Network Management Tool Architecture For Heterogeneous System Computer Science Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Network management is concerned with observing, monitoring, testing, configuring, and troubleshooting the different network components and users. The management process wraps all the network elements starting from the users, through the systems applications and data till the end systems and low network infrastructure. Three main aspects form the whole NM area namely: Human where human mangers define the organization approaches and its components, Methodology which draw the NM architecture and defines the functions to be run, and the Instrumentation aspect which contains system information collection, processing and reporting, and prediction of network probable problems. Figure 1 shows those three aspects.
NM aims to give the network managers a complete view of the whole network through a visualized network management tool (NMT). The management functions that can be managed by such tool are divided by the International Organization for Standards (ISO) [1] into five categories: Fault, Configuration, Accounting, Performance, and Security as shown in Table 1.
Table 1: FCAPS functions



Fault Management (FM)

Detection, isolation and correction of abnormal operations.

Configuration Management (CM)

Maintains an accurate inventory in a changing environment and consistency of operating parameters.

Accounting Management (AM)

Measures network usage and compute charges for that usage.

Performance Management (PM)

Concerned with gathering statistics, measuring utilisation of resources, tuning, etc.

Security management (SM)

Controls access to services, network infrastructure and the management system itself

An example of NM application is shown in figure 2 where a user interface used to manage all the functions for Software, Hardware, Firmware, and Users as well.

Most of the existing Network Management systems are developed in an individual fashion, where each system grows with minimum consideration of other systems, which creates many systems with low level of communication among them. The most used protocols in NM systems are:

SNMP (Simple Network Management Protocol) is an IETF defector standard known as RFC (Requests For Commands). SNMP is a framework for managing internet device using the TCP/IP protocol suite [4]. The SNMP management model contains a management workstation where the management process takes place, host devices, and all the available network peripherals with a network management Agent embedded inside as shown in Figure 3. The manager and agent style is used in SNMP where the manager monitors a group of agents (other network devices). The relationship between manager and agent is shown in figure 4 below, where a manager can check the agent behavior through a set of parameters, and can force it to behave in certain way by resetting those parameters values. Agents can alarm the manager of any odd situation. Different network managers can exchange information about each other networks as well.

In fact the SNMP works in cooperation with the SMI (Structure of Management Information) and the MIB (Management Information Base) protocols. The SMI is responsible for defining the general rules for naming objects, defining objects types, and show how to encode objects and values [5]. While the MIB is a conceptual database that can be distributed on different sites or assembled in single location. It creates all the objects needed and names them according to the SMI rules, then assigns them to their proper type. For each entity for example a device a set of defined objects will be created and kept in the database. Figure 5 shows the ISO MIB that consists of all the data which reflect the FCAPS that are requested by the different network management architectures.



TMN (Telecommunication Management Network). The TMN was defined by the ITU-T (International Telecommunication Union -Telecommunication Standardization Sector). It is a framework to achieve communication between heterogeneous telecommunication networks. TMN defines a standard interface for network elements that handle the communication process, in this way network elements can be managed by a single network management system regardless of their different manufacturers. The framework identify four logical layers of network management[6]:

  • Network Management Layer: Performs network resources distribution functions such as configuration and control.
  • Element Management Layer: Contains individual network elements handling functions ,this includes alarm management, information handling, backup, logging, and hardware/ software maintenance.
  • Business Management Layer: Performs functions related to business aspects, analyzes trends and quality issues, provide a basis for billing and other financial reports.
  • Service Management Layer:Performs functions for network services handling: definition, administration and charging of services.

FCAPS is a framework defined by the ISO and the name is a contraction of the five management categories that were mentioned before in Table 1: Fault, Configuration, Accounting, Performance, and Security. At the beginning the approach was to create a separate protocol for each one of those functions, but it turned out that those five protocols would be very similar so the ISO decided to make them all under one protocol, named then as CMIP (Common Management Information Protocol) [6].


There are number of network management tool that claim to integrate network management functions. Those tools came to manage network components from the same vendor or the same service provider. Therefore we are mainly interested in bulding a comprehensive network management tool through a single interface, which means we aim to provide a central control in a distributed management fashion. In this section we will highlight several integrated network management architecture.

