Channel-aware Decentralized Detection via Level-triggered Sampling

TL;DR

This paper proposes a channel-aware decentralized detection scheme where sensors use level-triggered sampling to send 1-bit messages over noisy channels, and the fusion center updates global LLRs for sequential detection, with asymptotic delay analysis.

Contribution

It introduces a novel channel-aware detection method with level-triggered sampling and derives asymptotic detection delay considering noisy channels.

Findings

Asymptotic detection delay characterized by KL information number.

Fusion rules derived for various channel types.

Delay analysis guides signaling scheme choices.

Abstract

We consider decentralized detection through distributed sensors that perform level-triggered sampling and communicate with a fusion center via noisy channels. Each sensor computes its local log-likelihood ratio (LLR), samples it using the level-triggered sampling, and upon sampling transmits a single bit to the FC. Upon receiving a bit from a sensor, the FC updates the global LLR and performs a sequential probability ratio test (SPRT) step. We derive the fusion rules under various types of channels. We further provide an asymptotic analysis on the average detection delay for the proposed channel-aware scheme, and show that the asymptotic detection delay is characterized by a KL information number. The delay analysis facilitates the choice of appropriate signaling schemes under different channel types for sending the 1-bit information from sensors to the FC.