p2pnet - rss feed: 3G tries to get indoors
p2pnet.net News Feature:- One of the major challenges of 3G deployment has been achieving sufficient coverage, particularly inside buildings, to make the economics work to full advantage.
New research from ABI highlights an increasing trend for 3G operators to instal UMTS pico stations in large indoor structures to boost coverage, with European shopping malls leading the way. Since pico stations can be an expensive option outside large commercial centers or enterprises, there is also an increasing focus on how integration of cellular and Wi-Fi can improve the overall service for business data users.
The weaknesses of cellular networks within walls have not been such a serious issue as long as networks have been voiceoriented, since most voice usage is outdoor and mobile, but with 3G and its update HSDPA focusing on data revenue, it becomes increasingly likely that users will spend much of the time on their smartphones inside their offices or homes, or on public transport, and that primary usage will be urban.
Also, in the 2.1GHz frequency, where it has mainly been launched in Europe, UMTS has reduced ability to penetrate walls. In the US, the problem will be less serious, since UMTS will be mainly deployed in 1.9GHz and even 850MHz, which have better penetration capabilities.
One of the reasons that Wi-Fi has been widely - and usually over optimistically - touted as a rival to 3G is that it was created for indoor use, and it cheaply and effectively fills some key market gaps that are intrinsically difficult for 3G. These include city hotzones, enterprise indoor networks and hotspots in public places. This means that indoor solutions for 3G are needed at a far earlier stage than with GSM, because of different usage patterns, creating hot areas of usage indoors that will skew the macro coverage.
Addressing 3G penetration indoors
Two approaches have been held up as the best way to address such hard-to-reach areas with 3G ? IP-connected picocells or distributed antenna systems. Increasingly, big operators like Vodafone and NTT DoCoMo are interested in pico stations within buildings to boost coverage in the market that all first tier cellcos are chasing, the enterprise. This will become even more critical with the accelerated roll-out of HSDPA, which will raise expectations among users of 1Mbps access any time, anywhere.
Indoor data users will be more sensitive to delivered data rates. The average base station RF path loss will be greater than for voice calls because 3G base stations will generally be collocated with GSM and GPRS ones, but this will no longer be an ideal topology as the weight of coverage moves indoors. Although cellular systems are designed to operate over a wide range of path loss, users will have high quality of service expectations for data in terms of rates, but rates will rise and fall according to conditions.
All 3G standards use some form of adaptive coding in which the data throughput rate is adjusted to match channel conditions.
If the received signal level is low or interference is high, the data rate will be slowed and perceived coverage will be
lower. Therefore the traditional cellular topology is inadequate to deliver high performance data rates that the user perceives to be as good as those from alternatives like WiMAX. Extending the network and shifting the balance to the urban indoors is a key success factor for 3G.
There are many lessons to be learned from the Wi-Fi world here. As voice over Wi-Fi and dual-mode cellular/Wi-Fi handsets emerge, enterprises are increasingly interested in systems where WLan is the primary network for voice and data, and users are automatically handed off to cellular when this is unavailable.
The main reasons are to reduce cellular bills and to take RadioFrame?s architecture advantage of Wi-Fi?s higher data rates. However, this could effectively exclude the cellular operator, and the maker of cellular pico stations, from the enterprise.
Combining WLan and 3G
This gives these companies a powerful incentive to provide combined WLan/cellular systems, so that they have control of the WI-Fi element ? and this picture will include WiMAX too from 2005 ? and can not only collect the revenues but also, to some extent at least, influence the pricing levels.
Key technologies will be software definable base stations and picocells, which support several wireless networks imultaneously and, in the future, will work with cognitive end user devices that switch intelligently to the best available connection according to user defined criteria such as cost or speed. We are seeing interesting developments in this area from companies like PicoChip, whose software defined base station already supports WiMAX as well as all the established networks.
Another key company is picocell provider RadioFrame, which has developed technology for easily deploying mix and match indoor networks supporting GSM, iDEN, CDMA, Wi-Fi, UMTS, EDGE and future technologies such as 4G and 802.16. It provides an Ethernet-connected base chassis unit with interfaces to the Lan, internet, cellular network and so on. It places multiple, low powered radios at the network edge and manages them as a unified platform.
The base chassis connects to a Network Chassis Unit, which supports integration with the PBX and wired data networks, and an
Airlink Chassis Unit, which provides wireless processing for up to eight RadioFrame Units. These units contain up to seven RadioBlades or transceivers, each of which can support a different wireless standard. These convert all traffic into Ethernet packets and so apply Ethernet functions such as Quality of Service.
The first customer is US operator Nextel which will offer a service to enterprises aiming to improve the coverage of its iDEN network within buildings while also enabling enterprise users to implement iDEN and Wi-Fi blades in the same RadioFrame base station, with the option to mix and match GSM, 802.11g, UltraWideband and other blades as they are developed. Nextel said the improvement in indoor quality of service over iDEN was ‘a five bar jump’.
The RadioFrame system clearly opens the way to voice over IP and seamless roaming between WLan and cellular and points to the most likely way forward for indoor coverage.
Another problem network builders face is getting their picocells indoors at all. Again, there are lessons to learn from Wi-Fi. Some businesses run parallel networks from their WLan base stations, one secured for internal use and the other open for use by visitors and even by the general public in the neighbourhood. This provision of a ‘hotspot’ by a business is particularly common in small towns where there are one or more mediumsized companies aiming to improve the facilities of the local community. Pico base stations supporting Wi-Fi and cellular, and the increasing integration of the two networks, could be installed on a similar model.
In any circumstances, while the media may portray 3G and Wi-Fi as rivals, it is clear that, for 3G networks to achieve their full potential, their operators will need to treat WLan as their ally.
Caroline Gabriel - Wireless Watch, UK
