Analysis of feedback prediction error on the downlink performance of OFDMA systems

In this paper we analyze the OFDMA downlink transmission performance of our proposed feedback (FB) reduction mechanism, where the base station (BS) predicts the FB (in terms of signal-to-noise ratio (SNR)) to reduce the FB overhead, hence increases the uplink throughput. In this mechanism, The BS uses well known recursive least square (RLS) algorithm as prediction tool which adapts with SNR variations over time and provides predicted SNR with minimized error. We define the error distribution from the error data thus, we derive the probability of error in each adaptive modulation level to define the average spectral efficiency, average bit error rate (BER) and average throughput with predicted FB and actual FB. Numerical examples confirm that the FB prediction mechanism creates minimum degradation of downlink performance.