|
||||||||||||||||||
New Zealand Engineering 1999 July IT & Networks 
Internet Generation Wireless Communications In this short article I present a brief overview of third generation (3G) wireless communications and related developments in the internet. Since the 1980s there has been an ongoing effort within the ITU to define a single standard for public wireless telecommunications that could form the basis for networks in every country. It is officially called IMT 2000 (International Mobile Telecommunications in the year 2000) and commonly called 3G. The difference with existing technologies such as GSM (Vodafone) and TDMA (Telecom NZ) is intended to be principally in four areas: coverage, capacity, quality and user bandwidth. Most focus in the media has been on the planned increase in user bandwidth – up to a maximum of 2 Mb/s in local areas with low-speed mobility (see table) - and generally 3G is associated with wideband. 3G has been summarised as two things. First, a flexible radio interface; one that supports realtime and non-realtime services, with bandwidth on demand. Second, a multimedia core network. One essential change to achieve this flexibility is to shift from narrowband circuit transmission and switching to packet mode. Work on IMT 2000 standards accelerated during 1996 with the announcement by ARIB of Japan that they would select the European wideband CDMA technology for a new 3G system. This was driven by their need for more wireless communications capacity than they consider can be delivered using their current second generation technologies: PDC, a TDMA derivative, and cdmaOne, a narrowband CDMA standard. The original intention was that the ITU would recommend one technology for the whole world, however, that was never likely to happen after the US licensed frequencies in the 1900 MHz band and 2G narrowband technologies were deployed. During 1997 technology decisions for the radio interface were made by ETSI (Europe) and the TIA (USA). ETSI chose wideband CDMA for wide area systems. They decided not to emphasise backwards compatibility with deployed 2G systems in order to give more freedom in choosing an optimised technology. However, they have the luxury of new spectrum in the 2 GHz band. In the US several technologies were proposed, the main ones being UWC-136 and cdma2000. In the US there is no additional spectrum for 3G systems. They must cohabit in the same spectrum as the narrowband systems and hence the key difference in the US approach – backwards compatibility. The (nearly) end result is that the ITU plans to prepare recommendations that include four modes for IMT 2000 (see table). There is a considerable effort now ongoing to harmonise WCDMA and cdma2000 so that they are – if not the same – at least compatible in the sense that it is easy and inexpensive to build mobile units that operate on either standard (dual mode terminals). Some of the barriers to harmonisation have been removed since the conflict over IPR between Qualcom of the US and Ericsson has been resolved by the two parties. To deploy WCDMA, an operator needs to secure spectrum in the 2 GHz band (also called the IMT 2000 band, or the UMTS band). Licensing of spectrum in Europe has begun with Finland. Most countries are expected to offer at least two licences, and some, like the UK, may offer up to five. It is uncertain if existing operators will get UMTS spectrum – it largely depends on the licensing process the administration chooses and on regulations that may control operator ownership. Therefore, there is a 3G alternative for GSM networks, called EDGE, that is compatible with operating GSM systems. EDGE also forms the basis of the North American UWC-136 standard and so there is a common 3G radio technology for GSM and TDMA. Hence we speak of TDMA and GSM converging. EDGE handsets will be able to operate on either network. Convergence of GSM and TDMA also concerns the core network. TDMA already offers packet data service (the CDPD service being used for the America’s Cup Challenge in Auckland). Packet data will be introduced in GSM networks during 2000 using the GPRS standard. When we move to 3G, GPRS will provide the core packet data network - both with EDGE and with WCDMA. At Ericsson we call this Phase 1 – best effort packet data based upon GPRS with a wideband radio interface. We will introduce products starting in 2001.
Before talking about Phase 2, we need to introduce the concurrent revolution in communications, the internet. Putting it simply, the world of telecommunications can be divided in two: telephony, and the internet. The internet may turn out to be the biggest thing ever to happen for telecoms. It is easy to understand why. In any form of communication it is essential the parties share a common language (and hence standards). Internet protocols provide in effect an open language that any computing device can use to communicate with any other. Right now a critical point is being reached or has been passed in many countries: the amount of telecoms traffic in the form of data exceeds that in voice. And the trend is accelerating. Simply put this means networks should be engineered to efficiently transport data and voice can ride along. It is tempting to think that voice is just another form of data (it’s all just bits in the digital world). However, not all bits are created equal. In a two-way conversation, voice bits are no use if they are delayed over about 0.5 sec. Experiment has proved conversation becomes intolerable. This raises a major problem for the internet – all the hype around voice over IP (VoIP) notwithstanding. The internet is designed and built as a best effort data transport medium. Guaranteed quality of service (including reliability) is a new concept in this domain. Telephony networks are engineered for extremely high availability and predictable and consistent quality of service. Up until the internet explosion, the fastest growing telecom technology was the mobile phone. Now the challenge is to make the internet mobile, first for best effort data (Phase 1) and then for realtime services, like voice (Phase 2). GPRS and CDPD provide mobile users with end to end, best effort, IP based data communications. In Phase 2 we will introduce architecture changes in the core networks to provide realtime services over the IP transport with guaranteed quality of service. At Ericsson, we call this Phase 2 network, the Internet Generation Network. It is the fusion of the best elements of the internet and telephony domains: the openness and innovation of the net with the quality and reliability of telephony. During the past two years we developed the core concepts and now we and our partners are leading efforts to standardise interfaces and associated protocols. We will introduce products starting around 2003. The Internet Generation Network has many benefits. It is being designed to facilitate an all connected and all communicating world. To enable users to access their own personal information and services using whatever technology they choose – mobile, or fixed. To be always connected, always online. Advances in mark- up languages and proxies will mean the presentation (look and feel) is as much as possible the same irrespective of the technology chosen. And it enables operators to reduce their costs by as much as ten times compared to today’s levels. 3G wireless networks will start to be deployed in Japan and the US during 2001, and in Europe in 2002. Initially, users will be offered innovative services based on best effort delivery. From around 2003, they will be offered multimedia with quality of service. What will they use it for? Look at the internet. That’s where the innovation will come from. 3G networks make it practical to connect everything. Repeat, everything. And everyone. The possibilities are only limited by imagination. Andrew Duxfield, manager, Business Development, Business Unit, TDMA Systems Ericsson Radio Systems |
||||||||||||||||||
 |
|
|
|
