Paper Title (use style: paper title)

International Journal of Electronics Communication and Computer Technology (IJECCT) Volume 2 Issue 3 (May 2012)

Analytical Model Analysis Of Distributed Cooperative Spectrum Sensing Method Ravi Prakash Shukla

Sandeep Singh

PG Student, SHIATS, Allahabad, India [email protected]

PG Student, SHIATS, Allahabad, India [email protected]

Abstract— Spectrum sensing is a key function of cognitive radio to prevent the harmful interference with licensed users and identify the available spectrum for improving the spectrum’s utilization. Various methods for spectrum sensing control, such as deciding which sensors should perform sensing simultaneously and finding the appropriate trade-off between probability of misdetection and false alarm rate, are described. However, detection performance in practice is often compromised with multipath fading, shadowing and receiver uncertainty issues. To mitigate the impact of these issues, cooperative spectrum sensing has been shown to be an effective method to improve the detection performance by exploiting spatial diversity. Keywords- cognitive radio, spectrum sensing, cooperative, distributed spectrum sensing I.

INTRODUCTION

In this paper, we focus attention on the particular task on which the very essence of cognitive radio rests: spectrum sensing, defined 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. The term B-spectrum holes[ stands for those sub bands of the radio spectrum that are underutilized (in part or in full) at a particular instant of time and specific geographic location. To be specific, the task of spectrum sensing involves the following subtasks: A. Detection of spectrum holes B. Spectral resolution of each spectrum hole; C. Estimation of the spatial directions of incoming interferes; D. Signal classification. The subtask of spectrum-hole detection is, at its simplest form, when the focus is on a white space (i.e., a sub band that is only occupied by white noise). Specifically, the detection of a white space may be performed by using a radiometer, which is well known for its energy-detection capability [4], [5]. Alternatively, we may resort to the use of cyclostationarity, which is an inherent property of digital modulated signals that naturally occur in the transmission of communication signals

ISSN:2249-7838

over a wireless channel [6], [7]. In both of these two approaches to spectrum sensing, the detection of a spectrum hole boils down to a binary hypothesis-testing problem. Specifically, hypothesis H1 refers to the presence of a primary user‟s signal (i.e., the sub band under test is occupied) and hypothesis H0 refers to the presence of ambient noise (i.e., the sub band is a white space). The cyclostationarity approach to detection has an advantage over the energy-detection approach in that it is also capable of signal classification and has the ability to distinguish co channel interference. The use of both of these approaches is confined to white spaces only, which limits the scope of their spectrum-sensing capabilities. In order to further refine the detection of white spaces and broaden the scope of spectrum sensing so as to also include the possible employment of gray spaces (i.e., sub bands of the spectrum that contain noise as well as interfering signals), we may have to resort to a sensing technique that includes spectrum estimation. II.

SPECTRUM SENSING:

In terms of occupancy, sub bands of the radio spectrum may be categorized as follows. 1) White spaces, which are free of RF interferers, except for noise due to natural and/or artificial sources. 2) Gray spaces, which are partially occupied by interferers as well as noise. 3) Black spaces, the contents of which are completely full due to the combined presence of communication and (possibly) interfering signals plus noise. Let Consider the commercial cellular networks deployed all over the world. In the current licensing regime, only primary users have exclusive rights to transmit. However, it is highly likely to find small spatial footprints in large cells where there are no primary users. Currently, opportunistic low-power usage of the cellular spectrum is not allowed in these areas, even though such usage by cognitive radios in a femto- or picocell with a small base station is not detrimental to the primary user [12]. Thus, spectrum holes may also be found in commercial cellular bands; naturally, spread of the spectrum holes varies over time and space. In any event, interference arising from conflict relationships between transmitters (base stations) of various radio infrastructure providers that coexist in a region

IJ ECCT | www.ijecct.org

84


must be taken into account [13]. Consequently, the spectrum holes found in cellular bands may also be gray spaces.

Negotiation messages, needed for general collaborative information sharing.

A cognitive radio is designed to be aware of and sensitive to the changes in its surrounding. The spectrum sensing function enables the cognitive radio to adapt to its environment by detecting spectrum holes [18]. The most efficient way to detect spectrum holes is to detect the primary users that are receiving data within the communication range of an xG user. In reality, however, it is difficult for a cognitive radio to have a direct measurement of a channel between a primary receiver and a transmitter. Thus, the most recent work focuses on primary transmitter detection based on local observations of xG users.

CCR is also a means to coordinate collaborative spectrum sensing between network nodes. Here coordination covers functions like: sharing of sensing effort, requests for spectrum sensing, or coordination of quiet periods for spectrum sensing. To share the spectrum sensing load between network nodes, the frequency band allowed for cognitive use is divided into subbands. Different nodes sense different frequency sub-bands. The frequency sub-band each node is sensing in each time instant is determined by a pseudorandom time-frequency code. The hidden node problem can be alleviated if at least two nodes measure the same part of spectrum at the same time; subsets of collaborating nodes for each spectrum sub-band are formed and always changed after a certain period of time. The effects of propagation, such as fading and shadowing, are effectively mitigated through diversity because the channels the signals experience can be assumed to be uncorrelated since the secondary users are displaced from each other. IV.

Figure 1: Classification of spectrum sensing techniques III.

