Home Asia-Pacific I 2001 Planning and Deployment of Third Generation UMTS Networks

Planning and Deployment of Third Generation UMTS Networks

by david.nunes
Stéphane DupontIssue:Asia-Pacific I 2001
Article no.:11
Topic:Planning and Deployment of Third Generation UMTS Networks
Author:Stéphane Dupont
Title:Product Manager
PDF size:20KB

About author

Not available

Article abstract

Before looking at the planning problems for UMTS networks it is important to review the history of mobile telephony that has led to the development and drivers of 3G technology standards. The earliest analogue systems (Ist Generation) were developed in the USA in the late 1970s. Cellular mobiles were developed more widely in the 1980s with Global System for Mobile (GSM) becoming the standard for the second-generation digital cellular system. The increased demand for high-bandwidth data services has led to a series of developments that enhance the functionality of GSM. These are known as 2.5G technologies and include: WAP – Wireless Application Protocol; HSCSD – High Speed Circuit Switched Data; GPRS – Generalised Packet Radio System; EDGE – Enhanced Data rates for GSM Evolution. These technologies will allow the current GSM networks to provide key benefits of higher data rate services and packet switched data applications. The current array of 2.5G technologies will provide an evolutionary bridge to the next phase of mobile telecommunications, 3G.

Full Article

The third generation of mobile telephone systems – or the Universal Mobile Telephone System (UMTS) as the standard will be known – will take cellular communications to new dimensions. Wireless communications to date have now moved alongside the advances in multimedia, digital television, Internet/ Intranet and Internet protocol (IP) based services. Third generation systems are set to integrate these technologies into a universal system, encompassing comprehensive, feature-rich entertainment, information and commerce at the subscriber terminal. The world market for physical users of terrestrial mobile services including multimedia will be 426 million by the end of 2000 rising to 940 million by the year 2005 (UMTS forum). This represents a doubling of users in only 5 years! A large share of this 940 million market will be linked with networked multimedia providing services including audio-on-demand, education, information services and real-time communication services like video telephony. Interactive UMTS internet/web access opens the door to a wide range of new mobile data applications such as electronic banking, telemedicine and location based services that are not possible with current wireless systems. Countries of the Asia-Pacific region were among the first to demonstrate a commitment to 3rd generation technologies. Many of these countries have already completed a licence auction or beauty contest and others are well advanced in the process. Countries issuing 3G licences include: Japan (Three licences granted on the 12th July 2000 by beauty contest); Thailand (Have issued two third generation licences); South Korea (Issuing 3 licences towards the end of 2000 through a beauty contest.); Hong Kong & Singapore; (Issuing 4 to 6 licences each towards the end of this year); Australia Issuing (4 to 6 licences through an auction that started on 14th July 2000); New Zealand (Issuing 4 licences through an auction starting December 2000). Challenges for UMTS network designers UMTS subscriber terminals will be multi-service capable and will enable users to select different bit rate service requirements. Each UMTS carrier frequency used across the air interface will therefore bear multiple services and the planning of concurrent services need to be considered. Different data rate services can be transported in circuit switched or packet switched format and, additionally, can be transported in real-time or non real-time. These basic characteristics complicate traffic engineering predictions, an initial phase of UMTS networks planning. The radio air interface of UMTS employs Wideband Code Division Multiple Access (W-CDMA) as its multiple access strategy. The Key to Successful Network Planning To plan and deploy a mobile network effectively, RF engineers face a number of obstacles, several of which can be overcome by selecting first-rate geographic data. Signal propagation predictions are only as good as the geographic data they are based on. As the sophistication of simulation technology increases, so does its dependency on quality mapping data. To update the existing network at all levels, UMTS network planners need to apply even more rigorous precision to their projects: compared to voice transmission, data transmission leaves even less margin for error. Planners thus require a variety of mapping products providing different features and resolutions, yet offering homogeneous quality. One of the main challenges facing RF engineers involves positioning micro cells in order to absorb larger amounts of traffic while improving signal penetration inside buildings. They need to compute signal propagation along streets and buildings very precisely, and this is only possible if accurate and up-to-date 3D building databases are available. Each database need to be computed from multiple views giving consistent restitution of built-up areas, building details, vegetation and bridges. It must provide ground surface elevation information and also represent the outline and elevation of built-up structures. In the long term, to dimension networks that provide ever-broader bandwidth, RF engineers will increasingly rely on high resolution, three-dimensional