Dissertation/Thesis Abstract

Compressive Sensing: Performance of Sparse Channel Estimation in Orthogonal Frequency Division Multiplexing
by Shenoy, Dinesh Nandakumar, M.S., California State University, Long Beach, 2018, 43; 10751778
Abstract (Summary)

Compressive Sensing is a signal processing technique that is based on the property that the sparsity concept of a given signal can be exploited to reconstruct the signal successfully while disobeying Shannon’s Nyquist Theorem with fewer samples than what is required.

Sparse estimation which is an application of Compressive Sensing is applied to an OFDM transmission, as the model provides a better performance along with controlled inter symbol interference (ISI).

The thesis is based on the assumption that a single OFDM block is essentially time invariant and the simulations are carried out based on these assumptions. This thesis uses training data to simulate the channel estimation and to successfully recover the transmitted information. Convex optimization techniques to minimize noise and obtain the transmitted signal are employed. The thesis estimates the bit error rate (BER) using a Rayleigh channel and AWGN to illustrate how sparse estimation can be used to increase the performance of the system. The BER is estimated using the output at the receiver end and then compared with the BER of Least Square Estimation technique.

Simulation results show that the BER results for a system based on Sparse estimation of the channel are much better than a system with Least Square Estimation at high signal to noise ratios (SNR).

Indexing (document details)
Advisor: Yeh, Hen-Geul (Henry)
Commitee: Hoang, Trong T., Jula, Hossein
School: California State University, Long Beach
Department: Electrical Engineering
School Location: United States -- California
Source: MAI 57/05M(E), Masters Abstracts International
Subjects: Electrical engineering
Keywords: Compressive sensing, Convex optimization, Least square estimation, Ofdm, Signal processing, Sparse estimation
Publication Number: 10751778
ISBN: 9780355863505
Copyright © 2019 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy