At the end of every school year, it’s normal for children to return home with a written report on their effort and attainment. We’ve given marks for the femtocell industry overall, with some suggestions for 2010. We've also revisited our own predictions to see if our forecast turned out as expected. Overall, I’d say the femto industry can claim to have made substantial progress this year, even if this hasn't translated into as many commercially installed units as some had predicted.
How did ThinkFemtocell do?
First let's mark our own report card for 2009 using the predictions we forecast a year ago:
|Prediction at end 2008||Accurate||Comment|
|Commercial launches, more in the second half of 2009||Correct||We've identified eight commercial femtocell networks now in operation, with the majority coming onstream later in the year.|
|SIP/IMS femtocell architecture adopted in US and Japan||Half right||Progress is being made for 3G CMDA femtocell standards based on IMS/SIP but this won't be approved until mid 2010. Meanwhile in Japan, NTT DoCoMo seems to have adopted a traditional architecture for their UMTS femtocell and news of SoftBank's femtocell strategy is difficult to decode|
|3GPP standards incorporate SIP/IMS femtocell protocols in Release 9||Wrong||With several alternative femtocell SIP/IMS architectures still under discussion it looks like Release 10 for this. The cellular industry overall needs to firm up its plans for Voice over LTE in the outdoor network first before it can really finalise this. Meanwhile, the femtocell industry is concentrating on its existing Iu-h based standard with many operators and vendors actively adopting it.|
|Prices reduce to $150 by the end of the year, but don’t reach the target $100 till 2010||Correct||I've heard that the $150 price point is now in range from some of the lower cost vendors (I don't have visibility of any prices or commercial quotes myself, so don't use this as a bargaining point with your supplier!). With Taiwanese and other Far Eastern volume manufacturers entering the fray, as well as more lower cost component vendors to choose from, this still looks an attainable goal.|
|Business case is better articulated. Several new applications unveiled. Tie-ups with iPhone, Google Phone and Windows demonstrate applications for “the connected home”.||Half-right||Operators still seem to be selling femtocells on the basis of coverage alone. A few vendors demonstrate applications, and the Femto Forum launches an API initiative but with little takeup. But the Apple iPhone eco-system is still much more accessible and promising for todays application developers.|
|“Over the top” WiFi-style services, such as Fring and TruPhone, will become more popular – driven by greater focus on cost cutting and improved 3G data rates/quality.||Half right||These services still seem to be in the "geek zone", with comparatively few users. Significantly, major networks approved the use of these services, notably Verizon allowing Google Voice and Apple/ATT allowing Skype. In Europe, we saw the operator "3" offer a SkypePhone which included free Skype-to-Skype calls for life (without any subscription, contract or ongoing fees). As mobile broadband data services improve, this is one to watch.|
|LTE femtocells won’t become reality until at least 2011, operators will continue to prepare and roll out initially using traditional macrocellular basestations.||Correct||The first commercial LTE service launched in December 2009 in Oslo and Stockholm. Many other network operators are planning their launches for 2010 onwards. But with limited number of LTE capable devices and many homes being constrained by broadband internet speeds, LTE femtocells would initially only be attractive for deployment by operators themselves.|
Overall: Very good progress and attention to detail.
Didn’t meet expectations given by some of the more enthusiastic commentators, but overachieved on many industry expectations. Special marks for achieving published standards which are actively being developed by most vendors.
There seems general consensus that the technology works. With many large networks now offering the products commercially, there are known workarounds to problems identified. SFR have even gone as far as listing the small number of handsets which are incompatible with their service, with their supplier NEC having tested many different types and adapted their system to cater for various quirks (I didn't say bugs) in handset devices.
The emphasis changed during the year towards addressing larger capacity femtocells targetted for enterprise use. Where typical domestic femtocells were restricted to four concurrent sessions, we saw the battle of super-femto vendors announcing 8 and 16 channel devices. Ubiquisys offered an alternative software solution they called grid femtocells, which load balanced tariff between cells in high traffic areas. Percello demonstrated 16 concurrent sessions on their chipset and data rates in excess of 10Mbit/s (limited only by the availability of suitable handset devices).
