A resource allocation algorithm for throughput maximization with fairness increase based on virtual PRB in MIMO-OFDMA systems

This paper deals with a new resource allocation algorithm in downlink MIMO-OFDMA systems. The objective is to maximize the system throughput with respect to fairness criteria since some users may experience bad channel conditions for a long time. Known to be NP-hard, the original optimization problem is divided into two sub-problems where radio resource allocation and power allocation are performed separately. Firstly, a recursive PRB allocation algorithm is performed aiming at maximizing the system throughput. In LTE systems, 41% of sub-carriers are considered unused which introduces spectral efficiency loss. As solution, the eNodeB aggregates the unused sub-carriers by each user to construct a “virtual” PRB to be allocated to seldom served user for fairness and throughput increase. Secondly, power allocation is performed to select a more appropriate MCS.