Brought to you by EarthWeb
IT Library Logo

Click Here!
Click Here!

Search the site:
 
EXPERT SEARCH -----
Programming Languages
Databases
Security
Web Services
Network Services
Middleware
Components
Operating Systems
User Interfaces
Groupware & Collaboration
Content Management
Productivity Applications
Hardware
Fun & Games

EarthWeb Direct EarthWeb Direct Fatbrain Auctions Support Source Answers

EarthWeb sites
Crossnodes
Datamation
Developer.com
DICE
EarthWeb.com
EarthWeb Direct
ERP Hub
Gamelan
GoCertify.com
HTMLGoodies
Intranet Journal
IT Knowledge
IT Library
JavaGoodies
JARS
JavaScripts.com
open source IT
RoadCoders
Y2K Info

Previous Table of Contents Next


2. BACKGROUND

A telecommunications network is a complex aggregate of switches and a transmission medium, which together provide a multiplicity of channels over which many customers' messages and associated control signals can be transmitted. The telecommunications network provides access to different types of circuits and services that use them. Network switch software is one of the most complex software systems in the world. These networks must be managed carefully for efficient and reliable operation.

According to the Telecommunications Management Network (TMN) framework, there are four layers of functionalities in a telecommunications management model: element management layer that manages network elements, network management layer that manages telecommunications networks, service management layer that deals with customers, and business management layer that handles business decisions (see Figure 1). The information required to make a decision is passed upward toward the higher layers, whereas control messages are always directed downward the lower layers.

Specifically, in the bottom element management layer, one deals with network elements such as switches, cell sites, transmission medium, signaling system components (e.g., SS7), or customer access facilities (e.g., PBX). Each element generates alarms that need to be monitored and filtered. Hardware problems must be corrected on the spot quickly. In the next network management layer, there are (1) fault management that is concerned with the detection, isolation, and correction of anomalous network conditions, (2) performance management that evaluates the quality of network services and determines the effectiveness of communication processes, (3) configuration management that allows technicians to view and modify the network configuration, (4) security management that authorizes use of system resources and protects network management data, and (5) accounting management that addresses costs associated with the use of network resources and allocates charges for the usage. Above that, the service management layer deals with customers including service setup, quality control and response, and billing. It often needs information retrieval. Finally, at the top layer, the business management makes decisions about network planning, market analysis, finance and budget, and resource allocation. Not all telecommunications companies follow the exact same management model, but TMN does summarize telecommunications management tasks.


FIGURE 1 The TMN framework.

Historically, expert system applications mainly focused on the network management layer (Liebowitz, 1988). Due to the complexity and diversity of telecommunications systems, telephone companies started to invest resource on artificial intelligence (AI) technology in the early 1980s to automate some of the operations. As a result, various rule-based expert systems were prototyped and fielded, especially in the areas of network repair and maintenance requiring monitoring and diagnosis. Some of them saved companies millions of dollars. Based on these initial successes, other AI techniques were also explored, including case-based reasoning, distributed AI, data mining, and machine learning. Cooperation between industries and universities on AI technology in that period was frequent. People had high expectation for AI in general.

Around the late 1980s and early 1990s, the hype of AI technology hit the wall. Meanwhile, the telecommunications industry also began to experience downsizing and focus on the bottom line. It was needed then to reaccess AI technology based on its value. Rule-based expert systems continued to serve important roles, either as independent systems or embedded within other systems. Yet, many other AI-based systems often stopped at the prototype level and did not become products. There was a gap between good concepts and hard reality. This, in some way, demonstrated the difficulty of telecommunications domains and the lack of understanding them when applying new AI techniques.

With continuing downsizing and early retirement of domain experts, AI technology was once again pushed to front in the mid 1990s. This time, deregulation spurred competition, automation is needed in all four management layers to cut cost and becomes critical to the lifehood of a company. New growth opportunities in wireless and digital services and international markets demand every company to do more with the same or less resources. In addition, many telecommunications software systems have been developed over the years. They input and output tremendous amounts of data that need to be interpreted and analyzed. This is similar to the "data overflow" of the information highway. Expert systems based on AI techniques can help to analyze this information, alleviate technicians from many routine tasks, and train them to be the experts of new technologies. More importantly, AI practitioners have learned from past lessons, and understood the potentials as well as the limitations of AI technology and what it take to develop AI-based products. Consequently, various types of expert systems have been fielded (see Section 5) and more are coming.


Previous Table of Contents Next

footer nav
Use of this site is subject certain Terms & Conditions.
Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Please read our privacy policy for details.