Browse Prior Art Database

Model-driven Configuration Management based on Directed Acyclic Graphs

IP.com Disclosure Number: IPCOM000130967D
Original Publication Date: 2005-Dec-10
Included in the Prior Art Database: 2005-Dec-10
Document File: 2 page(s) / 179K

Publishing Venue

Siemens

Related People

Juergen Carstens: CONTACT

Abstract

Telecommunication networks consist of various Network Elements of different types, e.g. MSC (Mobile Switching Center), SGSN (Serving GPRS (General Packet Radio Service) Support Node), or GGSN (Gateway GPRS Support Node). These Network Elements (NE, Network Element) have to be configured properly by the operator before the required service can be offered. The process of configuring the NE's properly is called Configuration Management (CM), which can be characterized as follows: • An NE implements a Finite State Machine where the current contents of NE Database reflect its current CM state. The network operator can change this CM state by using the given set of Operation & Maintenance commands (O&M commands). The set of O&M commands is given by the NE, in particular by the implementation of the O&M interface of this NE. The set of O&M commands is limited to certain language instructions. • The network operator requires a well-structured overview of the NE's CM state. A state-of-the-art representation is a table-based overview where the NE Database contents are displayed in tables. The instances stored in the NE Database are reflected as rows of the tables. The operator modifies these tables. Subsequently, these changes are required to be incorporated in the NE Databases. • The purpose of a CM application is to mediate between the tables required by the operator and the NE's O&M interface. The NE's CM state and the O&M commands have to be mapped to the table entries and vice versa. CM applications are offered to network operators to facilitate the CM process. But these products are hard-coded in such a manner that the mediating logic of the CM application is attuned to a certain NE interface, to a certain subset of O&M commands and to NE object classes. The main drawback of these approaches is that the relationship and dependencies between the O&M commands and the NE Database instances are hard-coded in the CM applications. In case of new NE types or a new NE interface version the source code of the CM application's mediating logic has to be extended. This usually leads to significant R&D (Research & Development) and maintenance costs for CM applications.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

S

Model-driven Configuration Management based on Directed Acyclic Graphs

Idea: Masoud Amri, DE-Berlin; Dr. Ralf Kattner, DE-Berlin; Aljaz Noe, SI-Kranj; Dr. Heiko Selber,

DE-Berlin

Telecommunication networks consist of various Network Elements of different types, e.g. MSC (Mobile Switching Center), SGSN (Serving GPRS (General Packet Radio Service) Support Node), or GGSN (Gateway GPRS Support Node). These Network Elements (NE, Network Element) have to be configured properly by the operator before the required service can be offered. The process of configuring the NE's properly is called Configuration Management (CM), which can be characterized as follows:

* An NE implements a Finite State Machine where the current contents of NE Database reflect its current CM state. The network operator can change this CM state by using the given set of Operation & Maintenance commands (O&M commands). The set of O&M commands is given by the NE, in particular by the implementation of the O&M interface of this NE. The set of O&M commands is limited to certain language instructions.

* The network operator requires a well-structured overview of the NE's CM state. A state-of-the- art representation is a table-based overview where the NE Database contents are displayed in tables. The instances stored in the NE Database are reflected as rows of the tables. The operator modifies these tables. Subsequently, these changes are required to be incorporated in the NE Databases.

* The purpose of a CM application is to mediate between the tables required by the operator and the NE's O&M interface. The NE's CM state and the O&M commands have to be mapped to the table entries and vice versa.

CM applications are offered to network operators to facilitate the CM process. But these products are hard-coded in such a manner that the mediating logic of the CM application is attuned to a certain NE interface, to a certain subset of O&M commands and to NE object classes. The main drawback of these approaches is that the relationship and dependencies between the O&M commands and the NE Database instances are hard-coded in the CM applications. In case of new NE types or a new NE interface version the source code of the CM application's mediating logic has to be extended. This usually leads to significant R&D (Research & Development) and maintenance costs for CM applications.

In order to reduce the costs, the following idea proposes to use Directed Acyclic Graphs (DAG) as the central data structure of CM application's mediation logic. This data structure reflects the current NE's CM state as follows:

* A node of the DAG generates an instance in the NE Database, e.g. an ATM (Asynchronous Transfer Mode) path instance.

* The arcs model the relationship and dependencies between the NE Database instances. For example, a path has to exist before an ATM channel can be created and assigned to an ATM path.

* Every connected NE is generated by one DA...