Balanced slices in wireless sensor networks

In this paper we formalize a problem called weighted balanced two-slice problem (WB2S) that calls for partitioning (slicing) a wireless sensor network (WSN) so as to support concurrent running of two independent applications. Our model associates a weight with each node to reflect the node´s relative importance to the applications. As well, the model imposes a degree constraint on each node to control its traffic load when it operates in a multi-level tree. The problem calls for computing two disjoint trees (sharing a common sink) whose minimum total weight is maximized among all possible feasible pairs of trees. We show that the WB2S problem is NP-complete, present a dynamic program to handle the problem, analyze the algorithm, and show its potential impact by simulation.