Rethinking network management: Models, data-mining and self-learning

Abstract

Abstract-Network Service Providers are struggling to reduce cost and still improve customer satisfaction. We have looked at three underlying challenges to achieve these goals; an overwhelming flow of low-quality alarms, understanding the structure and quality of the delivered services, and automation of service configuration. This thesis proposes solutions in these areas based on domain-specific languages, data-mining and selflearning. Most of the solutions have been validated based on data from a large service provider.We look at how domain-models can be used to capture explicit knowledge for alarms and services. In addition, we apply datamining and self-learning techniques to capture tacit knowledge. The validation shows that models improve the quality of alarm and service models, and enables automatic rendering of functions like root cause correlation, service and SLA status, as well as service configuration.The data-mining and self-learning solutions show that we can learn from available decisions made by experts and automatically assign alarm priorities.

ALACA: A platform for dynamic alarm collection and alert notification in network management systems

“…Wallin et al aim to establish a general state machine for the alarms and propose various interpretations of the alarm state based on the context information in Wallin 6 and focus on validating the alarm model based on the static analysis in Wallin. 22 Duarte et al propose an SNMP focused specific language called ANEMONA that focuses on alarm monitoring solutions. 7 Li et al propose a neural network and FP-tree approach to perform weighted association rule mining in the alarm domain.…”

“…In Wietgrefe et al, the authors consider finding the root cause alarm using neural networks. Wallin et al aim to establish a general state machine for the alarms and propose various interpretations of the alarm state based on the context information in Wallin and focus on validating the alarm model based on the static analysis in Wallin . Duarte et al propose an SNMP focused specific language called ANEMONA that focuses on alarm monitoring solutions .…”

Self-modeling based diagnosis of Software-Defined Networks

“…Such diversity and complexity implies a huge cost and effort to deploy new services and manage all the equipment, what prevents Operators from improving network and service quality and reliability to satisfy the market needs and cope with the aggressive competitiveness of the industry. One symptom of this complexity is the overwhelming and unmanageable amount of alarms received by medium-sized operational teams, estimated in millions per day [1]. This urgently requires automation and intelligence to keep this management under control.…”