Performance Analysis of Multiple-Antenna Ambient Backscatter Systems at Finite Blocklengths

TL;DR

This paper investigates the maximum data rate achievable in multiple-antenna ambient backscatter systems at finite blocklengths, providing bounds and analyzing how quickly capacity is approached as blocklength increases.

Contribution

It introduces finite blocklength bounds for ambient backscatter channels and demonstrates the accuracy of channel dispersion as a measure of capacity backoff.

Findings

Bounds converge quickly to channel capacity with increasing blocklength

Channel dispersion accurately predicts capacity backoff at finite blocklengths

Finite blocklength analysis is essential for practical ambient backscatter system design

Abstract

This paper analyzes the maximal achievable rate for a given blocklength and error probability over a multiple-antenna ambient backscatter channel with perfect channel state information at the receiver. The result consists of a finite blocklength channel coding achievability bound and a converse bound based on the Neyman-Pearson test and the normal approximation based on the Berry- Esseen Theorem. Numerical evaluation of these bounds shows fast convergence to the channel capacity as the blocklength increases and also proves that the channel dispersion is an accurate measure of the backoff from capacity due to finite blocklength.