<font color=red> RoamAD 80211.b network
See roamad.com
telecommagazine.com
< There are high expectations for high-speed W-LANs. Today's customer base of 400 to 500 million fixed-line internet users, combined with increasing laptop sales, could boost the global W-LAN market size to as much as a hundred million users by the second half of this decade (Figure 1 ).
Yet operators still have to consider whether there is a business plan to justify getting involved. Certainly, W-LANs could successfully sell on their bandwidth advantage over GSM, GPRS or even UMTS 'hotspots'. Company fixed-line LANs could also migrate to W-LANs, but questions of data security, compatibility of systems, and the reliability of the service provider would all have to be considered by business users. Making money from the W-LAN market will be no walk in the park for operators eyeing up this technology as a comparatively cheap route to offering high-speed internet access to business users.
Market developments
Although the W-LAN market is currently fragmented, the number of W-LAN installations is growing worldwide. Having said that, there is no big player visible on the horizon with large investments and service offerings at hand - it is small and medium-sized operators which are emerging. There are also access networks - free of charge - set up by vendors to promote W-LANs and so we have a diverse picture of different business models all competing in this new wireless internet/intranet market.
Depending on which business model wins out, there are three possible scenarios for W-LAN development over the next six years:
a 'niche' business-dominated scenario with laptop users in the order of 20-30 million by 2006;
a medium market scenario comprising both business users and consumers in the order of 30-50 million subscribers by 2006; and
a mass market scenario with over 90 million W-LAN users, including laptops and PDAs. This would happen through a large involvement of mobile network operators (MNOs).
The services rollouts today are in relatively few locations, although a number of operators are already providing public W-LAN coverage - Sonera, Telia and BT, for example. The failure of Metricom in the US last year, however, serves as a reminder that it will not be easy for start-ups to make money from providing W-LAN-only services.
For public operators, question marks remain over the implications of shared frequency spectrum use, both in public areas and in corporate buildings. By using the ISM band for IEEE 802.11b/g technologies, there is always the chance of interference with other technologies that may impact service quality. There are also some countries that do not even allow public use in these bands. There should, however, be more operator opportunities in the 5GHz band (more bandwidth and greater harmonised technology).
In addition, acceptance and use of new software with enhanced functionality - such as Windows XP - could influence the market development as an enabler for voice/video over IP. W-LAN can also be seen as a stimulator for the 3G market.
How fast can W-LANs go?
Although 802.11b W-LANs are generally promoted as having 11Mbps capability, it has to be remembered that this is gross throughput; it does not take into account the radio transmission protocol and the impacts of the 'hidden terminal' problem. The practical bit rates are therefore lower (55 per cent and less) and are dependent on the number of users and their environments.
Practical cell radii also vary depending on the product. Benchmark tests have shown that the direct sequence spread spectrum (DSSS) techniques may achieve bit-rates up to 6Mbps depending on transmit power, cell size, number of simultaneous users, application, and cellular environment.
W-LAN hotspots, according to the IEEE 802.11b standard, require 22MHz minimum bandwidth. As such, up to three parallel W-LANs can operate within the ISM band of around 80MHz. The capacity of a system is a non-linear function depending on environment, distance and number of simultaneous users. Bit-rates between 500Kbps and 2Mbps seem to be feasible, without considering the backbone network transmission capacity.
There is no doubt that W-LAN technology should not be used for universal coverage. Even in cities the number of access points would go up to the order of thousands if that was attempted. For example, a typical 3G/UMTS cell area of 1km2 would require 200-400 W-LAN sites. All these sites would have to be supplied with expensive wireline transmission backbone capacity.
Fig. 1 W-LAN users worldwide
W-LAN integration with 3G
By mid-2002 more than 110 frequency licences had been granted for 3G network deployments, guaranteeing exclusive frequency spectrum in the order of approximately 30MHz (duplex) per operator. In addition, CEPT has designated licence-exempt spectrum for the use of UMTS/TDD technology in the bands from 2010-2020MHz. It is expected that by 2005 UMTS coverage will reach 200-300 million people. Dual-mode 2G/3G terminal technology will further enable every user to access a mobile network with either UMTS or GSM/GPRS technology. This means nearly global 2G/3G coverage for the user - at least for basic services as there are now 500 GSM networks in operation.
Although UMTS can theoretically offer bit-rates up to 10Mbps depending on mobility parameters and up-/downlink directional traffic requirements, it is designed mainly for large capacity with medium bit-rates per users, say to several hundred Kbps. The continued development of compression techniques for audio/video will allow streaming for small screens from 40Kbps up to 384Kbps (high quality). As UMTS is already in operational use, the first 3G/UMTS terminals confirm bit-rates of up to 384Kbps on the downlink side.
UMTS bit-rates are specified for both operational modes - UTRA-FDD and UTRA-TDD. TDD is a comparable radio access technique to W-LANs, but it allows delivery of all services (eg, voice, video and data) and in all environments from pico to macro cells.
As far as W-LAN integration with UMTS/3G is concerned, the 3GPP standardisation project is discussing the following issues: complexity and degree of integration; technologies according to IEEE 802.11b (a, g); and timing of specifications.
The network integration needs to be investigated from a service point of view, namely mobility and security/authentication. On the 3G side, the USIM card is associated with the mobile backbone network and so a solution with a radius server and/or multiplexer switch to the internet will be required.
There is also some complexity in the field of authentication, billing and security. Naturally, the W-LAN user needs to have an ISP account linked directly to their corporate ISP as a matter of course. But as the business traveller using a laptop can be described as a mobile user, this should allow an extension of the MNO's subscription on his mobile account, but implies an MNO ISP service provisioning for intranet access.
A weakness of W-LAN and 3G integration could be lack of security as the WEP (wire equivalent privacy) algorithm used by 802.11 is not sufficient. As many access points run without default settings, the virtual private network approach cannot easily be applied in order to achieve end-to-end security.
...contd... > |
