Utility-Based Flow Control for Sequential Imagery over Wireless Networks

Wireless sensor networks provide a unique set of characteristics that make them suitable for building surveillance, including their small size and unobtrusiveness and their capabilities for rapid deployment, decentralized monitoring and control. Bandwidth utilization constrains the success of a wireless surveillance system, especially when the network serves to transmit video imagery. Creating an effective information management policy will help to obtain an allocation of bandwidth that is "best" from the end user\´s perspective. This paper presents a decentralized pricing-based flow control method to allocate bandwidth across multiple competing cameras that are simultaneously attempting to make use of a shared, low bit-rate wireless channel. At present, this flow control model is being integrated into a building surveillance system in which off-the-shelf cameras detect intruders and then start streaming data according to user-specified preferences for quality, where various components of the flow control system monitor bandwidth utilization and adjust the quality settings of image streams to maximize end-user utility.