Radio Network Planning and Optimisation for UMTS

John Wiley & Sons

Copyright © 2006 John Wiley & Sons, Ltd
All right reserved.

ISBN: 0-470-01575-6

Chapter One


Jaana Laiho, Achim Wacker, Tomáš Novosad, Peter Muszynski, Petri Jolma and Roman Pichna

1.1 A Brief Look at Cellular History

The history of mobile communications started with the work of the first pioneers in the area. The experiments of Hertz in the late 18th century inspired Marconi to search markets for the new commodity (to be). The communication needs in the First and Second World Wars were also supporting and accelerating the start of cellular radio, especially in terms of utilisation of ever higher frequencies. The first commercial systems were simplex, and the operator was required to place the call. In the case of mobile-originated calls the customer had to search for an idle channel manually. Bell Laboratories first introduced the cellular concept as known today. In December 1971 they demonstrated how the cellular system could be designed.

The first cellular system in the world became operational in Tokyo, Japan, in 1979. The network was operated by NTT, known also as a strong driver for cellular systems based on Wideband Code Division Multiple Access (WCDMA). The system utilised 600 duplex channels in the 800 MHz band with a channel separation of 25 kHz. Another analogue system in Japan was the Japanese Total Access Communication System (JTACS). During the 1980s it was realised that, from the user's point of view, a single air interface was required to provide roaming capabilities. A development study was initiated in 1989 by the Japanese government, and a new digital system, Pacific Digital Cellular (PDC), was introduced in 1991.

In 1981, 2 years later than in Japan, the cellular era also reached Europe. Nordic Mobile Telephone started operations in the 450 MHz band (the NMT450 system) in Scandinavia. The Total Access Communication System (TACS) was launched in the United Kingdom in 1982 and Extended TACS was deployed in 1985. Subsequently in Germany the C450 cellular system was introduced in September 1985. Thus, at the end of the 1980s Europe was equipped with several different cellular systems that were unable to inter-operate. By then it was clear that first-generation cellular systems were becoming obsolete, since integrated circuit technology had made digital communications not only practical but also more economical than analogue technology. In the early 1990s second-generation (2G) (digital) cellular systems began to be deployed throughout the world. Europe led the way by introducing the Global System for Mobile communications (GSM). The purpose of GSM was to provide a single, unified standard in Europe. This would enable seamless speech services throughout Europe in terms of international roaming.

The situation in the United States was somewhat different than in Europe. Analogue first-generation systems were supported by the Advanced Mobile Phone System (AMPS) standard, available for public use since 1983. There were three main lines of development of digital cellular systems in the US. The first digital system, launched in 1991, was the IS-54 (North American TDMA Digital Cellular), of which a new version supporting additional services (IS-136) was introduced in 1996. Meanwhile, IS-95 (cdmaOne) was deployed in 1993. Both of these standards operate in the same band as AMPS. At the same time, the US Federal Communications Commission (FCC) auctioned a new block of spectrum in the 1900 MHz band. This allowed GSM 1900 (PCS) to enter the US market. An interesting overview of the GSM and its evolution towards 3G can be found in.

During the past decade the world of telecommunications changed drastically for various technical and political reasons. The widespread use of digital technology has brought radical changes in services and networks. Furthermore, as time has passed, the world has become smaller: roaming in Japan, in Europe or in the US alone is no longer enough. Globalisation has its impact also in the cellular world. In addition, a strong drive towards wireless Internet access through mobile terminals has generated a need for a universal standard, which became known as the Universal Mobile Telecommunication System (UMTS). These new third-generation (3G) networks are being developed by integrating the features of telecommunications- and Internet Protocol (IP)-based networks. Networks based on IP, initially designed to support data communication, have begun to carry streaming traffic like voice/sound, though with limited voice quality and delays that are hard to control. Commentaries and predictions regarding wireless broadband communications and wireless Internet access are cultivating visions of unlimited services and applications that will be available to consumers 'anywhere, anytime'. They expect to surf the Web, check their emails, download files, make real time videoconference calls and perform a variety of other tasks through wireless communication links. They expect a uniform user interface that will provide access to wireless links whether they are shopping at the mall, waiting at the airport, walking around the town, working at the office or driving on the highway. The new generation of mobile communications is revolutionary not only in terms of radio access technology, and equally the drive for new technical solutions is not the only motivation for UMTS. Requirements also come from expanded customer demands, new business visions and new priorities in life.

