A Novel Approach to Guarantee Causal Message Ordering in Pre-planned Wireless Sensor Networks

摘要

Data fusion often depends on the time of occurrence of fused sensor readings, called as temporal ordering, which is defined with respect to a single message, or as causal ordering to guarantee dependency relationships between messages in many-to-many communication patterns. There are some temporal ordering protocols based on physical time synchronization in publish/subscribe (P/S) paradigm of wireless sensor networks, but there exist little research works on development of causal ordering protocols based on logical time in P/S of wireless sensor networks (WSNs). Causal message ordering is more useful for most distributed applications in which a large number of sensor nodes request cooperating to fuse their data in WSNs. Temporal ordering is not sufficient for these distributed applications because it is not defined for dependency relationships between these messages. Also, many-to-many communication patterns attempt to address the problem of providing scalability of data propagation, guaranteeing message delivery order and supporting overlapping multicast groups in WSNs. In this paper, we present a novel approach based on gossiping and firefly synchronization instead of physical time synchronization, guaranteeing the causal message ordering property in P/S of WSNs. In the proposed protocol, every sensor broker disseminates the multicast message including the latest time-stamped information that represents the gossip round in which the message is generated to subscribers for causal message ordering. The latest time-stamped information is represented using colors. Its scalability feature might be highly suitable for the area of the applications requiring only the minimum causal information of message delivery with flexible consistency by cooperating to fuse their data.