dissertation · Sabanci University · 2008
Distributed detection algorithms for parallel and hierarchical wireless sensor networks
Cited by 0Abstract and details only
Abstract
Wireless Sensor Networks (WSNs) have recently attracted a lot of attention in various potential applications in military, health, environment and commerce due to their detection, processing and communication capabilities. In this thesis, we consider distributed detection problem for both parallel and hierarchical topology in which sensor decisions are sent over non-ideal wireless channels. We first investigate optimal fusion rules in Neyman-Pearson sense for all considered network configuration. We then, suggest suboptimal fusion rules to decrease computational complexity of the optimal fusion rules. Thirdly, multi-bit distributed detection is investigated both analytically and numerically to increase the detection performance. Finally, we propose fusion center diversity by employing multiple antennas at the fusion center to improve the detection performance of the network and derive optimum fusion rules accordingly. Simulation results suggest that fusion center diversity increases the probability of detection for a given constant false alarm probability.
Cite this paper
Eritmen, K. (2008). Distributed detection algorithms for parallel and hierarchical wireless sensor networks. Sabanci University.
Eritmen, Kayhan. “Distributed detection algorithms for parallel and hierarchical wireless sensor networks.” Sabanci University, 2008.
Eritmen, Kayhan. 2008. “Distributed detection algorithms for parallel and hierarchical wireless sensor networks.” Sabanci University.
Eritmen, K. (2008) ‘Distributed detection algorithms for parallel and hierarchical wireless sensor networks’, Sabanci University.
@phdthesis{eritmen2008distributed,
title = {{Distributed detection algorithms for parallel and hierarchical wireless sensor networks}},
author = {Eritmen, Kayhan},
school = {Sabanci University},
year = {2008}
}Made from the index record. Check names, capitalisation and page numbers against the paper. Need another style? AI for Research has over 10,000.