Research On Scheduling And Control In Multi-service Wireless Cellular Networks

As the widely deploying of mobile applications, the demand for providing multi-service in wireless cellular networks becomes more imperative. The next generation of networks is expected to eventually carry multimedia traffic - voice, video, images, or data, or combinations of them. Different traffic has diversified quality of service (QoS) requirements, for example, the requirements of time delay, error bit rate, and transmitting rate for different traffic are differentiated. Future wireless communication systems trend to provide multi-services to mobile users. There are two main objects for wireless network design, one is to guarantee QoS requirements of different traffic, and the other is to make resource utilization maximized. Because of the unique characteristics of wireless channel, such as location dependent, channel error and time varying channel states, the problems faced by future wireless traffic control mechanisms are greatly different from traditional wireline networks and much more challenging.This dissertation studies some key issues in traffic control including scheduling, admission control and congestion control mechanisms for multi-service wireless cellular networks.As the first step of the research work, the author studies scheduling algorithm for multi-service wireless cellular networks. In chapter 2 of this dissertation, the author focouses on opportunistic scheduling with multiple QoS constraints for downlink in multi-service wirless cellular networks. For different QoS reqirements, the author proposes some differnent opportunistic scheduling algorithms. The proposed algorithms optimize the system throughput, while guaranteeing multiple QoS reqirements for differnt traffic. For multiple long-term QoS constraints, the author proposes a QoS normalized based opportunistic scheduling algorithm. For short-term QoS constaint of multimedia traffic, the author proposes a real-time multimedia opportunistic scheduling algorithm.In chapter 3 of this dissertation, the author focouses on scheduling problem that maximizes system throughput, while guaranteeing multiple QoS constraints for multi-serice HSUPA system. Fisrt, the author gives feasible rate region with the character of HSUPA. Then, the author proposes a fair opportunistic scheduling for multi-service HSUPA system. The proposed algrithm exploits channel condition to improve the system throughput, while guaranteeing fairness. Finally, the author proposes a multi-service opportunistic schedling algorithm, with the coexitence of non-real-time data traffic and real-time multimedia traffic in HSUPA system.With the development of network technonigy and the requirement of real-time application, protocol with cross-layer design exchanges information between defferent layers to optimize network performance. In chapter 4 of this dissertation, the author studies cross-layer shceduling algorithm with traffic information. The author uses traffic information in cross-later design, such as file size information and mobility information. First, the author proposes a fair scheduling algorithm with file size information for uplink. Then, the author proposes a cross-layer scheduling algorithm with file size and mobility information for downlink.In chapter 5 of this dissertation, the author studies CAC (call admission control) for multi-service CDMA cellular networks. First, the author proposes a compensation policy based CAC algorithm and develop a Markov model for the proposed algorithm. The proposed algorithm uses degradation policy to accommodate new users, when system is overload and use compensation policy to improve fairness performance when systm has residual bandwidth. Then, the author proposes a dual threshold admission control algorithm in voice/data integrated CDMA cellular networks.Chapter 6 of this dissertation studies the problem of jointly cross-layer design of congestion control and opportunistic scheduling. The optimal design aims to maximize the system performance, while guaranteeing multiple QoS constraints for wireless multi-services networks. The author proposes a congestion control algorithm with queue and channel information to avoid under utilization and packet losses. And the author incorporates the queue information to opportunistic scheduling to stabilize the queues in the network.Chapter 7 of this dissertation summarizes the simulation software. Simulation software is developed to verify and evaluate the proposed algorithms in this dissertation. Chapter 5 presents the main parts of the software flow diagram.Chapter 8 of this dissertation concludes this dissertation, presents some suggestions for future research on some issues.