The deployment and overwhelming success of
commercial cellular communication systems, based on GSM, IS-95 and recently 3G
standards, spurred a worldwide intensive research effort to assess the ultimate
theoretical performance limitations of cellular systems. In general, 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 separately processes the signals related to
its mobile users, treating other cell sites related signals as additive noise;
we use a novel approach in which a finite or infinite cluster of adjacent cells
are jointly processing their mobile users signals. We demonstrate, in various
setups, 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. In
order to enable an analytical framework, a class of simple system setups based
on a model introduced by Wyner in 1994 is used. Although this simple model is
hardly realistic, it serves as a tractable model providing considerable insight
into complex and analytically intractable real-world cellular communications.
Shannon-theoretic limits on the achievable throughput for the uplink channel of
Wyner's model are presented. Both TDMA and Wide-Band transmission schemes are
considered. For both schemes, performance enhancement as compared to single
cell-site processing is demonstrated. In addition, it is observed that under
certain conditions, the presence of fading increases the maximum reliable equal
rate of the system. Next, chip-interleaved randomly spread DS-CDMA scheme,
employed in the uplink channel of Wyner's model, is considered. Focusing on the
asymptotic setup in which both the number of users per cell and the processing
gain go to infinity while their ratio converges to some finite constant, the
spectral efficiencies of both the optimum and linear MMSE joint multi-cell
receivers are investigated. This asymptotic setup allows the use of recent
results from the theory of random matrices. Performance enhancement as compared
to single cell-site processing is demonstrated. Finally, the downlink channel
sum-rate capacity, of a Wyner like setup, is considered. The duality between
the Multiple Access Channel and the Broadcast Channel is used to derive an
analytical expression for the sum-rate capacity of the system in the non-faded
setup. Introducing fading, a multiuser diversity gain of the order of O(log
log(K)) is observed for the sum-rate capacity when the number of users per-cell
K is large. The impact of restricted cell-site cooperation is also considered.
