Read Microsoft Word - NOCPaper JM3.doc text version

The future of access network provisioning, planning and design

Mark Rose and John Mellis Evolved Networks Phoenix Place, Adastral Park, Martlesham, Ipswich IP5 3RE, U.K. E-mail: [email protected]; [email protected] Abstract This paper describes how the processes of telecom access network provision, planning and design can be radically automated in the next five years. It discusses the issues facing service providers engaged in these activities, and how advanced software tools can play a key role in driving down infrastructure costs and generating revenue, in an area of telecoms operations which is often deemed to be complex, costly, unfashionable, and difficult-to-automate. Introduction In today's telecoms companies, the increasing competition drives the need to provide service more quickly, increase utilisation of capital assets, and drive down operational costs. Particularly in the access network, this is a problem, because many telecom operators have little understanding of where their assets are, what spare capacity is available, how best to provision service orders or where capacity growth is needed in the future. To try to address these issues, companies are purchasing commercial-of-the-shelf (COTS) inventory systems to record their network assets, to provide a basis to speed up the design and provisioning processes, and to drive down costs. This leads to two major problems: · as with any software system, high-quality results need high-quality input data in computer-readable format. In most instances where there is legacy network inventory data, converting the data manually to digital formats is very expensive, extremely time consuming, and prone to error. · even when the inventory data is digitised, using the data effectively to automate the design and provisioning processes is a major challenge. Skilled, highlyqualified users are still required to operate the systems, taking the same time (or even longer) to plan and design networks compared to traditional pen-and-paper methods. For these reasons, many COTS operational support systems could be said to computerise rather than to automate the design and provisioning process. Therefore the first need is for a system which can convert and stream data from legacy data-stores into modern inventory database systems. This data can then be used to automate the access design and provisioning process. The next sections describe three novel systems which address these requirements, and their likely future evolution.

Automated Data Conversion Most telecoms companies hold logical inventory data, describing the connectivity of their physical access bearers (copper pairs and optical fibres). Most physical inventory data (describing the geographical layout of ducts and cables) is held in non-computerreadable formats, e.g. bit-map images or even paper maps. Since this data is critical for the automation of the provision, planning and design processes, it must be captured and migrated to new systems. Traditionally telecoms companies have had no choice but to convert all the physical data manually, by the use of Data Conversion Vendors (DCVs) or in-house labour using manual point-and-click methods on digitisation tables. Evolved Networks has addressed the problem of converting the legacy data through the use of software tools which automate the processes of data extraction, verification, and quality assurance, reducing substantially the costs of converting this data. The geographical diagrams describing the content of the plant, for example, duct and cable layouts, can be generated by inference from a combination of the legacy logical data and basic converted geographical civils data. It has been shown in early trials with BT that the generated cable layout data can have such a high degree of accuracy that any further improvements in accuracy through manual conversion may not be cost-effective. The automated system can also produce detailed reports which validate the captured DCV geographical data, describing the layout of trenches and ducts. Exceptions between the physical and logical data are reported, and certificates of quality can be produced and assigned to each dataset before loading it to the target COTS inventory system. Figure 1 shows the user interface to the system, which aids the visual highlighting of data exceptions and anomalies. An additional operational advantage of this system is that it can be easily coupled with knowledge of where there are network areas of high demand or revenue generation. The automated data conversion process can then be targeted and prioritised to further reduce data conversion costs.

