A load adaptive subcarrier and bit allocation algorithm for non-real time services in an OFDMA system

Orthogonal Frequency Division Multiple Access (OFDMA) is an efficient access technique that allocates different subcarriers to multiple users where each subcarrier is loaded with a number of bits belonging to that user. Allocating subcarriers and bits to multiple users has been the focus of many algorithms that attempt to maximize the throughput or minimize the total transmit power used. Non-real time applications such as web browsing are dominant applications in most of the emerging wireless networks. A key performance measure of such applications is the average packet delay provided by the network. In this paper, we propose a novel Load Adaptive (LA) algorithm for non-real time services that allocates the subcarriers and bits to each user according to its individual load and the available transmit power so as to minimize the overall average packet delay for all users. The algorithm takes into account queueing theory and the quality of service requirement of each user. Numerical results show that the overall average packet delay of the proposed algorithm is significantly reduced compared to other algorithms.