NeTS – Next Generation Network Operations and Management
Start Date: 2010 End Date: 2013
PIs: Ricardo Morla (FEUP/INESC Porto), Hyong Kim (CMU)
Dual Degree Ph.D. Students: Tiago Filipe Rodrigues de Carvalho (Electrical and Computer Engineering) and Cheng Wang (Electrical and Computer Engineering)
Teams: INESC Porto, Instituto de Telecomunicações (IT), Faculdade de Engenharia da Universidade do Porto (FEUP), Faculdade de Ciências da Universidade do Porto (FCUP), Instituto Superior Técnico da Universidade Ténica de Lisboa (IST/UTL), Carnegie Mellon University (CMU)
Company: Portugal Telecom
Keywords: Enterprise Network Operations and Management; IP/TV Access Network; Network Abstraction; Probabilistic Graphical Models
It is well known that networks are difficult to manage and operate. Companies are spending more resources on the daily management and operations of their networks than investing on developing and launching new IT services. Studies have shown that scheduled maintenance and upgrades can account for more than 30% of network outages in Tier 1 ISPs, and operator errors are common and can be the root causes for more than 50% of failures in computer systems and networks. The abstractions provided in the Internet are not appropriate for network operations – they are combined, in an ad-hoc manner and mostly ineffectively, to derive information necessary to operate a network. This mismatch plagues day-to-day network operations with a number of challenges. The excessive cost of network management is mainly due to the large scale of networks, the heterogeneity of different technologies, and their subtle interactions. The goal of NeTS is to develop a novel network operation and management framework that departs from conventional approaches through a cross-disciplinary research collaboration based on hierarchical network abstraction modeling, structure learning of probabilistic graphical models for machine learning, and wavelet and kernel-based signal processing technologies. Our approach to this research problem is significantly different from conventional approaches that focus on monitoring and correlation analysis. We bring together insights from network abstraction and statistical and heuristic machine learning methods and closely tie them to build an intelligent network abstraction model. The network configuration dictates the behavior of the network control plane and the network control plane dictates the behavior of the data plane. We bring configuration, control and data plane abstraction models together with probabilistic graphical models to derive an intelligent and predictable network abstraction model of networks. The proposed models will be deployed in an European ISP in 3 phases for evaluation of performance and cost efficiency. The deployment experiments will further our understanding of abstract modeling.
Articles published:
New Network Management Software will Improve the Quality of IPTV Service (CMU Portugal Website, June 2013)
Joining Universities and Industries Can Create Value to Portugal (Ciência Hoje, March 2011)
