Design of real-time medium access control and the optimization of end-to-end delay using genetic algorithm
Item
-
Title
-
Design of real-time medium access control and the optimization of end-to-end delay using genetic algorithm
-
Identifier
-
d_2009_2013:7685ae254ba8:11249
-
identifier
-
11615
-
Creator
-
Yoon, June Seung,
-
Contributor
-
Myung J. Lee
-
Date
-
2012
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Electrical engineering | End-to-end delay | MAC | Optimization | QoS | Real-time
-
Abstract
-
The Quality of Service (QoS) of wireless ad hoc and sensor networks is identified as one of the most important research issues due to the difficulty of overcoming the unstable and unforeseeable characteristics of wireless medium. The difficulty is amplified for the networks that handle real-time data such as audio/video stream, vital sign monitoring, and factory process control. These are so delay-sensitive that usually become inefficacious if belated. Therefore, the most prevailing measure for the real-time data applications is the end-to-end delay. This dissertation addresses two medium access control protocols that can manage the QoS required for wireless body area networks and wireless industrial sensor networks, and a metaheuristic scheduling algorithm for multi-channel, multi-hop wireless sensor networks. First, the preemptive slot allocation and non-preemptive access medium access control (PNP-MAC) protocol, designed for wireless body area networks is introduced in chapter 1. Wireless body area network applications, especially medical applications are very exacting in delay; thus, PNP-MAC equips diverse QoS features to meet the QoS requirements of the applications. Second, chapter 2 introduces Enhanced Guaranteed Time Slot (EGTS), developed for industrial applications that require capabilities to support large scale multi-hop networks and deterministic delay. EGTS's multi-channel over the Time Division Multiple Access (TDMA) enables multiple simultaneous data transmissions in a single time slot, and the three-way-hand shaking allocation mechanism allows nodes to use non-overlapped slots and channels, leading the data forwarding with minimal end-to-end delay. Third, chapter 3 introduces Genetic Algorithm Assisted Scheduling (GAS), a scheduling algorithm designed based on the genetic algorithm, one of the most popular metaheuristic algorithms. Genetic algorithm is shown to provide a near-optimal solution at a reasonable computation time for our slot and channel allocation problem given an end-to-end delay bound. Newly designed features such as scheduling order of flow, pre-channel assignment, and sifting along with enhanced genetic algorithm operations expedite search process and lead to the optimal solution.
-
Type
-
dissertation
-
Source
-
2009_2013.csv
-
degree
-
Ph.D.
-
Program
-
Engineering
