The budding femtocell market has so far been dominated by specialist start-ups, but as the trials and even contracts from major operators pile up, the baton will pass to the major vendors that can deliver the volumes and pricing the cellcos will demand. The problem for the traditional mobile equipment suppliers is that they are unlikely to be those majors – the femtocell logically fits far better with the business models of consumer and internet product suppliers, like those that have dominated Wi-Fi access points and DSL modems, than with the proprietary, high margin norms of Ericsson and Alcatel-Lucent. So while most of the mobile infrastructure makers have partnered with a femtocell pioneer to be able to provide these devices, the way of the future is pointed far more clearly by a reported AT&T plan to spend $500m on femtos from the UK’s ip.access, significantly backed by Cisco; and by Nokia Siemens’ plans to expand its supply chain beyond specialists to the masters of the commodity network equipment markets, the Taiwanese ODMs.
According to a research note from San Francisco-based analysts at ThinkPanmure, ip.access beat Motorola, Alcatel-Lucent/Sagem, Nokia Siemens, Airvana and 2Wire to the AT&T contract for 7m units, which has not been officially announced but follows a lengthy evaluation by the US cellco that has proved most enthusiastic about femtocells. Sprint Nextel has actually got to market first, with its launch of the Samsung Airave in limited markets, and recently Verizon Wireless expressed its interest, particularly when it starts to move to LTE from 2009, while T-Mobile USA remains primarily focused on its Wi-Fi convergence strategy, Hotspot@Home. According to the research note, the five-year AT&T deal calls for ip.access eventually to price its Oyster3G units at less than $100 each, the magic number initially thrown onto the public table by Vodafone CEO Arun Sarin last year. Such a tag will be essential to make a femtocell convergence/coverage strategy price competitive, especially for US operators, with the Wi-Fi/dual-mode handset approach favored by T-Mobile.
As in Wi-Fi access points, there is likely to be a shake-out among the start-ups as the femtocell market matures and heads for volume and for the $100 price tags the operators are asking for. As in Wi-Fi, and other networking markets, we might just see one pioneer surviving as an independent, and will almost certainly see one bought by Cisco (in the light of the AT&T news, highly likely to be ip.access) and one or more by other majors. However, the big cellco suppliers remain ambivalent about femtocells ‘ a $100 miniature base station, even sold in quantities of multiple millions, is highly threatening to the business assumptions of Ericsson, which remains understandably cagey, even head-in-the-sand, about the femto craze. Most of the mobile infrastructure vendors are working with partners in order not to be excluded from operator deals, but may see little reason to acquire an operation of their own, since it could be difficult to integrate into their existing structures and would have a negative impact on overall corporate margins. For instance, Alcatel-Lucent is selling devices from Sagem, which for all its recent financial pressures certainly understands how to run a supply chain geared to mass market, low priced products rather than complex, high margin base stations. Nokia Siemens looks set to take this logic a step further, working with even more efficient deliverers of the price sensitive consumer gadget, the Taiwanese ODMs. According to reports from the island state, NSN is in talks major ODMs Gemtek Technology and Accton Technology for a femtocell venture to reduce the costs of delivering the tiny base stations. NSN has already spread its femtocell favors among several partners ‘ Ubiquisys, directly and via its Netgear deal; Airvana; Thomson, which is also working with Airvana on gateways; and RadioFrame (the latter mainly for picocells). Now NSN is said to be looking to the Taiwanese companies to help it drive down femtocell prices to the operator from about $250 to under $100 in readiness for major contracts that are expected from later this year.
Meanwhile, Motorola is working with Airvana in CDMA and Ubiquisys in its home networks business; NEC is also partnering with UK-based Ubiquisys; ZTE, Huawei and Samsung are currently pursuing their own routes. For Ubiquisys, whose femtocells will be used in one of the first European commercial roll-outs this year, by O2 in the UK, its most significant partnership, given the likely realignment of the market, may yet prove to be Netgear, which has incorporated the company’s Zonegate 3G design in a Femtocell Voice Gateway – a residential gateway with integrated ADSL2+ modem, router, 10/100 wired Lan switch, wireless access point, VoIP, firewall and 3G femtocell. These integrated devices, controlling the entire networking requirements of the multimedia, broadband home, are what will appeal to the consumer and therefore allow the cellco to infiltrate the home via the femto, gaining enhanced account control and the chance to provide the hub technology, and the primary point of contact, for the residential user base. This type of control has been threatened by Wi-Fi and other approaches controlled by the user, but cellcos have recently been showing renewed determination to fight back, with indoor base stations and cellco-driven home media network standards, such as the Orange-backed Soft at Home venture, which also includes Thomson and Sagem (see Wireless Watch February 26 2008).
