Hats off to Hay Systems Limited, the Scottish independent SMS service and equipment vendor. Having gone against the mainstream trend of 3G femtocells, they've developed their own 2G GSM one and are intent on creating end user demand for it. Prospective customers who want to solve their indoor coverage problems can register their interest at a special website setup for the purpose. Where enough registrations are submitted, HSL can then use this to justify sales of their solution to the specific network operators.
HSL's femtocell is GSM
This approach differs from mainstream 3G technology by using the well known and understood 2G GSM technology. This provides not only voice and SMS, but can support data services such as MMS (picture messaging) and internet browsing/email as well. Read the full product spec for more detail.
At 384kbit/s maximum, this will be somewhat slower than the 6 or 7Mbit/s achievable with the latest 3G devices (such as the iPhone) and 3G femtocells, but adequate for many types of service such as blackberry email. HSL argue that even higher data rates would be available to such devices that require it through WiFi instead.
By using a single GSM carrier, up to 14 simultaneous voice calls can be handled (based on using the GSM half-rate codec). This capacity is shared with the data services, so you can't have 384kbit/s and 14 voice calls at the same time.
Range is quoted at 30 metres (or up to 50m if outdoors)
End-user pricing is not specified but has been suggested before to be comparable with the 3G femtocells from other vendors. I suspect this would be mainly dependent on volume of product shipped.
The argument for using 2G GSM rather than 3G is that it is compatible with every GSM and 3G phone on the market (all mainstream 3G phones have 2G built in, except perhaps some Japanese models). This means there is no need to upgrade any phone in a household to take advantage of the femtocell.
I guess this is comparable to Sprint's 2G CDMA femtocell- known as Airave - which has been successfully launched across the US. It's also remarkably similar to ip.access nanoGSM products (and those of similar vendors), who have been touting this technology for quite some time and carreied billions of minutes of call traffic through their installed base.
The main difference in terminology (calling it a femtocell rather than picocell) is presumably that its aimed for self-installation by the domestic consumer rather than being directly installed by the network operator.
The initial product is designed for use in the 1800MHz band. Here in the UK, all operators now have some 1800MHz spectrum in addition to the 900Mhz that was used for initial GSM service. All phones made in the last 10 years or so are dual-band, so can operate at this frequency.
A 900Mhz version of the product is planned for 2010.
Remote management of the device is based on the TR-069 standard as used by many set-top boxes and broadband modems.
GPS is optionally included in the device, presumably to help in both aspects of timing/synchronization (which can be derived from the highly accurate GPS clock signals) and location (so the operator knows where it is and can use this for emergency location information, validating it is inside licensed territory etc).
Some issues which might concern end users
As with today's 3G femtocells, handover of calls when entering the femtocell area is unlikely to work - the handset won't know to search for it. Whilst this problem will be fixed for 3G femtocells when handsets incorporate some of the changes specified in the latest standards, this is unlikely to be addressed for 2G systems.
Handover of calls when leaving the femtocell area should work seamlessly, assuming the femtocell can detect the signals from external masts.
Relatively slow data rates - some 384kbit/s versus 7Mbit/s. HSL argue that users could opt for WiFi access from their smartphones for high speed data access instead, leaving the femtocell for voice, text and simpler data usage.
Some issues which might concern operators
Limited choice of vendors for 2G femtocells
The femtocell standards have concentrated on 3G technology, and specifically enable operators to buy them from multiple suppliers. This will foster price and feature competition, keeping prices down and functionality up.
There is no fully implemented 2G GSM equivalent of the Iu-h interface. In theory, the Abis interface defines the protocols between basestations (BTS) and basestation controllers (BSC). But this doesn't run over broadband internet yet. In practice, most BTS vendors have developed their own variation of Abis and operators would buy their sets of BSC and BTS from the same vendor.
There is perhaps a smaller distinction between 2G picocells and femtocells, so you might well argue that ip.access nanoGSM and other similar products would be the direct competition here.
Integrating with other systems
As we've said before, launching a full femtocell solution to end consumers requires a lot of staff training, systems updates and procedures to be put in place. Everything from customer care, dealers, billing and procurement can be affected.
I commend HSL for going direct and appealing to the end customers for support in launching femtocells as a coverage solution. It should help demonstrate and validate the demand for coverage solutions in many places.
Whilst I can see the benefit of being compatible with all existing GSM/3G phones (an issue that was brought home to me when I got my own 3G only femtocell), I think some of the issues raised above may direct operators to 3G products.
There have been some good 2G GSM only products available for some time, such as the ip.access nanoGSM , which have succeeded in addressing specific markets in limited volumes (airplanes, maritime, enterprise). I don't see why the HSL product would be radically different from what is already available.