Publikation

Discrepancy and Multipath Routing in Wireless Sensor Networks - Semantic Scholar

Goce Trajcevski, Oliviu Ghica, René Schubotz, Marco Zuniga, Manfred Hauswirth, Peter Scheuermann

Technical Report Northwestern University Northwestern University Technical Reports NWU-EECS-09-18 2009.

Abstrakt

This work addresses the problem of balancing the spatial distribution of the routing-load among the nodes in a given sensor network, in multipath settings. It has been conjectured in a number of works that in dense networks, the field-based multipath routing paradigms such as electrostatic fields, uniformly distribute the traffic load throughout the $2D$ space of deployment. However, when the distribution of the nodes deployment is not uniform, two major shortcomings of such approaches become apparent: (1) paths-merging due to decrease in available neighbors; and (2) paths-merging tendency near the physical boundary of the network. Together, these two effects significantly distort the initial energy distribution of the network. We postulate that an important parameter that needs to be taken into consideration when designing routing algorithms is the discrepancy of the nodes distribution. Motivated by this, we propose a novel multipath routing approach that enables a better load balancing, in the sense of reducing the spatial deviation of the energy consumption, when a single or multiple sources are transmitting data towards a given sink. Our experiments demonstrate that, in addition to improving the discrepancy of the network-wide energy distribution, our techniques also prolong the network lifetime. Abstract This work addresses the problem of balancing the spatial distribution of the routing-load among the nodes in a given sensor network, in multipath settings. It has been conjectured in a number of works that in dense networks, the field-based multipath routing paradigms such as electrostatic fields, uniformly distribute the traffic load throughout the 2D space of deployment. However, when the distribution of the nodes deployment is not uniform, two major shortcomings of such approaches become apparent: (1) paths-merging due to decrease in available neighbors; and (2) paths-merging tendency near the physical boundary of the network. Together, these two effects significantly distort the initial energy distribution of the network. We postulate that an important parameter that needs to be taken into consideration when designing routing algorithms is the discrepancy of the nodes distribution. Motivated by this, we propose a novel multipath routing approach that enables a better load balancing, in the sense of reducing the spatial deviation of the energy consumption, when a single or multiple sources are transmitting data towards a given sink. Our experiments demonstrate that, in addition to improving the discrepancy of the network-wide energy distribution, our techniques also prolong the network lifetime.

Weitere Links

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence