IEEE Project Abstract

With a growing number of connected devices in theInternet-of-Things (IoT), multi-user detection (MUD) becomes acritical issue in the IoT gateway at the edge. Thanks to the featureof activity sparsity in the IoT devices, compressive sensing (CS)is a promising solution for MUD to handle massive devices underlimited resources. For practical IoT applications, the CS-basedIoT gateway needs to support an unknown number of activedevices up to 10% of total devices and should complete a timeslotdetection within 0.5 ms. However, most existing CS detectionengines cannot fulfill above requirements simultaneously. In thispaper, we proposed a modified sparsity adaptive matching pursuit(MSAMP) detection engine for the IoT gateway at the edge.With decomposition-free architecture, our proposed engine canefficiently operate multiple and arbitrary indices updating thatcan achieve adaptive sparsity realization and index set backtrackingoptimization with no need of prior information. Moreover,coarse-searching and turbo step-size techniques are designed toenhance 2.86× of throughput rate gain. The efficiently implementedMSAMP detection engine in TSMC 90-nm technologycan complete a time-slot detection in 0.192 ms (<0.5ms). Thus,the proposed detection engine can provide a constructive realtimedesign for practical IoT applications.