1.2 Evolution of Radio Network Planning

There is very little published on the Radio Network Planning (RNP) process itself. An integral approach is proposed in, but this is more related to the functionalities of an RNP tool than to the overall planning process. This paper challenged the existing practises in RNP by listing the following weaknesses:

planning was based on hexagonal network layout;

traffic density was assumed to be uniform;

radio wave propagation was considered independent from the environment;

base station locations were chosen arbitrarily, while in practice fixed sites were used;

traffic region boundaries usually were not taken into account.

The discussion continued in, which for the first time accounted for the impacts of quality requirements in radio network planning. This paper starts to have a process approach, and capacity enhancement with base station sectorisation to support network evolution is investigated. The challenges of non-uniform traffic conditions are identified, and cell splitting as one solution is proposed.

It can be noted that radio network planning and its development through time can be easily mapped to the development of the access technologies and the requirements set by those. The first analog networks were planned based on low capacity requirements. Radio network planning was purely designed to provide coverage. Omnidirectional antennas were used and positioned high in order to keep the site density low. The Okumura-Hata model was and still is widely used for coverage calculation in macro-cell network planning. Certain enhancements and tailoring by the COST231 project have finally resulted in the still widely used COST231 Hata model, also applicable to third-generation radio networks. The latest COST developments of this area can be found in. Walfisch-Ikegami is another model often referred to. This model is based on the assumption that the transmitted wave propagates over the rooftops by a process of multiple diffractions (and). Although the Walfisch-Ikegami model is considered to be a micro-cell model, it should be used very carefully when the antenna of the transmitter is below the rooftops of the surrounding buildings. More about propagation models can be found in Section and conclusions about their applicability in and.

During the course of time, together with the evolution of 2G systems, the site density got higher due to increasing capacity requirements. Furthermore, the initial assumption that cellular customers would mostly be vehicular turned out to be incorrect. Thus the maximum transmit power levels of the user equipment were reduced by at least a factor of 10, causing a need to rebalance the radio link budgets. All this forced the cellular networks to omit the omnidirectional site structure and lead to the introduction of cell splitting - i.e., one site consisted of typicall y three sectors instead of just one. Owing to increased spectral efficiency requirements, the interference control mechanism became more important. In addition to the sectorisation antenna tilting was also introduced as a mechanism for co-channel interference reduction. Furthermore, the macro-cellular propagation model was no longer accurate enough; new models were needed to support micro-cellular planning. Sectorisation, antenna tilting and link budgets are discussed later in Chapter 3.

Higher site densities also necessitated a more careful management of the scarce frequency resources. As the frequency planning and allocation methods were widely based on predicted propagation data, the propagation models had to undergo a further refinement. Examples of such more accurate models are the ones based on ray tracing. Some ray-tracing models can be found, for example, in [19] and [20].

In addition to the propagation model development it was noticed that the increasing capacity demands could only be met with more accurate frequency planning. The frequency assignment together with neighbour cell list (for handover purposes) planning and optimisation were the main issues when planning GSM networks. In the case of GSM, frequency hopping was introduced to further improve the spectrum efficiency. Advanced frequency allocation methods can be found in the literature, one example based on simulated annealing is in. In a method for automatic frequency planning for D-AMPS is studied. In advanced features for Frequency Division Multiple Access/Time Division Multiple Access (FDMA/TDMA) systems are introduced. These features include improving the frequency reuse by applying:

frequency hopping;

adaptive antennas;

fractional loading;

hierarchical cell structures.

