ONTOLOGY ADAPTER

Network Management System Interface Model

Lingli Meng, Lusheng Yan

Network Management Research Center, Beijing University of Posts and Telecommunications

Beijing, Popular Republic of China

Wenjing Li

Department of Computer science and technology, Beijing University of Posts and Telecommunications

Beijing, Popular Republic of China

Keywords: Ontology Adapter, Interface Model.

Abstract: This paper proposes a new method to define the interface model of network management system, that is

ontology adapter. This model includes three parts, which are ontology agent, ontology knowledge base and

ontology resource description. We can realize the uniform presentation of different network resource

interface information using them. Therefore, we can take this model as a common data platform to offer the

interface information to the network management system.

1 INTRODUCTION

The telecommunication network and business have

such a great speed development, which cause more

network resource needed to be managed in the

managed network system (Luoming,2001). However,

the traditional network management interface model

of the TMN proposed by the ITU-T is not suitable

for changes of the network technology, and can only

satisfy the static management of the managed

network resources, adjusting badly to all kinds of the

existing interface technology. Therefore, the

interface model is encountering the great challenges

(Luoming ,2003)(Zhipeng,2007):

1) variety of managed network resources. For the

complicated managed network resources are

included under the background of network

management integration, it is an important problem

how to describe these managed resources in the

interface layer, and give the uniform description of

them before offering to the NMS(Network

Management System).

2) diversity of interface description. At present,

there exactly exists the diversity definition of private

interfaces among all equipment vendors, although

the ITU-T has standardized these interface

descriptions. It can cause the inconsistence

presentation of the data, and not share the managed

resources with different systems, as well as be

deficient in flexibility when NMS collects the

information from different interface equipments.

Therefore, how to find a better interface model to

integrate all these different descriptions of network

resources, realize the uniform data presentation,

which will be a creative breakout of interface model

in NMS.

So as to solve these problems above, we proposal

an interface adapter model using ontology language.

The rest of this paper is organized as follows:

Section 2 introduces the existing method of interface

definition in NMS. In section 3, we give a glimpse

of ontology language, presenting the ontology

language OWL. We describe the detail of the

ontology interface adapter model in section 4.

Finally section 5 would conclude our paper and give

the future work.

2 INTERFACE ANALYSIS

ITU-T proposed five layers architecture of TMN in

1990’s. The five layers are composed of business

management, service management, network

318

Meng L., Yan L. and Li W. (2008).

ONTOLOGY ADAPTER - Network Management System Interface Model.

In Proceedings of the Fifth International Conference on Informatics in Control, Automation and Robotics - ICSO, pages 318-321

DOI: 10.5220/0001491903180321

 SciTePress

management, network element management and

network element, which are managed by NMS.

At present, it has been defined that several methods

of the interface between NMS and managed

resources, Such as SNMP, CORBA and XML. They

design the interface model distinguishingly

according to different prospective.

2.1 SNMP

SNMP is a standard protocol to manage the IP

network. It can monitor and control the variable

set,as well as monitor the two data format of

equipments, SMI and MIB. SNMP adopts the model

of “Management Process---Agent Process”. When

the Management Process sends an order to the

Managed Object, the Managed Object receives the

order by the Agent Process. Then the Agent process

also sends back the response to the Management

Process. That is, the communication between

Management Process and Managed Object is not in

direct, but using the Agent Process.

SMI is a method of defining the management

object using SNMP. It based on object-oriented, not

using encapsulation, inheritance, polymorphism. So

it simplifies the management information models. It

is on behalf of an instance of a class by using table

to represent the data type, table, column, and row.

(Jorge Vergara, 2003)

2.2 Corba

CORBA is proposed by the OMG. It is an object-

oriented software development as well as application

platform, with the function of portable, reusable and

connection, under distributed heterogeneous

circumstance.

IDL, which ITU-T use to define the TMN and

UTRAD. It is an object-oriented description

language. By defining of the IDL, classes are

interfaces between different models. They have

attributes and methods. IDL can be mapped to

different languages, such as C++, JAVA, Ada,

Cobol and SmallTalk.

2.3 XML

XML is given by the W3C. It is an information

processing tool, which has on relevant to platform,

software and hardware. XML has the feature of

flexibility, easy-realization, easy-extension, as well

as low cost. At present, we only use this tool to

define format of the configuration files and the

performance files in NMS.

