Quality Queue Management for Future Wireless Networks

Abstract

Wireless networking protocols and mobile devices are key contributors to the ever-growing demand for realtime services. These large throughput services are often prioritised and managed using specialised traffic policies. Today’s wireless networks will be replaced by their newer, faster, less expensive counterparts in the very near future. These Very High Throughput networks will deliver a significant increase in throughput and a reduction in effective cost. Realtime services are sensitive to packet loss and delay, and their successful delivery demands a high standard of Quality of Service (QoS). The quality perceived by the final user is measured as the Quality of Experience (QoE). Hence, Telephone Companies (TELCOs) aim to provide realtime services with the best QoE possible. This thesis aims to extend the provision of QoE to realtime services on future wireless networks. The first contribution of this work is a novel metric, the expected Quality of Service (eQoS), that estimates QoE on future wireless networks using combinatorics. eQoS provides an almost instantaneous estimate of the QoE at the node and is inferred from the traffic flows crossing the node. It provides a statistical estimate of the end user perception of the network’s quality. The second contribution is a theoretical model to capture the probability of successful and unsuccessful channel access on a future wireless network. This theoretical model uses combinatorics to calculate the probabilities that (i) a collision occurs, (ii) a packet is transmitted and (iii) the channel is idle. The third contribution is to converge eQoS and the theoretical model into the Quality Queue Management (QQM) system for future wireless networks. QQM is designed using fuzzy logic controllers, it manages wireless resources with the goal of providing real time traffic flows with the best QoE possible in the network. QQM incorporates eQoS and is a robust, efficient, scalable means of managing realtime services on future wireless networks. The novelty of the QQM system is to use eQoS to manage network parameters. The QQM system optimises the throughput between the flows, manages the queues and provides the maximum number of realtime services with the best quality possible.