A fast convergence feed-forward automatic gain control algorithm based on RF characterization of Software Defined Radio

High throughput SDR networking waveforms require fast convergence Automatic Gain Control algorithms due to frequent transmissions in short time slots. In this paper, we propose a novel algorithm for feed-forward AGC with fast convergence and reduced computational complexity. Firstly, the RF front end of Software Defined Radio (SDR) is characterized through the use of vector signal generator. An RF characterization table consisting of Received Signal Strength Indicator (RSSI) values is constructed for selected values/ranges of gain and received power. Secondly, a comprehensive table is constructed through piece-wise analysis of the data obtained in first step and applying interpolation and curve-fitting techniques offline. By using the current RSSI and previous gain values, the received power is found using the constructed table. The working gain value, obtained from the same table, is then applied corresponding to the current received power. The actual convergence time of the proposed AGC algorithm is one clock cycle but it can be controlled through the use of loop filter.