The deployment and overwhelming success of commercial cellular communication systems, based on GSM, IS-95 and recently 3G standards, encouraged a worldwide intensive research effort to assess the ultimate theoretical performance limitations of cellular systems. In general, a cellular system divides a geographical area into variant sized cells (depending on the expected number of users). In each cell there are mobile users that communicate with other users (not necessarily members of the system) through the cell-site transceiver. The cell-site terminals are connected to each other by special communication links, where the connection to the outside world is done through gateways. The communication channels included in the cellular systems can be divided into two main categories: the uplink channel, where messages are transmitted from the users to the cell-cite receiver, and the reversed downlink channel. In addition to the ambient noise, both channels suffer from various phenomena such as multi-path fading, intra- and inter-cell interferences, and Doppler shifts.

The research deals with information theoretic aspects of joint multi-cell site processing of cellular communication models in the presence of fading. Unlike conventional system models, in which each cell independently processes the signals related to its mobile users, treating other cell-sites' signals as additive noise; a novel approach, in which a finite or infinite cluster of adjacent cells are jointly processing their mobile users' signals, is studied. This approach forms a compromise between the advantage of incorporating additional information from other cell-sites on one hand, and the associated excess processing complexity, on the other. It is demonstrate, in various setups [1][2][3], that the joint multi-cell site processing approach yields a non-interference limited behavior and provides a significant performance enhancement over the conventional single cell-site processing approach. Currently, sub-optimal scheduling schemes for the cellular system downlink channel are under investigation. Different aspects of system performance, such as the system throughput, the amount and shape of the required feedback channel-state-information (CSI), and "fairness" of service between users, are evaluated for simple yet insightful cellular system models.

The multi-cell processing approach seems like a strong candidate method for increasing cellular systems performance. Along with other performance increasing methods, such as the use of antenna arrays (MIMO), it is most likely to be included in some form by future cellular standards.