Performance Study of Distributed Power Control Algorithms Under Time-Delays and Measurement Uncertainty

This paper is aimed to evaluate different distributed power control algorithms for code division multiple access (CDMA) systems. The evaluation scenarios consider three practical aspects: i) time-varying channels, ii) feedback time-delays, and iii) uncertainty in the measurement of the signal to interference-noise ratio (SINR). First, dynamic power control is formulated in its general form as a tracking problem for an objective SINR and by assuming a feedback structure. Next, seven distributed power control algorithms are described: fixed step, Foschini-Miljanic, proportional-integral-derivative (PID) control, robust control H_infinity, robust Smith predictor, variable structure control and linear quadratic Gaussian (LQG) control. The analysis and implementations were performed in simulation by using MATLAB. In our evaluation, the LQG control outperformed other strategies in terms of SINR tracking precision, robustness and complexity.