Using Machine Learning to Maintain Pub/Sub System QoS in Dynamic Environments

Free registration required

Executive Summary

Quality-of-Service (QoS)-enabled publish/subscribe (pub/sub) middleware provides powerful support for scalable data dissemination. It is hard, however, to maintain specified QoS properties (such as reliability and latency) in dynamic environments (such as disaster relief operations or power grids). For example, managing QoS manually is often not feasible in dynamic systems due to slow human response times, the complexity of managing multiple interrelated QoS settings, and the scale of the systems being managed. For certain applications the systems must be able to reflect on the conditions of their environment and adapt accordingly.

  • Format: PDF
  • Size: 123.08 KB