Probabilistic Network Management

The Probabilistic Network Management (PNM) paradigm specifies non-deterministic methods in which decisions are based on probabilistic objectives and richer statistical monitoring information, rather than on strict performance guarantees and measurements.

Compared to current network management technology, PNM approaches provide new effective means to resource-efficient and flexible network management solutions. This is achieved by combining statistical modeling with adaptive mechanisms and decentralized operation. Decentralized probabilistic management enables prediction of network performance, adaptivity to network variations, robustness and improved failure-resilience, thereby allowing for optimized and cost-efficient network operation and management. 

We currently develop PNM approaches within the following areas:

  • Mobility - prediction of mobility patterns and paging sequences
  • Network Monitoring - autonomous detection and localization of network faults and anomalies, change detection and event correlation
  • Goal Translation - mapping of high-level goals to low-level parameters with probabilistic feedback to maintain SLAs
  • Distributed Data Management - efficient caching and methods targeting management of big data
  • Data Analysis - mining of event processes; detection of faults, anomalies and trend changes; event correlation

Approaches developed under the PNM paradigm target the challenges in future networked systems (e.g. scalability, adaptivity, controllability, etc), and encompass features that are highly valuable for dynamic networks such as mobile systems, sensor networks, Internet of Things, and cross-linked systems such as network clouds. 

For more information, please contact:

Daniel Gillblad, dgi [at] sics.se

Rebecca Steinert, rebste [at] sics.se

 

Publications
Number of items: 7.

Steinert, Rebecca and Gillblad, Daniel (2012) Performance Evaluation of a Distributed and Probabilistic Network Monitoring Approach. In: International Conference on Network and Service Management, 22-26 October 2012, Las Vegas, Nevada, USA.

Görnerup, Olof (2012) Scalable Mining of Common Routes in Mobile Communication Network Traffic Data. In: Tenth International Conference on Pervasive Computing, 18-22 June 2012, Newcastle, UK. (In Press)

Steinert, Rebecca and Gestrelius, Sara and Gillblad, Daniel (2011) A Distributed Spatio-Temporal Event Correlation Protocol for Multi-Layer Virtual Networks. In: Globcom 2011.

Bjurling, Björn and Steinert, Rebecca and Gillblad, Daniel (2011) Translation of Probabilistic QoS in a Decentralized Hierarchical Setting. In: 13th Symposium of APNOMS 2011.

Gonzalez Prieto, Alberto and Gillblad, Daniel and Steinert, Rebecca and Miron, Avi (2011) Towards Decentralized Probabilistic Management. IEEE Communications Magazine, 49 (7). pp. 80-86. ISSN 0163-6804

Steinert, Rebecca and Gillblad, Daniel (2010) Long-term adaptation and distributed detection of local network changes. In: IEEE Global Telecommunications Conference GLOBECOM 2010, 6-10 December 2010, Miami, FL, USA.

Steinert, Rebecca and Gillblad, Daniel (2010) Towards Distributed and Adaptive Detection and Localisation of Network Faults. In: 2010 Sixth Advanced International Conference on Telecommunications, 9-15 May 2010, Barcelona, Spain.

This list was generated on Wed Aug 23 02:49:41 2017 CEST.