Adaptive Gradient-Based Methods for Adaptive Power Allocation in OFDM-Based Cognitive Radio Networks

摘要

A gradient-based method is designed for power allocation in orthogonal-frequency-division-multiplexing (OFDM) -based cognitive radio networks. The resource allocation problem subject to a mutual interference constraint is considered. We utilize the gradient descent approach to allocate power to subcarriers in cognitive radio (CR) networks. The proposed gradient-based power allocation method with a well-designed step size can approximate the optimal solution within a few iterations. Due to the derived equation for power allocation in an adaptive manner, the proposed method is feasible for adaptive power allocation in time-varying channels. The analysis for the selection of the step size is presented in this paper. For comparison purposes, a greedy power-loading method requiring numerous iterations is also designed for this power allocation problem. The proposed gradient-based method and the greedy power-loading method both have a computational complexity of $O(N)$, but the proposed gradient-based method requires far fewer iterations. As demonstrated in the simulation results, the proposed gradient-based method with the adaptive step size has a fast rate to achieve a near-optimal solution within an extremely small number of iterations and has quite a low computational complexity of $O(N)$.