M.Comm. / Most modern and even not so modern businesses are currently investing in the hire or purchase of private access branch exchange (PABX) equipment. Usually the main concern is one of reduced costs to the company, since the internal calls between users on the same PABX are not routed via the public telecommunications operator (Telkom) and therefore free of charge [ James, 1997: 17 ]. Larger companies are experiencing the problem that their offices are distributed countrywide. They can install PABX equipment at each regional office, but this will only allow the employees within the regional office to communicate via internal calls. These organisationssoon realise the need for a voice network that provide them with the capability to communicate internally on a country-wide and in some cases, even a worldwide scale. A private voice network can be established by following two routes. If the company is relatively small and cannot afford its own transmission infrastructure in the form of openwire routes. coaxial cables. microwave routes, satellite routes and the associated multiplexers and flexible multiplexers. then they can lease transmission infrastructure from the public telecommunications operator (Telkom) and pay rental and other associated charges for usage thereof If the company can afford the capital investment in its own transmission infrastructure, it will be capable of communicating on a countrywide basis without relying on the service provided by the public telecommunications operator. In most cases, however, the latter option is illegal, due to regulatory protection of the national telecommunications operator. as stipulated by the telecommunications act (103/1996). In some cases, like the Transnet group, legislation requires the use of a private telecommunications network. The reason for this was that a major catastrophe in the transport industry might result due to problems with the telecommunications facilities. In that case, the public telecommunications operator would not accept any responsibility for the consequences of the break in communications. Transtel is the business unit that acts as a "private telecommunications operator" for the Transnet group . Transtel provides both voice and data communications to its clients within the Transnet group . In this case the main concern is not one of reduced costs, and therefore Transtel is implementing a tariff structure similar to that of the public telecommunications operator. The problem of providing voice communications between geographically distributed offices must be solved reliably and efficiently by the private voice network implemented by Transtel to serve the Transnet group. This implies that Transtel have the challenging responsibility to implement a managed voice network, in order to ensure better service levels than can be obtained from Telkom. The next problem that faces the private telecommunications operators for organisations with several business units and a plethora of departments, is that the costs for telecommunications facilities must be broken down into separate amounts per business unit and per department. This breakdown of costs must also reflect actual usage of the communications network and the billing systems must have the capability to generate detailed reports on network usage, as requested. In organisations like Transnet, where lower costs were not the primary objective of the private communications network, usage of the private network as well as usage of a combination of the private and public network, is being billed back to the department that uses the service. In short, the private telecommunications operator must rely on highly efficient and reliable billing systems to keep track of network costs and clients, since it operates as a separate business unit within the group and have to show a return on the investment on network technology [ Wasnick, 1998 : 26 ]. Inevitably, the private communications operator will need to ensure its clients of superior telecommunications services, and in many cases formal service level agreements will define the acceptable service level from the client's point of view. In order to place the private operator in a position to honour the service level agreements, its networks and especially its voice network must be managed by a global, integrated network management system. The network management system must be global in the sense that it manages all the resources in the voice network as well as resources in other networks that are utilised by the voice network. In order to facilitate optimal service levels, the network management system must be fully integrated with the rest of the operator's information technology infrastructure. It is, however, not enough to have a network management system in place that can cater for all the technical network management needs of the organisation. The organisational structure, management philosophies and business processes must also be adapted, in order to fully utilise the potential benefit of the technical network management functionality [ Sutherland, 1998 : 8 ]. A multi-faceted approach to voice network management will ensure that bi-directional information flow and a balance is established between the technical network side, the business side and the organisational structure, as depicted in Figure 1.1. A major disadvantage will be overcome by focusing on the total telecommunications environment, rather than emphasising the business environment, the technical environment or the organisation. No modern telecommunications operator will be able to survive in the highly competitive environment without a strong business focus integrated with technical excellence and supported by an appropriate organisation structure.

The increasing competition between telecommunication service providers (TSPs) and the large amounts of human resource costs incurred to provision telecommunication services to small, medium and micro enterprises (SMMEs), highlight the need for effective self-service provisioning methods to reduce costs. Existing service provisioning methods do not effectively support SMMEs in the service provisioning process. The current methods used are manual and Web-based. The manual service provisioning method is labour and time-intensive. TSPs are being forced to cut human resource costs as the profit margins for provisioning services are narrow due to competition. The current web-based systems do not allow SMMEs to effectively purchase services as they provide long textual lists to consult and limited information about services. The objective of this research was to develop an intelligent web-based self-service system to support SMMEs in telecommunication service provisioning, using a Design Science Research methodology. A constraint-based, knowledge-based recommender system was selected as the most appropriate intelligent system to support telecommunication service provisioning. A prototype called SelPro was developed as a solution to address the shortcomings of the current ineffective service provisioning methods available for SMMEs. SelPro incorporates an interactive wizard-style user interface with dynamic recommendations to support novice users as well as an expert interface which provides less user support. An expert review and a user study were conducted to determine if SelPro was usable, useful and effective. SelPro was perceived as being usable for SMMEs wishing to purchase telecommunication services. Expert participants were satisfied with the usability of SelPro and perceived it as being useful, easy to use and easy to learn. SelPro was considered effective because of the high task success ratings provided by the participants using the two interfaces. The research determined that intelligent web-based self-service provisioning can be used to effectively support SMMEs in telecommunication service provisioning. General design recommendations for similar systems were proposed and future work will involve using intelligent systems to support large enterprises in telecommunication service provisioning.

An analytical framework is developed for distributed management of large networks where each node makes locally its decisions. Two issues remain open. One is whether a distributed algorithm would result in a near-optimal management. The other is the complexity, i.e., whether a distributed algorithm would scale gracefully with a network size. We study these issues through modeling, approximation, and randomized distributed algorithms. For near-optimality issue, we first derive a global probabilistic model of network management variables which characterizes the complex spatial dependence of the variables. The spatial dependence results from externally imposed management constraints and internal properties of communication environments. We then apply probabilistic graphical models in machine learning to show when and whether the global model can be approximated by a local model. This study results in a sufficient condition for distributed management to be nearly optimal. We then show how to obtain a near-optimal configuration through decentralized adaptation of local configurations. We next derive a near-optimal distributed inference algorithm based on the derived local model. We characterize the trade-off between near-optimality and complexity of distributed and statistical management. We validate our formulation and theory through simulations.

M.Comm. / In today's society everyone is dependent on information for decision making, and a big part of the information is provided via telecommunication lines. Telecommunication is the unseen and unsung hero when it comes to information transfer. It is no wonder that telecommunication service providers are striving to give better and faster service almost daily. Furthermore because of the competitive nature of the telecommunications arena, service providers are striving to get the best performance from their telecommunications links. It has thus become more important to provide excellent service and guaranteed performance above anything else. The purpose of this study is to determine what kind of information is needed by the management of a telecommunications service provider to make business or strategic decisions. Questionnaires and interviews were used in this study to determine what kind of information is needed by the managers and the decision makers in the management of the telecommunication network. The basis of the study was provided through the examination of what a telecommunication network consists of, and the information that is needed to run such a network. An additional feature was that there is already a network monitoring system working practically. This system was used to measure the success of telecommunication network management and to determine what other factors should be catered for when a total telecommunication management network is evaluated.