PASTA: Deriving Rich Presence for Converged Telecommunications Network Applications

Presence has rapidly evolved from its roots in instant messaging status to become a standard event mechanism for aggregating context about individuals, devices and abstract entities (such as meetings, activities and location coordinates). Scalable tracking and dissemination of presence attributes, across multiple communication channels, is a critical requirement for future IMS-based converged network architectures. This paper presents PASTA (presence advanced services for Telco applications), an enhancement to a presence server that employs scalable, rule-based event correlation middleware and machine-learning engines, to automatically derive higher-level, "semantically richer" presence information about individual or multiple presentities. By augmenting the event store of the presence server with such "rich presence", PASTA enables applications and users to subscribe to presence notifications at higher-levels of abstraction. Such abstractions could significantly reduce the volume of network traffic resulting from low-level presence updates, but also promote privacy preferences by permitting presentities to control the exposure of raw presence information. We describe the design, implementation and performance analysis of PASTA and discuss open future issues associated with "rich presence".