DESIGNING PROVABLY GOOD HEURISTICS FOR RAPID INFORMATION DISSEMINATION PROBLEMS

Abstract

A fundamental role of complex networks is to propagate information so that agents in one part of the network are kept informed of the status of nodes in other parts with a low latency. Information dissemination problems in networks capture this key requirement by specifying the underlying model of information transmission, and the demand requirements for transmission such as the set of important nodes from which information freshness is vital. The main goal of this project is to develop a comprehensive theory of provably near-optimal approximation algorithms for network information dissemination problems encompassing both different models of transmission and a variety of dissemination demand requirements. Additionally, a new generalization of these problems is proposed using a vector clock model arising from distributed systems and social networking, and relate them to the more classical models. The proposed research will investigate the computational relationships and reductions between the various models and develop new ideas for designing approximation algorithms both for the newly proposed models and for improving upon the current best performance guarantees. A key outcome of this work will be the design of better centralized schedules for information flow in arbitrary networks, with the ability to specify the importance of different dissemination requirements by different weights in the generalized models that are developed.

Document Details

Document Type
DoD Grant Award
Publication Date
Aug 12, 2021
Source ID
FA95502010080

Entities

People

  • Ramamoorthi Ravi

Organizations

  • Air Force Office of Scientific Research
  • Carnegie Mellon University
  • United States Air Force

Tags

Fields of Study

  • Computer science

Readers

  • Computer Networking
  • Geospatial Intelligence and Artificial Intelligence Analytics
  • Neural Network Machine Learning.

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms