Distributed Source Detection with Dimension Reduction in Multiple-Antenna Wireless Networks

Bo Wang,Robert C. Qiu,Yanping Zhao
DOI: https://doi.org/10.1109/tvt.2016.2587361
IF: 6.8
2017-01-01
IEEE Transactions on Vehicular Technology
Abstract:We consider the problem of multiantenna source detection with dimension reduction in distributed wireless networks. Traditional strategies typically collect and perform raw data at the fusion center. In moderate-to-large-scale networks, however, the schemes create the bottleneck of computation and communication for high-dimensional data. The goal of this paper is to design a distributed algorithm in multiple-antenna wireless networks to project the raw data into the low-dimensional data to reduce the communication and computation burden while maintaining high detection performance. In this paper, a pseudosketching matrix is constructed to transform the raw data of each multiple-antenna node into the low-dimensional data. Furthermore, it is transformed into the vector by using the subspace method. By gathering the data vectors of all nodes at the fusion center, the eigenvalue-based detection methods, such as the generalized likelihood ratio test (GLRT), can be applied directly to determine if the source signal exists or not. Moreover, using the concentration inequalities of subgamma random variables, the theoretical analysis is derived to support the claim that the proposed algorithm has high detection performance. The simulations are presented to demonstrate the effectiveness of our proposed algorithm.
What problem does this paper attempt to address?