A near-optimal stochastic MIMO signal detection with a mixture of t-distribution prior

Multiple-input multiple-output (MIMO) systems are a promising key technology for future wireless communication. However, improving their signal detection performance is still challenging to further increase the wireless transmission efficiency. To address this challenge, we propose to intentionally extend the discrete signal detection problem in MIMO systems to a continuous one and to utilize the Hamiltonian Monte Carlo method, a type of efficient Markov chain Monte Carlo (MCMC). We already presented the use of a mixture of normal distribution for the prior distribution based on the same approach. This paper proposes the application of a mixture of t-distribution that further improves detection performance. We show that the proposed method can achieve near-optimal signal detection with a polynomial order computational complexity through theoretical analysis and computer simulation. The proposed high-performance and pragmatic MIMO signal detection should significantly contribute to developing the 6th-generation mobile network.