Prediction-based power control for distributed cellular communication networks with time-varying channel uncertainties

Abstract

Fast changing radio channels in a CDMA based cellular network have detrimental effects on the control efforts required to regulate the Signal to Interference plus Noise ratio (SINR) to the desired level, especially for highly mobile terminals (MTs). The motivation behind introducing a prediction-based power control algorithm is to meet the problems associated with rapid changes in the channel gain (by orders of magnitude between power update intervals) influenced by fading and exacerbated by the MTs coming out of the ‘deep faded’ zone. For a fast fading channel, a reliable prediction of the channel coefficient is required for accurate control design. For this purpose, we propose to use a linear prediction filter to estimate the channel fading parameter, and this information is fed to the controller. The controller uses local SINR measurements from the current and neighboring cells to maintain the SINRs of all the MTs present in the acceptable communication range. A Lyapunov based analysis is provided to explain the bound that the SINR error reaches, the size of which can be reduced by choosing appropriate control gains. The power control algorithm is simulated on a cellular network with distributed cells and the results indicate that the controller regulates the SINRs of all the MTs with low outage probability.