Meta-heuristic Optimization based Parameter Adaptation in Cognitive Radio Systems

Abstract

Cognitive Radio (CR) has emerged as an enabling technology to dynamically use the unused or underused spectrum, thereby increasing the spectral efficiency. The task of adapting or reconfiguring the system parameters is of utmost importance to enhance the overall performance of the CR system. In CR based system, the decision engine is an important module that not only embeds the features of observation and cognition of the radio but also responsible for parameter adaptation. As meta-heuristic algorithms offer numerous advantages over classical mathematical approaches, the performance of these algorithms is investigated to design an efficient CR system that possesses the ability:  To adapt the transmission parameters for efficiently minimizing the power consumption, bit error rate and adjacent channel interference while maximizing the throughput of a secondary user (SU).  To maximize the number of transmission opportunities for a SU by optimizing the sensing period of a licensed channel.  To improve the reliability of a physical layer based sensing by enhancing the probability of detection of a licensed user using cooperative spectrum sensing (CSS) scheme. In this research work, the performance of various parameter-less meta-heuristic techniques such as ant lion optimizer (ALO), grey wolf optimizer (GWO), grasshopper optimization algorithm (GOA), moth flame optimization (MFO) and whale optimization algorithm (WOA) is investigated to reconfigure the transmission parameters for minimizing the total system power consumption at CR transmitter (CR_Tx) operating with Class B power amplifier. A simple system model is considered where one primary transmitter communicates with a primary receiver on a TV band. Upon finding this band as vacant, i.e. on the occurrence of TV white space, SU_Tx communicates with the secondary receiver in an interweave manner. A mathematical formulation of total system power consumption is shown and the problem is solved for data transmission scenario with constraints on data rate, bit error rate and adjacent channel interference. Simulation results show the effectiveness of WOA in minimizing the system power consumption by parameter adaptation in a multicarrier CR system while satisfying different QoS constraints.Further, the application of ALO, GWO, MFO, WOA and Jaya algorithm is investigated to reconfigure the parameters for various transmission scenarios of a CR based IoT device. In each scenario, different user requirement, i.e. an IoT application and radio‟s battery level are considered. Constrained multi-objective optimization problem is solved by employing the weighted sum method where each weight vector emphasizes different communication objective such as minimize power consumption, minimize bit error rate and maximize throughput. The constraints for ACI and total transmit power of a SU are incorporated into a given problem by using a novel exponential penalty function. After adapting the transmission parameters, the performance of a MAC layer based sensing is improved by optimizing the sensing period of a licensed channel. The recently proposed Jaya algorithm is employed to maximize the number of transmission opportunities for a SU while constraining the sensing overhead and interference time within a user-defined limit using a penalty function. Jaya algorithm is found to achieve quick convergence and better optimal value; making it a preferable choice for real-time applications. This thesis also focuses on improving the detection performance of a physical layer based sensing. CSS is an effective technique to improve the probability of detection of a primary user. The performance of CSS can be enhanced by optimizing the weight vector of the observation statistics obtained from different SUs at the fusion center. A novel integrated technique, i.e. opposition based grey wolf optimizer (OBGWO) is proposed and tested on several benchmark problems by comparing its performance with existing algorithms: MFO, GWO and sine cosine algorithm. It is observed that OBGWO provides better solutions and improved convergence characteristics when compared with these techniques. Further, the performance investigation of these algorithms is done for optimizing the weight vector of a CSS scheme that improves the detection probability of a PU. OBGWO scheme is then employed to study the effect of variation in the number of secondary users, sensing channel noise and control channel noise on the proposed CSS model.