I am interested in design and implementation of distributed algorithms and protocols for wireless sensor networks (WSN) -- large scale distributed systems of small, untethered, low-power nodes capable of sensing, processing and communication. The applications of WSN envisioned include large scale scientific studies like microclimate and earthquake monitoring, machine control and diagnosis in factories, traffic monitoring in highways, and battlefield awareness for militery interests, to name a few. These applications pose a different set of challenges in algorithm and protocol design not found in a centralized setting; i) Algorithms need to be energy-aware since nodes in a WSN is battery-operated and communication consumes order of magnitude more energy than local computation. ii) How does each node collaborate with others nodes to extract useful information? iii) Which nodes store what information? iv) How does a node find other nodes that have useful information? v) How does a user read information of his/her intnerst out from a WSN? My dissertation research aims to answer these questions in a specific application of estimating locations and identities of mobile object like people or vehicles in a WSN. In terms of application point of view, locations and identities of mobile objects like people or vehicles provide rich contextual information that can be useful in many applicatoins ranging from surveillance, traffic monitoring and ubiquitous computing and the problem has been studied actively in WSN community. In my dissertation, I propose a cross layer architecture - a set of algorithms and protocols for network and application layers - to solve the problem in a fully distributed manner.