Cross-layer adaptive wireless resource allocation scheme

Abstract

A new cross-layer adaptive resource allocation scheme for real-time services under a time-varying fading wireless communication system was proposed,in considevation of not only the random service arrival,limited buffers and QoS requirements at MAC layer,but also the wireless channel characteristics,and end-to-end system performance was improved the thanks to adaptive process.The algorithm(CLARA) was composed of two parts:adaptive scheduling and adaptive subcarrier-power allocation.The adaptive scheduling controlled the order in which packets from different users were served according to traffic QoS requirements,such as packet delay,buffer status as well as CSI.The second part dynamically allocated subcarriers and the corresponding transmission power to users according to scheduling results,traffic QoS requirements,such as BER and the BS power constraint.Thus,the system performance could be improved by sufficient exploiting of time diversity,frequency diversity and multiuser diversity.Simulation results show that the CLARA algorithm has about 7dB gain when the throughput,which was defined as successfully received packets in a frame,began to stablize compared to fixed subcarrier-power allocation.The performance of normalized packet delay,which was defined as the ratio of total packets delayed in simulation with the total simulation times,was also significantly improved by the algorithm.