What is the scope of NEC's femtocell activities?
NEC has created a market leading femtocell solution incorporating partner products and NEC's own products. NEC offers a complete end-to-end Femto solution including integration of the solution to the IT and provisioning systems in the operator's network. We have conducted extensive interoperability testing with many handsets and with several core networks.
To date, we have two major contracts in Europe. One is for residential market and the other for enterprise customers. Outside Europe, we are actively engaged in Middle and in the Far East with several trials underway and have a major customer in Japan.
Our Japanese customer uses SIP/IMS based architecture for Femtocell and we expect a commercial launch in late 2009. All our other customers are using the standard UMTS architecture today - our solution is software upgradeable to the standard based Iuh solution next year.
We've found that most of our customers are starting with a coverage play to address the problem at hand. We believe they will quickly move on to capacity and application based business case when other barriers, such as the Access Point price is addressed.
NEC has recently won the Femto Forum award for "progress in commercial deployment" based on its customers engagements around the world.
What issues have you come across during your customer trials?
We don't see any major technical barriers - nothing that should prevent operators launching service today. Over the last year we've worked with operators and dealt with issues including:
- Interference - would femtocells adversely affect the existing macro-cellular network?
- Plug-and-Play - can femtocells really be foolproof and instantly installed by customers out-of-the-box?
- Handset interoperability - are there problems with specific handsets where they won't work properly?
The results of extensive tests have convinced operators that the interference issues with femtocells won't be a problem and any interference is seen as manageable.
Plug-and-Play Access points
We've proven that femtocells really can work in this way. End users can self-install and the system automatically configures the device ready for use. It typically takes 10-15 minutes to initialize when first installed without any operaor intervention. This has been a common approach for network operators with CPE equipment (such as wireline broadband providers/ISPs), but has taken more effort to convince mobile network operations staff who are new to this approach.
We've conducted tests with something like 150 different models and software versions of handsets to ensure our femtocell solution is compatible. Although all these devices should conform to the 3GPP standards, there are sometimes quirks, bugs or unusual behavior involved. Since we can't change the handset software, we've had to accommodate these issues through software workarounds in our solution.
In the future, we expect operators to mandate to their handset vendors a requirement to test new models against femtocells.
How do you see the femtocell market growing in the coming year?
Globally, I'd expect the industry will ship 200 to 300K units by the end of 2009. Several operators will have launched, but many more will already have run trials and made their supplier decisions by that point. NEC is visible in femtocell trials in Europe, but we are also active in LATAM, Middle East and the Far East. We have a competitive roadmap which responds to market requirements.
Which partners do you work with?
Our main equipment partners are Ubiquisys (who provide the femtocell access points), Kineto Wireless (who provide the main part of our Femtocell Gateway) and Motive (who provide the access point management system based on TR.069). We supplement these with NEC's own products including our Security Gateway which forms part of the femtocell gateway.
As the industry moves towards implementing Iuh, we will add other femtocell access point vendors into our solution.
When will you adopt the Iuh standard?
We plan to introduce Iuh based solution in the first half 2010. We are heavily involved in the Femto Forum's "plugfest" planned for March 2010. We have a transition strategy to Iuh compliance based around software upgrade.
Part of the "value add" from NEC is our plug and play capability through our ecosystem. This avoids problems with standards, because we have pre-integrated and pre-tested all our partner components. By working with several femtocell vendors, this will ensure diversity in the supply chain including the underlying chipset/components to reduce risk.
Why do you think operators will launch femtocells soon?
There is no longer any reason to delay. There are coverage problems today, and poor coverage is one of the main reasons for churn. This coverage issue gives the femtocell market a reasonable size today. Today we'd expect operators to focus on their top revenue customers.
Different operators may choose to package the femtocell in different ways, but the technology is available and proven today.
The next step is to address the capacity crunch. There are no simple answers, but the industry thinking is currently looking at bundling and femtocell applications.
What do you think of Metro-Femto?
[Ed Note: Metro-Femto is where operators install femtocells themselves outdoors and in public areas to achieve high capacity and lower cost. It's particularly promoted as a solution for 4G rollout].
We do see demand for higher power and higher capacity femtocells (16 to 24dBm with 16 to 32 channels). These would be used for enterprises addressing larger coverage areas whilst carrying more traffic.
Any thoughts on IMS/SIP based femtocells?
We've seen major operators such as ATT and Vodafone provide a strong lead for IMS in the Femto forum. They see really big benefits in the longer term.
In the short/medium term, initial femtocells will be based on today's Iu-h standard with a transition from pre-Iu-h, through standard Iu-h and then IMS.
Of course, this will need IMS femtocell standards to be completed and finalized. The Femto Forum is currently debating several proposed IMS based architectures. This makes it unlikely that a complete IMS femtocell standard will be finalized in the 3GPP Release 9.