The 1-Byte Wireless QoS Solution: Scheduling Algorithms for QoS over Wireless Channels

For bursty traffic with a large peak-to-average ratio and a stochastic channel, is it possible to minimize the response time of every frame while maximizing the effective channel utilization and maintain fairness? This is the question fee resolve in this paper. In wireless networks with a single shared channel, channel access is a core issue for flows with throughput and timeliness requirements on the uplink. This paper presents a detailed MPEG-4 packet-level simulation comparing two Medium Access Control (MAC) architectures and four scheduling strategies that cover a range of classes: Dynamic-TDMA, Proportional Share Algorithms, Real-Time Scheduling Algorithms, and Size-based Scheduling Algorithms. Our results demonstrate that only a single byte is necessary for Quality of Service (QoS) support in infrastructure-based wireless networks. Our adaptation of the Shortest Remaining Processing Time (SRPT) algorithm, Fair-SRPT, is a superior scheduling algorithm and outperforms other algorithms in terms of QoS performance, channel utilization efficiency and response time under all utilization levels and error rates. In addition, a MAC protocol based on dynamic superframes and group polling scales well with higher data rates and under all error rates. An attractive feature of the proposed approach is that it can be implemented with no modifications to the IEEE 802.11e standard. We have extended the ns-2 network simulator to model centralized wireless MAC protocols (e.g. IEEE 802.11e and high-rate 802.15.3), and a modular framework to plug-in scheduling algorithms.