On the Uplink Performance of Cooperative LEO Satellite Networks Under Realistic Assumptions

In this paper, we investigate the uplink performance of Low-Earth Orbit (LEO) satellite constellations, focusing on the role of Inter-Satellite Links (ISLs) in enhancing communication performance. The study examines two configurations of the proposed Soft Handover (SH) strategy: 1) Single-Satellite Relay (SSR), where one target satellite assists the serving satellite, and 2) Multi-Satellite Relay (MSR), where multiple satellites cooperatively forward the received signal from ground user (GU) through ISLs to the serving satellite. The results obtained under SH are compared with those from conventional Hard Handover (HH) schemes. Using realistic 3GPP channel models for both ground-to-satellite (G2S) and ISL, we derive a framework based on information theory to estimate the system block error rate (BLER) and conditional capacity, employing two relaying techniques: Amplify-and-Forward (AF) and Decode-and-Forward (DF). We also compare the impact of using Maximal Ratio Combining (MRC) and Selection Combining (SC) to aggregate the signals received from the GU and the relaying satellites at the serving satellite. Furthermore, we investigate the impact of imperfect Channel State Information (CSI) and beam misalignment within ISLs and propose mitigation techniques to counteract these effects. Our analysis demonstrates that the proposed SH scheme—under both SSR and MSR reception—significantly outperforms traditional HH methods, reducing the required signal-to-noise ratio (SNR) on the G2S link and the required transmission power on the ISL to meet given quality-of-service requirements. For instance, a constellation comprising 100 LEO satellites requires an SNR of 21.5 dB to achieve a BLER of 10−4 with HH, whereas the proposed SH approach achieves the same BLER with only 12 dB. These results highlight the effectiveness of the proposed SH techniques in improving the robustness and efficiency of future LEO satellite communication networks.