On Lattice Sequential Decoding for The Unconstrained AWGN Channel

TL;DR

This paper analyzes the performance and computational complexity of lattice sequential decoding for the unconstrained AWGN channel, highlighting its potential as a low-complexity alternative to traditional lattice decoders.

Contribution

It characterizes the performance-complexity tradeoff of lattice sequential decoding and derives the cut-off volume-to-noise ratio for effective low-complexity decoding.

Findings

Identifies the error exponent and decoding complexity as functions of the bias parameter.

Derives the cut-off volume-to-noise ratio for achieving good error performance.

Shows the potential of sequential decoding as a low-complexity alternative to NP-hard lattice decoding.

Abstract

In this paper, the performance limits and the computational complexity of the lattice sequential decoder are analyzed for the unconstrained additive white Gaussian noise channel. The performance analysis available in the literature for such a channel has been studied only under the use of the minimum Euclidean distance decoder that is commonly referred to as the lattice decoder. Lattice decoders based on solutions to the NP-hard closest vector problem are very complex to implement, and the search for low complexity receivers for the detection of lattice codes is considered a challenging problem. However, the low computational complexity advantage that sequential decoding promises, makes it an alternative solution to the lattice decoder. In this work, we will characterize the performance and complexity tradeoff via the error exponent and the decoding complexity, respectively, of such a decoder as a function of the decoding parameter --- the bias term. For the above channel, we derive the cut-off volume-to-noise ratio that is required to achieve a good error performance with low decoding complexity.