A Neural Network Prediction Based Adaptive Mode Selection Scheme in Full-Duplex Cognitive Networks

TL;DR

This paper introduces a neural network-based adaptive mode selection scheme in full-duplex cognitive networks that enhances secondary user throughput and minimizes collision with primary users by switching between transmission modes based on prediction.

Contribution

It presents a novel neural network predictor and adaptive mode selection scheme that improves throughput and reduces collisions in full-duplex cognitive networks.

Findings

Achieves higher secondary user throughput than traditional transmission-reception mode.

Reduces collision probability close to transmission-and-sensing mode levels.

Demonstrates effective prediction performance in adaptive mode switching.

Abstract

We propose a neural network (NN) predictor and an adaptive mode selection scheme for the purpose of both improving secondary user's (SU's) throughput and reducing collision probability to the primary user (PU) in full-duplex (FD) cognitive networks. SUs can adaptively switch between FD transmission-and-reception (TR) and transmission-and-sensing (TS) modes based on the NN prediction results for each transmission duration. The prediction performance is then analysed in terms of prediction error probability. We also compare the performance of our proposed scheme with conventional TR and TS modes in terms of SUs average throughput and collision probability, respectively. Simulation results show that our proposed scheme achieves even better SUs average throughput compared with TR mode. Meanwhile, the collision probability can be reduced close to the level of TS mode.