Compressed Sensing Based Block Type Multi-User Detection for Sporadic IoT Communications

Abstract

Over the last decade, a significant opportunity for wireless networks has been recognized as the Internet-of-Things (IoT). IoT is a massive device-interconnected platform that enables seamless communications among objects. Recently, a variety of diverse IoT applications have been developed to improve the human life. However, among those applications there are still many major challenges, such as the massive access issue (MAI) and coverage. In this dissertation, we study the MAI in massive IoT applications. By exploiting the block sparsity nature of sporadic IoT communications, we propose a compressive sensing (CS) based block type multi-user detection (CS-BT-MUD) algorithm to address the challenge of MAI. In particular, our algorithm employs a decoupling operation to deal with complex-valued data, which enable to solve complex-valued problems via conventional CS algorithms. Based on two types of data traffic models, numerical evaluations demonstrate that our proposed CS-BT-MUD algorithm is very effective in addressing the MAI and thus offers benefits to block type multi-user detection in IoT communications.