Multiconnectivity for SAGIN: Current Trends, Challenges, AI-driven Solutions, and Opportunities

TL;DR

The paper addresses the challenge of enabling reliable, high-capacity multiconnectivity (MC) in space-air-ground integrated networks (SAGIN) across heterogeneous non-terrestrial and terrestrial links. It proposes an agentic reinforcement learning framework with an actor–critic architecture to optimize link selection and scheduling as a dynamic, partially observable decision problem, balancing capacity, latency, and power. A case study demonstrates that learning-based MC can outperform baselines in capacity and latency, with a moderate power increase reflecting multi-link usage. The work also surveys standardization progress, identifies key challenges, and outlines AI-driven research directions (e.g., ISAC, synchronization, security) essential for scalable SAGIN-enabled MC.

Abstract

Space-air-ground-integrated network (SAGIN)-enabled multiconnectivity (MC) is emerging as a key enabler for next-generation networks, enabling users to simultaneously utilize multiple links across multi-layer non-terrestrial networks (NTN) and multi-radio access technology (multi-RAT) terrestrial networks (TN). However, the heterogeneity of TN and NTN introduces complex architectural challenges that complicate MC implementation. Specifically, the diversity of link types, spanning air-to-air, air-to-space, space-to-space, space-to-ground, and ground-to-ground communications, renders optimal resource allocation highly complex. Recent advancements in reinforcement learning (RL) and agentic artificial intelligence (AI) have shown remarkable effectiveness in optimal decision-making in complex and dynamic environments. In this paper, we review the current developments in SAGIN-enabled MC and outline the key challenges associated with its implementation. We further highlight the transformative potential of AI-driven approaches for resource optimization in a heterogeneous SAGIN environment. To this end, we present a case study on resource allocation optimization enabled by agentic RL for SAGIN-enabled MC involving diverse radio access technologies (RATs). Results show that learning-based methods can effectively handle complex scenarios and substantially enhance network performance in terms of latency and capacity while incurring a moderate increase in power consumption as an acceptable tradeoff. Finally, open research problems and future directions are presented to realize efficient SAGIN-enabled MC.

Figures (5)

• Figure 1: Overview of SAGIN-enabled MC.
• Figure 2: The high-level depiction of the proposed agentic RL framework for resource allocation in SAGIN-enabled MC.
• Figure 3: System architecture of the proposed algorithm. The agent performs perception (environment modeling, data preprocessing), planning/decision (actor–critic), and memorization (replay buffer, target critic).
• Figure 4: The illustration of the learning behavior and convergence performance of the agent: (a) episode reward, (b) switching rate, (c) average latency, and (d) normalized power consumption.
• Figure 5: Performance comparison of the proposed scheme with baseline methods in terms of (a) capacity, (b) latency, and (c) power consumption.