MAR 01 2017
SELFNET Architecture
The main objective of SELFNET is to design and implement an autonomic network management framework enabling status awareness of the network. This framework will deliver self-organizing capabilities to manage network infrastructures by automatically detecting and mitigating a range of common (mobile) network problems that are, for the moment, manually addressed by network operators.
This approach will significantly reduce operational costs. In support of these aims SELFNET considers the following key management tasks:
Automated network monitoring: SELFNET is capable of deploying NFV applications to monitor the network. These applications are considered sensors that can be spread across the mobile network to enable context awareness in both traditional low-level metrics and a set of high-level Health of Network (HoN) metrics. In particular, HoN metrics will facilitate greater control by delivering direct and precise knowledge of network status. These capabilities will enable network administrators to reduce significantly the time spent in identification and diagnosis of these concerns.
Autonomic network maintenance: SELFNET defines high-level tactical corrective and preventive measures to enable autonomic reactive and proactive network maintenance. These tactical measures define the reactive actions in response to detected network problems, and define how to proactively configure, protect, improve and repair the network capabilities. The combination of reactive and proactive strategies with advanced intelligent algorithms drives the autonomic response of the system.
Automated deployment of network management tools: The management of a set of automatic reactive and proactive actions against existing or potential network problems respectively, implies automatic deployment of various distributed services in the network. For example, intrusion protection tools and honeynets could be automatically deployed in order to mitigate different kinds of security threats.
Automated network service provisioning: This capability is related to the efficient management and optimization of the usage and deployment of NFV applications. Efficiency will be achieved by dynamic smart selection of the best location where the services should be deployed (or migrated to) for optimization of network performance, health and security, and by timely release of the resources occupied by the services once they have completed their jobs.
One of the main results achieved by the project was the specification of an architecture that enables de implementation of the previously mentioned tasks:
The SELFNET architecture is partitioned in six layers in order to logically separate the different areas of concern. These layers are (i) the Infrastructure Layer, (ii) the Virtualized Network Layer, (iii) the SON Control Layer, (iv) the SON Autonomic layer
