Opportunistic Resource Management to Improve Network Service Performance in User-created Networks

Internet service providers have always struggled to provide sufficient last-hop connectivity to end-users. Even as new broadband technologies are beginning to meet end-user demands for indoor access, end-users are now demanding always-on high-speed mobile connectivity where ever they are. Wireless network operators are challenged by limited wireless resource availability that prevents mobile broadband speeds similar to the levels enjoyed at indoor static locations. However, as the use of wireless technology has increased, the availability and proliferation of Wi-Fi enabled devices has exploded to the point that ample idle bandwidth, storage and computation resources are available in end-user owned and operated devices. By collaboratively pooling these idle resources into user-created networks, the network capacity to mobile end-users can be greatly increased. Doing so is challenging because of the dynamic and unreliable availability of end-user resources. The thesis of this work is that in spite of the dynamic and disconnected nature of user-created networks, high-throughput, high-availability and low-latency network services for mobile end-users can be enabled through localized and location-based solutions using only end-user contributed resources. This is demonstrated through the design and implementation of four services including i) node-based local congestion control that improves message delivery rate for store-carry-and-forward routing in disconnected networks, ii) a location-based data overlay running on top of mobile devices, iii) localized flow scheduling to increase bandwidth for mobile clients connecting to external networks and iv) local, distributed authentication with location-dependent certificate revocation list segments to reduce authentication delays. These services are evaluated through deployments in real testbeds and through simulations of large-scale mobile networks.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Computer Science

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2010

Author dc:creator

• Thompson, Nathanael A.

Contributors dc:contributor

• Kravets, Robin H.
• Abdelzaher, Tarek F.
• Gupta, Indranil
• Zerfos, Petros

• user-created networks
• local algorithms
• location-based algorithms
• opportunistic resource management

Statement dc:rights

• Copyright 2010 Nathanael A. Thompson

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/16901

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/16901