Power control in a multicell CDMA data system using pricing

Abstract

The wireless communications community has a well established understanding of voice services in contrast to the emerging data services. Traditionally, voice service quality is regarded acceptable as long as it exceeds some subjective level of signal-to-noise ratio (SIR). While this approach works well for voice, it may not be an appropriate measure of QoS for data services. In earlier work, we have applied microeconomic principles to model the QoS requirements of data services in which the terminal satisfaction (utility) is mapped onto a function of the SIR obtained at the base station (BS) and the power expended to achieve that SIR. We have investigated distributed power control in a single-cell system where each terminal maximizes its utility measured in bits/Joule by adjusting its transmit power. In this work, we extend the previous work to a multicell CDMA system. We show that an equilibrium vector of powers emerge as a result of independent utility maximizing behavior by each user under an arbitrary base station assignment. We also discuss a pricing scheme by which each terminal ends up transmitting lower power in return for increased utility. A distributed algorithm is provided by which terminals achieve their equilibrium power levels. Joint transmit power control and BS assignment that maximizes utility is also studied.