Semantic Reasoning in Cognitive Networks for Heterogeneous Wireless Mesh Systems

The next generation of wireless networks is expected to provide not only higher bandwidths anywhere and at any time but also ubiquitous communication using different network types. However, several important issues including routing, self-configuration, device management, and context awareness have to be considered before this vision becomes reality. This paper proposes a novel cognitive network framework for heterogeneous wireless mesh systems to abstract the network control system from the infrastructure by introducing a layer that separates the management of different radio access networks from the data transmission. This approach simplifies the process of managing and optimizing the networks by using extendable smart middleware that automatically manages, configures, and optimizes the network performance. The proposed cognitive network framework, called FuzzOnto, is based on a novel approach that employs ontologies and fuzzy reasoning to facilitate the dynamic addition of new network types to the heterogeneous network. The novelty is in using semantic reasoning with cross-layer parameters from heterogeneous network architectures to manage and optimize the performance of the networks. The concept is demonstrated through the use of three network architectures: 1) wireless mesh network; 2) long-term evolution (LTE) cellular network; and 3) vehicular ad hoc network (VANET). These networks utilize nonoverlapped frequency bands and can operate simultaneously with no interference. The proposed heterogeneous network was evaluated using ns-3 network simulation software. The simulation results were compared with those produced by other networks that utilize multiple transmission devices. The results showed that the heterogeneous network outperformed the benchmark networks in both urban and VANET scenarios by up to 70% of the network throughput, even when the LTE network utilized a high bandwidth.