UC San Diego

Driven by the improvements in wireless technology and proliferation of available content, wireless networks have become extremely popular. Having evolved significantly in recent years, wireless networks now serve as a gateway to a diverse set of rich applications and functionality. Applications range from standard Internet services to real -time services, such as multimedia video and voice over IP. Although wireless networks have become popular, there are many challenges that remain. In addition to the inherent problems of wireless connectivity, which include varying channel conditions, interference, and propagation loss, upcoming networks must support applications and devices that place a greater demand on network resources. Users will also have higher requirements for future wireless networks, expecting guaranteed problem-free wireless access and a better overall user experience. This dissertation addresses several key problems in wireless networks: the need for improved resource efficiency, both in terms of energy and network resource allocation, the lack of Quality of Service (QoS) support, and the inability of applications to scale based on channel or device limitations. This dissertation proposes network and application layer adaptation techniques to enhance resource efficiency and Quality of Service. It introduces the concept of dynamic adaptations at various layers of the network protocol stack and considers four protocol adaptation techniques. The first targets improving energy efficiency by dynamically changing link layer parameters. The second focuses on ensuring differentiated channel access by adapting related channel parameters. The third technique uses network scheduling to ensure appropriate network resource allocation and improved channel access. The final adaptation technique investigates scaling at the application layer to improve network resource efficiency and achieved application quality. The presented experiments demonstrate that the developed network and application layer adaptation techniques can significantly improve the wireless user experience. By providing a more efficient utilization of energy and network resources, as well as providing QoS guarantees, the protocol adaptation schemes show remarkable improvement in several metrics compared to existing approaches. The adaptation techniques described in this dissertation will enable wireless networks to support higher quality applications for a greater and more diverse set of users