Event based fairness for video surveillance sensor networks

Abstract

With their ease of installation, infrastructureless mode of operation and flexible deployment style, Video Surveillance Sensor Networks (VSSNs) provide more opportunities than legacy surveillance methods for applications such as habitat monitoring and border surveillance. We argue that events created in the coverage area of a VSSN are the application level messaging units and propose to employ Event Based Fairness (EBF) which aims at a fair distribution of nodes' resources according to the event flows. We carried out simulation experiments to compare the application level performances of two different EBF implementations with that of FCFS based queueing. We observe that EBF enhances the VSSN performance in two ways: Firstly, when the video traffic due to the events created exceed the total capacity of the network, EBF increases the overall number of events properly reported to the sink. Secondly, EBF reduces the initial event reporting delay, thus decreases the response time to the occurring events within the network. © IFIP International Federation for Information Processing 2009.