Complexity Adjusted Soft-Output Sphere Decoding by Adaptive LLR Clipping

TL;DR

This paper introduces an adaptive LLR clipping scheme for soft-output sphere decoding in MIMO systems, reducing complexity in favorable conditions without sacrificing BER performance.

Contribution

It proposes a novel adaptive method that adjusts LLR clipping based on BER tracking, optimizing complexity-performance trade-offs in MIMO receivers.

Findings

Significant complexity reduction in favorable channel conditions

Maintains BER performance comparable to full APP decoding

Adaptive scheme outperforms fixed clipping approaches

Abstract

A-posteriori probability (APP) receivers operating over multiple-input, multiple-output channels provide enhanced bit error rate (BER) performance at the cost of increased complexity. However, employing full APP processing over favorable transmission environments, where less efficient approaches may already provide the required performance at a reduced complexity, results in unnecessary processing. For slowly varying channel statistics substantial complexity savings can be achieved by simple adaptive schemes. Such schemes track the BER performance and adjust the complexity of the soft output sphere decoder by adaptively setting the related log-likelihood ratio (LLR) clipping value.