Cross layer routing and scheduling for multi-channel Wimax mesh networks

Abstract

Broadband wireless networks are becoming increasingly popular due to their fast and inexpensive deployment and their capabilities of providing flexible and ubiquitous Internet access. Due to the limitation of shared resources in wireless mesh network such as bandwidth, spatial reuse is introduced for concurrent transmissions. The simultaneous transmissions face many challenges regarding interference on the ongoing transmission. To maximize the network performance of mesh networks in terms of spatial reuse, it is essential to consider a cross-layer for resource allocation in different layers such as the routing network layer, the scheduling resource allocation Media Access Control (MAC) layer and physical layer. Therefore, this thesis focuses on improving the spatial reuse for resource allocation mechanism including routing tree construction by taking into consideration the reliable path, channel assignment and scheduling algorithms. Firstly, a Fuzzy based Constructed Routing Tree (FLCRT) is proposed to incorporate fuzzy logic with routing to enable cognitive capability in packet forwarding for uplink or downlink communication. Secondly, the link-aware routing path is proposed to satisfy the connection lifetime and better routing stability for successful requirements of transmission using multi sponsor node technique. Then, a better understanding of reliability analysis is pursued in the context of homogeneous wireless network. Ultimately, heuristic resource allocation including channel assignment and centralized scheduling algorithms are proposed based on the cellular learning automata to enhance the number of concurrent transmissions in the network by efficiently reusing the spectrum spatially. The attempt of heuristic resource allocation algorithms is to find the maximal number of nodes that could transmit data concurrently. The numerical and simulation results show that FLCRT, Learning Automata Heuristic Channel Assignment (LAHCA), and Learning Automata Heuristic Centralized Scheduling (LAHCS) perform better in terms of scheduling length, channel utilization ratio, and average transmission delay as compared with the existing approaches. The proposed FLCRT scheme with respect to the number of subscriber station (SS) nodes performs better in decreasing the scheduling length, average transmission delay, and channel utilization ratio by 38%, 19%, and 38% compared with Interference-Load-Aware routing. LAHCA algorithm improves the number of channels in comparison with random selection algorithm by 8%. LAHCS algorithm using multi channels proposed by LAHCA can reduce the scheduling time, average transmission delay as well as enhance channel utilization ratio versus number of SS nodes by 7%, 8%, and 6% respectively compared with Nearest algorithm in higher traffic demands.