Impact of user mobility on resource allocation schemes in cellular radio systems

Next generation wireless cellular radio systems are being designed to provide anytime, anywhere communication capabilities to serve a range of applications. The ability to support mobility is a key reason for the increasing demand for such systems. To accommodate this demand, efficient resource allocation schemes that can operate over the harsh wireless channel environment need to be devised. User mobility has a significant influence on the design and performance of these schemes. The focus of this dissertation is the analysis of the impact of mobility on such resource allocation schemes.; What impact mobility has depends on the scheme under consideration. We first analyze the impact of user mobility on the performance of a link adaptation scheme that employs the recently proposed no-transmission mode. In this scheme, users adapt their modulation and coding for transmitting data packets based on their estimates of the link condition and suspend transmissions when link quality is very poor. Based on a simplified system model, we derive expressions for the system performance as a function of the basic, system-defining parameters. We show that for a stable system, the channel correlation, a function of user speed and feed-back delay of estimates, is an important factor that determines the optimal link adaptation thresholds. We then study a packet based multiple access scheme called Packet Reservation Multiple Access (PRMA), which can simultaneously handle the different traffic requirements of periodic, delay intolerant (voice) and bursty, delay tolerant (data) users. An approximate technique is developed to analyze the impact of user mobility as well as channel fading and interference-induced packet errors on PRMA. Both these effects lead to a premature loss of reservation and, consequently, more dropped packets for voice users. Finally, we look at dedicated channel assignment schemes that assign an entire channel to a user for the duration of his conversation. We investigate heuristic prediction based techniques that take into account mobility traffic statistics to modify the new call access criteria. This is done so as to introduce prioritization for hand-off requests in hitherto unprioritized channel assignment schemes.