Toward a Robust Sparse Data Representation for Wireless Sensor Networks
This work addresses data sparsity issues in wireless sensor networks, offering incremental improvements for efficient data processing in this domain.
The paper tackles the problem of transforming non-sparse sensor data into a sparse representation for wireless sensor networks, introducing an unsupervised neural network method that outperforms conventional approaches in real-data analysis.
Compressive sensing has been successfully used for optimized operations in wireless sensor networks. However, raw data collected by sensors may be neither originally sparse nor easily transformed into a sparse data representation. This paper addresses the problem of transforming source data collected by sensor nodes into a sparse representation with a few nonzero elements. Our contributions that address three major issues include: 1) an effective method that extracts population sparsity of the data, 2) a sparsity ratio guarantee scheme, and 3) a customized learning algorithm of the sparsifying dictionary. We introduce an unsupervised neural network to extract an intrinsic sparse coding of the data. The sparse codes are generated at the activation of the hidden layer using a sparsity nomination constraint and a shrinking mechanism. Our analysis using real data samples shows that the proposed method outperforms conventional sparsity-inducing methods.