A robust Channel Prediction Algorithm in Coordinated Networks with Backhaul Delay
Zezhong Mao, Jin Xu, Xinyao Zhang, Xiaofeng Tao · 2018
Coordinated Networks have attracted tremendous attention since its ability to provide large increment in spectral efficiency as well as reduce inter-cell interference. In order to achieve accurate scheduling, real-time channel state information (CSI) is crucial to assign appropriate resource blocks (RBs). However, backhaul delay exists in practical coordinated networks, which may cause outdated CSI and performance degradation. To eliminate the influence caused by backhaul delay, a robust channel prediction algorithm combined of echo state network (ESN) and genetic algorithm (GA) is proposed in this paper. In order to evaluate the performance of the proposed algorithm, simulation results are provided for channel prediction error, which show that the proposed algorithm is much more accurate than traditional linear prediction algorithm with different back-haul delays, Furthermore, simulation results are also given for the system throughput with scheduling based on predicted CSI. Compared with the traditional linear prediction algorithm, the proposed algorithm can achieve 10.91% and 15.86% additional gain in average user throughput and edge user throughput respectively, with 15ms backhaul delay.