Algebraic optimization of data delivery patterns in mobile sensor networks

Database-like query processing over a network of sensors has become an attractive paradigm for building sensor applications. A sensor query is characterized by data streams among participating sensor nodes with possible in-node data filtering or aggregation, and can be described as a tree-like data delivery pattern. The data delivery pattern can also be considered as a query execution plan, or query routing tree. We propose an algebraic optimization of the query routing tree construction and reconfiguration. In particular, we aim at generating query trees that maximize collision-free concurrent data transmissions hence reducing energy and time wastes due to retransmissions. Towards this, we introduce a data transmission algebra (DTA) and apply it for efficient generation of such query trees.