| dc.description.abstract | In this thesis, we study the performance of cellular networks under various resource allocation schemes. Radio resources in cellular networks need to be managed efficiently in order to achieve good system performance. In channelized cellular systems (e.g., time division multiple access (TDMA) and frequency division multiple access (FDMA) systems), the radio resources to be allocated to different users are channels, which could be time slots or carrier frequencies or combinations of both. We study the performance of dynamic channel allocation in channelized cellular systems with voice?only traffic. The performance measure of interest is the blocking probability of calls. In code division multiple access (CDMA), the radio resources are the spreading codes. Since CDMA systems are interference?limited, allocation of resources could be based on code availability as well as interference conditions. We study both voice?only and mixed voice/data cellular CDMA systems with admission control based on interference measurements. The performance measures of interest are the outage probability for voice calls, average system throughput, and mean delay for data traffic. System performance in cellular CDMA depends on the transmit powers and transmission rates of users in the system. By suitably allocating transmit powers and transmission rates to users, it is possible to utilize system resources efficiently. We study the power and rate allocation problem in cellular CDMA.
In the first part of the thesis, we study the performance of channelized cellular systems with distributed dynamic channel allocation (DCA). We present an analytical approach to compute the blocking probability in cellular systems with DCA. Most approaches in the current literature to evaluate the blocking probability in channelized cellular systems with DCA do not take interference measurements into account. We model the channel occupancy in a cell by a two?dimensional Markov chain, which can be solved to obtain the blocking probability in each cell. The transition rates in the Markov chain take into account both channel availability and interference measurements. We first apply our analytical model to linear highway systems and then extend it to two?dimensional cellular systems. We show that for linear highway systems, the distributed DCA scheme performs similarly to the centralized DCA scheme in terms of blocking probability. However, for two?dimensional cellular systems, the improvement in performance is significant, and the reduction in blocking probability in systems with distributed DCA is by one to two orders of magnitude compared to systems with centralized DCA.
In the second part of the thesis, we analyze the performance of a signal?to?interference ratio (SIR)?based admission control strategy on the uplink in cellular CDMA systems with both voice and data traffic. Most studies in the current literature to estimate CDMA system capacity with both voice and data traffic do not take into account admission control based on SIR constraints. We present an analytical approach to evaluate the outage probability for voice traffic, the average system throughput, and the mean delay for data traffic in a voice/data CDMA system that employs an SIR?based admission control. We make two main approximations in the voice?call outage analysis-one based on the central limit theorem (CLT) and the other based on Fenton抯 method. We apply the Fenton抯 method approximation to compute the retransmission probability, the mean delay for data traffic, and the average system throughput. We show that for a voice?only system, a capacity improvement of about 30% is achieved with SIR?based admission control compared to code?availability (CA) based admission control. We also show that for a data?only system, an improvement of about 25% in both Erlang capacity and mean delay performance is achieved with SIR?based admission control compared to CA?based admission control. For a mixed voice/data system with 10 Erlangs of voice traffic, the improvement in mean delay performance for data is about 40%. Also, for a mean delay of 50 ms with 10 Erlangs of voice traffic, the data Erlang capacity improves by about 50%.
In the third part of the thesis, we present a utility?function?based approach for power and rate allocation in cellular CDMA. On the downlink, we study power allocation with and without pricing. We propose a pricing policy to charge users. We present the power allocation to users that maximizes the effective system throughput while incorporating pricing. We also study the asymptotic behavior of the system, i.e., the behavior of the system with a large number of users and a large amount of resources. We derive an expression for the asymptotic spectral efficiency of the system, where spectral efficiency is defined as the effective system throughput per unit system bandwidth. On the uplink, we present a class of utility functions for power and rate allocation. Most approaches in the current literature on optimal power and rate allocation consider utility functions that require complex dynamic programming approaches to solve. Here, we propose a class of utility functions that depend on the user rate and transmit power, and obtain closed?form expressions for the optimal powers and rates. We maximize the utility function such that the SIR and delay requirements are satisfied for all users. When SIR constraints alone are considered, the proposed utility function results in approximately the same utility for all users, irrespective of their locations. When delay constraints are also taken into account, system throughput increases, but at the cost of non?uniform utility among different users. | |
