OPTIMIZED MIMO DETECTION TECHNIQUE

Computer Engineering

For both downlink and uplink cases, linear detectors such as Zero Forcing (ZF) or Minimum Mean Square Error (MMSE) are critical. These linear detectors, on the other hand, necessitate matrix inversion is a computationally intensive operation in big structures. We point out in this study that finding the ZF/MMSE solution does not need computing an exact inverse, and that an approximate inverse might yield equivalent results. When the quantized answer computed using the approximate inverse and the precise inverse are the same, this is achievable. We calculate how much approach may be permitted in order for this to happen. As a result, we recommend that existing iterative approaches be used to obtain approximate inverse with a low level of low-complexity. We demonstrate that inversion utilizing iterative approaches can produce equivalent error performance after a reasonable number of rounds. Furthermore, we demonstrate the benefits of utilizing an estimated inverse aren’t restricted to sequential detectors, but may also be applied to non- sequential detectors like sphere decoders (SD). In the case of any SD that necessitates matrix inversion an approximate inverse can be employed. We show that using the approximate inverse results in a lower radius, which minimizes the search space and thus the complexity. The numerical results back up our contention that utilizing approximate matrix inversion in large/massive MIMO systems reduces decoding complexity without sacrificing performance owing to mistakes.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.