Rate distance and MST-based multiratecasting in wireless sensor networks

In the multiratecasting problem in wireless sensor networks, source sensor should report to multiple destinations at different rates for each of them. Two existing localized solutions have drawbacks. One selects best neighbour serving the highest rated destination and is suboptimal when rates are close to each other. The other has high computational time for testing many subsets of neighbours. We present two new localized algorithms. MST-based multiratecast routing protocol (MSTRC) examines only one set partition of destinations at each forwarding step. A message split occurs when the locally-built minimum spanning tree (MST) over the current node and the set of destinations has multiple edges originated at the current node. Destinations spanned by each of these edges are grouped together, and for each of these subsets the best neighbor is selected as the next hop. In rate-over-distance first (RoDiF) algorithm, we repeatedly select neighbour with maximal sum of rate · (distance reduction) toward destinations with progress. We also add a novel face recovery mechanism to deal with void areas, when no neighbor provides positive progress toward destinations. It constructs MST of current node and destinations without progress via neighbors, and, for each set partition of destinations corresponding to an edge e in MST, traverses (with maximal rate among covered destinations) face containing e until a node closer to one of these destinations is found, to allow for greedy continuation, while the process repeats for the remaining destinations similarly. Our experimental results demonstrate that MSTRC and RoDiF are highly rate-efficient in all scenarios, and, unlike exiting solutions, are adaptive to destination rate deviations.