Dynamic Neural Network-Based Resource Allocation Framework for Latency-Aware and Energy-Efficient Optimization in Wireless Communication Systems

Abstract

In wireless communication, efficiency is a crucial factor due to the inherent constraints of the radio frequency spectrum and the escalating demand for connectivity. With a limited spectrum available, efficient communication protocols are vital to maximize its utilization and accommodate the increasing number of devices. This study introduces a dynamic resource allocation system for wireless communication, encompassing essential components such as Base Stations (BS), Mobile Devices (MD), and a Centralized Controller (CC). BSs, serving as access points, establish wireless communication with MD within predetermined coverage areas. The CC acts as the neural centre of the system, collecting real-time data from both BSs and MDs to facilitate an adaptive resource allocation process. The comprehensive mechanism involves a scheduling algorithm for optimized data transmission, considering priority and Quality of Service (QoS) requirements, alongside power control mechanisms that dynamically adjust power levels to optimize energy efficiency while maintaining communication quality. MDs communicate with nearby BSs, relaying data transmission information to the CC, which strategically coordinates resources based on the collected data. This integrated system enhances resource utilization, minimizes latency, and improves energy efficiency, collectively contributing to the robustness and reliability of the wireless communication infrastructure.