Data Dissemination and Collection Algorithms

Broadcasting and gossiping are common conditions which have already been extensively researched for many years. In broadcasting, one source node wants to deliver a message to each and every other node, on the other hand in gossiping, each node possesses a message which they want to send to everybody else. Both of these are probably the most fundamental problems arising in communication networks. Our objective in this dissertation is to research problems which generalize gossiping and broadcasting. For instance, the source node could possibly have a number of messages to transmit or multicast. Most of the works on broadcasting in the literature are centered on homogeneous networks. The algorithms created are more relevant to managing data on local-area networks. However, large-scale storage systems usually include storage devices grouped on the wide-area network. Locating a suitable model and developing algorithms for broadcast which recognise the heterogeneous nature of the communication network is actually a vital element of this project.

Contents

1 I1 Introduction
1.1 Preliminaries
1.2 Organization and Overview of Contributions
1.2.1 Data Dissemination Problems
1.2.2 Coordinated Data Collection
2 Related Work
2.1 Computing Data Layout
2.2 Generalized Broadcasting and Gossiping
2.3 Broadcasting in Two-tier Communication Networks
2.4 Coordinated Data Collection
3 Single-Source Multicasting
3.1 Problem Speci cation
3.1.1 Model
3.2 Algorithm Single-Source Multicast
3.3 Details of Phase I
3.4 Analysis
4 Multi-Source Broadcasting
4.1 Problem Speci cation
4.2 Algorithm Multi-Source Broadcast
4.3 Analysis
5 Multi-Source Multicasting
5.1 Problem Speci cation
5.1.1 Background
5.2 Algorithm Multi-Source Multicast
5.2.1 Analysis
5.3 3 + o(1)-approximation Algorithm
5.4 Allowing Bypass Disks
5.5 Bounded-Size Matching Model
5.6 NP-hardness Result
6 Broadcasting in Two-tier Communication Networks
6.1 Problem Specification
6.2 Broadcasting
6.2.1 Analysis
6.2.2 Bad Example
6.3 Multicasting
6.3.1 Analysis
6.4 Bounding Global Transfers
6.4.1 Bounded Degree Model
6.4.2 Bounded Degree Model: Multicasting
6.4.3 Bounded-Size Matching Model
6.4.4 Bounded-Size Matching Model: Multicasting
6.5 Postal Model
6.5.1 Analysis of LCF
6.5.2 Interleaved LCF
6.6 Experiments
6.6.1 Results
7 Coordinated Data Collection
7.1 Problem Speci cation
7.2 Overview of Data Collection Approaches
7.2.1 Direct Methods
7.2.2 Non-coordinated Methods……..

Source: University of Maryland

Leave a Comment