city models. The ideal solution is to have access to a range of complementary datasets that can be integrated in a single database. These datasets are available from companies such as ISTAR (www.istar.com). Computed from satellite or aerial stereo-photographs, they provide precise building information, at 10m resolution for extended urban areas, suburbs and industrial zones, down to 0.50cm for city centres. In the race to deploy UMTS networks, the availability of off-the-shelf, recent mapping products can make a significant difference to the timing of a project by allowing re-engineering to start without delay. For a telecommunications company, purchasing quality 3D databases is key to improving the efficiency, reliability and cost-effectiveness of network planning. When planning a GSM network it can be assumed that the coverage of the network is dependent only on the configuration and location of the transmitter stations. For this reason GSM network planning follows an almost linear, sequential process of coverage, capacity then optimisation analysis. As demonstrated above, the coverage from a UMTS network is also dependent on factors such as the number of users, their locations and the services they are using. Conventional linear planning techniques can no longer be applied. As coverage, capacity and quality of service become interdependent it is necessary to consider all factors simultaneously. Monte Carlo Simulation for UMTS Network Planning. A variety of different planning tools and techniques have been proposed to overcome the considerable challenges presented to UMTS network planners. The most powerful and preferred solution makes use of a Monte Carlo Simulator. Put simply, a Monte-Carlo simulator generates multiple snapshots of a network – each showing the likely operation at any point in time. For each individual snapshot, active subscriber terminals (users that are connected to the network) are scattered intelligently throughout the geographic area under analysis. The scattering of users is essentially random, although the planner may decide to weight the importance of certain areas such as airports, certain types of land, roads or urban areas. By doing this, an array is built up with the probability that a terminal will be active in any given location. This is known as a traffic array, and must be created for each different type of subscriber terminal that will be used on the network. Additionally, it is necessary to define the telephony/data services that are going to be available on the network. With UMTS, different services will often require different data rates, and may only be supported on certain types of subscriber terminals. Any Monte-Carlo simulation must consider these complexities in order to achieve the necessary planning accuracy. Once the subscribers have been scattered geographically across the coverage area, the performance of the network at that moment of time can be predicted and analysed. A wealth of important engineering information can be gathered and logged for each user including. Can that user connect to the network? How much power does that user need to transmit, and therefore how much interference does that user cause to other terminals on the network? If the user cannot connect, why not? What signal strength does each user receive from the base station? Which base transmitter stations does each user connect to, and could the subscriber simultaneously connect to several base transmitter sites? By repeating the snapshot process it is possible to build up a meaningful statistical view of the performance of the network. The number of multiple snapshots required is very dependent on the size of the area under analysis. Thousands or even tens of thousands of individual snapshots may be required to build up an accurate picture of the performance of a large UMTS network. Using the results provided by each of these snapshots, a great deal of information about the network can be extrapolated automatically and displayed graphically. Figure 3 shows a plot produced for a UMTS network deployment on the island of Jersey. This shows the probability of being able to make a connection to the network for subscribers across the island. Many other engineering parameters can be calculated and quantified using Monte-Carlo simulation planning. Typical examples would be; most probable reasons for call failure in a given location. Average interference level in a given location. Which base transmitter stations are most likely to be the dominant servers at each location. By viewing/analysing these and other such parameters, a UMTS planner can determine the expected performance of the network and take forward actions to resolve any problems that may occur at the design stage. Certainly, powerful Monte-Carlo based planning techniques are essential to achieve this critically important end-goal! Conclusion Summary Without doubt, time and experience have demonstrated that expert, powerful network planning tools can ensure that complex modern communications systems are designed and implemented so as to achieve peak performance. With the roll-out of UMTS infrastructure and commercial service just around the corner, it is essential to maximise the use of intelligent, tailored planning processes for UMTS. It is expected that Monte-Carlo simulation techniques will form the cornerstone of all design processes for UMTS. Obviously, using the latest, innovative planning techniques will deliver a critical edge in an ever-increasing competitive environment for mobile operators.

Related Articles

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More