Alcatel [7] offers a complete suite of network management systems for all sub networks in the MSP's (Mobile Service Providers) environment. They can be integrated with Alcatel's mobile OSS (Operating Support System) solutions to provide full end-to-end service management. Alcatel Network Management Systems (NMS) are based on a coherent architecture that benefits from ALMAP's common features, common user interfaces and scalable architecture. To help MSPs to meet these new challenges, Alcatel NMS and OSS have been designed as open solutions that facilitate the smooth integration of multi-technology and multi-vendor elements. Alcatel integrated fault management based on MicroMuse Netcool® to consolidate real-time alarm information across all network elements and application servers, providing real-time alerts relating to problems that affect service, thereby reducing the time to repair. Alcatel performance management based on Metrica™/NPR includes the collection of performance data and its consolidation across the entire network and all application servers to provide end-to-end QoS indicators. Alcatel problem management based on Action Request System® from Remedy® integrates with previous applications, supporting network and service problem resolution by handling the workflow of problem resolution. It reduces network and service downtime and thus improves QoS and customer satisfaction. It can provide repair time and reaction time metrics as a basis for evaluating Service Level Agreements (SLA). The Evolium™ OMC-CN provides efficient cross-network element configuration capabilities and extensive performance monitoring functions tuned to the core network technologies. The Evolium™ OMC-R combines the integration of RNO (Radio Network Optimizer) with Alcatel's worldwide expertise in 2G mobile networks and its UMTS experience in Japan, to realize the best radio network management system on the market.

In the CORDS project [8], several management tools have been developed independently. The tools are NetMod (Network Modeling Tool), NEST, NETMATE (Network Management, Analysis, and Testing Environment), Hy+ and Shoshin Distributed Debugger. NetMod is a software package that predicts the performance of new network technologies in a large-scale local network environment. NEST is a graphical environment for distributed network systems rapid prototyping and simulation developed at Columbia University. The NETMATE was developed at Columbia University to provide a unified and comprehensive software environment for network management to oversee and orchestrate the operation s of diverse devices and protocol. Hy+ is a visual database system being developed at the University of Toronto. The system is capable of manipulating data by visually expressed queries on large complex systems such as a computer network. Shoshin Distributed Debugger is being developed at the University of Waterloo to support debugging of distributed and parallel applications.

Netdisco (Network Discovery and Management) [9] is an Open Source web-based network management tool hosted by Sourceforge. This project was headed by Eric Miller, Bill Fenner, and Max Baker. Netdisco is a network management application targeted at large corporate and university networks. It integrated devices and software into the one integrated system. The integrated devices are Airespace, Allied Telesyn, Aruba, Asante, Bay, Cisco, Dell, Enterasys Networks, Extreme Networks, Foundry Networks, HP, Juniper, Net-SNMP, NetScreen, Nortel, Proxim, Sun, Synoptics and Zyxel. These devices provide hardware like switch, router and hub. The integrated software are Perl, Mason, Net-SNMP, PostgreSQL, Apache 1, GraphViz and MIBs. Table 1 below summarizes all three network management architecture and tool as been describe above.




Same Service Provider

Different Service Provider


Mobile network

i. Alcatel The Evolium™ OMC-CN

ii. Alcatel Evolium™ OMC-R

i. MicroMuse Netcool®

ii. Metrica™/NPR

iii. Action Request System® from Remedy®


Large-scale local network

Columbia University: NEST and NETMATE

University of Toronto: Hy+, University of Waterloo: Shoshin Distributed Debugger




Devices : Airespace, Allied Telesyn, Aruba, Asante, Bay, Cisco, Dell, Enterasys Networks, Extreme Networks, Foundry Networks, HP, Juniper, Net-SNMP, NetScreen, Nortel, Proxim, Sun, Synoptics and Zyxel.

Software: Perl, Mason, Net-SNMP, PostgreSQL, Apache 1, GraphViz and MIBs

The proposed NM archetcure

In this paper we propose a network management tool architecture which supports heterogeneous NM system in many different architectural platforms. The proposed system should meet the following objectives:

  • Easy to use: to minimize the level of expert needed too manage the network.
  • Easy to access: web based architecture is the best way to achieve that.
  • The network management tool should have a standardized set of FCAPS function objects. This is needed to eliminate the data inconsistency that can be resulted from different standards in different networks.
  • Be able to interpret between definitions in different managed devices.

