Title: Ising Machines' Dynamics and Regularization for Near-Optimal Large and Massive MIMO Detection

Abstract: Optimal MIMO detection has been one of the most challenging and computationally inefficient tasks in wireless systems. We show that the new analog computing techniques like Coherent Ising Machines (CIM) and Oscillator-based Ising Machines (OIM) are promising candidates for performing near-optimal MIMO detection. We illustrate a fundamental problem with using classical, optical or quantum mechanical Ising Machines for MIMO detection: the error floor problem, which is a major bottleneck to practical deployments of Ising machine-based MIMO detectors. We propose a novel regularized Ising formulation for MIMO detection that mitigates the error floor and further evolves it into an algorithm that achieves near-optimal MIMO detection. Massive MIMO systems, that have a much larger number of antennas at the Access point (AP) than at the users, allow linear detectors to be near-optimal. However, the simplified detection in these systems comes at the cost of overall throughput, which can be improved by supporting more users. We show that our methods allow us to add more transmitter antennas/users and increase the overall throughput of the cell by several folds. We further show that our methods allow us to operate using more aggressive modulation and coding schemes and hence achieve much higher throughput. We demonstrate that, for a $16\times16$ large MIMO system, our methods achieve around 2.5$\times$ more throughput in mid-SNR regime ($\approx 12 dB$) and 2$\times$ more throughput in high-SNR regime( $>$ 20dB) than MMSE.