There is a lot of industry momentum building up for LTE femtocells. Much of this is from the vendor community, who have invested to create the chipsets and technology software platforms which enable it. For example, picoChip have already developed and proven an LTE femtocell prototype last year using the emerging standards.
Some operators have made it clear that they plan to adopt LTE quickly. Both CDMA and GSM operator communities have chosen the technology, with the likes of China Mobile, Vodafone, Telefonica, Verizon and others strongly supporting it. Indeed, some operators have even said they will skip HSPA+ and instead invest in LTE so that they can more quickly arrive at higher performance levels.
LTE is a new radio interface, similar to WiMAX, using OFDM (Offset Frequency Division Multiplexing) which can cram more bits per second through the same expensive radio spectrum. Other benefits include shorter latency (end to end turn around for data applications), higher data rates and better reception in difficult urban conditions.
Some vendors are suggesting that it may be more cost effective for operators to rollout LTE using femtocells, using an "inside out" approach.
First, we would ask whether LTE femtocells are indeed appropriate for domestic use. We think LTE will initially be deployed for use by data devices. Operators will want to migrate as much of the heavy data traffic away from their (by then) heavily congested networks. USB dongles working with laptops are likely to be a prime target. The benefit for users may be not just improved performance, latency as stated above, but more capacity before the new network fills up. As LTE becomes more widespread, we can expect smartphones to incorporate the technology with voice only phones following after an extended period during which product costs drop.
Before 4G LTE becomes widespread, we expect 3G HSPA femtocells to become more widely deployed - certainly by anyone who would be interested in a 4G LTE model. These 3G femtocells are likely to provide more than enough individual capacity to satisfy the most demanding mobile device consumer when in their home. For laptops or other extremely heavy data usage, WiFi may be seen as a more accessible alternative. The limiting factor in many cases will be the speed of the connecting broadband line, which is typically around 4Mbit/s today - much less than the peak rate that a 3G femtocell can deliver. Depending on the country (or state/county) you live in, it may take some years before mass deployment of fibre achieves the higher speeds which would deliver visible benefits for an LTE femtocell.
Having already got a 3G HSPA femtocell, we don't expect many customers to want to upgrade - even if they have the higher speed wireline broadband. It would require new mobile devices and/or require to support both 3G and 4G simultaneously - which makes it inherently much more expensive.
Instead, we are hearing suggestions that 4G LTE femtocells would be used directly by network operators (which we would classify as picocells) as a means of rolling out LTE. By deploying femtocells with very small cell sizes, these devices can increase capacity much more substantially than by changing the radio interface alone. The capability to self-manage, self-optimise and self-organise means that they would adapt to the surrounding environment and automatically configure themselves - saving cost and complexity of developing, modelling and maintaining a radio frequency plan for every cell as is done today.
This would also mean that the initial products would be LTE only (no need to support both 3G and 4G), with mobile broadband USB dongles capable of both 3G and 4G (and possibly 2G) being used with laptops to connect to them.
There's no doubt that smaller cells combined with the higher performance of LTE and the self-managing/optimising techniques developed for femtocells can deliver a massive amount of capacity in targetted areas.
Perhaps we'll see one of these at a lamp post near you from 2010 onwards.