3 ONTOLOGY DESCRIPTION

Ontology is an explicit specification of a

conceptualization. The term is borrowed from

philosophy, where an ontology is a systematic

account of existence. In 1990’s, ontology was

brought into the field of the knowledge

representation, knowledge sharing, semantic reason

and so on.

In recent years, it has been great development of the

applications and research of ontology, which NMS

has been focusing on. There may exist the

possibility of the application of ontology in network

management information model (Jorge, Vergara,

2003), the relations among ontology mapping,

network management operation (Wong, 2005), as

well as special management information model

based on ontology. However, it is still short of the

research to design interface adapter model using

ontology language in NMS. We will propose an

ontology interface adapter model next section.

3.1 OWL

Generally speaking, we use a special ontology

language to describe ontology. At present, the main

ontology description languages are DAML, OIL,

KIF, and OWL, among which, the OWL proposed

by the W3C has the great effect.

The OWL (Web Ontology Language) is designed

for processing the content of information instead of

just presenting information to humans. OWL

facilitates greater machine interpretability of Web

content than those supported by XML, RDF, and

RDF Schema (RDF-S), by providing additional

vocabulary along with a formal semantics. OWL has

three increasingly-expressive sublanguages: OWL

Lite, OWL DL, and OWL Full.

OWL has been applied widely to module in most

kinds of domain ontology: W3C has used OWL in

the ontology model of Web Service; Using OWL

defined the analysis model, which can be adapted to

test circumstance (Bodkin, 2005).

4 ONTOLOGY ADAPTER

DESIGN

4.1 Ontology Adapter Model Analysis

We propose and discuss the ontology adapter model

in the interface definition below. We divide

ontology adapter model into three parts: ontology

ONTOLOGY ADAPTER - Network Management System Interface Model

319

agent, ontology knowledgebase and ontology

resource description. Figure 1 shows the model.

Now we will introduce what functions the three

parts have, and how to realize the uniform

description of the interface information when NE

management Layer sends the network resource to

the NMS. First of all, the network resource

information has been defined well in the NE

Management Layer.

Step1, ontology agent receives resource

information, represented by the SMI, IDL or XML

from the NE Management Layer. Then they will be

transformed to the uniform representation using

OWL. If just the

Figure 1: Ontology Adapter Model.

information describe static resources, which can be

dealt with by the ontology agent; otherwise

including dynamic information, which would be

forwarded to the Ontology Knowledgebase. It will

give a further deal.

Step2, ontology knowledgebase is used to

transform the complex interface information. There

exists a database, which stores more transduction

rules between OWL and other interface description

languages, all that the ontology agent can’t deal

with, will be sent to the ontology knowledgebase.

Step3, ontology resource description, in which, we

can complete the interface information

representation finally. After receiving the static

information sent by ontology agent and the dynamic

ones dealt by ontology knowledgebase, ontology

resource description module will integrated both of

them, and present a uniform format to NMS.

4.2 Ontology Adapter based on XML

This section we give a simple realization of ontology

adapter model. We have defined some

transformation rules and a simple ontology resource

description frame using OWL.

4.2.1 Ontology Agent

Figure2 shows the process how ontology receives

the interface information. Firstly, read interface

information; Secondly, analyze language type and

judge whether it present dynamic information;

thirdly, transform the format, finally, the final OWL

representation is formed.

In transform mode 1, take IDL as an example,

there are some rules below:

Rule 1:

module IDL, →

xmls:

NameSpace OWL

Figure 2: Ontology Agent Flow Chart.

Case study:

module TypeDemo { …… } →

xmlns="http://www.w3.org/TR/2004/REC-

owl-guide-20040210/TypeDemo#"

xml:base="http://www.w3.org/TR/2004/REC-

owl-guide-20040210/TypeDemo#"

Rule 2:

Interface IDL, → class OWL

Case study:

interface Father { ……. }; →

<owl:Class rdf:ID="Father"/>

Rule3:

Attribute IDL, →

property OWL

Case study:

attribute string name; →

<owl:DatatypeProperty rdf:ID="name">

<rdfs:domain rdf:resource="#Father"/>

<rdfs:range rdf:resource="&xsd;string"/>

</owl:DatatypeProperty>

Role4:

fundamental datatype IDL, →

xsd:

datatype OWL

Case study: String → xsd:string

Note: there are some data types that can’t be

mapped directly, therefore, you must do some

changes. Such as, char → xsd:byte.