If this is what the operators want, they can hardly avoid beating a path to the door of Cisco/Linksys, a company that has understood and exploited the shift to IP networks and cultivated a highly effective set of distribution channels and supply chains for the enterprise and home markets, which it now aims to apply to carriers. Linksys is the undisputed leader in home wireless equipment, and has all the structures in place to get femtocells into the market in volume, and integrated with other home network technologies. Cisco has never been very successful in adopting the channels and strategies that work with large carriers, and its recent acquisition of WiMAX base station maker Navini was widely greeted with scepticism that the company could change its spots now. But in an industry that is about to change dramatically, Cisco should find that the cellco base starts to move towards its sales model, rather than having to leap the chasm itself to reach the mobile carriers.
This is Cisco’s market to lose, and it will be operating alongside some familiar competitors, from the Taiwanese to Netgear, and some new ones, depending which cellco suppliers manage to adapt their own models to embrace femtocells effectively. In the first phase, it seems that Cisco will take ip.access along with it. The giant already has a stake in the UK start-up, and will now fulfil the $500m deal AT&T has made for ip.access’ 3G femtocells. Acquisition would be a logical next step, given Cisco’s normal pattern of behavior in new markets (and, as the larger company points out, would actually follow the purchase of two previous femtocell-style pioneers, admittedly before the term had been coined. At the turn of the decade, Cisco acquired Jetcell, a company making ‘GSM mini-picocells’, as it called them; and a similar start-up in the CDMA space called Exio. The Jetcell product was briefly marketed as part of the Cisco MNet mobile office network range). “We believe Cisco is looking to increase its profile in wireless, and that ip.access is likely next in this strategy as Cisco’s femtocell play,” wrote ThinkPanmure analysts Anton Wahlman and Eric Kainer. The company would fit logically with Cisco’s growing range of acquisitions targeted at control of the home multimedia network, including Linksys and Scientific Atlanta, as well as helping it penetrate carrier markets by the back door.
More strategic partnerships could follow a significant Cisco push into the carrier market via femtocells, especially as operators start to look to LTE and WiMAX. Many cellcos in developed mobile economies are interested in building out 4G-class networks and services initially in limited area ‘hotzones’ covering high value, well targeted areas of user demand ‘ keeping 3G and its evolutions for the foreseeable future as the main engine of wide area mobile coverage. So while femtocells help with the latter by enhancing coverage, indoor penetration and capacity at relatively low cost, they can have a far more fundamental role in the hotzone approach, becoming a major component of early stage deployment of new networks (hence the wariness of Ericsson, which has been making assiduous efforts to ensure its own iron grip on first generation LTE systems). As femtocells move from tactical to strategic ‘ to reduce cost and risk for 4G and to give cellcos the guiding hand in the home network ‘ Cisco may find, if it has achieved the success if should do in this sector, that it is being courted by would-be partners from the cellular equipment world. Cisco has often been expected to pursue an alliance or even merger with one of these suppliers to further its own cause among cellcos (Nortel being the most frequently tipped), but in the 4G era, where key elements of the access network are coming from the consumer electronics and off-the-shelf internet model, the position may be reversed, with a top tier mobile vendor discovering it needs Cisco’s help after all.
Femtocell chips:
Deals like AT&T’s will help convince the market that femtocells are a real opportunity rather than a flash in the pan, and that should lure some of the major chipmakers into the fray. So far, they have generally steered clear of real commitment to femtocells, and the early work has been done by start-ups ‘ most notably the UK’s picoChip, which is currently the only choice for companies that choose not to design their own silicon, though others like NextWave and, in WiMAX, Sequans, have promised femtocell products. picoChip has focused mainly on W-CDMA and WiMAX, and recently announced an LTE reference design based on the same physical platform as the WiMAX product (see Wireless Watch February 5 2008). Its chips are used by Ubiquisys, ip.access, Sagem and their partners, and the next challenge will be to evolve the designs to meet the needs of commodity vendors such as the ODMs.
As the market gains momentum, all eyes will be on the chip players in the cellular and Wi-Fi sectors, and as in the base stations themselves, it may be the companies that are accustomed to supporting mass market devices that find themselves best positioned to move, rather than those that have mainly worked with complex macro infrastructure.
Texas Instruments has done most of its femtocell experiments to date within its infrastructure division but has said several times that the economics may well work better within its handset unit. Qualcomm has expressed interest in femtocells but generally chosen to take the side of the macro base station vendors when pushed, and its most likely strategy would seem to be to enter at the LTE stage. Qualcomm will be relying on its ability to get silicon working very early in new technologies to give it the headstart it will badly need in LTE, to compensate for the eventual loss of its CDMA business, and is unlikely to ignore the opportunity to put its chips into femtocells, since these are likely to be a major component of early phase LTE deployments.
STMicro pulled back from its own femtocell developments last year, perhaps surprisingly given its steadily rising profile in 3G and likely early move to LTE. Among the PC/internet/Wi-Fi companies fighting to gain greater share in the cellular market, Intel remains quiet, and Broadcom is probably the one to watch (and perhaps Atheros). Broadcom urgently wants to increase its tiny market share in the mobile sector, and although these efforts have focused on handsets – with some success following its anointing by Nokia as a preferred EDGE supplier – there would be a clear logic to turning its strong showing in home gateways, Wi-Fi and DSL to the femtocell space.