Major operators continue to be committed
Following on from ATT and TIM the previous day, where their extensive testing programs are bearing fruit, T-Mobile and SFR shared their experiences. Many of the technical and business issues presented will be familiar to readers of ThinkFemtocell. T-Mobile re-iterated the enormous growth in data traffic, ranging from +20% for Nokia N95 users, to +100% for iPhone and +200% for iPhone 3G. Primary operator motivations for the femtocell includes providing coverage, reducing churn and increasing data speeds.
T-Mobile said that femtocells must be standardised - although this need not delay initial commercial launch, since models can be software updated to comply next year. This is required to stimulate competition between vendors and drive down costs. They see no real technical "show stoppers" although some technical challenges remain such as handset battery life, lack of hand-in when entering the femtozone and a short service outage of 5-10 seconds which can affect non-allowed handsets entering femtozones. Many of these issues have been addressed in Release 8.
T-mobile believe the long term solution has to be an integrated gateway and envisage a phased implementation:
1) Indoor mobile voice and data coverage
2) Integrated fixed/mobile offer
3) Integrated home services
They will launch during 2009 but anticipate limited volume during the first year - it won't be in the hundreds of thousands of units. Femtocells will also be used to replace picocells and repeaters. With LTE coming up in the next 1-2 years, femtocells need to be available some 12-18 months later.
In summary, T-Mobile very much believe in the femtocell concept. Some outstanding technical issues, high prices and reluctance of customers may prevent a rapid mass market adoption. They plan a phased introduction leading to integrated home gateway solutions.
SFR France (also a fixed and mobile operator and part owned by Vodafone) with some 20 million mobile customers (and 4M fixed broadband) again re-iterated the enormous growth of data traffic on their network, which grew 500% last year alone. Their iPhone users consume an average of more than 100MBytes of data per month. They raised the issue of safety concerns about radiation, which have caused concerns in France generally (libraries in Paris turned off their WiFi for a few weeks due to complaints).
Cellcom Israel surprised me even further (I need to recheck these numbers) by saying that the ratio of data vs voice traffic on their network was 13x more. Around 80% of this traffic comes from indoor locations, probably helped by a tariff plan of unlimited data (fair use policy applies) of 24 Euros a month. The operator is rapidly expanding capacity on their 3G network with new RF sofwtare licences, baseband processor cards, backhaul transmission and RNC capacity being added almost daily.
Analysts predict a more cautious uptake
Several analysts gave their views. Whilst again confirming the enormous growth in data (ABI suggesting up to 800MBytes per month per user might become the norm in the long term), iDate presented a more cautious view. They predict stronger takeup in North America (51%), Asia (26%) than Europe (23%) with 9 million units shipped by 2013 - arguing that the difference is closely linked to the market structure of each region as well as the different ways it would be used in each region. In the US, dropped calls are frequent and femtocells will address coverage issues especially indoors. In Asia and Western Europe, integrated triple play set top boxes and converged services are seen as the route to market. Even this more cautious forecast still predicts a booming long term market for femtocells once the perceived added value is unlocked by end users.
Designing and developing a femtocell is difficult
Will Franks from Ubiquisys reminded me that it's taken years for the main industry players to design their femtocells. The chipsets which picoChip and Percello have created (Ubiquisys demonstrated live femtocells using chips from both chipset suppliers) themselves take many man-years of effort in design and extensive testing. The basic femtocell software stacks can be bought, but the bulk of development work is the specialist technology such as required for self-organising and optimisating/adapting to the environment.
This led me to believe that in general, the leading mainstream femtocell vendors (Airvana, ip.access, Ubiquisys, Huawei) will retain high levels of market share, rather than the technology being redeveloped by many independent home gateway vendors.
Integrated Home Gateways are available
Analysts and field trial results showed that customers would really prefer a complete home gateway which includes a femtocell, rather than a separate device. Initial sales and commercial launches will use standalone devices, but long term we can see this is the way the industry will go. Some operators are asking for what seemed to me like unrealistically low price points (Virgin Media from the UK want it to be only $50 extra), vendors would argue that the business case is more complex yet still valid at higher pricing levels.
Airvana said they see strongest activity from CDMA operators in both North America and Asia. These markets now appears to be a two-horse race between Samsung (who provide all the CDMA femtocells used by both Sprint and Verizon) and Airvana (who would claim technical superiority supporting both 2G and 3G CDMA in the same box). They've provided the femtocell module for several home gateway vendors in addition to their standalone offering.
Volumes are increasing
Nigel Toon of picoChip was prepared to say they had shipped over 100,000 femtocell chipsets as operators pre-stock units prior to launch. With the company providing chipsets to the majority of the market today, it's a good indication that critical mass has been reached and rollouts will start to ramp up in the coming months.
Rosum continue to promote their innovative clock and location solution
Femtocell costs will continue to be under pressure, and innovative ways to reduce it will be a constant ongoing issue. Rosum presented their innovative solution which not only saves money but makes femtocells technically feasible in more environments. By combining several potential sources of timing data, including TV signals (which are surprisingly strong even indoors) and assisted GPS, they can derive a very accurate clock timing without needing to have a very accurate (and expensive) crystal. They can also determine location by calculating the range from the different TV signals picked up, essential both for emergency call location and to determine what spectrum the femtocell should use (or if it has been taken abroad).
With this solution destined to be developed on an integrated chip, Rosum suggest a price point of around $11.50 including separate GPS receiver and crystal, compared with around $20 today for a traditional clock design (ignoring the 20m cable to connect to an external GPS antenna). It provides accurate timing/clock (suitable for all types of femtocell technology) to 20ppb/1uS, accurate location to <50m at lower cost. Moreover, by removing the phase noise found in traditional designs it also improves signal quality and allows the femtocell to be located in the centre of a building rather than needing to be near a window.
Epitiro show off their real-time broadband network monitoring system
With femtocells connected by uncontrolled 3rd party broadband lines, mobile operators are going to need to have the ability to view the status and quality of those connections so they can determine where and when problems occur. Epitiro provide a wide range accurate statistics measured from live broadband networks by monitoring tens of thousands of end-users and aggregating the results. They demonstrated this with live read-outs of the status of several networks, measured independently (i.e. at the customer premises).
Femto Awards Ceremony
This is the first year that the femto forum have run an awards scheme.
Congraulations are due to the winners:
1) Femtocell or Femtocell Network Element design and technology innovation: ip.access (for nano3G)
2) Femtocell service: ip.access (for Facebook virtual fridge notes)
3) Progress in commercial deployment: NEC (Gaining momentum with commercial deployments)
4) Significant progress or commercial launch by a large carrier: Sprint (First national commercial deployment)
5) Significant progress or commercial launch by a small carrier: Cellcom USA
6) Contribution to femtocell standards: Taka Yoshizawa of Thomson (TR-196 management protocol standard)
7) Enabling technology: picoChip (optimised system-on-chip femtocell solution)
8) Social vision on use of femtocells: Softbank (Niimi project - femto brings truly affluent life)
9) Award for individual contribution to Femto Forum: Chris Fenton (Architectural consensus)
There was also a strong vote of thanks to Simon Saunders for his work as chairman of the Femto Forum.
Surprisingly, there wasn't an award for Best Femtocell Website - surely just an oversight - perhaps because there would be only one obvious choice ;)