A Dynamic Channel Assignment Using MNN-RS and FPGA Implementation

Abstract

In cellular mobile communication systems, a channel assignment problem has been extensively studied. In order to avoid exhaustive search, the number of reallocatable channel is limited only one in TPB method which is one of dynamic channel assignment proposed by Chang. In this paper, a dynamic channel assignment by using the maximum neural network with reinforced self-feedback (MNN-RS) is proposed. From numerical experiments, it is confirmed that the proposal improves blocking probability and reduces the probability that blocking probability exceeds arrowable rate. Lastly the proposed model is implemented on Xilinx FPGA XCV300E and scalability is discussed.