Technische Universität Darmstadt
Institut für Nachrichtentechnik
Rundeturmstraße 12 (S3/19, 7)
+49 (0)6151 16 75 308
+49 (0)6151 16 72 138
Bio-inspired signal processing techniques for detection in wireless sensor networks
In the context of adaptive cooperative wireless sensor networks, bio-inspired algorithms are investigated and developed for distributed detection. These algorithms mimic the complex behavior commonly manifested by self-organizing biological networks. Bio-inspired techniques for wireless sensor networks have shown significant promise in satisfactorily handling typically encountered issues of scalability, communication and computational resource efficiency and robustness to node and link failure. One such technique for distributed detection is diffusion adaptive algorithms, where sensors rely solely on local interactions and in-network processing, and are able, on the fly, to learn environmental statistics, and adapt their dynamics to changing conditions. Furthermore, it is desirable that these algorithms be endowed with robustness as a preemptive measure against departures from nominal statistical design assumptions.