Figure 1. Automated data generation system

The future evolution of this system may be greatly influenced by its ability to receive and send input and output data as self-contained data documents (typically formatted in eXtensible Mark-up Language, XML). This capability, combined with the `off-line' nature of the data processing, means that the system can be easily accessed, viewed and operated remotely (for example via Web-browser) by the various stakeholding partners in the data-conversion process. Automated Order Provision The order provisioning process is critical to provide service to customers and constitutes most of the access network expenditure, (in the order of 36% for fixed line provisions as opposed to 12% for new plant and upgrading of existing network). Fully automating this activity to optimise best use of existing network assets is critical in reducing costs and providing service on time. If we assume that converted physical network data is accessible to the system, then we can use business rules to best allocate network capacity to a customer order, by: · selecting the best network transmission medium, for example fibre, copper, or radio, · automatically planning and designing rearrangement of existing capacity, and/or extending network where duct already exists. If suitable spare network capacity exists at a convenient access node, then the system assigns the required capacity (e.g. a copper pair) to the customer order. Many in-house network operational support systems have this capability today. However, if there is no spare capacity at an adjacent customer access node, the advanced provisioning system interrogates the logical and physical inventory databases for the best alternative solutions, which could provide service through a re-configuration of the access network. The various alternative solutions are then ranked according to a number of configurable factors, including: the business rules governing the prioritisation of customer orders; the planning rules which dictate which technical solutions are allowed; and the relative costs of the possible network re-configurations. Once the optimal valid solution is found, the customer order is automatically progressed through the order management process to the field force, and all relevant database systems are updated. Figure 2 shows a screen from the service provisioning system showing summary reports of customer service orders, and the ranking of the best network provisioning solutions. Again, an advantage of this system is that all orders and tactical provisioning solutions are recorded, and can be used to make strategic forecasts of where and when network capacity growth is needed. In cases where there is under-utilised network capacity, the telecoms operator's marketing department can target win-back campaigns and other revenue generation initiatives. The natural future evolution of this system is to allow customers directly to check the availability of a particular service. For example, if a customer requires Broadband service, and the system identifies that the ADSL capacity is available, then the standard provisioning lead time could be reduced, thus preventing churn, increasing customer satisfaction, and bringing forward service revenues. Conversely, if the capacity is not immediately available, and a work-around solution can be found, then the system will provide the customer with the expected time-of-delivery.

Figure 2. Automated service provisioning system In those cases where customer service orders cannot be fulfilled from existing network capacity, or when the customer is located on a `greenfield' site beyond the reach of existing network, then we need to design the optimal new network layout to provide service. This in turn requires an automatic network planning and design system, which is described in the next section. Automated Planning & Design The network design system aims to plan the optimal new capacity beyond the last existing node, by using configurable business rules and optimisation algorithms to determine best routes, physical layouts, and equipment selections. For example, the system selects the optimum size and type of cable required, makes sure that end-toend performance (e.g. transmission loss) is within service limits, determines the best duct/trench paths within existing network. The automated planning & design system then progresses the order to the field, with relevant work instructions, to install the cable network which will provide service to the customer. To achieve productivity benefits the automated planning and design system must be capable of use by lesser-skilled technicians rather than highly-qualified planning and design technical staff. The Evolved Networks planning & design system (shown in Figure 3) identifies the best network solution based on configurable planning, design and optimisation rules, rather than the user's accumulated domain experience and training, as at present. Therefore, both a high degree of automation and radical simplification of the network design process can be achieved. The user simply identifies possible geographical cable paths to the customers, and the system then automates the planning and design process. It identifies the best possible cable paths, automatically

allocates correct cable and plant equipment, and issues a `job pack' including cable installation and jointing instructions, to the field force. The system automatically updates legacy systems to prevent expensive double-keying of data. This system potentially allows the future option for telecom service operators to outsource the network planning and design function, since the system guarantees that the in-house planning, costing, equipment and safety rules are met at all times. The need for quality-assurance checking of scheme designs by managers will be greatly reduced or be no longer required.

Figure 3. Automated planning and design system Conclusions The automation of the processes of network design and provisioning is a crucial challenge for telecom operators, particularly in the access network. We have identified three critical areas which are capable of radical automation through the application of modern, expert software systems: · · · The generation and maintenance of high-quality network infrastructure data The provisioning of access network resources to customer service orders The planning and design of new network infrastructure and extensions to existing networks.

Systems to automate these processes have been designed and described, and have been proven in network operator implementations or service trials. The business

benefits derived from these early system implementations will be described in the conference presentation. Acknowledgements The authors would like to thank all their colleagues in Evolved Networks for their help in the development of the ideas and solutions described, and in particular thank David Mortimore and Walton Teasdale for their help in the preparation and review of the paper.


Microsoft Word - NOCPaper JM3.doc

6 pages

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate


Notice: fwrite(): send of 201 bytes failed with errno=104 Connection reset by peer in /home/ on line 531