Using Least Mean Square Estimation and Graph Structure for Sampling of Graph Signal

This thesis contains two main techniques for sampling graph signals. The first method uses the Least Mean Square (LMS) method for sampling the nodes of a graph. The second is the graph structure sampling method used for sampling group nodes instead of the individual node. The graph signal is considered as k-bandlimited. These studies contain the feasibility of the suggested solution and include various insights into the implementation of useful sampling methods for graph signals. For LMS estimation, by using different strategies, simulation experiments show the sampled nodes. Moreover, the results show that the reconstruction of the signal is similar to the original. Meanwhile, the steady-state property of the strategies we used is also shown in the experiment. For the structure sampling method, we sample groups of nodes related to each other in the k-bandlimited signal to embed stably. We will also design some sampling distributions to make the number of sampled groups as small as possible. The experimental results provide data and validate our theory.

Author

• Yang, Juli

Language

• en

Keyword

• Sampling
• Least Mean Square
• Graph Structure
• Graph Signal Processing

Date created

• 2021

Type of Work

• Master's Thesis

License

• All rights reserved

Rights statement

• In Copyright

GW Unit

• Electrical Engineering

Degree

• M.S.

Advisor

• Doroslovacki, Milos

Committee Member(s)

• Dehghanian, Payman
• Aslani, Amir

Persistent URL

•  https://scholarspace.library.gwu.edu/etd/0g354f86x