Integrated Resource Allocation and Adaptive Antennas in Mobile Cellular Systems
Space, Telecommunications and Radioscience Laboratory
Electrical Engineering Department
The recent and continuing exponential growth of mobile communications demonstrates
great demand for such services. This ever-increasing thirst for higher capacity has fueled an interest in
finding more efficient ways of allocating the limited resources available to the system and incorporating
adaptive antenna techniques in order to enhance capacity on a per-channel-per-base-station basis. Over the
last several years, different resource allocation schemes including Dynamic Channel Allocation (DCA), Power
Control (PC) and adaptive antenna techniques have been proposed and intensively studied. Separately, all these
schemes have been shown to increase system capacity. However, little research has been done to quantify the
capacity gain of their integration and to investigate the interaction among them in realistic mobile environments.
The high complexity of the interactions renders analytical solutions intractable. We investigate this integrated
performance using large-scale Monte Carlo computer simulation.
We show that more than 20 times of capacity gain, compared to the conventional fixed-power Fixed Channel Allocation (FCA)
system, can be achieved by integrating resource allocation and adaptive antennas. With the help of Directed Retry (DR),
systems using efficient interference cancellation in the adaptive antennas can achieve an effective channel reuse factor
smaller than one. In addition, we discover that the separate capacity gains achieved by power control and adaptive antennas
are not directly additive due to the significant overlap in their ability to reduce interference. Furthermore, the relative
effectiveness of different channel allocation schemes depends on whether or not adaptive antennas are included. We also show
that due to shadowing, only about half of the first-tier co-channel users will be dominant interferers.
Moreover, when evaluating the performance of systems with adaptive antennas, it is important to include certain spatial
information about the propagation channel. Unfortunately, most commonly used models for angle spread cannot be used here
due to the extra computation burden incurred on our already computation intensive large-scale system-level computer simulation.
We propose a simplified channel model to characterize the angle spread and evaluate its impact on performance. Under this model,
we show that the impact on capacity is minimal unless the angle spread is large. Nevertheless, the angle spread does cause an
increase in the required transmit power. The system with optimal interference cancellation in the adaptive antennas achieves
the highest capacity gain, but it also suffers the largest power increase when angle spread is taken into consideration.