A Cost-Quality Tradeoff in Cooperative Sensor Networking

Wireless sensor networks consist of a large number of sensor nodes, each of which senses, computes and communicates with other nodes to collect and process data about the environment. Those networks are emerging as one of the new paradigms in networking with great impact on industry, government and military applications. A sensor network attempts to collect sensing data from the entire domain of its deployment, to process this data to understand phenomena and activities going on in this domain, and finally to communicate the results to the outside world to enable actuators to execute the necessary reactions. However, a sensor node is only capable of sensing events within its limited sensing range, so it has only a localized information about its environment. Hence, to provide the coverage of the entire domain, sensors need to collaborate and share their information with each other. Such sharing increases the knowledge of each sensor about the environment, but it also brings extra communication cost and increases the network operation complexity. In other words, cooperation and data sharing invokes a cost-quality tradeoff in the network. In this paper, we study two different sensor network applications: (i) finding an efficient sleep schedule based on sensing coverage redundancy, and (ii) adjusting traffic light periods to optimize traffic flow. In both applications the cost-quality tradeoff arises. In the paper, we study how fast network functionality increases when the level of cooperation raises and how much this increased functionality is offset by the raising cooperation costs. We simulated both applications with different level of cooperation and without it and demonstrated significant improvements in the overall system quality resulting from the properly selected levels of cooperation between the network's nodes.