Real-Time Radio Map Construction and Distribution for UAV-Assisted Mobile Edge Computing Networks

The radio map has emerged as a promising tool for optimizing spectrum resource utilization and shaping the future landscape of intelligent wireless networks. However, the deployment of radio maps across the network introduces computational and latency challenges, restricting their real-time applications from the user’s perspective. In this article, we introduce an innovative scheme for constructing and distributing radio maps in unmanned aerial vehicle (UAV)-assisted mobile edge computing (MEC) networks. Initially, we transform the distribution of radio maps into a collaborative process between UAV server and smart mobile devices (SMDs), proposing four distribution modes tailored to different network conditions. This ensures that each SMD can access radio maps with the lowest cost. Additionally, our scheme integrates a deep reinforcement learning (DRL) framework, fostering seamless coordination between UAV server and SMD to enhance overall system performance and operational efficiency. Simulation results validate the efficiency and efficacy of our proposed scheme in optimizing radio map distribution strategies and resource allocation, further confirming the potential real-time applications of radio maps in future wireless networks.

I. Introduction

The extensive adoption of ubiquitous networks and smart devices has heightened the demand for spectrum resources, resulting in a complex radio environment marked by discernible conflicts over spectrum allocation. To confront these challenges, the radio map emerges as a pivotal tool, facilitating a nuanced understanding of radio environments, and enabling a comprehensive analysis of network performance [1]. By integrating geographic information, it provides an exhaustive depiction of the radio environment across various dimensions [2]. Beyond comprehending the communication environment, the radio map offers a simplified and adaptable means of utilizing and managing multidimensional spectrum data. As a result, it has found widespread application in diverse domains, such as cognitive radio [3], resource allocation [4], mobility management [5], and path planning [6]. The construction of a radio map involves the collection of spectrum signals, achievable through dedicated sensors or smart mobile devices (SMDs) [7]. Subsequently, this data undergoes meticulous processing to generate a detailed representation of the radio environment. This process empowers operators with valuable insights into the behavior and characteristics of the radio spectrum, facilitating informed decision making and optimal resource allocation [8].