Dynamic Topology Estimation and Resource Allocation for Power Line Communication


Zhang, C (2016) Dynamic Topology Estimation and Resource Allocation for Power Line Communication. PhD thesis, University of Liverpool.

[img] Text
200687388_Jan2016.pdf - Unspecified
Download (1MB)

Abstract

Power line communication (PLC), which uses existing infrastructure of power delivery for data transfer, is regarded as an economical, pervasive and extensive communication solution for smart grid and home broadband applications. One of the challenges of applying communication technologies to power line network lies in acquirement of channel state information (CSI), which is dependent on network topology. Moreover, the knowledge of topology provides a basis for the design of routing protocols and power flow optimization. Therefore, efficient approaches for dynamic topology estimation are highly demanded. While dynamic routing and resource allocation enable high-speed and multi-tasking communication services over power lines. In this thesis, a dynamic topology estimation scheme for PLC is investigated, and a cross-layer routing and resource allocation scheme assisted by dynamic topology estimation is developed to improve the system performance. In the first contribution, a high-resolution and low-complexity dynamic topology estimation scheme for time-varying indoor PLC networks is proposed. The scheme consists of three parts: a) a time-frequency domain reflectometry (TFDR) based path length estimation method, which requires measurement at a single PLC modem and achieves a much higher resolution than the frequency domain reflectometry (FDR) based method; b) a node-by-node greedy algorithm for topology reconstruction, which is much more computationally efficient than the existing peak-by-peak searching algorithm; c) an impulsive noise assisted dynamic topology re-estimation method, which results in a significant complexity reduction over fixed-frequency re-estimation. In the second contribution, a cross-layer routing and resource allocation (RA) scheme assisted by dynamic topology estimation is proposed to optimize the system throughput of indoor PLC network with heterogeneous delay requirements. The proposed scheme provides a multi-layer solution, which conducts the network layer routing based on the result of PHY layer resource allocation which is constrained by the MAC layer queuing delay. With the dynamic topology estimation proposed in the first contribution, the routing can be solved centrally at the source, which is more robust against topology changes compared to distributed solutions. The proposed cross-layer RA scheme consists of subcarrier allocation (SA) to multiple users and power allocation (PA) to subcarriers satisfying heterogeneous delay requirements. It is demonstrated that the proposed centralized routing strategy achieves a much lower packet loss rate (PLR) than a distributed routing scheme; while with optimal RA, the system throughput is significantly improved compared to the routing schemes without considering RA.

Item Type: Thesis (PhD)
Divisions: Faculty of Science and Engineering > School of Electrical Engineering, Electronics and Computer Science
Depositing User: Symplectic Admin
Date Deposited: 01 Aug 2016 10:29
Last Modified: 19 Jan 2023 07:37
DOI: 10.17638/03001025
Supervisors:
URI: https://livrepository.liverpool.ac.uk/id/eprint/3001025
Dimensions
Altmetric
CORE (COnnecting REpositories)