Indoor Surveillance with Multimodal Wireless Networks: Multi-Cycle Detection and Performance Analysis

Abstract

This paper addresses the problem of using a multi-node indoor wireless network as a distributed sensor network that can detect illegal intrusion/trespassing by detecting changes in the propagation environment caused by the intruder, thereby providing the functionality of an indoor surveillance system. The authors propose a detector that takes advantage of multiple scanning cycles, resulting in a very promising detection capability. The authors also investigate the theoretical performance of such systems by analyzing the asymptotic distribution of the log-likelihood ratio and provide guidelines for setting the system parameters for the desired performance. The experiments demonstrate that the theoretical results predict the detector performance very well, and they can be used for configuring the detectors.