Resource allocation in WiMAX relay networks

Abstract

WiMAX is a promising wireless technology to provide high-speed, reliable communications in large areas. Relay stations can be deployed in a WiMAX network to extend its coverage and improve its capacity. Orthogonal Frequency Division Multiplexing Access (OFDMA) enables better channel utilization and exploits the benefits of channel diversity and user diversity. A smart adaptive antenna provides multiple Degrees of Freedom (DOFs) for intended communications and interference suppression, which results in more efficient spatial reuse and higher throughput. In this dissertation, we discuss how to combine those aforementioned technologies in a WiMAX relay network (WRN), and study the routing, scheduling, channel assignment, and antenna DOF assignment problems. It has been shown by previous research that the performance of wireless scheduling algorithms usually depends on the interference degree. Therefore, we study the interference degree in WiMAX networks and show that it is at most 4 in any 2-hop WRN and at most 14 in any general WiMAX network. Next, we consider routing and scheduling in WRN with smart antennas. We formally define the Interference-aware Tree Construction Problem (ITCP) for routing, which offers full consideration for interference impact and DOF availability. We then present an algorithm to optimally solve it in polynomial time. As for scheduling, we first present a polynomial-time, optimal algorithm for a special case in which the number of DOFs in each node is large enough to suppress all potential secondary interference. An effective algorithm is then presented for the general case. We also study a scheduling problem for throughput maximization in OFDMA-based WRN with consideration for multi-user diversity, channel diversity and spacial reuse. We present a Mixed Integer Linear Programming (MILP) formulation to provide optimum solutions. Furthermore, we show that both the simple greedy algorithm and our proposed weighted-degree based greedy algorithm have approximation ratio of 1/5 for 2-hop WRN and 1/15 for any WiMAX networks. In addition, we present 3 other efficient algorithms, namely, the maximum weighted independent set (MWIS) algorithm, the sequential knapsack algorithm and the LP rounding algorithm. Extensive simulations are conducted to evaluate the performance of all proposed algorithms.