Optimal Non-coherent Data Detection for Massive SIMO Wireless Systems with General Constellations: A Polynomial Complexity Solution

Massive MIMO systems can greatly increase spectral and energy efficiency over traditional MIMO systems by exploiting large antenna arrays. However, increasing the number of antennas at the base station (BS) makes the uplink noncoherent data detection very challenging in massive MIMO systems. In this paper we consider the joint maximum likelihood (ML) channel estimation and data detection problem for massive SIMO (single input multiple output) wireless systems, which is a special case of wireless systems with large antenna arrays. We propose exact ML non-coherent data detection algorithms for both constantmodulus and nonconstant-modulus constellations, with a low expected complexity. Despite the large number of unknown channel coefficients for massive SIMO systems, we show that the expected computational complexity of these algorithms is linear in the number of receive antennas and polynomial in channel coherence time. Simulation results show the performance gains (up to 5 dB improvement) of the optimal non-coherent data detection with a low computational complexity. Keywords—ML detection, channel estimation, massive SIMO, maximum likelihood, sphere decoder