Compressed Sensing Based Active User Detection in MIMO Systems With One-Bit ADC

Abstract

We investigate an active user detection for massive machine-type communication systems, in which a massive number of users with sporadic transmission communicate with a base station (BS) with multiple antennas. Due to hardware cost and power consumption, it is assumed that each receiver antenna at the BS is equipped with one-bit analog-to-digital converter (ADC). This problem can be formulated as a new compressed sensing problem with one-bit quantized multiple measurement vectors (termed one-bit CS-MMV). In this paper, we propose a novel algorithm (named Turbo-QISTA) by properly incorporating Turbo principle into quantized iterative shrinkage threshold algorithm (QISTA). Our major contribution is to design two key metrics (i.e., extrinsic and a priori information) of Turbo principle suitable for one-bit CS-MMV problem. Via simulations, we demonstrate the superiority of the proposed Turbo-QISTA.