Compression Optimization via Dynamic Programming-Based Decoding Node Selection in Multi-Hop LEO-ISL Networks

Abstract

This paper proposes a dynamic programming (DP)-based cooperative decoding framework for low earth orbit (LEO)-enabled Internet of Things (IoT) networks. Multiple LEO satellites sequentially receive the signal transmitted by a ground IoT device, and compressed observations are forwarded through inter-satellite links (ISLs) to improve decoding reliability. To reduce ISL resource consumption, we formulate a joint decodingsatellite selection and compression design problem that minimizes the total required ISL rate under a target signal-to-noise ratio (SNR) constraint. The compression design is reformulated using the forwarded SNR contribution, which enables a sequential DPbased solution. Simulation results show that the proposed method adaptively determines the decoding satellite according to the accumulated SNR and ISL forwarding cost, avoiding unnecessary forwarding when early decoding is feasible.