A conceptual view of the proposed architecture is shown in figure 7 below. The local network management system in each network gather the information needed about all the devices, terminals, software, and users in that network using its own protocols. The gathered data has to be send to the Agents Interface in the Console part. The Agent Interface communicates all the agents then stores those data in the database which is MIB and is explained in Figure 5. The Management Functions combines all the FCAPS functions (Fault, Configuration, Accounting, Performance, and Security) all those functions can be accessed by the user through the User Interface which is a web based interface that visualize and show all the attached networks management information. In this architecture agents on different networks interact with a gateway to communicate with any other agent. Networks such as like LAN, GSM, ATM, and ISDN are local networks and each one has their network management systems. Through gateway agent each local network has an agent and they speak to gateway which leads to centralized control with distributed management.

There are four requirements for proposed management envionment:



Data description

Common Information Model (CIM)

Transport Encoding


Operations to manipulate the data

Hyper Text Transfer Protocol(HTTP)

Common Information Model (CIM):

This is an object oriented information model for specifying management information in a way that is independent of applications, platforms, protocols and implementations.CIM defines a way to exchange the data from any source and network. The data represented using CIM can be understood and analyzed by and any network management tool or applications which understand CIM. The heterogeneous networks can make correlations between information coming from different locations in the network.

Extensible Markup Language:

The Desktop Management Task Force (DMTF) recently advanced the CIM encoding to standard Extensible Markup Language (XML) to show actual management data in a standard way in order to interoperate with each other for applications it must be possible to represent actual management data in a standard way. Extensible Markup Language (XML) is a markup language for representing information in a standard format in order to make heterogeneous platforms and applications able to interoperate with each other.

CIM operation over HTTP:

This concept stands for mapping of CIM operations onto the Hyper Text Transfer Protocol.All management functions can be accessed by the user through the User Interface which is a web based interface that visualize and show all the attached networks management information.


A gateway stands for communicating between two networks that use different protocols, or which have the same protocols but do not otherwise communicate. Gateways integrate various data sources and create the appearance that all data resides in a single, logical relational database. It is a combination of hardware and software that translates between two different protocols and acts as the connection point to the Internet.

The block diagram of a gateway application is shown in Figure8.It consists of 5 components: Message Accepting Rely Agents, Message Extractor, Message Translator, Message Rebuilder and Message Dispatching Relay Agents. Information after receiving to gateway from different networks would be processed as follows in Figure 7.In this proposed architecture, all the data coming from agents of different types of networks such as GSM, ATM, LAN, and ISDN are translated in the gateway agent. The information in gateway agent are translated from actual to generic or vice versa.


Diversity and complexity are the titles of the coming communication technologies, this situation caused by the increased production of the communication devices and systems without laying on one standardized concepts or common language. Those systems and devices nowadays forming heterogeneous networks .That raised the intensive need to find a one station to control and manage those networks, since controlling them separately brings a lot of difficulties and inconsistency. In this paper we proposed a new network management tool architecture for managing heterogeneous networks with a web based interface


[1]. Alex Gorod, Ryan GoveDr. Brian SauserDr. John Boardman,” System of Systems Management: A Network Management Approach”, ISBN: 1-4244-1160-2.

[2]. Celia A. Joseph Kurudi H. Muralidhar, “Integrated Network Management in an Enterprise Environment”, IEEE

[3]. Mohd Nazri Ismail and , Abdullah Mohd Zin, “Comparing the Accuracy of End-to-End Network Performance Measurement Testbed and Simulation Model for Data Transfers in Heterogeneous Environment”, Second Asia International Conference on Modeling & Simulation, IEEE , pp,

[4]. Do-Hyeon Kim, You-Ze Cho, “An Efficient Integrated Network Management for Heterogeneous LANs and WANs”

[5]. Book

[6]., 19-Feb -2009.

[7]. P-H Gross and N. Mercouroff. Integrated Network and Service Management for Mobile Networks,