Synchronous timing

In synchronous distributed systems the processing of all the events occurs at time which corresponds to pulses defined by the local clocks of the system's nodes. The basic principles of the synchronization which are necessary for the validation of its simulation are the following:

• The notion of local clock at each node. This means that each node should sense the time in pulses defined by its local clock.
• All the clocks have the same pulse duration.
• The link propagation delay of the messages is fixed.
• The processing of more than one events in a single step is allowed.
• The messages received by a node are processed in its local clock's next pulse. This implies the need for simulation of some local memory operation at each node of the system.
• The nodes may spontaneously wake up at random multiples of their step.

Synchronous systems are ideal according to the transfer delay on their links. Considering the pulse duration of the node's clock as unit, the link propagation delay is forced to be less than or equal to this unit. This way, it is guarranted that a process will receive the messages sent to it in time and process them at its clock next pulse.

The main difference between synchronous and asynchronous systems is that while in asynchronous systems event time determines system time, in synchronous systems, system time determines event time.

SEE ALSO: Asynchronous and Archimedean timing, Synchronizers (Section 4.3.1) and Making new networks (Section 6).