Research On Dynamic Resource Allocation For OFDM Systems

In order to achieve high speed, high efficiency and low power requirements of the modern wireless communication systems, dynamic resource allocation technology becomes more and more important. Compared with the static resource allocation, dynamic allocation's performance is better in fully utilizing system resources.In OFDM systems, the wireless channel is divided into a number of orthogonal sub-carriers, and the high speed data are assigned to those sub-carriers to transmit with the relatively low transmission rate. OFDM technology, due to its high ability to combat multi-path fading and support for a high data rate (HDR), has been widely adopted in many wireless systems, such as IEEE 802.11a/g wireless LAN standard and IEEE 802.16a/d/e wireless MAN standard.Each sub-carrier of OFDM systems can use different modulation, and also can be arbitrarily assigned to different users to achieve multiple access. This results in the higher flexibility of the system. Making use of those flexible multi-carrier characteristics of OFDM systems, we can allocate power, bits and sub-carriers for OFDM systems dynamically, which will improve spectrum utilization and system capacity.The principle and implementation of OFDM technology are briefly introduced at the beginning of this thesis. Then the description of convex optimization problem for the dynamic resource allocation of OFDM systems is analyzed. Last but not the least, the suboptimal, low complexity and high proportionality algorithms are studied for OFDM systems, including water-filling algorithm, greedy algorithm and Chow algorithm for single user OFDM systems and united and sub-step algorithms for multi-user OFDM systems.The main achievements of this thesis are summarized as follows. A low complexity suboptimal algorithm, based on study of water-filling algorithm, greedy algorithm and Chow algorithm, is proposed for single user OFDM systems. Simulation results show that the proposed algorithm achieve almost the same performance as the optimal algorithm with lower complexity, which is approximately one-tenth of that of water-filling algorithm and one-hundredth of that of greedy algorithm. A simple and intuitive proportional fairness factor is defined for the condition that proportional user rates are required. A novel sub-step allocation algorithm for multi-user OFDM systems is proposed. Different from normal sub-step algorithm, the proposed algorithm assigns bits at first and then allocates the sub-carriers with the bit allocation result. Simulation results show that the proposed algorithm achieve a good balance in proportional fairness, complexity and system capacity.