New mobile technologies support new products, which become new sources of revenue. The latter play a crucial role in how operators are responding to downward pricing pressures from a growing number of competitors – including upstarts using new technologies such as WiFi and voice-over-Internet protocol (VoIP). It is essential, therefore, that mobile operators evolve and upgrade their network technologies on a continuous basis.
Downward pricing trends also mean that operators must be able to evaluate the costs and benefits of implementing new technologies. During the initial build-out phase and at various network upgrade points, operators have tended to focus on growth and adding capacity at the expense of cost-efficiency. As a result, cost-inefficiencies have built up in networks over the years. This has been compounded by a tendency for operators to use separate processes to decide on network growth and how that growth is undertaken – which can lead to costly and inefficient implementations.
Unnecessary network costs could quickly become a significant financial burden. Indeed, a recent study of European and North American operators revealed that network costs constitute about 30% of the total operating expenses. Of that 30%, transmission costs account for 34%, and site costs account for another 15%.
As average revenue per user (ARPU) stagnates, operators must pay more attention to the optimization of their transmission and site costs and ensure that overall network architectures are as efficient as possible. For example, a 5% saving in transport costs can yield 0.5% increase in earnings before tax, depreciation and adjustments.
Four aspects of network operations must be balanced and tuned to optimize the overall cost structure of a network. These are forecasting and demand alignment; backhaul; network capacity management; and achieving technology improvements from operational efficiency. Balancing these four domains requires a robust information architecture, analytical framework, operational process integration, and operational process efficiency.
When considered independently, each of these operational domains can generate significant savings. However, their inter-dependency must be considered and this could result in a domain being operated sub-optimally to achieve a better global result. Flexibility is also necessary, as real-world constraints such as personnel skills, process/culture changes and financial/contractual commitments often determine the rate at which changes can be made.
Network inventory problems
Operators have been so focused on growth and build-out that they have not been careful in the management of information pertaining to network assets. As a result, the quality and completeness of network inventory information is somewhat deficient. Specific problems include fragmented information; numerous private databases; design and provisioning information still in paper format; and systems limitations that have allowed process exceptions to creep in. Improving the quality and completeness of inventory information – as well as integrating different data types – is a longer-term challenge that operators must address during their ongoing programme of optimization.
The current generation of optimization tools and technologies tend to be home-grown and are limited in their ability to employ the complex algorithms needed to address the problems of cost-effectiveness. While some commercial software packages provide limited capabilities, none so far have been able to solve the multi-carrier, complex mixed integer program (MIP) that results when the problem is modelled correctly. This class of problem has numerous local optima, which must all be determined before the global optimum can be found.
MIPs are computationally intensive and can be solved using heuristics. Recent advances in mathematical programming and genetic algorithms have given rise to a new class of approaches to reducing the computational complexity. Only recently have tools that embody these new approaches become available. In most cases, none of these tools has been deployed for mobile telecoms applications.
Another limiting factor is that any constraints on the optimization process are usually considered beforehand, which limits the scope of optimization and encourages sub-optimization. Most optimization systems in use today have been tailored to fit processes that are operationally focused, and meet financial criteria for acceptance of business value at a circuit level. They do not usually consider an entire market or geography. Some systems employ well-defined market planning processes, while others use more ad hoc processes. Data integration and rationalization capabilities often require improvement.
Current tool-sets are limited in scope and are tailored to meet well-established design and operational criteria of the past – criteria that do not consider today`s real-world limitations. This limits the system`s ability to consider non-traditional scenarios.
Schema has addressed these challenges by taking a life-cycle view of network optimization that covers analysis to implementation phases. It addresses the major fixed-network domains and it is based on years of practical experience. Schema has developed a platform that fulfils many of these requirements, while solving the key network problems. The system provides an environment that eases data integration and achieves the quick identification of data-quality issues, while maintaining data consistency throughout the optimization process.
Figure 1 (**see link below) shows a typical optimization process and how it can be implemented to deliver real economic benefits. The process is based on a divide-and-conquer modular approach. However, most real-world optimization problems are computationally intractable because a near-infinite number of network configurations must be evaluated.
To address this challenge Schema has joined forces with the telecoms consulting firm Accenture. Schema`s Transport Network Optimization Suite (TNO Suite) has been combined with Accenture`s analysis system, to create a modular approach to transport network optimization. After a network design problem is defined, it is broken down into manageable components. A network planner, with the help of the TNO Suite`s advanced modules, uses a combination of heuristic algorithms and analytical modelling techniques to arrive at optimum configurations for individual network scenarios.
Meticulous development and comprehensive field trials ensure that the modules in the Schema TNO Suite operate and interact in a coordinated and consistent manner. Each module adheres strictly to the properties and definitions as applied to the entire suite. Formats, data and operational protocol requirements are transferred seamlessly between the modules.
Modular format
Schema`s TNO Suite comprises several modules that can be used either independently or integrated in various combinations. The suite also includes a graphical user interface that controls all data, parameter definitions and module manipulations. Each module is highly versatile and can be configured to solve a wide variety of logically equivalent problems.
Once the optimization is completed for various scenarios, the operator must develop detailed implementation plans with supporting business cases, which justify the need for implementing the changes recommended by the optimization. To achieve this, Accenture has developed algorithms that use business rules to analyse and optimize the network to create provisional plans for change. These algorithms cover most of the operating considerations that determine when and how network changes can occur.
The key to success in on-going cost optimization – while migrating towards 3G – is a system that combines capacity and asset optimization. Strong linkage and integration between real-time processes and optimization tools is essential. In short, integration between Schema TNO tools and inventory/provisioning systems is part of a seamless process. This helps in striving for continuous improvement in the cost structure through better optimization linked to the execution. This enables the planner, with an easy-to-use optimization tool, to be shielded from complexities of various types of technologies, services and network arrangements.
Credit: WirelessWeb
