Token Bucket Fair Scheduling Algorithm with Adaptive Rate Allocations for Heterogeneous Wireless Networks

In this paper, we address the problem of uplink multi-traffic sources scheduling in heterogeneous wireless networks (HWNs) under real time traffics that requires fairness constraint and strict per-packet delay bounds and under non-real time traffics as well. The wide spread of HWNs, such as integrated of WiMax–WiFi–Relay wireless networks, enforces networks operators to design an efficient radio resource management (RRM) in order to tradeoff between network resource utilizations and user satisfactions. RRM concerns with traffic scheduling and admission control amongst different traffic sources to guarantee the Quality of Service. This paper participates to RRM developments by proposing and analyzing an efficient packet scheduling scheme with two modes of operations for the uplink channel of WiMax–WiFi–Relay station HWNs. These two modes of operations are known as Fixed Shared Flow and Adaptive Sharing Flow (ASF). The main aim of this scheduling is to allocate shared WiMax resources in a fair manner and to balance between maximizing the bandwidth utilization and users satisfactions. The performance bounds in terms of delay and throughout are presented. Then, the performance figures of the two modes of operations of the proposed scheduling scheme are compared. The results show that bounded delay can be provisioned for direct WiMax users, users via WiF and users via Relay Stations sharing the same WiMax bandwidth using both operations. Moreover, the simulation results show that the ASF scheduling scheme improves both system utilization and average delays.