Opportunistic scheduling and cross-layer design exploiting multiuser diversity in wireless networks

Abstract

With the development of advanced wireless communication technologies, the demand for high throughputs and quality of service (QoS) guarantees has been explosively increasing in modern wireless networks. However, available radio resources are scarce and limited. To efficiently utilize the available radio resources and provide high throughputs and QoS guarantees, in this dissertation, we consider adaptive transmission, opportunistic scheduling and admission control. Different from wired channel, wireless channel is time-varying (i.e., fading). To increase spectral efficiency of the time-varying wireless channel, adaptive transmission operates at the PHY layer of a wireless network and adapts constellation sizes and/or coding rates to the time-varying wireless channel. Due to the time-varying nature of the wireless channel, different users in a wireless network experience peaks in their channel quality at different times, from which multiuser diversity comes. Opportunistic scheduling exploits multiuser diversity by allocating the shared wireless medium to only the user with the best channel condition at each time, and can maximize the total information-theoretic capacity of a wireless network. The capacity advantage of opportunistic scheduling can be obtained at the penalty of the following two problems. The first problem is related with the fairness: opportunistic scheduling can cause unfair scheduling by letting users with the strongest channel conditions in average monopolize the shared wireless medium. The second problem is related with the channel state information acquisition: opportunistic scheduling necessitates the instantaneous channel qualities of all users, which are estimated by each user and then fed back to the scheduler. Hence, as the number of users increases, the feedback load becomes significant, and feedback bandwidth expended by the non-scheduled users gets wasted. Therefore, there is a strong motivation for the design of opportunistic...