Rules above, which just aim at static information.

There are still methods to represent dynamic

ICINCO 2008 - International Conference on Informatics in Control, Automation and Robotics

320

information, such as operation and context. We do

such analysis in ontology knowledgebase. Simply,

we use OWL text to descript these methods. Here,

we only offer a sample:

e.g. Method1 (var1, var2)

IDL →

Method1 has var1, var2

OWL.

In transform mode 2, we translate XML into OWL,

we should distinguish the differences between the

two kinds of data types, the fundamental data type

and the user-defined ones. Because the user-defined

data type is so complex that we must avoid the

semantic loss at our best, while transforming such

information.

4.2.2 Ontology Resource Description

In this part, we represent a uniform network resource

description framework using OWL. When ontology

agent and ontology knowledgebase send the

transformed information to ontology resource

description, it will integrate these two parts, and give

a uniform representation using OWL. We now give

a sample description, as follows:

The definition of class:

<owl:Class rdf:ID=

Equipment

<rdfs:subClassOf rdf:resource=

PhysicalME

<rdfs:subClassOf>

<owl:Restriction>

<owl:hasDatatypeProperty rdf:resource=

equipType

<owl:ConstraintRequirement

rdf:datatype="&xsd;string">

</owl:ConstraintRequirement>

</owl:hasDatatypeProperty>

<owl:hasObjectProperty

rdf:resource=hasInterface/>

</owl:Restriction>

</rdfs:subClassOf>

<rdfs:subClassOf>

<owl:Restriction>

<owl:onProperty rdf:resource=

“

#hasInterface

”

<owl: minCardinality

rdf:datatype="&xsd;int"> 1

</owl: minCardinality>

</owl:Restriction>

</rdfs:subClassOf>

<rdfs:subClassOf>

<owl:Restriction>

<owl:onProperty rdf:resource=

"# hasInterface "/>

<owl:haslValuesFrom rdf:resource=

"# Interface "/>

</owl:Restriction>

</rdfs:subClassOf>

</owl:Class>

The detail definition of attributes for the above:

<owl:DatatypeProperty rdf:ID=" equipType ">

<rdf:type rdf:resource=

"&owl;FunctionalProperty"/>

<rdfs:domain rdf:resource="#Equipment"/>

<rdfs:range rdf:resource="&xsd;int"/>

</owl:DatatypeProperty>

<owl:ObjectProperty rdf:ID="hasInterface">

<rdfs:domain rdf:resource="#Equipment"/>

<rdfs:range rdf:resource="#Interface"/>

<owl:inverseOf

rdf:resource="#belongEquipment "/>

</owl:ObjectProperty>

5 CONCLUSIONS

It is a challenging research of the integration of

network resource interface information. The

ontology adapter model we presented has solved

how to integrate these different interface

informations in architecture. Our future work will

concentrate on the research of offering a specific

method and realization, using which we can success

to transform existing interface presentation language

into OWL. If possible, a tool which can be used to

produce this transformation automatically is

expected.

REFERENCES

Luoming, Meng. Qi feng. Modern Network Management

Technology ,M. Beijing: Press of BUPT, 2001.

Luoming, Meng. Network Management: Problems,

Progress and Prospect, J.Journal of BUPT, 26-2: 1-8,

2003.

Zhipeng,Gao. Ontology-based Modeling Methods,Model

and Application of Shared Management Information,

D. Beijing: Doctoral Dissertation of BUPT, LAB. of

BUPT,2007.

Jorge.E López de Vergara, Víctor A. Villagrá,Juan I. and

Julio Berrocal. Ontologies: giving semantics to

network management models, J. IEEE Network, 17-

3:15-21, 2003

Wong, A.K.Y. Ray, P. Parameswaran, N. Strassner,

J. Ontology mapping for the interoperability problem

in network management, J. Selected Areas in

Communications, IEEE Journal on. 2005, Volume: 23

Issue: 10: 2058 – 2068.

Bodkin, M., Harris, M., Helton, A. Use of ontologies for

decision support generation and analysis,A.

Autotestcon, 2005. IEEE. 2005:684

ONTOLOGY ADAPTER - Network Management System Interface Model

321