Name
Title | ||
|---|---|---|
Multimodal Wireless Networks: Communication and Surveillance on the Same Infrastructure |
Abstract | ||
|---|---|---|
In this paper, we propose a new family of wireless networks-the multimodal wireless networks. These networks offer multiple functionalities realized on the same infrastructure. A multimodal wireless network has two modes of operation: 1) the communication mode, when the network is used as a traditional wireless communication network, and 2) the surveillance mode, when the network is used as a distributed sensor network that can detect illegal intrusion. The surveillance functionality is realized by analyzing the properties of the received signals, and the change of the propagation environment caused by the intruder serves as the basis for intrusion detection. We derive maximum likelihood estimators and detectors based on the generalized likelihood ratio test that detects changes in the propagation environment: the single-cycle detector, which makes decisions at the end of each scanning cycle, and the multicycle detector, which combines information from multiple scanning cycles prior to detection. We also analyze the performance of these detectors by deriving the Cramer-Rao lower bound on the variance of the parameter estimators and determining the distribution of the log-likelihood ratio under both detection hypotheses. This will allow us to compare the theoretical performance of the single-cycle and the multicycle detectors and obtain analytical results to determine the decision threshold for a given probability of false alarm. To prevent performance degradation due to slowly drifting parameter values, we propose different strategies to track and update the parameter estimates. The experimental results obtained from an implemented prototype surveillance system show very promising detection capabilities. For example, the state of a door (open or closed) could be detected with a probability of 0.99 at a probability of false alarm 10-5. |
Year | DOI | Venue |
|---|---|---|
2007 | 10.1109/TIFS.2007.904944 | IEEE Transactions on Information Forensics and Security |
Keywords | Field | DocType |
multimodal wireless network,traditional wireless communication network,multicycle detector,multimodal wireless networks,promising detection capability,parameter estimate,false alarm,detection hypothesis,propagation environment,sensor network,intrusion detection,maximum likelihood estimate,detectors,generalized likelihood ratio test,maximum likelihood estimation,wireless network,cramer rao lower bound,log likelihood ratio,wireless networks,security,wireless sensor networks,wireless communication,sensor networks,parameter estimation | Wireless network,Telecommunications,Wireless,False alarm,Likelihood-ratio test,Computer science,Real-time computing,Artificial intelligence,Detector,Intrusion detection system,Cramér–Rao bound,Computer vision,Wireless sensor network | Journal |
Volume | Issue | ISSN |
2 | 3 | 1556-6013 |
Citations | PageRank | References |
5 | 0.50 | 72 |
Authors | ||
3 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Jianjun Chen | 1 | 39 | 12.52 |
| Z. Safar | 2 | 241 | 13.32 |
| J. A. Sorensen | 3 | 79 | 7.83 |