Lower Complex Resource Allocation Methodology based on Graph Theory and Index Matrix Rearrangement for OFDMA/MIMO-OFDMA Systems

Resource allocation assigns subcarriers and power at the Base Station (BS) to different users, and is an important aspect in multiuser OFDMA downlink (DL) transmission. Various resource allocation approaches differ in the objectives and constraints in defining the problem and in the algorithms developed to solve the respective problems. Complexity of resource allocation mainly depends on the algorithm developed. This paper intends to fill gaps in literature by introducing graph based lower complex methodology to solve a per-symbol rate maximization problem for three different scenarios in which perfect Channel State Information (CSI) is available. Firstly, a Single Input Single Output (SISO) OFDMA system with Sub-band Carrier Assignment Scheme (SCAS) which supports only a single modulation scheme per user is considered. A SISO-OFDMA system with Generalized Carrier Assignment Scheme (GCAS) which supports multiple modulation schemes within a users’ sub-band is considered next, and finally a Multiple Input Multiple Output (MIMO) OFDMA system is considered. A new parameter is introduced to evaluate fairness of resource allocation, and extensive performance evaluation of developed algorithms based on average SNR values and fairness is carried out. MATLAB simulations show that algorithms perform well and allocate resources fairly even when the channel is extremely frequency selective.