Solving the mobile broadband capacity crunch will require reinforcing the macrocell network with approximately 10 times as many small cell base stations as there are macrocells. There are several competing options for the technology of choice to implement this small cell network, but the top three candidates would be:
- Picocell base stations
- Carrier Wi-Fi access points
- Femtocell Access Points
In their public statements, femtocell industry players make little distinction between residential femtocells and carrier picocells. This likely reflects an effort to extend the femtocell concept to the carrier arena, but given the strength of Wi-Fi, it is likely to turn into a reorientation of the femtocell industry towards carrier solutions.
The news from the Femto Forum has been quite encouraging lately. For instance, in their latest industry report, the point out several interesting statistics:
- An estimated 2.3 million femtocell access points have been deployed, versus 1.6 million HSPA macro base stations, as of Q4 2010.
- 8 of the top 10 mobile operator groups (by revenue) now offer femtocell services, including AT&T, France Telecom/Orange, Telefonica, T-Mobile/Deutsche Telecom and Vodafone among others.
- As of June 2011, there are 31 commercial services and a total of 43 deployment commitments (compared to from 19 deployments and 34 commitments in February 2011).
However, this spurt of growth comes after years of trials and smaller deployments, and still only counts 31 operators among the hundreds of operators worldwide.
In the meantime, carriers are moving towards Wi-Fi as an alternative small cell technology, which specifically addresses the issue of data traffic offload that represents operators' current challenge, not lack of coverage, which was the original motivation for femtocells. Wi-Fi is seen as a win-win proposition for broadband on the go: it will help operators manage cost by offloading traffic, while offering subscribers a cost-effective high-bandwidth service at home as well as in many heavily-used locations.
AT&T is already operating a large carrier Wi-Fi network in the U.S., with 24,000 access points. KDDI recently announced plans to deploy 10,000 Wi-Fi AP's in the short term, with a goal of 90,000 APs deployed throughout Japan by mid-2012. In neighboring Korea, the country's major mobile service providers plan to jointly set up 1,000 new Wi-Fi hotspots (which sounds kind of paltry) around the country, in train stations, bus terminals, and public offices. In terms of an ecosystem however, these public deployments are a drop in the bucket compared to private Wi-Fi hotspots (residential and enterprise). For example, Skyhook (the company that pioneered Wi-Fi based positioning services) has over 250 million access points in its database. That kind of number makes the 2.3 million femto access points deployed pale by comparison, and calls into question whether femtocells can ever challenge Wi-Fi for residential access. Given the installed base, it seems more likely that security, voice services, and mobile handover will be added to Wi-Fi (all of these being efforts that are underway) rather than femto cells being used to offload data services.
The other technology on the horizon that is looking for a piece of the small cell pie is carrier picocells. Although they are currently available for 3G, the market is looking forwards to LTE picocells as a major portion of LTE deployments, in part because the primary bands for LTE deployment is likely to 2.6 GHz, where cell sizes are inevitably small, while capacity requirements continue to skyrocket, which again drives cell size down.
Picocells avoid many of the issues carriers have identified with femtocells:
- They are carrier owned and operated, ensuring quality and predictability of deployment. Picocells will never show up where the operator least expects them.
- They will be coordinated to the macrocell layer to avoid capacity degradation
- They maintain operator control of the traffic
Picocells do address one of the key issues that operators will face going forward: controlling CAPEX as capacity demands continue to climb.
In many ways picocells represent a sweet spot for many players in the mobile broadband ecosystem:
- For carriers, they are cost-effective solution that remains entirely under their control
- For chipset vendors they are small and simple enough, and there is sufficient volume, to be addressable and profitable
- For system vendors there are sufficient revenue and margin that a reasonable business case can be built.
For example, several small cell chipset vendors interviewed by Maravedis for our upcoming special report on the small cell market indicated that the BOM costs for a low power picocell (of the type used for enterprises or small outdoor cells) are likely to be in the range of $200 (small indoor cell) to $500 (small outdoor cell). These types of cells could handle a few hundred subscribers, well beyond what traditional femtocells can manage, and putting them firmly into carrier class territory. Moreover, the projected costs compare favorably to the selling prices of a carrier-class Wi-Fi AP, which vendors of that technology indicated to us would be in the range of $1,500 for an outdoor MIMO AP.
So, the small cell industry faces a fairly simple choice:
- Continue to pursue the "sub-$100" residential femtocell, in the face of a massive residential Wi-Fi installed base
- Or, focus their efforts on the indoor and outdoor picocell opportunity, where prices in the thousands of dollars per AP can be justified, while taking advantage of rapidly falling silicon costs and rapidly increasing performance.
Who is likely to benefit from a reorientation of the femtocell industry towards carrier picocells? Obviously the femtocell providers themselves, who have a large and lucrative market at their doorstep, and also the chipset suppliers that enable their products. Somewhat surprisingly, the Wi-Fi suppliers are also likely to gain from this reorientation. Carrier Wi-Fi and carrier picocells are likely to be complementary technologies - once a carrier acquires a small cell site and brings backhaul capacity to that location; it is very likely that they will deploy both Wi-Fi and LTE technology. In effect, the cost of deployment of both technologies is reduced by sharing infrastructure. In addition, in our research we found that carrier Wi-Fi vendors are also somewhat more advanced integrating HSPA and LTE into their products, compared to the effort made by femtocell makers to integrate Wi-Fi into their products. The two independent Wi-Fi market leaders, Belair Networks and Ruckus Wireless, have already announced dual-mode Wi-Fi / picocell products and are moving towards field trials.
Network OEM's could be challenged by the appearance of very cost-effective picocell gear. They have been busy developing their own picocell base stations, but the equipment announced by vendors such as Nokia Siemens Networks and Huawei are fairly large devices, with transmit power of 2 X 5W (2 X 37dBm), more akin to a WiMAX macro cell than a picocell. These types of base stations are likely to be very useful in 2.6 GHz LTE rollouts, but smaller picocells from independent vendors, co-located with Wi-Fi access points may prove to be a more affordable, more aesthetically acceptable solution.
Factors that could drive the reorientation of the femtocell industry towards carrier picocells include an accelerated takeoff of LTE, with its focus on small cell form factors, as well as continued carrier focus on solving data capacity issues, as opposed to improving coverage for voice services (the so-called 5-bars issue).
Fernando Donoso is the 4Ggear Team Leader at research firm Maravedis Inc. Maravedis is a leading analyst firm focusing on 4G and broadband wireless technologies and markets.