Most 3G UMTS operators have been allocated 10Mhz of paired frequency spectrum (i.e. 10MHz uplink+10MHz downlink), typically in the 2100Mhz range allocated worldwide. This is used with two separate paired WCDMA carriers of 5Mhz each. Wideband CDMA differs from the original mobile CDMA system by using fewer, wider frequency range transmissions – CDMA typically uses carriers with a spread of 1.5Mhz.
Apart from the name, this provides benefits in poor or difficult transmission areas (e.g. built up, urban areas). WCDMA also differs by not requiring the basestation transmissions to be phase synchronised with each other (I think this was to get around a Qualcomm patent), making it more complex but in the long term easier to manage.
In the US, where spectrum has been sold off in smaller chunks of 1.5MHz bands, it’s more difficult to manage and accumulate blocks of 5MHz for 3G UMTS carriers.
LTE, the 4th generation mobile standard due to rollout out around 2010, takes these limitations into account, and can be deployed with carrier sizes of anything from 1.5Mhz to 20Mhz. The wider the carrier frequency, the higher the peak data rate which can be achieved.
So how do 3G UMTS operators allocate their spectrum
In the early launch phase, most operators just used a single 3G UMTS carrier frequency. This allows handsets to receive signals from more than one basestation at the same time, in a procedure known as soft handover (or handoff). The same call can be transmitted from more than one cellsite with different CDMA scrambling codes. The handset decodes both signals simultaneously, determines which signal is of higher quality (has fewer errors) and presents that as the incoming voice or data.
As traffic capacity has grown, most operators have “lit up” their second carrier (or third if they have one). They may also have sectorised basestations, like cutting a cake into 6 pieces, where the same frequency is reused three or six times, once for each segment. From both of these practices many cellsites now have dramatically increased their peak capacity. For cost and power efficiency purposes, operators have started switching off those second carriers during off-peak hours (e.g. overnight). There is a primary and secondary (overload) carrier.
What about frequency allocation for femtocells
With only two carriers to play with, there simply isn’t enough to allocate a dedicated frequency for femtocells - the ideal option. Instead, operators are allocating their second (or overload) frequency for this purpose.
Femtocells do not use soft handoff between the macrocell and the femtocell (or between femtocell access points). Handsets report the measured signal strength and quality to the network, and use this to decide if and when to switch to the femtocell. This would normally happen when the handset is in idle mode (i.e. not processing a call or data session), so is simply reported as a “location update” to the network – any incoming calls can then be directed via the femtocell because the network knows where to find and contact the handset. If a call is in progress, than a hard handover is used to switch the call between the outdoor network and the femtocell. The user is often unaware of the switch, except that the call quality may radically improve (or degrade if leaving the femtocell area).
There is a case where mobile users who are not authorised to use the femtocell but are also served by the same network operator, are using their handset near to the femtocell. For example, if you are visiting a friend, but are not on the list of authorised users of his/her femtocell, your calls will continue to be served by the external macrocell network. Tests have shown that when operating at the same 3G frequency, there can be an isolated poor coverage area around the femtocell for these users. This is known as the “near far” problem, and will be familiar to students of CDMA technology.
The solution to this problem is to switch these visiting users to the other carrier frequency (or even more likely to 2G, which penerates in building better due to the lower frequency commonly used). This process is handled automatically by the handset, which will have been monitoring external 2G and 3G cellsite reception.
Of course, if the operator has poor national coverage and their signals don’t penetrate indoors (perhaps a reason why the customer bought the femtocell in the first place), then this won’t be much help. But in this case, the visitor wouldn’t have had good indoor reception beforehand anyway. It’s probably best for the visitor to ask to be added to the closed subscriber group for the femtocell, or for the femtocell owner to consider opening up access to all visitors as a benefit to all their friends and others who might need to use it.
For those who are served by other mobile networks, the femtocell will make virtually no difference because it operates only at the frequencies licensed to its own network and will ignore/be ignored by handsets from other providers.
A research paper entitled “Uplink Capacity and Interference Avoidance for Two-Tier Femtocell Networks ” from a Research project sponsored by Texas Instruments at the University of Austin, Texas.
Authors:Vikram Chandrasekhar and Jeffrey G. Andrews