Capacity and Fairness of Distributed Antenna Systems in Multi-Cell Environments with User Scheduling, Power Control and Imperfect CSI

Abstract

Distributed antenna systems (DASs) have attracted
lots of attention as a method to improve the performance of future
wireless networks. Capacity analysis and optimum power allocation
for the physical layer of DASs have been extensively explored
in the literature. However, the study of cross-layer issues, such
as channel-aware scheduling and fairness evaluation, is relatively
scarce. This paper partially fills this gap by addressing the downlink
capacity and fairness analysis of a DAS assisted by joint
user scheduling and transmit power control. The algorithm is
evaluated in a multi-cell environment assuming imperfect channel
state information. The algorithm exploits the spatial diversity
provided by the distributed antennas in order to schedule over the
same frequency band a different user attached to each one of the
distributed nodes. The objective is to optimize the power levels
to control the interference created between the transmissions
of the selected users, thereby multiplexing as many of them as
possible while maximizing capacity. To achieve this goal, a sumrate
capacity optimization with respect to the power levels is
here proposed by using a gradient descent iterative technique.
The result is the set of optimum user-antenna pairs to be
scheduled and their optimum power levels. Inter-cell interference
is calculated by reusing the results of previous simulation runs
in the transmission parameters of outer-cells, thereby efficiently
replicating system-level behavior. The algorithm is also evaluated
in terms of fairness by using the spatial distribution of the user
capacity. Capacity and fairness of the algorithm considerably
outperform previous solutions, particularly in scenarios with
good line-of-sight and optimum node location.