Much of the 3G UMTS standards work had been done by the end of last year but it took until April before the 3GPP Release 8 which incorporated these specifications was formally approved. There were a few outstanding aspects, such as security mechanisms that have since been addressed. In the CDMA world, the focus was on a 3G architecture using SIP and IMS which is on track for approval during the first half of 2010. The emphasis was due to move on to the long term IMS/SIP architecture for 3G UMTS femtocells too, with an original target of 3GPP Release 9. This has slipped out to Release 10, with several alternative architectures still under debate. I believe further progress needs to be made elsewhere (such as industry consensus for Voice over LTE) before this can be put to bed. Meanwhile, the industry is starting to demonstrate that the standard has been implemented, with initial interworking between vendors and a "plugfest" planned for March 2010.
The range of femtocell vendors continued to grow during the year, with a mix of RAN vendors and smaller startups claiming the first commercial contract announcements. Far Eastern manufacturers are entering the fray, with the expectation of high volume/low price integrated products appearing next year.
Excellent progress. We believe the price point for a standalone femtocell has dropped from around $250/200 to nearer $150 in volume during 2009 and continuous price pressure will see further reductions towards the $100 target during next year. Previously high cost components such as crystal oscillators, which had started off using OCXOs at $75 in the very early designs have been replaced by much lower cost TCXOs that are closer to $5 in volume. Growing competition for picoChip for baseband chipsets from Percello and soon Qualcomm have seen significant progress from all vendors towards more integrated, higher performance and lower system price designs.
Some progress but everyone is not yet converted. Despite publishing an extensive study identifying a range of potential benefits which could be tailored to each operator's circumstances, there remains public skepticism about the approach. At this stage, the commercial launches focus on solving indoor coverage problems (sometimes as a customer retention tool), with no operator yet achieving substantial data offload. We believe this is still a strong driver and that operators suffering from capacity issues will be seriously considering and/or trialling the effects of data traffic offload during 2010.
Femto Vendors 8/10, Femto Operators 2/10. While the vendors have managed to continue to maintain a lot of the original hype and buzz around the technology, the various operator launches have generally been remarkably quiet affairs with few outside the industry being aware that the product is available. I continue to be surprised at how few in the telecoms industry have heard of the technology, and fewer that are aware these are commercially onsale at their local store. Whether operators are actively choosing to be restrained about marketing femtocells (perhaps they don't want to highlight any deficiencies in their coverage, or are concerned about scaling up quickly etc.), or just haven't allocated the marketing budget I don't know. Hopefully we can expect to see a bit more attention to this during 2010, with video adverts such as the one from SFR.
I'd argue that for voice, femtocells will probably see off UMA (which uses WiFi). This might change if VoLGA is adopted for LTE. The main reason being that end users want to be able to use any handset and not have any special options or actions to take when making/receiving calls. The growing threat is still WiFi, but for data traffic. With many operators now mandating that all smartphones have WiFi, and this working well on devices such as the iPhone, operators will need compelling propositions for customers to continue to use 3G femtocell data rather than WiFi.
Femtocell Report Card Summary
|Technical||A||Overcome technical concerns and now actively deployed commercially by large network operators|
|Standards||A||Formally published 3GPP Release 8, good progress with arranging public interoperability (plugfest) and adoption by the industry.|
|Commercial||A||Eight live network operators worldwide, with many actively working towards this goal for 2010. Growing range of vendors joining the marketplace.|
|Cost||A+||The aggressive $150 price point now in reach, competition actively driving prices down|
|Business Case||B||There are still doubts about the overall business value outside the coverage and enterprise. Industry investment suggests this battle is being won.|
|Marketing||B||Femto Industry gets an A, but the operators deserve a C for keeping it so quiet.|
|Competitive||B||With most smartphones incorporating WiFi, will need to promote the benefits of femtocells to end users and ensure data pricing packages encourage use.|
Predictions for 2010
- Commercial launches continue with 20 operators live by end 2010. At least 5 operators offer more than one type of femtocell.
- SIP/IMS femtocell standards published for 3G CDMA and actively deployed in the US
- Prices reduce towards $100 by the end of the year, with integrated modem/router/WiFi/femtocell products for less than $150
- Operators more actively promote their femtocell offerings. A range of more attractive pricing options.
- "Unlimited" outdoor data tariff plans replaced by a wider range of pricing options, including premium prices and encouraging greater data offload
- Business case is better understood including the competitive threat (to femtocells) from WiFi for data use on smartphones.
- Some consolidation in the industry with at least one industry startup being taken over or dropping out
- LTE femtocells won’t become reality until at least 2011, operators will continue to prepare and roll out initially using traditional macrocellular basestations.