COOPERATIVE SPECTRUM SENSING

This section relates to spectrum sensing control, awareness networking and cooperative spectrum sensing in Figure 1. Cooperative spectrum sensing is a powerful concept to leverage the spatial separation of multiple spectrum sensing nodes in a wireless network. The optimal fusion of sensing results, acquired by distributed network nodes, allows to alleviate the hidden node problem and/or to share the sensing load between network nodes. The optimal fusion of decentralized observations has been studied since a long time, see e.g., [20] and the references therein. It has been shown already in [8], [9] that the optimal fusion rule is to compute the joint likelihood ratio of the distributed observations. Cooperative spectrum sensing requires a networking solution to communicate sensing results (sensing messages) between nodes. Using spectrum sensing individual network nodes, as well as the whole network by virtue of collaboration, becomes aware of the local radio spectrum situation. Consequently the distribution of spectrum sensing results can be understood as Awareness Signaling. Within E3 (End-to-End Efficiency project) an awareness signaling solution, namely Cognitive Control Radio (CCR) has been developed [10], [11].

SYSTEM MODEL

In this model, cognitive radio CRs operate in distributed cooperative manner; that divide CRs population into groups, each of which select the node with the best reporting channel gain as a fusion node. The CRs conduct local sensing based on maintained energy detection and forward their binary detection decision to fusion node where the processing and fusion of local spectrum observation for candidate nodes is made, the modeling flow is shown in Fig. 2. The flow chart as shown in Fig. 3 illustrate the formation of DCS network architecture and the selection of fusion nodes based on the reporting channel SNR to fusion centre. It also show the possible actions for node leaving and joining the network, however we consider no change in architecture during Sensing period and no node mobility. The DCS is modeled with a standard parallel fusion network. A schematic representation of distributed cooperation is illustrated in Fig. 4; each fusion node calculates its group decision. It then sends the result to the fusion centre through a best control channel. The fusion centre computes the global decision, from the outputs of the fusion nodes.

The CCR is targeted for sharing spectrum sensing and use related information between Cognitive Radio networks. The CCR network provides information mainly for the secondary users, which form local wireless networks. Thus, it can be seen to complement CPC, which is mainly targeted for providing information to primary users. CCR is an awareness signaling solution that supports the exchange of Information, Query, and

ISSN:2249-7838


85


Figure 2: Modelling diagram

Figure 4: Schematic representation of distributed cooperation for spectrum sensing V.

DETECTION MODEL

The basic problem concerning spectrum sensing is the detection of a signal within a noisy measure. We assume that prior knowledge of the primary user signal is not known. Therefore, optimal detector based on matched filter is not an option since it would require the knowledge of the data for coherent processing. Instead a suboptimal energy detector is adopted, which can be applied to any signal type. We assume the noise is additive white Gaussian with zero mean and power spectral density. We consider a low-mobility environment, so we assume that during the course of the transmission, or for each sensing period, each user observes only one fading level towards the fusion node/fusion centre. Due to the spatial separation between users, the channels corresponding to different cognitive users are assumed to be independent. All channels are assumed to experience Rayleigh fading. Therefore, the received signal at the secondary receiver has the following simple form, s[n] = hx[n] + w[n]

Figure 3: Network formation and fusion node selection

(1)

Where x[n] is the signal to be detected, w[n] is the additive white Gaussian noise (AWGN), h is the channel fading coefficient, and n is the sample index. Note that x[n] = 0 when there is no transmission by a primary user. The received signal at cognitive radio has one of the following hypotheses, busy channel, H1, which indicate primary user present and White space/Spectrum hole/Idle channel, H0 , that indicate primary user absent H1 : s[n] = hx[n] + w[n] H0 : s[n] = w[n]

(2)

The energy detection based sensing metric can be obtained as [10], M=

ISSN:2249-7838

𝑁−1 𝑛=0

s[n]

(3)


86


VI.

ANALYSIS AND CONSIDERATION:

(i) By help of this model we investigate of ROC (receiver operating characteristics), only for group members, for the 16 user groups (=16 j G), with different SNR (5, 10 and 15) for the reporting/control channel from the fusion node to fusion centre is done. Average sensing SNR is 10 and detection threshold is set with maintained probability of false alarm. (ii) By the impact of number of nodes within the group. We see that better performance is achieved while considering the 2 user group compared to the group with large number of cognitive radios (16 nodes). Direct reporting is plotted for each group, it can be confirmed that increasing the number of cooperative users exponentially can obtain gain in detection probability and it is clear that DCS outperforms CS corresponding to the case. (iii) Also by this model the reporting Bit Error Rate (BER) for the proposed method calculated for different number of sensing group. The analytical result is given for 3 sensing group with 2, 4, 8 users. As more number of group incorporated in sensing the probability of error reduced, the results shown that the probability of error in the reporting stage for the same SNR is decreased when number of groups increase. This indicates that the selection of the reporting channel by the mean of fusion node (selection diversity) in the group sensing is achieved.

REFERENCES [1] [2]

[3]

[4] [5]

[6]

[7] [8]

[9]

The number of users in these groups was varied and the effect on radio sensitivity for a 90% and 95%, probability of detection was observed.

[10]

(iv) This method creates a loss in the probability of false alarm due to the large „no decision region‟, additionally this scheme may eliminate a user with real detection information from reporting the decision which increase probability of interference to primary system.

[12]

[11]

[13]

[14]

VII. CONCLUSION The main focus of this paper is to examine the effects of distributed decision fusion based on maintained probability of false alarm and best reporting channel selection on the cooperative spectrum sensing employed by cognitive radios. We adopted a dynamic distributed architecture for cooperative sensing based on the link quality and found condition on the channel quality for cooperation to be beneficial. Using probability of detection, and BER metrics we evaluated the performance improvement of distributed cooperation over direct cooperation and non cooperative sensing. We used analytical formulation with possible candidate selection criteria to investigate and maximize the cooperation gain. By employing such distribution and selection technique, the reporting error due to the fading channel is reduced. Results show that the method effectively improve performance of sensing, it increase the probability of detection up to 0.9 at