Printer Friendly

AGENT-ORIENTED TECHNOLOGY FOR TELECOMMUNICATIONS.

The vision of the telecommunications market today is "information at any time, at any place, in any form." Within this open market of information services, the aspects of service customization and instant service provision are of fundamental importance. Moreover, the provision of telecommunications services will be a key component in future business successes. In this context the new paradigm of agent-oriented programming is gaining momentum. The main actors in this scenario are agents, which are very suitable to address a lot of the problems related to the diversity of access devices, the limitedness of current procedures, the complexity of inherently distributed problems, and the open market in general.

The world of telecommunications is changing at a rapid pace, and these changes are taking place in the technological as well as in the regulatory arenas. Software companies are entering the market offering their products as service solutions. Additionally, market forces are at work on an unprecedented scale. Given these circumstances, it will no longer be sufficient for private network operators to solely provide the network infrastructure, instead the challenge is to evolve to full-service providers. This implies the increasingly complex telecommunications infrastructure needs to be managed more efficiently while, conversely, new types of telecommunications services need to be developed and provided. It is in particular such future services that need to satisfy a broad range of requirements, including personalization, support for user mobility, on-demand combination of different services, and both online and offline service usage.

It is useful to distinguish basic telecommunications applications from telematics services. By the notion of a telecommunications application we generally denote any kind of information system necessary to run, to manage, and to administrate computer networks. Additionally, all facilities in the area of internal administration within a telecommunication or network provider--customer service systems, for example--will be subsumed by the notion of telecommunications applications. On the contrary, the notion of a telematics service emphasizes the mostly external information service aspect. This includes the fact that the customer of telematics services has to pay for them. That is, the service must represent an appropriate equivalent value, transparent accounting models must be invented, and the specific roles of vendor and customer have to be modeled by the system, especially with respect to the generally expected rights and duties that come along with these roles.

Future telematics services will have to meet a number of requirements resulting from rapidly changing markets and technologies. These requirements include service personalization, support for user mobility and network convergence, on-demand service integration and combination, security, manageability, transparency, robustness, and delegation. This list is not comprehensive; depending on the type and level of detail of user requirements on specific services, it might be arbitrarily extended.

The basic notion of the domain of software agents is the agent itself. This is a piece of software that acts on behalf of a user or assists a user. More specifically, an agent is a self-contained software element responsible for performing part of a programmatic process. The exact definition of an "agent" is still under discussion, and varies from author to author. One way agents might be used as a structural element of systems design, commonly called Agent-Oriented Techniques (AOT), is shown in the article "An Agent-based Approach for Building Complex Software Systems."

In order to be applicable for telecommunications applications and service provisioning agents must have certain specific capabilities, especially in the areas of agent management, security, personalization, and mobility. Further, different basic types of agents can be identified: manager agents for administration of the runtime platform, stationary agents as service providers, and mobile agents for mobility support and service provisioning. For efficient development and deployment of agent-based systems it is helpful to utilize toolkits. An example of a toolkit that has been specifically designed for the telecommunications domain is described in "Agent-based Telematic Services and Telecom Applications."

There are a number of telecommunications scenarios that are suitable candidates to be realized using agent technologies. Telecommunications providers have been changing from voice carriers to suppliers of differentiated services and need to leverage their existing network infrastructure to manage it more efficiently and enhance it for service provisioning of today's and future services. The new services have to be quickly created, combinable from basic services to high-value services, reliably provided, and efficiently administered. Customers want to use personalized value-added services, thus these services are expected to be dynamically adaptable to customers' needs: they must satisfy increasing requirements in quality of service, availability, and personalization. Decentralizing network management activities using AOT can be very effective, due to the characteristics of mobility, cooperation, and autonomy of the agents in distributed management applications. AOT is an approach to fulfill the prerequisites of active nodes and mobile program code in active networking. It is expected that AOT will lead to new frontiers in the areas of intelligent networks and active network management. The articles "Community-based Service Location" and "Service Creation and Management in Active Telecom Networks" are set in this scenario.

