Quasi-synchronous Random Access for Massive MIMO Based LEO Satellite Constellation

Low earth orbit (LEO) satellite constellationenabled
communication network is considered to be an indispensable
part to realize the Internet of Things (IoT) due to its unique
advantages in providing seamless global coverage. In this paper,
we investigate the random access problem in massive multipleinput
multiple-output based LEO satellite communication systems.
To deal with grant-free random access in IoT, a training
sequence padded multi-carrier system is designed with tolerance
to imperfect synchronization. Specifically, we construct a multisatellite
system where a training sequence is utilized to perform
joint activity detection and channel estimation (JADCE) at the
edge satellite nodes. Considering the sparse feature of terrestrialsatellite
link and sporadic transmission of user terminals (UTs),
we propose a compressed sensing-based algorithm to estimate
the delay tap and UT activities. To further utilize the structured
feature of the receive array, a 2-D ESPRIT algorithm is performed
for augmented parameterized channel estimation. Finally,
enhanced activity detection and data detection are performed
at the central node by leveraging the aggregated information
from edge nodes. To achieve reliable information transmission, we
propose a centralized interference cancellation and data detection
method, where both the high spatial correlation among UTs and
quantized backhaul are taken into account. Simulation results
verify the effectiveness of our proposed scheme in terms of
channel estimation, activity detection, and data detection for
quasi-synchronous random access satellite systems