Collaborative Resource Allocation in Wireless Sensor Networks

Sensor networks have the potential to greatly aid the broader scientific community by enabling dense spatio-temporal monitoring of phenomena of interest. Adeptly managing network resources, including energy, bandwidth, and storage, can increase the lifetime of the system and the quality of the data collected. Collaborative resource management in wireless sensor networks permits different approaches than other contexts previously studied, as nodes are not expected to be strictly self-interested and the potential for adversarial behavior is limited. My poster examines techniques for understanding and managing collaborative resource usage across deployed networks. We discuss the use of resource vectors as a primitive for representing the resource cost of various operations and reasoning about their network-wide impact, and examine techniques for shifting resource consumption away from valuable and/or heavily-loaded nodes.