Parameterized subscriptions in publish/subscribe systems

  • Authors:
  • Yongqiang Huang;Hector Garcia-Molina

  • Affiliations:
  • Department of Computer Science, Stanford University, Stanford, CA 94305, United States;Department of Computer Science, Stanford University, Stanford, CA 94305, United States

  • Venue:
  • Data & Knowledge Engineering
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Traditional publish/subscribe systems commonly deal with static subscriptions, whose event filtering criteria stay fixed once defined. Although systems with static subscriptions are simpler to implement, there are cases where the subscription criterion involves state that changes frequently over time. Rather than having the user re-submit his/her subscription repeatedly, we propose parameterized subscriptions as a systematic solution for adaptive subscriptions. Parameterized subscriptions depend on one or more parameters, which are state variables stored and maintained automatically by the publish/subscribe servers. In this paper, we modify traditional publish/subscribe protocols in order to deal with parameterized subscriptions. We then look at certain optimizations that improve the efficiency by controlling where and how much state is allocated in the system. Finally, we present a simple evaluation framework to illustrate the fundamental operating differences between several schemes.