A Distributed Energy-Aware Routing Framework for Load Balancing in LEO Satellite Networks via Synergy of Power Optimization and Fuzzy Logic
Abstract
Low Earth orbit (LEO) satellite networks constitute a critical infrastructure for achieving seamless global Internet coverage. However, existing routing schemes generally suffer from substantial signaling overhead and a lack of dynamic energy regulation mechanisms. Therefore, we propose a distributed power allocation fuzzy logic load balancing routing (DPAFL-LBR), which integrates power optimization with fuzzy logic to jointly optimize quality of service (QoS) and satellite energy consumption. First, a two-hop neighbor information update mechanism is designed to reduce storage and query complexity while mitigating the impact of delayed state updates. Second, satellite transmission power allocation is formulated as an energy optimization problem, where intra-group power budget optimization is employed to reduce queuing delay, and an efficient solution is obtained via best square approximation. Then, a truncated bisection update and a power compensation scheme are developed to allocate more power budget to hotspot satellites under a fixed total power constraint. Finally, a joint decision mechanism combining power and fuzzy logic is constructed, enabling satellites to adaptively determine forwarding directions via neighbor states, together with six types of membership-function updates. Numerical results demonstrate that DPAFL-LBR achieves superior performance in both load balancing and energy management, and can effectively satisfy QoS requirements, particularly under high traffic loads.