Adjacent-Channel Emissions and "Dead Zones" in W-CDMA Networks
Webb on Wireless
17 July 2002
William Webb reveals that adjacent-channel emissions resulting from the way spectrum is allocated could lead to "dead zones" in W-CDMA networks.
Detailed study of the 3G W-CDMA specifications reveals a potential problem. Take the situation where operator A has been allocated a particular frequency, and operator B the adjacent frequency. If both operators locate their base stations in the same geographical area but at different sites, then subscribers to operator A who are close to operator B's base station will experience interference that could cause their calls to drop. This would occur because the restrictions on adjacent-channel emissions are not stringent enough to prevent base stations from emitting significant levels of radiation into neighbouring channels. These "dead zones" could make 3G calls less reliable than those in 2G.
It is difficult to predict just how much of the coverage area this might affect. It depends largely on whether the manufacturer's equipment actually produces less adjacent-channel interference than the specifications demand. It also depends on individual sites and the loading on the network.
Dead zones are not an issue in 2G because guard bands are allocated between the operators. These bands ensure that adjacent-channel interference falls to acceptable levels. However, because of the much wider channel bandwidth in 3G this kind of guard band allocation has not been possible. Furthermore, in 2G, operators have many channels and so can reassign frequencies to work around any problems that emerge. In 3G, operators typically only have two channels and both might be next to those of other operators.
This is a particularly awkward problem to solve. As an operator, improving the adjacent-channel emissions on your base station will benefit your neighbour (and competitor) and not yourself. Hence, only if operators decide to work together, or regulators intervene to demand improved performance, will it be worth tackling the problem at source.
One option is for operators to co-locate most of their base stations with those of their neighbour. In this way interference would only be experienced in areas where the desired signal was strong. However, this approach is difficult for those with two neighbours. Alternatively, operators could liaise and design their cell plans to increase the signal strength around their neighbour's sites. In the future, smart antennas may be able to help by boosting signal levels into dead zones. None of these alternatives, though, sounds particularly appealing.
Solving this potential problem together as an industry is certainly a better way to proceed, but this requires major coordination between operators, manufacturers and regulators. Otherwise, achieving even the same level of call quality as experienced with 2G could be a challenge for 3G.
Author: William Webb is a managing consultant with PA Consulting Group's Cambridge Technology Centre.
wireless.iop.org |
