Consistent update diffusion in mobile ad hoc networks

Abstract

Applications of mobile ad hoc networks (MANETs) occur in situations, where networks need to be deployed immediately but network infrastructures are not available. If MANET nodes have sensing capabilities, they can capture and communicate state of their surroundings, including environmental conditions or other nodes in proximity. If the sensed state information is propagated and collected in a database, this allows for a variety of promising automatic monitoring, tracking and navigation applications, using global state information to built up models of reality. Since state changes represent events happening in reality, applications are typically interested to see the most recent state. Also the order of state changes should be consistent with the corresponding order of events in reality. In particular, preserving consistency becomes a challenging research problem if there are multiple nodes sensing the same object, either subsequently or even concurrently. In this paper, we introduce a generic model of state propagation in MANETs and propose two consistency levels for this model. For each of these consistency levels, we define a state propagation algorithm based on information diffusion. Our simulations show, that for typical scenarios the additional synchronization overhead for achieving the proposed consistency levels is low. In terms of communication overhead and state propagation latency one of the proposed algorithms shows a similar performance as the underlying flooding mechanism.