The management of computer networks becomes more and more difficult with an increasing variety of hardware devices, protocols and services. Agent technology has been proposed to support service and network management in several ways. The resulting area of agent-based management emerged as one of the few acknowledged application domains for intelligent and mobile software agents. The typical characteristics of intelligent agents, like autonomy and high-level communication, can for instance lead to more flexible and general network control structures. Furthermore, mobility of agents can result m reduced communication delay and bandwidth if computation involved with managing a network node can be performed closer to the node or even directly on the node. The autonomous character of agents can also give support for better fault tolerance in case of connectivity problems.
Because of the cost and complexity of building bespoke service management systems, telecommunications management developers are moving toward the use of off-the-shelf componentware to satisfy their management requirements. However, a crucial problem with such an approach is the ability to integrate components to realise integrated management solutions. This paper identifies the technology requirements for the development of systems which support telecoms management business processes which are constructed from reusable components. It sets out requirements on component integration technologies that are expected to be key to the development of future operational support systems. The current status of relevant component integration technologies is reviewed in the context of these requirements. The paper outlines two telecommunications management case studies which are being conducted to evaluate these integration technologies. The paper draws some conclusions about the relative merits of the different technologies examined and make some suggestions for further work.
The process of developing agent-based applications requires at least two tasks that are usually tackled separately by programmers. On one hand, programmers need to develop business rules and other support tasks for agents. On the other hand, programmers need to develop user interfaces (UI) for agents in order to enable end-users (not only owners but also other third parties) to interact directly with them. This paper focuses on this second task (developing user interfaces) and describes the solutions offered by the AgentSpace mobile agent system. Basically, we show and discuss two complementary ways to gather user-interfaces with mobile agents. On one hand, mobile agents don’t provide any UIs. This situation promotes the separation of the UI and the backend (i.e., the agent) which allows flexibility and reuse. On the other hand, mobile agents provide by default UI components, which consequently promotes agents as better units of development and management. This situation can be very suitable in the context of dynamic and large-scale applications such as those found in electronic commerce domains. This paper also shows the relationship between the application of these mechanisms and the model-view-controller architecture used currently to build user interfaces in modern object-oriented frameworks. Finally, this paper presents some concrete examples with source code based on the AgentSpace system in order to validate and clarify the discussed mechanisms.
One of the major efforts for standardising of service capabilities rather than standardising the services in UMTS is to achieve service differentiation and system continuity. While pre-UMTS systems thrived on the ability to access services in a same manner when user roamed, these new deregulation efforts of services will pose some problems in provisioning of services across networks. The concept of the virtual home environment (VHE) addresses problems like this but does not rigidly specify how services can be deployed. This contribution will discuss how agent technology can be applied for designing a service provisioning architecture based on the VHE concept. More importantly, the paper will focus on the importance of service interworking for supplementing service functionality and discuss how agent communication based on FIPA(2) ACL standard can be used to overcome this issue. Finally, the paper will discuss the design and implementation of a middleware known as a common communication module (CCM) used by service components on the network and the mobile nodes to perform ACL communication.
The deployment of distributed applications has been mainly a manual task, despite of the proprietary solutions that already exist in the market. Due mainly to the heterogeneity of the operational environment and the software to be deployed, deployment is something that is usually done by writing installation and configuration scripts in an appropriate scripting language. With the wide acceptance of Enterprise Java Beans and the arrival of the CORBA 3.0’s Components, an extraordinary opportunity for automating the deployment process has emerged. This is the basis for this paper, in which we define the concept of deployment, present a list of assumptions that would enable the automation of the process and establish a list of requirements on the deployment process itself and on deployment means.
Since ATM switches are intented to be simple and inexpensive, a significant part of the network cost is in the cost of the links. A way of increasing the traffic we can send over these expensive links is to transmit compressed ATM cells. This idea, although it seems very simple, is a new one for ATM and as we show it can signicantly increase the useful bandwidth of a typical ATM network. In this paper the CAC algorithm that can be effectively used in a network which uses this compression scheme is also described and the results that can be achieved when this algorithm is used in a real network are presented. This CAC algorithm is based on Large Deviation Theory: the large deviation rate-function (entropy) of compressed (and thus more bursty) ATM traffic can be estimated from measurements of trffic activity. The entropy can be used to determine the bandwidth requirement of the traffic.
Network Control is currently carried out mainly by means of signalling protocols. Although these protocols are robust and facilitate standardisation, they present several drawbacks such as the inability of the Service Providers to adapt network resource control to the particular needs of their services, or the difficulty in deploying advanced traffic control schemes in order to achieve increased utilisation. An open distributed software architecture for Network Control based on concepts such as Intelligent Agents can provide a solution to this problem. This paper highlights the major issues of this problem and presents the architecture and implementation of a system that enables open and flexible network control and resource management using intelligent agents. The description of the architecture is complemented by experimental results as well as a generic implementation framework for deploying open programmable control capabilities in existing networks.
