Cost Minimization for Space-Air-Ground Integrated Multi-Access Edge Computing Systems

TL;DR

A multi-agent deep deterministic policy gradient (MADDPG)-convex optimization and coalitional game (MADDPG-COCG) algorithm is proposed to optimize the continuous temporal decisions for heterogeneous nodes in the partially observable SAGIN-MEC system.

Abstract

Space-air-ground integrated multi-access edge computing (SAGIN-MEC) provides a promising solution for the rapidly developing low-altitude economy (LAE) to deliver flexible and wide-area computing services. However, fully realizing the potential of SAGIN-MEC in the LAE presents significant challenges, including coordinating decisions across heterogeneous nodes with different roles, modeling complex factors such as mobility and network variability, and handling real-time decision-making under partially observable environment with hybrid variables. To address these challenges, we first present a hierarchical SAGIN-MEC architecture that enables the coordination between user devices (UDs), uncrewed aerial vehicles (UAVs), and satellites. Then, we formulate a UD cost minimization optimization problem (UCMOP) to minimize the UD cost by jointly optimizing the task offloading ratio, UAV trajectory planning, computing resource allocation, and UD association. We show that the UCMOP is an NP-hard problem. To overcome this challenge, we propose a multi-agent deep deterministic policy gradient (MADDPG)-convex optimization and coalitional game (MADDPG-COCG) algorithm. Specifically, we employ the MADDPG algorithm to optimize the continuous temporal decisions for heterogeneous nodes in the partially observable SAGIN-MEC system. Moreover, we propose a convex optimization and coalitional game (COCG) method to enhance the conventional MADDPG by deterministically handling the hybrid and varying-dimensional decisions. Simulation results demonstrate that the proposed MADDPG-COCG algorithm significantly enhances the user-centric performances in terms of the aggregated UD cost, task completion delay, and UD energy consumption, with a slight increase in UAV energy consumption, compared to the benchmark algorithms. Moreover, the MADDPG-COCG algorithm shows superior convergence stability and scalability.

Figures (8)

• Figure 1: The architecture of the proposed SAGIN-MEC system in a remote area. In this hierarchical system, a set of UAVs in the air layer are deployed as mobile MEC servers to provide low-latency computing services for ground UDs. Meanwhile, LEO satellites in the space layer act as relays to forward tasks from UDs to a remote GS for cloud computing, thereby ensuring wide-area service coverage.
• Figure 2: The proposed MADDPG-COCG algorithm addresses hybrid and varying-dimensional action spaces by decoupling the decision variables. The MADDPG component learns continuous decisions, namely, the task offloading ratio and UAV trajectory planning. To enhance learning stability, the COCG module deterministically handles the remaining decisions: deriving a closed-form solution for computing resource allocation via convex optimization , and obtaining a stable UD association structure via a coalitional game
• Figure 3: System performance comparison among different algorithms.
• Figure 4: Training performance.
• Figure 5: System performance with the number of UDs.

Theorems & Definitions (10)

• Remark 1
• Remark 2
• Theorem 1
• proof
• Theorem 2
• proof
• Definition 1
• Definition 2
• Theorem 3
• proof