A weighted average-based external clock synchronisation protocol for wireless sensor networks

A model for the classification and survey of clock synchronization protocols in WSNs

“…In the second approach, the protocols [21,19] are fully distributed and provide only internal clock synchronization. The protocols [16,22] which use multiple reference node clocks are intrinsically fault-tolerant to failure of some reference nodes in comparison to the protocols which use only a single reference node clock. However, since the protocols [20,23] using only a single reference node need to synchronize with a single reference clock only, synchronization accuracy in these protocols would be much better.…”

“…The technique usually assumes that the synchronization packet is received by all the neighbors nearly at the same real time even though there may be extremely small differences between reception times of different nodes depending upon the distance from the sender. In case of senderto-receiver synchronization approach [22], the sender node sends a synchronization packet timestamped with its local time to the receiver. After receiving this request packet, the receiver timestamps it with its own clock and sends it back to the sender.…”

“…synchronous, and asynchronous. In synchronous clock synchronization protocols [22,23], the clocks are synchronized with each other periodically. For a fixed period of real time, all the nodes perform the synchronization process, and during the rest of the time, nodes can either engage themselves in monitoring the environment or go to sleep mode to reduce energy consumption.…”

“…In contrast, in post-facto synchronization, the clock synchronization protocol runs depending upon the requirement. A-priori clock synchronization protocols [22,23] in WSNs have the advantage of synchronizing the clocks over the entire WSNs, and most of the clock synchronization protocols in WSNs are a-priori based. In post-facto synchronization [20], clock synchronization is carried out whenever it is required to be done, and therefore, post-facto clock synchronization process is usually useful for synchronizing clocks locally over a small area of a WSN.…”

“…During this switchover period, depending on the protocol, it may or may not provide synchronized logical clock time. If a clock synchronization protocol uses multiple reference nodes [22], the protocol is inherently fault-tolerant to the failure of a few reference nodes (at least one reference node should be alive). If a clock synchronization protocol uses a hierarchical structure [24], and an internal node of the hierarchical structure fails, the protocol needs to be made fault-tolerant specifically to deal with such failures.…”