Managing Heterogeneous Data Flows in Wireless Sensor Networks Using a "Split Personality' Mote Platform

Remote sensing applications are increasingly making use of sensors that generate demanding data flows (e.g. digital imaging devices). This paper presents the design of GridStix 1.5, a novel wireless sensor network platform that offers alternative high-performance and low-power 'personalities'. This two-tier approach allows GridStix 1.5 to provide power efficient support for sensors which produce simple data flows, and rich computational and network support for more demanding sensor types. Our approach is evaluated in the context of a deployed flood monitoring and warning system.