In this paper we propose an integrated proactive routing and mobility management strategy that makes feasible the realization of a flat single-tier routing architecture in multihop packet radio networks. The integration of the routing and mobility management functions is achieved via the use of the geographic position reporting mechanism and the generalization of the zone routing concept. The underlying principle behind the proposed routing strategy is to reduce the network routing overhead by making the accuracy of the routing information in each node “inversely proportional” to its distance from any other node in the network. The proposed integrated scheme provides for a flat routing architecture with no hierarchical entry/exit points, where every node can act as a router, therefore increasing the network’s routing flexibility and robustness. Finally we show through modeling and simulation that the proposed routing protocol is a bandwidth-efficient routing mechanism that can be applied across large-scale networks.
This paper examines load balancing issues relating to a distributed CORBA-based Service Control Point. Two types of load balancing strategies are explored through simulation studies: (i) A novel ant-based load balancing algorithm, which has been devised specifically for this type of system. This algorithm is compared to more traditional algorithms. (ii) A method for optimal distribution of the computational objects composing the service programs. This is based on mathematically minimising the expected communication flows between network nodes and message-level processing costs. The simulation model has been based on the recently adopted OMG IN/CORBA Interworking specification and the TINA Service Session computational object model.
Telecommunications network management has attracted a lot of attention in terms of research and standardisation in the last decade. TMN and TINA architectural frameworks try to address the management needs of broadband networks and services. They both cater for multi-domain, multi-operator environments while they also allow for electronic customer access to management services. These services, though, are fixed in the sense that new features can only be added after a full research-standardisation-deployment cycle and are static as far as their use: they execute according to their in-built logic and clients may customise their operation only through tuning standardised operational parameters before service execution. The advent of mobile agent technologies opens new possibilities, allowing dynamic and customisable services to be offered to clients. In this paper we explain how mobile agents can enhance traditional connectivity management services, presenting the relevant architecture and design of a customisable network management system based on mobile agents.
This paper proposes an Accounting Management architecture, based on TINA principles. Accounting management is defined and positioned in the service provision chain. A relevant computational architecture is proposed and some of the involved computational entities are modeled as agents. A discussion follows on the merits of introducing agents to cater for certain accounting management activities. The ACTS AC325 MONTAGE project provides the main source of the presented work and results.
This paper presents the issues related to the implementation of electronic commerce of services. To illustrate the main features of electronic commerce of services, we have selected legal and administrative services. We describe in the paper the actors who take part in these services, the system architecture and the workflow that defines the interactions and interchange of information among actors. With these elements we have implemented an electronic commerce application inside the TRADE project 1 that demonstrates how these services can be offered in an electronic manner.
The efficiency of interactive Video on Demand systems (iVoD) heavily relies on the performance of the video servers. This paper presents methods to increase the iVoD quality of service and to simultaneously increase of the number of video clients by the addition of simple distributed Near-Client-Caches. This paper discusses intentionally cheap and simple algorithms and asserts their performance for systems in the transient state because this better reflects real-life situations than the usual steady state assumptions. The examined algorithms permit an increase of the number of clients at a video server by a factor of 3 compared with conventional iVoD systems.
Liberalisation of telecommunications encourages competition between the various actors in the Open Service Market (OSM). In this highly competitive context, Connectivity Service Providers (CSPs) and Value Added Service Providers (VASPs) are investigating opportunities to provide differentiated Service Quality related Service Layer Agreements (SLAs) to their customers. The services provided will span several administrative domains which makes their management complex. The key element for end users when choosing a particular service is the guarantee of support to be provided when using the service and the desire to interact with as few actors as possible. On the other hand, key issues for network operators and service providers are the cost-effective maintenance of equipment and services. The aim of this paper is to present a novel architecture that provides the necessary infrastructure, models and mechanisms to help VASPs and CSPs to rapidly introduce customer care services for user quality assurance in a Multi-Domain environment. The architecture aims at integrating TINA, TMF and TMN concepts as well as established legacy in-house customer care and help desk systems. This work is being undertaken within the Assurance part of the CEC ACTS project FlowThru.
The growth of commercial activities across networks has led to the network itself becoming a competitive marketplace with a multitude of vendors, operators and customers. In such an environment, users will 'shop around' for the best deals in terms of network financial costs and services. Intelligent, autonomous and mobile software agents introduce an alternative approach that facilitates expertise-brokering activities on behalf of the users. In this paper we present such an agent supported brokering scenario. The scenario involves interactions between Java-based mobile agents and an interactive model of the network. The paper describes how this prototyping environment can be used to examine the impact of mobile agents for brokering network resources in the future open network marketplace.
The demand for a great variety of sophisticated telecommunications services with multimedia characteristics is increasing. This trend highlights the need for the efficient creation of distributed programs with multimedia data exchanges over distributed processing environments. Therefore, it is necessary to support the object-oriented development of distributed multimedia applications in a flexible manner. This paper recognises Microsoft's Distributed Component Object Model (DCOM) as a key potential technology in the area of service engineering and examines a structured approach to enhance it for the handling of continuous media streams through the design and implementation of a collection of suitable multimedia support services. The proposed approach focuses on the modelling of continuous media communications in DCOM and is validated through the design and implementation of a multimedia conferencing service.