It can be concluded based on several papers (for example,) that the prediction of propagation is of limited accuracy due to the fact that the propagation environment is very difficult to model and thus generating a generic model, which is applicable in multiple cells, is by nature accuracy-limited. This is especially applicable when the fading characteristics (both fast and slow) need to be considered. The latest radio network control activities concentrate on the closed-loop optimisation of the plan. The initial planned configuration is (semi-) automatically tuned based on statistics collected from the live network. Proposals for handover performance improvement in terms of correct neighbour cell lists can be found in and. The important aspect with this method is that neighbour relations that are initially based on propagation prediction are autotuned based on real measurements. Thus inaccuracies can be compensated in the optimisation phase. A similar measurement-based concept can be utilised also for WCDMA intra- and inter-system neighbour relations; this is discussed in more detail in Chapter 7.

Recently, methods for GSM frequency planning based on mobile station measurement reports have been introduced and implemented, see and. The possibilities offered by these reports in GSM and WCDMA should be more utilised in the network control process (planning, optimisation and integration of those two).

Another new trend in radio network planning research is plan synthesis, meaning automatic generation of base station site locations depending on a cost function output. This is briefly discussed in Section

In cellular networks network utilisation control requires such functionality that can utilise the measured feedback information from the network and react correctly based on that. Therefore, it is crucial that the planning phase is tightly integrated into other network control functions and the network management system. This is especially important in the case of WCDMA, owing to the fact that there will be a multitude of services; that is, customer differentiation will set a multi-dimensional matrix of Quality of Service (QoS) requirements. Planning such a network very accurately is not feasible due to limited accuracy of the input data (propagation, traffic amount, traffic distribution etc.). An example of the integration of a network management system and planning for 2G systems can be found in.

Integrating the network management system and advanced analysis and optimisation methods for effective configuration parameter provisioning and 'pre-launch network performance' estimation are the next challenges in the radio network development and optimisation area. An example of the effective integration of the planning tools' functionalities into the Network Management System (NMS) is, for example, visualisation of statistical performance data on cell dominance areas. Furthermore, the adjacency relations can be directly generated in the NMS based on the base station coordinates and simple distance-based rules. These initial lists can be later autotuned based on statistics collected from the live network. Also WCDMA scrambling code allocation can be done in the NMS without interfacing to external planning tools by utilising the mobile station measurement reports required by both the GSM and WCDMA standards. These reports contain information that can be used to complement the information generated traditionally by the planning tool (propagation, traffic density, etc.). When mobile station positioning methods are fully in use, another huge new dimension to optimisation tasks will be opened.

Together with the introduction of the all-IP mobile world, QoS provisioning becomes very important for the operators. This directs the network control increasingly away from radio access network control to service control. In practice this means an increased abstraction level for the operator and a new era for network management.

This book concentrates on the challenges with WCDMA networks. Furthermore, one of the main motivations is to move away from the 'analytical' control of the network, and enhance the modelling and tools to give a picture as realistic as possible of the actual network performance. Network functionalities can no longer be considered as individual entities, but their interactions must be accounted for. In the analytical, ideal world this has no relevance, but in the true cellular world the understanding of these interactions and network element algorithms and their limitations is essential.

It can be stated that radio access evolution towards third generation is the first big evolutionary step after the birth of cellular systems. The large step in radio access development, the great interest in applications and services also forces the radio network planning and optimisation process to improve to fully support the offered possibilities.

1.3 Introduction to Radio Network Planning and Optimisation for UMTS

The mobile communications industry throughout the world is currently shifting its focus from 2G to third-generation (3G) UMTS technology; that is, it is investing in the design and manufacturing of advanced mobile Internet/multimedia-capable wireless networks based on the Wideband Code Division Multiple Access (WCDMA) radio access platform. While current 2G wireless networks, in particular the extremely successful and widespread global GSM-based cellular systems, will continue to evolve and to bring such facilities as new Internet packet data services onto the market, more and more radio network planners and other wireless communication professionals are becoming familiar with WCDMA radio technology and are preparing to build and launch high-quality 3G networks. This book has been written in particular for those RF (Radio Frequency) engineering professionals who need to thoroughly understand the key principles in planning and optimising WCDMA radio networks, though it should also prove useful to others in the industry.


Excerpted from Radio Network Planning and Optimisation for UMTS Copyright © 2006 by John Wiley & Sons, Ltd . Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.