Timing-based Connection Management,

Marios Mavronicolas

A connection management protocol establishes and handles a connection between two hosts across a wide-area network to allow reliable message delivery. It is common, in such networks, for messages to be duplicated and reordered, and and for nodes to crash. Under such failure circumstances, two performance parameters of particular interest are the message delivery time, which is the time required to deliver a message, and the quiescence time, which is the time that elapses between periods of quiescence, in which the receiving host deletes all earlier connection records and returns to an initial state.

In a distinguished work, Kleinberg, Attiya and Lynch (Proceedings of the 3rd Israel Symposium on the Theory of Computing and Systems, pp.\ 258--267, January 1995) initiated the study of trade-offs between message delivery time and quiescence time from the perspective of the amount of synchrony provided by the processors' clocks. Their work was subsequently continued and extended by Mavronicolas and Papadakis in several directions.

In this talk, we present a survey of these all of these results and their proof techniques; we shall pay particular attention on timing models and their inherent impact on the efficiency of connection management protocols. Towards the end, we shall mention interesting possibilities for further research on timing-based connection management.