IDEALS Home University of Illinois at Urbana-Champaign logo The Alma Mater The Main Quad

Leveraging clustering for efficient communication in opportunistic networks

Show full item record

Bookmark or cite this item: http://hdl.handle.net/2142/30983

Files in this item

File Description Format
PDF Bakht_Mehedi.pdf (802KB) (no description provided) PDF
Title: Leveraging clustering for efficient communication in opportunistic networks
Author(s): Bakht, Mehedi
Director of Research: Kravets, Robin H.
Doctoral Committee Chair(s): Kravets, Robin H.
Doctoral Committee Member(s): Abdelzaher, Tarek F.; Nahrstedt, Klara; Ammar, Mostafa
Department / Program: Computer Science
Discipline: Computer Science
Degree Granting Institution: University of Illinois at Urbana-Champaign
Degree: Ph.D.
Genre: Dissertation
Subject(s): Opportunistic Networking Clustering Delay-tolerant Networking (DTN)
Abstract: The overarching goal of my research is to design protocols for efficient communication in delay tolerant networks (DTNs), with a particular focus on pocket switched networks (PSNs), an emerging class of ad hoc networks made up of smartphones and other portable mobile devices carried by humans. Communication in such an infrastructure-less scenario is inherently opportunistic since it relies extensively on detecting, as well as utilizing, unplanned encounters between nodes. While existing solutions look at the nodes in isolation, we propose that the clustering of nodes, which is a common phenomenon observed in different types of DTNs including PSNs, can be leveraged for significantly improving the efficiency of different components of opportunistic communication. The first step in enabling opportunistic communication is neighbor discovery. In this context, we have developed Searchlight, an asynchronous neighbor discovery protocol that uses systematic probing to considerably decrease discovery latency while allowing nodes to operate at low duty cycles. However, for an individual node, performing continuous neighbor discovery can still be too expensive with a high-power radio like 802.11. On the other hand, relying only on a low-power, short-range radio for detecting neighbors will result in significantly fewer available contacts. To mitigate this problem, we have developed a scheme for more efficient neighbor discovery that leverages the clustering of nodes as well as the radio heterogeneity of mobile devices. The basic idea is that coordination over a low-power, short-range radio can help clustered nodes distribute the load of scanning over the high-power, long-range radio. We have implemented the protocol successfully on a testbed of Android phones. Clustering can be also leveraged at a higher level for efficient forwarding of messages. Most routing protocols for DTNs only focus on one-hop encounters. However clustering creates islands of connectivity where path-based routing can be more efficient. Based on this insight, we have developed a lightweight clustering-based routing protocol that performs well in both partitioned and clustered environments by integrating store-carry-forward techniques with path-based techniques when appropriate.
Issue Date: 2012-05-22
URI: http://hdl.handle.net/2142/30983
Rights Information: Copyright 2012 Mehedi Bakht
Date Available in IDEALS: 2012-05-22
Date Deposited: 2012-05
 

This item appears in the following Collection(s)

Show full item record

Item Statistics

  • Total Downloads: 227
  • Downloads this Month: 1
  • Downloads Today: 0

Browse

My Account

Information

Access Key