Recently, various multiple-input multiple-output (MIMO) signal detectors based on deep learning have been proposed. Especially, deep unfolding (DU), which involves unrolling of an existing iterative algorithm and embedding of trainable parameters, has been applied with remarkable detection performance. In this study, we attempted to construct a simple DU-based trainable MIMO detector with the simplest structure without using matrix inversion. The proposed detector based on the Hubbard–Stratonovich (HS) transformation and DU is called the trainable HS (THS) detector. It requires a constant number of trainable parameters and only quadratic training and execution costs per iteration. Numerical results show that the detection performance of the THS detector is better than that of existing algorithms of the same complexity and close to that of a DU-based detector, which has higher training and execution costs than the THS detector.
We present a multiscale solution algorithm for dense rarefied gas flows. In this method the general conservation laws of compressible fluid dynamics are solved with a flux‐corrected‐transport‐based finite volume algorithm using physically accurate stresses and heat fluxes calculated with the Direct Simulation Monte Carlo method. The resulting computational method is expected to produce better solutions for moderate to high Knudsen number (transition‐regime) flows than solutions to the Navier‐Stokes equations at a much lower computational cost than DSMC and other particle‐based solution methods. The algorithm is tested on a low‐speed, planar Couette flow at two different Knudsen numbers. The effect of the lengths of time spent in the finite volume and DSMC portions of the algorithm are examined.