# Quasi-Synchronous Random Access for Massive MIMO-Based LEO Satellite Constellations

**Authors:** Keke Ying, Zhen Gao, Sheng Chen, Mingyu Zhou, Dezhi Zheng, Symeon Chatzinotas, Björn Ottersten, H. Vincent Poor  
**arXiv:** [2304.04484](https://arxiv.org/abs/2304.04484)  
**Date:** 2023-04-10  
**Summary page:** https://www.summarizepaper.com/en/paper/2304.04484v1/

## Key points

- Low Earth Orbit (LEO) satellite constellation-enabled communication networks are important for IoT deployments
- Random access problem in massive multiple-input multiple-output-based LEO satellite systems addressed
- Proposed training sequence padded multi-carrier system to detect devices' activity and estimate channels
- Consideration of sparsity of terrestrial-satellite links and sporadic traffic feature of IoT terminals
- Use of orthogonal approximate message passing-multiple measurement vector algorithm for delay coefficient and user terminal activity estimation
- Two-dimensional estimation of signal parameters via rotational invariance technique for channel estimation accuracy enhancement
- Majority voting scheme at central server node to improve activity detection by aggregating backhaul information from multiple satellites
- Proposed multi-satellite cooperative linear data detection and Bayesian dequantization data detection schemes for perfect and quantized backhaul respectively
- Simulation results show improved channel estimation accuracy, device activity detection, and data detection performance

## Summary

Low Earth Orbit (LEO) satellite constellation-enabled communication networks are becoming increasingly important for Internet of Things (IoT) deployments due to their capability to provide seamless global coverage. This paper focuses on addressing the random access problem in massive multiple-input multiple-output-based LEO satellite systems, where a multi-satellite cooperative processing mechanism is considered. To overcome the issue of imperfect synchronization at edge satellite nodes, the authors propose a training sequence padded multi-carrier system that uses a training sequence to detect devices' activity and estimate their channels. The authors also take into account the inherent sparsity of terrestrial-satellite links and the sporadic traffic feature of IoT terminals. They employ an orthogonal approximate message passing-multiple measurement vector algorithm to estimate delay coefficients and user terminal activity. To enhance channel estimation accuracy, a two-dimensional estimation of signal parameters via rotational invariance technique is performed by leveraging the structure of the receive array. At the central server node, a majority voting scheme is proposed to improve activity detection by aggregating backhaul information from multiple satellites. Furthermore, two data detection schemes are proposed: multi-satellite cooperative linear data detection and multi-satellite cooperative Bayesian dequantization data detection which are designed to cope with perfect and quantized backhaul respectively. Simulation results demonstrate that these techniques can effectively improve channel estimation accuracy, detect device activity and enhance data detection performance for quasi-synchronous random access in satellite systems.