3G Femtocells: Who Makes What?

It’s great to see AT&T start marketing their 3G (WCDMA) femtocell, the 3G Microcell. Unlike the 2G femto offerings from Sprint and Verizon in the US, this not only promises a premium voice experience but also full speed 24/7 mobile broadband.

Following the launches from SoftBank, StarHub and Vodafone, the AT&T Microcell is a clear signal that the femto market is taking off across the world.

And on the face of it, there are around 20 3G femto vendors competing to supply operators with 3G femtocells. But that’s only half the story.

Firstly, you can’t always go by the brand you see on the box. For example, the AT&T Microcell is supplied by Cisco, but the femtocell inside the Cisco device was developed by picocell/femtocell developer ip.Access.

Secondly, the only true measure of a femtocell product in this hyped-up market is a commercial launch. Taking this measure, there are only four 3G femtocell vendors in the market today (in alphabetical order):
- Alcatel-Lucent at Vodafone, UK
- Huawei at StarHub , Singapore
- Ip.Access at AT&T, USA
- Ubiquisys at SoftBank, Japan.

There are some interesting similarities here. All of these companies have been doing intensive femtocell development for at least 3 years, and all have been working with their operator customers for at least two years. They also have in common 3G Radio Access and Core Network expertise as well as significant knowledge and experience working with mobile handsets.

This expertise is vital, because femtocells are extensions of the Carrier’s Macro Network. They are an integral part of a highly complex, high-reliability carrier-grade network with all that brings in terms of capabilities, features, network IOT, management systems, handset IOT and cognitive radio functions. The result is a software system that presents an extremely high barrier to market entry for any prospective femtocell vendor, which is why there are so few in reality. It is also a reason why no femtocell reference design has ever made it to market, despite having being around for years.

But femtocells are also consumer products, and they have to meet the demanding volume and price points of this industry. This is a well trodden path by home gateway manufacturers and the component companies that supply them – it just takes time.

There are other 3G femtocells under development, but today it’s all about four industry pioneers.

IMS femtocells: the basis for a new generation of IP-PBX?

Next Generation Networks (NGNs) featuring IP Multimedia Subsystem (IMS) architectures have been steadily taking shape at many operators over several years. The justification and suitability of IMS, particularly for mobile operators, is still hotly debated.

The simple reason for the debate is that everyone agrees the business case has to work, but every operator is starting from a different point with a different set of objectives. So IMS can be seen as a moving target but it is increasingly being deployed, particularly with LTE edging towards reality. The following discussion is not intended to debate IMS’s pedigree but more to consider what benefit it brings to femtocell deployments and why IMS femtocells are an increasingly hot topic.

Most 3G femtocell deployments, either launched or pre-commercial, involve some sort of integration with the legacy core network using a femtocell gateway, which essentially treats femtocells as multiple base stations. This approach has been standardised to a large extent with the 3GPP’s Iuh interface, which will eventually allow any femtocell to work with any gateway.

The potential of IMS femtocells
But femtocells are not base stations. In homes and businesses, they are IP devices that sit on a LAN, and are therefore peers of a diverse range of other IP devices, such as home media servers and office IP-PBXs. Interaction between phones and other IP devices on the LAN is one of the major attractions of femtocells. But legacy networks were not designed to deal with serving media from the edge of the network, and although this can certainly be achieved, the solutions tend to be proprietary and single-service based rather than providing a set of standard-based compatible ingredients that can be mixed-and-matched in innovative ways.

Putting an IMS client inside a femtocell opens up new opportunities for local media interaction. Because in IMS signalling and media delivery are handled separately, the core network can stay in control without having to handle the media, whose control is delegated to the femtocell. Of course the IMS client is only a small part of what is needed to IMS-enable a femtocell. Whereas a legacy Iuh-based femtocell only needs a part of the UMTS stack to operate because the femtocell gateway runs part of the RNC function connecting into the core network, in the case of IMS a femtocell must terminate the whole call, effectively representing all RAN and core functions in a UMTS network.

So with the flexibility IMS brings it’s easy to see how this would work for say streaming video from a home media centre, but perhaps the single most exciting opportunity is in business, by using femtocells to integrate mobile phones into a company’s IP-PBX.

Benefits of femtocell & IP-PBX integration
Here the mobile phone becomes an extension of the company’s phone system, giving mobile users access to all the advanced IP-PBX features like a single corporate dial plan, single voicemail, abbreviated dialling and user hunting/follow service. Users can seamlessly switch between their mobile phone and their desk phone while making or receiving calls, and can be reached by a single number. What makes this possible is the IMS controlling the call and connecting the media directly between the femtocell and the in-office IP-PBX.

For operators, voice offload to an IP network reduces the burden on the network. It extends the ability to support more subscribers on the existing macro network infrastructure. It relieves the core network by routing calls locally. These cost savings can be passed to customers in the form of more competitive tariffs.

Femtocells as part of an IMS deployment
As I mentioned earlier, most IMS deployments today are being rolled out alongside legacy networks. That is typically reflected in the phone population too, which comprises a mix of IMS-enabled and legacy devices. This can be a frustration for operators who would like to offer IMS-based services to all of their customers. Femtocells can help. Legacy phones connected to a femtocell can make use of its IMS client to access previously unavailable IMS services.

Note: this is an often misunderstood part of IMS-connected UMTS femtocells. The mobile phones used with these femtocells are completely standard off-the-shelf models and do not need an IMS client.

The Worlds first UMTS femtocell deployment was announced by SoftBank Mobile last year using our Ubiquisys femtocell. They are now ahead of the curve, having decided to start from day one with IMS-supported UMTS femtocells rather than the Iuh based approaches. There is additional complexity and a different set of problems to solve when implementing an IMS-based solution compared to an Iuh-based one. However, SoftBank Mobile is now benefiting from the increased flexibility and new revenue generating opportunities that IMS brings to the mix.

I’m just as enthusiastic about IMS femtocells as I am about Iuh femtos connecting to legacy networks. That’s why at Ubiquisys we set out from the start to develop one femtocell that could handle both flavours. Iuh femtos will provide strong market adoption through many operators. IMS femtos will open new convergence application opportunities for homes and businesses.