Tiny Multiagent DRL for Twins Migration in UAV Metaverses: A Multileader Multifollower Stackelberg Game Approach

The synergy between unmanned aerial vehicles (UAVs) and metaverses is giving rise to an emerging paradigm named UAV metaverses, which create a unified ecosystem that blends physical and virtual spaces, transforming drone interaction and virtual exploration. UAV twins (UTs), as the digital twins of UAVs that revolutionize UAV applications by making them more immersive, realistic, and informative, are deployed and updated on ground base stations, e.g., roadside units (RSUs), to offer metaverse services for UAV metaverse users (UMUs). Due to the dynamic mobility of UAVs and limited communication coverages of RSUs, it is essential to perform real-time UT migration to ensure seamless immersive experiences for UMUs. However, selecting appropriate RSUs and optimizing the required bandwidth is challenging for achieving reliable and efficient UT migration. To address the challenges, we propose a tiny machine learning-based Stackelberg game framework based on pruning techniques for efficient UT migration in UAV metaverses. Specifically, we formulate a multileader multifollower Stackelberg model considering a new immersion metric of UMUs in the utilities of UAVs. Then, we design a Tiny multiagent deep reinforcement learning (Tiny MADRL) algorithm to obtain the tiny networks representing the optimal game solution. Specifically, the actor-critic network leverages the pruning techniques to reduce the number of network parameters and achieve model size and computation reduction, allowing for efficient implementation of Tiny MADRL. Numerical results demonstrate that our proposed schemes have better performance than traditional schemes.

I. Introduction

Unmanned aerial vehicles (UAVs) have garnered significant prominence in contemporary times owing to their multifaceted utility in expanding the scope of mobile communications, enhancing surveillance capabilities, and contributing to the development of intelligent urban environments [1]. The expeditious advancement of state-of-the-art technologies, encompassing blockchain, artificial intelligence (AI), and eXtended reality (XR), coupled with the ubiquity of mobile devices, portends the emergence of metaverses as a transformative paradigm for the forthcoming generation [2]. This evolution is anticipated to facilitate interpersonal engagements and interactions with virtual entities within virtual environments [3]. UAV metaverses are considered hybrid immersive realms that combine XR technologies with real-time sensing data collected by UAVs. The metaverses offer customized and intelligent services, such as smart farming and UAV-virtual games, to the users who engage with them, known as UAV metaverse users (UMUs). UAV twin (UT) plays a crucial role in UAV metaverses, serving as an essential component that encompasses precise and comprehensive digital replicas of UAVs [4]. The UTs are instrumental in ensuring synchronization and real-time updates for UMUs, bridging the gap between physical and virtual spaces [3]. Consequently, users can seamlessly access up-to-date information and immerse themselves in captivating experiences within UAV metaverses.