Abstract

The use of wireless services is rapidly spreading around the world and many of the world population no longer know how to cope without their cell phones; the feel of always being connected offers a great sense of flexibility and security. So far, voice has been the primary wireless application. However, with the Internet continuing to influence our daily lives, the demand for wireless data is extensively increasing. Already, in the countries that have cellular-data services readily available, the number of cellular subscribers taking advantage of data services has reached significant proportions. In this thesis, we investigate resource management techniques for cross layered star and mesh wireless data networks. In particular, we investigate several aspects related to resource management techniques over the reverse packet data channel in cdma2000 1xEV star networks. We provide an upper bound for the reverse packet data channel throughput as a function of the number of mobile stations that are allowed to transmit instantaneously on each time slot. We also provide a lower bound for the average sector throughput based on the number of users per sector and propose several autonomous rate assignment, and scheduling techniques that provide a significant throughput improvement relative to other published techniques. We also develop analytical models for lowest-rate-first, highest-rate-first priority scheduling techniques, and two round-robin fair scheduling techniques over the reverse data channel in cdma2000 1xEV star networks. For these four scheduling techniques, the distribution of the mobile stations among the possible data rates is modelled as a Markov process. An analytical expression for the steady state system throughput is derived from the steady state distribution of the above Markov process. The above model is extended to evaluate the performance of the cross layered design between the hybrid ARQ, rate assignment, and time slot scheduling over the reverse packet data channel in cdma2000 1xEV. Expressions for the steady state system throughput and file transmission delay are derived from the steady state distribution of this model. Fountain codes are a class of erasure codes with the property that a potentially limitless sequence of encoding symbols can be recovered from any subset of size equal to or slightly larger than the number of source symbols. In this thesis we relate the parameters of the higher layer Fountain codes to those of the physical layer codes to present a cross layered coding technique. Based on the bit error rate of physical layer codes, the upper layer Fountain code's parameters are designed to obtain a prespecified performance. The performance of the proposed cross layered coding technique is found to be comparable to that of physical layer based Hybrid ARQ technique. As an application we studied the performance of WiMAX backhaul mesh networks where cashing is allowed at the service stations and proposed cross layered coding technique is used as its main coding scheme. In particular, the performance of the above network is modelled as a Markov process and analytical expressions for the steady state system performance are derived from its associated steady state distribution. All the above analytical models are validated through simulations.