AbstractsComputer Science

Among different wireless technologies, wireless ad hoc networks and cellular networks account for a large part of wireless communications. In this work, we take an extensive study at the management of the two of very important resources in both wireless ad hoc networks and cellular networks: power and code (bandwidth). In the first part, we focus on the power management in IEEE 802.11 ad hoc networks. We first aim at the mismatch between the power management protocol and the ever-increasing transfer rate. We propose modifications on the power management operations of IEEE 802.11 which introduce very little overhead but bring a big boost on the performance in both power efficiency and data throughput by combining a scheduling algorithm with the power management. Then, we concentrate on the new challenges faced by the power and synchronization protocols when pushing the IEEE 802.11 ad hoc mode into the multi-hop environment. We study the relationship between clock synchronization and power management in Mobile Multi-hop Ad Hoc Network (MANET). After reaching the conclusion that clock synchronization is vital for not only efficient power management, but many other network operations as well, we present a protocol to generate a globally synchronous system from synchronized sub-networks. We discuss the correctness of the protocol, and show the power efficiency brought by such a synchronous system in simulations. Lastly, we present a framework, which takes advantage both of the two main approaches in energy conservation, i.e. power management and power control, to maximize power-saving. Because the goals that each scheme seeks contradict each other, we study the balance and the trade-off between them, and use them as the guideline on building our framework. In the second part of this work, we shift to the code (bandwidth) assignment for multimedia traffic in Code Division Multiple Access (CDMA) networks. We propose several algorithms to handle the jitters of compressed video transmissions, and try to get a better playback quality while reducing the effects which video traffic has on other users in the system. Our simulations show the improved performance from those algorithms, especially at heavy traffic scenarios.