleigh faded sensing channels to the selected best-relay of the adjacent cluster. Now the previous best-relay acts as a source and present bestrelay act as a destination. Thus the message signal is forwarded from source to destination through the clusters in the cooperative cognitive radio network.

V. SIMULATION RESULTS
In this section, we will discuss the performance of the proposed system using MATLAB simulations. A multihop cooperative cluster based network with equally spaced M number clusters between the source S and the destination D is considered. Each cluster is equipped with N randomly generated relays. Throughout the simulations the S and the D are located in the XY plane in between 3000 meter area.  Fig. 3. Shows the energy detection analysis of the proposed system model in terms of detection probability(P d )and false-alarm probability (P f ) for inter-cluster communication at SNR=-10dB. Here we take M=4 and M=6 in between S and D to compare with direct communication (S→D). We can see that the detection probability (P d ) increases when the number of the cluster increases from M=4 to M=6 and the direct communication from source to destination gives the lowest P d . So, the energy detection performance can be raised significantly with the increasing number of clusters in the cooperative network.  Fig. 4. Shows the energy detection analysis of the proposed system model in terms of detection probability (P d ) and false-alarm probability (P f ) for intra-cluster communication at SNR=-10dB. Here, we take M=4 and M=6 in between S and D to compare with direct communication (S→D). We can notice that the detection probability (P d ) increases when the number of the cluster increases from M=4 to M=6 and the direct communication from source to destination gives the lowest P d . Also, it gives better performance in comparison to inter-cluster communication as the communication distance between two cognitive relays are less in case of intra-cluster network. So, the energy detection performance may be enhanced with the increasing number of clusters in the network in case of intra-cluster communication than the inter-cluster communication. In Fig. 5. We provide the energy detection performance for different number of relays in each cluster like N=20, N=30, N=50 in between the Source and Destination in 4 inter-cluster communication. We notice that N=20 gives better performance than the N=30 and N=50. So, the optimal number of cognitive relays in each cluster for 4 inter-cluster communication are 20 which gives the better energy detection performance.  Fig. 6. We provide the energy detection performance for different number of relays in each cluster like N=20, N=30, N=50 in between the Source and Destination in 6 inter-cluster communication. We notice that N=30 gives better performance than the N=20 and N=50. So, the optimal number of cognitive relays in each cluster for 6 inter-cluster communication are 30 which gives the better energy detection performance.  Fig. 7, We show the comparison of the energy detection performance for different number of relays in each cluster like N=20, N=30, N=50 in between the Source and Destination in 4 intra-cluster communication. We can see that N=30 gives better performance than the N=20 and N=50. So, the optimal number of cognitive relays in each cluster for 4 intra-cluster communication are 30 which gives the better energy detection performance. In Fig. 8. We give the comparison of the energy detection performance for different number of relays in each cluster like N=20, N=30, N=50 in between the Source and Destination in 6 intra-cluster communication. We can see that N=20 gives better performance than the N=30 and N=50. So, the optimal number of cognitive relays in each cluster for 6 intra-cluster communication are 20 which gives the better energy detection performance. From Fig. 5, Fig6, Fig. 7 and Fig. 8 we can say that optimal number of relay selection increases the energy detection performance. Fig. 9. Comparative Energy Detection Analysis in between intra-cluster, inter-cluster and direct communication for M=6 clusters and N=20 cognitive relays.