A multi-layer hyper-graph routing with jamming-awareness for improved throughput in full-duplex cognitive radio networks

Abstract

There are differences between routing in traditional and in Cognitive Radio Networks (CRNs). The availability of channels, periodic spectrum sensing, jammers activity, are some of these differences, etc. CRNs may contain a large number of Internet of Things (IoT) devices, hence, modeling such networks with a flat graph model generates a huge graph size. Thus, the motivation behind this work is clear bearing in mind that modeling such networks with a Multi-layer Hyper-Graph (MLHG) is the idea presented in this paper, in which each hyper-edge represents a group of CR devices, and each layer represents a licensed channel. In this paper, the main aim is to consider four key factors to maximize end-to-end network throughput, which are, transmission rate, licensed users activity, jammers activity, and needed number of time sharing. In literature, considering both jamming activity and time sharing for routing exists in few papers till now. This work proposes and develops three cutting-edge routing protocols that are aware of the four aforementioned factors, namely, Jamming-Activity-Sharing-Rate-Aware protocol, Jamming-Activity-Rate-Aware protocol, and Jamming-Activity-Sharing-Aware protocol. We compare these protocols with three recent proposed protocols in CRNs. Simulation results indicate that when jammers activity and time sharing are considered, network throughput is significantly enhanced. Specifically, our proposed jamming-aware protocols significantly outperform the existing three jamming-unaware protocols by up to 127%.