Achieving Optimal Schedule by Network Flow Techniques
No Thumbnail Available
Authors
Meeting name
Sponsors
Date
Journal Title
Format
Thesis
Subject
Abstract
Scheduling plays an important role in protocol design and performance optimizations for wireless or wired networks. This dissertation studies how to achieve optimal performance for two fundamental scheduling problems using network flow techniques. The first problem is Maximum lifetime scheduling for roadside networks. In this problem, sensors are deployed along a road to monitor traffic activities. A set of sensors can form a network if their sensing ranges cover the entire road. Since battery-driven sensors have limited lifetime, we need to schedule each sensor between active and sleep modes such that active sensors form a network to monitor the road while sleeping sensors reserve their energy until their turns to become active. At that time, another sensor network is formed to continue to monitor the road. A challenging problem is how to achieve the optimal schedule that the road can be monitored by such sensing networks, one after the other continuously, for a longest time. The previous research only solved a simple case where all sensors are identical initially. This dissertation presents a network flow model to solve the general case where each sensor may have different initial energy. Furthermore this dissertation presents a novel technique to obtain optimal scheduling for the road surveillance problem with survivability k in which the road needs k independent networks to monitor in order to make sensing data reliable. Second, the problem is Maximum Traffic Time-slot Scheduling (MTTS) problem in SS/TDMA satellite systems. This is a historical scheduling problem proposed decades ago. A two-step approach was developed to solve it. However, the Incremental Time-slot Assignment (ITSA) Problem induced in this two-step approach was proved to be NP-hard, which makes the original MTTS problem seem to be NP-hard. In the past decades, the optimal solution of the MTTS problem remains open. In this dissertation, a preprocessing with network flow technique is developed to achieve the optimal scheduling, which optimally solves this open problem.
Table of Contents
Introduction -- Maximal lifetime scheduling for roadside sensor networks with survivability K (K=1) -- Maximal lifetime scheduling for roadside sensor networks with survivability K (K >1) -- Achieving optimal schedule for time-slot assignment SS-TDMA satellite systems -- The MTTS problem extension: deadline guaranteed scheduling -- Conclusions and future work
DOI
PubMed ID
Degree
Ph.D.