Performance evaluation of MIMO downlink WiMAX for different schedulers

WiMAX performance evaluation considering different scheduling techniques has received considerable attention by WiMAX researchers and also WiMAX operators. The standard does not define MAC scheduling although it affects the overall QoS performance of users as well as the achievable total cell goodput. The primary goal of this paper is to compare the performance of maximum total goodput including the average delay for UGS (Unsolicited Grant Service) users employing channel-aware and queue-aware downlink schedulers for a group of users. The paper initially considers the goodput behaviour and its peak across different modulation and coding schemes for SISO and MIMO (both Space Time Block Code (STBC) and Spatial Multiplexing (SM)) connections for a single user. The achievable maximum goodput is shown to be between 94.5% and 97.0% of the theoretical data rates. The analysis is then expanded to multiple users considering Proportional Fair (PF) with different observation window, t
c
, Greedy and also Weighted Fair Queuing (WFQ) schedulers to serve downlink QoS flows. The PF with a higher t
c
= 50K exceptionally outperforms WFQ as well as the PF with a smaller t
c
= 500 for a higher number of users and shows a greedy like performance. The channel-aware scheduler generally performs better than the queue-aware scheduler.