Analytical Modelling and Performance Evaluation of a Prediction based EH-Cooperative CRN under Erlang Distribution

Abstract

Underutilization of the radio spectrum and energy scarcity are the twin concerns that need to be resolved for the upcoming generation of wireless communication systems. The wireless communication networks today thrive for attaining an improved spectrum utilization and better energy scheduling. Cognitive Radio Networks (CRN) incorporating energy harvesting proves to be a viable solution. One of the main objectives of a cooperative CRN is to improve the spectrum efficiency and to provide an improved quality of service to the primary network. Communication nodes with cognitive capability and incorporated with energy harvesting ability (EH-CRN) can harness energy from both RF sources and non-RF resources. The functionality of a network model greatly relies on the type of distribution function used. In this work, we present Erlang distribution in order to describe the Primary User (PU) activity model and assessed the performance of prediction based EH-CRN under cooperative scenario. The result shows that Erlang family offers more tractability when modelling network traffic than the exponential family. In real time scenarios, the Erlang family seems to offer better flexibility in accommodating the distribution to the real time data. In this paper, the effect of number of frames delivered, prediction error and sensing period along with the collision constraint, splitting factor on the average network throughput and harnessed energy are examined. Numerical expressions for detection probability, normalized throughput and harvested energy are derived using OR fusion rule in a cooperative EH-CRN. A thorough comparison of the results are also provided in this paper. © 2019 IEEE.