Survival guide for dense networks

High density affects the communication in an energy-constrained network since the energy-saving protocols are not designed to cope with the high contention expected in dense networks. While increased delay and reduced throughput are predictable results of contention from background traffic, high contention increases collisions, resulting in increased packet retransmissions and dropped packets. Two factors are mainly responsible for this high contention in the emerging networks. First, the networks are getting denser with the proliferation of smartphones and tablets. Second, the communication protocols used in these energy-constrained devices interact negatively with high density increasing contention even more. As a result, instead of saturating the bandwidth as expected with high traffic, the dense network starts thrashing resulting in a steep increase in delay and a drastic fall in throughput. The devices also end up consuming more energy since sleeping during the idle times is no longer an effective strategy to save energy. None of these effects is desirable. The goal of our research is to improve communication in energy-constrained dense networks. While the increased density typically results in poor communication quality and high energy consumption, knowing that the network is dense often enables opportunities for optimization. Our research reveals that the increased number of devices and the frequent traffic, the characteristics that cause high contention in dense networks, can also be leveraged to improve the communication and energy-efficiency in these networks. With more devices within the communication range of a device, the neighbors can share the responsibility of finding an awake receiver in a wireless sensor network (WSN), or acquiring information about nearby access points (APs) to decide which AP to join in an infrastructure-based WLAN. The redundancy inherent in dense networks provides opportunities to reduce the delay in a multi-hop WSN by forwarding data packets to those neighbors which wake up the soonest. Interestingly, the security of the network can also be improved by leveraging the high connectivity. Neighbors can collaborate to eliminate any differentiability or predicatability in WSN communication such that the entire neighborhood appears to the jammer as a single node with a big battery. Finally, local observation about the neighborhood can provide information about the contention in the network and enable the devices to adapt themselves to cope with high density. The key to achieving efficient communication in an energy-constrained high density network is to embrace the high density. Our thesis shows that although high contention provides a harsh environment for communication, by leveraging the inherent characteristics of dense networks such as high connectivity, redundancy, frequent communication, devices can survive in such networks by successfully improving their quality of communication and energy-efficiency.

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

2013

Author dc:creator

• Ashraf, Farhana

Contributors dc:contributor

• Kravets, Robin H.
• Vaidya, Nitin H.
• Hu, Yih-Chun
• Abdelzaher, Tarek F.
• Chandra, Ranveer

• dense wireless networks
• energy-constrained networks
• opportunistic improvement

Statement dc:rights

• Copyright 2013 by Farhana Ashraf

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/44387

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/44387