ENERGY EFFICIENT COOPERATIVE SPECTRUM SENSING IN COGNITIVE RADIO NETWORKS

Abstract

Cognitive or intelligent radio has emerged as a promising technology to ameliorate the spectrum utilization by exploiting the unused spectrum bands. Cognitive radio is proposed in literature as an intelligent wireless communication system that is aware of its surrounding environment. It allows unlicensed or secondary user, to utilize the vacant spectrum at any time with no or less interference to the licensed users. Generally, cognitive radio creates networks in order to better identify spectrum vacancies, avoid induced interference and accordingly, boost their revenue. One of the major challenge in cognitive radio communication, now a days, is high consumption of energy, resulting in limitation of implementation specifically in battery-powered devices. The cognitive radio communication deals with three main steps such as spectrum sensing, allocation and reliable transmission. The initial step involves the detection of licensed user signals in order to determine the availability of a spectrum band for transmission. The spectrum sensing which deals with single primary and secondary node is known as single node spectrum sensing. It has been reported by previous studies that fading, shadowing and hidden node problem may affect the detection quality of single node sensing. It is therefore possible to improve the performance in such cases by employing multiuser diversity. In cooperative spectrum sensing (multiuser sensing), multiple secondary nodes work in collaboration with each other. The locally generated sensing results are exchanged in order to make a global decision regarding spectrum occupancy. However, due to overhead of multiple nodes, cooperative sensing consumes a significant amount of energy, which is a challenge for the cognitive users. Moreover, increasing number of sensing channels and periodicity of cooperative sensing further complicates the problem. Hence, energy efficiency in cooperative spectrum sensing have led to a proliferation of research in cognitive radio. In this thesis, multiple energy efficiency maximization algorithms/schemes for cooperative spectrum sensing are proposed. There are three models of collaboration viz. centralized, distributed and hierarchical, here, we consider centralized model. The proposed work comprises of optimization of energy efficiency at local sensing, reporting and transmission stage. This includes the optimization of number of secondary users, sensing time and transmission time. The optimization is carried out in fading and nonfading environment and at different SNR levels. The performance of the proposed work is evaluated in terms of energy efficiency and detection accuracy. The first objective of this dissertation is to optimize the sensing and transmission time to increase the bits transmitted per frame so as to achieve maximum energy efficiency for single cognitive radio. To maximize the energy efficiency a Sub Optimal Iterative Search Algorithm (SOISA) is proposed. This problem also considers the i interference occurred due to secondary user transmission to the primary user. The proposed algorithm shows the superiority in terms of complexity and high value of energy efficiency as compared to the other existing algorithms. The effects of change in sensing time, transmission time and transmission power on energy efficiency are shown by the simulation results. Following this, the maximization of energy efficiency by optimizing the hard decision fusion in case of multiple users cooperative spectrum sensing is considered as another problem. A method to improve the energy efficiency by optimizing the number of secondary users is proposed and the results are compared in frequency selective fading and non fading environments. Three hard decision fusion rules viz. OR rule, AND rule, and k-out-of-N rule are compared with respect to energy efficiency and number of secondary users. It is clear after simulation that k-out-of-N rule outperforms others in term of energy efficiency and Global probability-of-false-alarm. The above two problems optimized the sensing and transmission time and number of secondary users, separately. It is, however, possible to jointly optimize the above three parameters to achieve maximum energy efficiency in the cooperative spectrum sensing scenario. Therefore, joint optimization of sensing, transmission time and number of secondary users is proposed with the protection of primary user from secondary user transmission. In order to solve the problem, first the optimal expression for the number of secondary users is obtained and then an iterative sub optimal algorithm is proposed to achieve optimal sensing and transmission time. The effectiveness of this work is demonstrated by extensive simulation results and illustrations. Herein, comprehensive approaches for energy-efficiency maximization are proposed with different algorithms. The aim of these comprehensive approaches is to improve energy efficiency and provide consistency of the proposed algorithms. The simulation results reveal that the energy efficiency is achieved as high as 10:4957 bits/Hz/Joule, which is maximum as compared to other algorithms and methods. The Non Orthogonal Multiple Access (NOMA) enabled cognitive radio technique will increase the application of cognitive radio into future 5 G systems. The NOMA enabled cognitive radio is proposed to improve the energy efficiency of single cognitive radio networks. Here, the energy efficiency maximization problem with down link NOMA technique is studied. A base station is assumed which is equipped with two antennas, one for primary user and another for secondary user. The energy efficiency maximization problem is formed as the ratio of maximum achievable sum rate and total power consumed. Further, the energy efficiency is maximized for the NOMA and compared with other existing conventional multiplexing techniques.