To Reduce Channel Interference in Mobile Ad-Hoc Networks through LSTM and GRU Prediction Algorithms

Abstract

The demand for the applications that mobile ad-hoc networks (MANETs) can provide is rising along with their popularity. Two important features of MANETs are communication interference and connection. In MANET environments, the movement of mobile nodes has been rising. Co-channel or adjacent interference congestion is especially common due to the increasing mobility of mobile nodes. Due to the severe degradation in MANET performance caused by these interference events, they are becoming critical difficulties. There is a lack of accuracy in MANET's capacity to predict the resources that are required and accessible to prevent disrupting traffic patterns. The flow of traffic is one of its main drawbacks. One kind of deep learning is called the "deep channel," which learns sequentially using encoders and decoders. This model predicts variability in wireless signal quality based on past signal intensity. Two Deep channel versions share a core cell structure: long short-term memory (LSTM) and gated recurrent unit (GRU). A brief description of MANETs and their applications is given in this work. This study also presents the application of machine learning and deep learning in MANETs. The study demonstrated how signal interference issues may be resolved by using neural networks.