An Application-Tailored MAC Protocol for Wireless Sensor Networks

We describe a data management framework suitable for wireless sensor networks that can be used to adapt the performance of a medium access control (MAC) protocol depending on the query injected into the network. The framework has a completely distributed architecture and only makes use of information available locally to capture information about network traffic patterns. It allows nodes not servicing a query to enter a dormant mode which minimizes transmissions and yet maintain an updated view of the network. We then introduce an Adaptive, Information-centric and Lightweight MAC (AI-LMAC) protocol that adapts its operation depending on the information presented by the framework. Our results demonstrate how transmissions are greatly reduced during the dormant mode. During the active mode, the MAC protocol adjusts fairness to match the expected requirements of the query thus reducing latency. Thus such a data management framework allows the MAC to operate more efficiently by tailoring its needs to suit the requirements of the application.