DMT optimality of LR-aided linear decoders for a general class of channels, lattice designs, and system models

  • Authors:
  • Joakim Jaldén;Petros Elia

  • Affiliations:
  • ACCESS Linnaeus Center, KTH Signal Processing Lab, Royal Institute of Technology, Stockholm, Sweden and Institute of Communications and Radio-Frequency Engineering, Technical University of Vienna;Mobile Communications Department, EURECOM, Sophia Antipolis cedex, France

  • Venue:
  • IEEE Transactions on Information Theory
  • Year:
  • 2010

Quantified Score

Hi-index 754.84

Visualization

Abstract

This paper identifies the first general, explicit, and nonrandom MIMO encoder-decoder structures that guarantee optimality with respect to the diversity-multiplexing tradeoff (DMT), without employing a computationally expensive maximum-likelihood (ML) receiver. Specifically, the work establishes the DMT optimality of a class of regularized lattice decoders, and more importantly the DMT optimality of their lattice-reduction (LR)-aided linear counterparts. The results hold for all channel statistics, for all channel dimensions, and most interestingly, irrespective of the particular lattice-code applied. As a special case, it is established that the LLL-based LR-aided linear implementation of the MMSE-GDFE lattice decoder facilitates DMT optimal decoding of any lattice code at a worst-case complexity that grows at most linearly in the data rate. This represents a fundamental reduction in the decoding complexity when compared to ML decoding whose complexity is generally exponential in the rate. The results' generality lends them applicable to a plethora of pertinent communication scenarios such as quasi-static MIMO, MIMO-OFDM, ISI, cooperative-relaying, and MIMO-ARQ channels, in all of which the DMT optimality of the LR-aided linear decoder is guaranteed. The adopted approach yields insight, and motivates further study, into joint transceiver designs with an improved SNR gap to ML decoding.