Ji Wu

Graduation Semester and Year




Document Type


Degree Name

Doctor of Philosophy in Electrical Engineering


Electrical Engineering

First Advisor

Qilian Liang


Compressive sensing (CS) is an emerging field based on the revelation that a small collection of linear projections of a sparse signal contains enough information for stable, sub-Nyquist signal acquisition. It provides a potential way to acquire the sparse data efficiently, or equivalently, highly accurate recovery of sparse data from undersampled measurements. Huffman coding and compressive sensing are adopted to compress real-world wind tunnel data. Both uniform and non-uniform Huffman coding are evaluated in terms of the number of quantization levels, mean square error, codeword length and compression ratio. The main drawback of Huffman coding is that it requires calculating the probability of each symbol before encoding. It means it may not be appropriate for real-time compression. We applied CS to wind tunnel data and compared its performance against the theoretical error bound.Due to limited energy and physical size of the sensor nodes, the conventional security mechanisms with high computation complexity are not feasible for wireless sensor networks (WSNs). A compressive sensing-based encryption is proposed for distributed WSNs, which provide both signal compression and encryption guarantees, without the additional computational cost of a separate encryption protocol. The computational and information-theoretical secrecy of the compressive sensing algorithm is also investigated. For the proposed distributed WSNs, if only a fraction of randomizer bits is stored by an eavesdropper, then the eavesdropper cannot obtain any information about the plaintext.We studied a compressive sensing-based Ultra-WideBand (UWB) wireless communication system. Compared with the conventional UWB system, it can jointly estimate the channel and compress the data. No information about the transmitted signal is required in advance as long as the channel follows the autoregressive model. The performance of compressive sensing-based data encryption scheme shows that the original data could never be reconstructed when the measurement matrix is not available. Hence, compressive sensing can be implemented as a data encryption scheme with good secrecy.


Electrical and Computer Engineering | Engineering


Degree granted by The University of Texas at Arlington