Driven by the current growth of networking capabilities and network infrastructures for computer systems, both in business and private domains, business contracts and transactions mediated by means of electronic communications become more and more viable. This trend has been mainly influenced by the advent and success of the Internet. Corresponding to the distinction between business and private customers, the e-commerce market differentiates business-to-business and consumer markets. More specifically, the notion of e-commerce is often used to denote that particular part of the consumer market in which private customers use the Web to supply themselves with goods or services ordered online. The articles "Communication Management Experiences in E-Commerce" and "Agent-Oriented Technology in Support of E-Business" shed more light on the area of agent-supported e-commerce.

Another promising area for agent technology is the intelligent home, enabled by intelligent devices being enhanced with software technologies such as Jini for networked control and actuation. Current drawbacks of insufficient communication infrastructure will be resolved with the advent of wireless communication technologies like Bluetooth and the deployment of the Internet Protocol Version 6 for a sufficient range of addresses. Applications in this context allow for a more flexible management and use of the home-based equipment. In the future this area will grow together with the aforementioned e-business. Because of the heterogenous and decentralized structure of this area the utilization of agent technologies promises to be effective.

Scenarios from the area of mobility supporting services are characterized by asynchronous communications and the need for highest efficiency. Services must be personalizable and independent from time, place, and the access devices or operating systems that are used. Of special interest in this domain is the ability of the service-providing agents to keep track of their cost-effective activities--necessary for a reliable accounting of the called-on services--and to guarantee a high standard of security restrictions, going beyond the infrastructure properties of currently available platforms, such as authorization, authentication, or contracting.

Traffic telematics is a prominent representative in the area of mobility supporting services. In a traffic telematics scenario, several intelligent and personalized services are supplied and integrated to be used from within a car. Typical services comprise dynamic routing in consideration of traffic-related information, location-based services informing the driver about nearby sightseeing locations or shopping facilities, monitoring of the fuel level in combination with information about nearby fuel stations, diagnosis of machine failures combined with notification of a breakdown service, and emergency call in the case of an accident. An agent-based solution would provide different agents for the particular tasks, for example, content provider agents for accessing and bundling distributed information sources, service provider agents for service provisioning, as well as mobile user agents for requesting and using services. In order to provide only useful information to the driver all services should be personalizable by using profile data.

An agent-based realization would lead to a system architecture as illustrated in Figure 1. Information services may be installed on a traffic telematics portal or located elsewhere on the Internet. Several kinds of agents are used for enriching the raw data of the accompanied information sources into value-added, personalized services. The article "Agent-Oriented Technology in Support of E-Business" describes a typology of different agent types in a business setting.

[Figure 1 ILLUSTRATION OMITTED]

In the example traffic telematics scenario, a runtime environment for agents is installed in the car's on-board computer. This infrastructure enables software agents to run safely and to communicate with each other, and provides access to the periphery. Specialized agents collaboratively monitor and diagnose the operations and activities of the car equipment and display status messages. Most of the work is done in the background without requiring driver intervention. The autonomy property allows an agent to react flexibly according to a given situation. For example, in case of a malfunction the agent will not just notify the driver but additionally try to solve the problem in cooperation with other agents or seek assistance.

Other services also act in a push-like manner, even if they are initiated by the driver (see Figure 2). For instance, a traveler wanting to obtain information about places of interest would initiate this service from within the car. Consequently a mobile agent will be instructed to fulfill its principal's requirements. The agent would then move to a location where such a sightseeing service is offered by carrying the driver's route plan and personal interests with the agent, and then deliver all relevant data to a service-provider agent and authorize it to start working. During the service usage the car will regularly send its GPS coordinates to the mobile agent, accessing and filtering the suggestions of the service provider agent according to the traveler's interest profile. As a consequence, the mobile agent will communicate the appropriate results back to the car where they are presented to the driver.

[Figure 2 ILLUSTRATION OMITTED]

We begin our journey into the promising world of software agents with the articles presented here, and will monitor the horizon for more applications appearing down the road.

SAHIN ALBAYRAK (sahin@cs.tu-berlin.de) is the founder and scientific head of the DAI-Laboratory at the Technical University of Berlin.
COPYRIGHT 2001 Association for Computing Machinery, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2001 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Industry Trend or Event
Author:ALBAYRAK, SAHIN
Publication:Communications of the ACM
Geographic Code:1USA
Date:Apr 1, 2001
Words:1682
Previous Article:On Bias, Science, Equal Protection, and Liability.
Next Article:AN AGENT-BASED APPROACH FOR BUILDING COMPLEX SOFTWARE SYSTEMS.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters