Date of Publication

5-2020

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Electronics and Communications Engineering

Subject Categories

Electrical and Computer Engineering | Electrical and Electronics

College

Gokongwei College of Engineering

Department/Unit

Electronics and Communications Engineering

Thesis Adviser

Lawrence Y. Materum

Defense Panel Chair

Aaron Don M. Africa

Defense Panel Member

Felicito S. Caluyo
Ann E. Dulay
Edison A. Roxas
Ryan Rhay P. Vicera

Abstract/Summary

Propagation channel modeling is essential in the design, simulation, and planning of wire- less communications systems. The performance of the wireless systems can be tested even before the construction of the communications network. Many propagation channel mea- surements show that multipath components are distributed as clusters. Existing clustering approaches have low accuracy and give only the number of clusters without considering the membership of the clusters. In this paper, the results of Simultaneous Clustering and Model Selection Matrix Affinity (SCAMSMA), Deep Divergence-based Clustering (DDC), and Modified SCAMSMA in clustering multipaths of eight-channel scenarios generated by the COST 2100 channel are presented. The clustering approaches group the COST 2100 datasets by determining simultaneously the number of clusters and the membership of the clusters. SCAMSMA and DDC cluster multipaths in indoor scenarios decently but they give low accuracy, as evidenced by the Jaccard indices, in semi-urban scenarios. Modified SCAMSMA improved the clustering accuracy in all channel scenarios, hence, the new clustering approach is better suited in multipath clustering.

Abstract Format

html

Language

English

Format

Electronic

Keywords

Wireless communication systems; Agricultural innovations

Upload Full Text

wf_yes

Embargo Period

8-26-2022

Download

Share

COinS