Abstract
The scientific field of information theory provides a mathematical framework which aims to quantify the maximum achievable data rate over a communication channel. The underlying mathematical concepts to predict the capacity of a single communication link were developed by Shannon more than half a century ago. The first important attempt to study the capacity of a cellular system was carried out in the last decade. That work by Wyner introduced the concept of Base Station co-operation. The available information theoretic findings are not directly usable to provide realistic estimates for the capacity of practical systems since they cannot model the effect of changes in physical parameters in the environment like path loss exponent and the distance between the adjacent cell sites. The objective of this work is to extend the known formulations for the capacity of the Gaussian Cellular Multiple Access Channel (GCMAC) with joint processing by incorporating path loss and other channel conditions that represent a real communication system and thus to provide a more realistic analytical upper bound for the capacity of the wireless cellular network. In this direction, the available GCMAC model with joint processing and small scale fading is extended by adding multiple antennas at the Base Stations (BSs) and User Terminals (UTs) and by removing the assumption that UTs are co-located at the BS position. Since it is concluded that the available GCMAC model is not sufficient to analyse more complex systems a new model is built which enables the evaluation of the achievable capacity of more realistic systems. Based on this model a new geometric and mathematical model is also developed which enables the study of fairness and user rate distribution in joint processing systems with small and large scale fading. The analysis provides several achievable capacity formulae and some very useful insights on the behavior of the sum rate as well as the fairness, when certain system parameters change, are derived. The formulae can be used to evaluate the achievable sum rate of practical systems employing full BS co-operation, given the parameters that control that capacity.