Non-Coherent Massive Access in Massive MIMO-Based LEO Satellite Internet-of-Things

In this paper, we investigate the grant-free massive access problem for the low earth orbit (LEO) satellites equipped with massive multiple-in-multiple-out (mMIMO) antenna array and adopt a non-coherent data transmission scheme to reduce the access latency. In such transmission scheme, each user terminal (UT) is assigned with a unique codebook and the message is encoded by the indices of the transmitted codewords. The LEO satellite performs non-coherent data detection with joint codeword detection and channel estimation (JCDCE). We propose the orthogonal approximate message passing for generalized multiple measurement vector (OAMP-GMMV) algorithm to solve the JCDCE problem. Compared with prior works that only leverage the sparsity of angular domain channel, the proposed OAMPGMMV algorithm exploits the clustered sparsity of mMIMO channel and the structured sparsity across different subcarriers simultaneously. Since the indices of active codewords are not necessarily the same in different carriers, it poses a challenge for us to utilize the multi-carrier common support set property of the channel. To deal with the issue, we extract intermediate estimation result from the OAMP-GMMV algorithm and predict the indices of active codewords. Simulation results are provided to show the superiority of our proposed algorithm under the scenario that time resources are limited.