Five years ago, it became somewhat fashionable to write off Qualcomm as a spent force whose leadership would be broken once the industry moved from CDMA-based technologies, where it dominates in IPR and innovation terms, towards the more level playing field of OFDM. In 2014, few would be so negative. The firm is still in the midst of a huge transition, and will have to make the difficult adjustment to being one among many in mobile patents, rather than the firm that could call all the shots. It has amassed a strong IPR position in LTE, and has consistently been first to market with the latest technical advances, but it faces intense pressures from Chinese rivals, amid the shift of the balance of revenues in handsets eastwards.
Yet Qualcomm is weathering the storm by facing up to the new realities, particularly that the modem, the main source of its unassailable power in 3G, is no longer the main driver of innovation and success. Yes, ubiquitous cellphones and the internet of things will increase volumes for modem and system-on-chip devices, but increasingly the margins and the competitive edge ‘ and the patents, as the Apple/Samsung battles show ‘ are concentrated in other aspects of the increasingly complex mobile device. The display, the graphics processors, the arrays of sensors ‘ all these are important elements of the modern device, and Qualcomm is seeking a position in all of them, and others.
But actual wireless connectivity remains the heart of its business, and much of its R&D, and its acquisitions strategy, will still focus on this. Reluctantly, Qualcomm dropped its old hostility to unlicensed spectrum and embraced WiFi with the purchases of Airgo and Atheros (though the leopard cannot quite change its spots, and Qualcomm is the biggest backer of LTE-U, which harnesses the licence-exempt 5GHz frequencies for cellular standards).
But it knows that WiFi is becoming an integral part of the mobile landscape as it already was in home and hotspot networking. It is not enough for Qualcomm to sell a lot of WiFi chips and integrate them tightly (still its key competitive strength) with 3G, LTE, Bluetooth and the processor. It needs to be the firm which drives the next generation of platforms, establishing the influence of an IPR base (even when, in the changing world of mobile patents, these may be royalty-free, or placed in patent pools or open source foundations). Patents have generated revenues and driven most of Qualcomm’s profits but they are also a source of power and a reflection of the firm’s engineering heart and its relentless pace of innovation.
This is where its latest reported acquisition comes in. The firm is reported to be close to finalizing the purchase of Wilocity ‘ the pioneer in WiGig, which will deliver multi-gigabit speeds in unlicensed 60GHz spectrum with WiFi-like technology. This would enable Qualcomm Atheros to take an early position in a emerging 802.11ad standard as this starts to penetrate mobile and home devices and will strengthen the budding alliance with Cisco. But it would also bring expertise, inventions and IPR in high frequency spectrum technologies, which will be important in the near future for small cells ‘ another area of keen interest to Qualcomm ‘ and in the medium term as an element of ‘5G’.
Qualcomm is on the point of buying Israel-based Wilocity for about $300m, according to media reports in that country. The deal which could give it a headstart in multimode, ultra-high speed devices; even closer ties to Cisco in enterprise small cells; and an early IPR play for next-wave wireless networks, which will be heavily focused on high frequencies.
Wilocity was founded in 2007 by former engineers from Intel’s Centrino WiFi group and was the first company to deliver commercial silicon supporting preliminary drafts of the 802.11ad standard, branded as WiGig. This extension to the 802.11 family of specifications is WiFi-compatible and achieves multi-gigabit speeds (up to 7Gbps in theory) by harnessing the high capacity of the 60GHz licence-exempt band.
Initially a competitive effort to very high speed efforts within the main 802.11 groups, the 60GHz technology gradually came closer to the WLAN mainstream as the 802.11ad specifications matured, and its organization was finally merged with the Wi-Fi Alliance, which kicked off a certification program in September (though still retaining some separation by keeping the distinct WiGig name and logo). Now it is widely seen as complementary to the other new high speed standard, 802.11ac, which runs in conventional 5GHz Wi-Fi spectrum and can achieve speeds of over 300Mbps using MIMO and wide bands (and theoretically get up to 1Gbps).
The short-range technology has initially appeared mainly in peripheral connectivity applications like Wireless USB but its main potential is seen in home video networks, enterprise or even hotspot systems, and multimedia mobile devices. Its limitations in range mean it will often appear in triband combinations alongside 2.4GHz/5GHz 802.11ac radios, or will be targeted at short-distance applications such as smartphone/TV data transfer.
Qualcomm cast off its long-standing enmity towards unlicensed wireless technologies when it acquired Airgo and then Atheros, the two firms which contributed most to the 802.11n generation of fast WiFi. Now it could take a similarly powerful position – in terms of standards influence and IPR ‘ in 802.11ad. But this is not one of the acquisitions where Qualcomm only gains expertise and patents ‘ Wilocity has actual commercial silicon, which could help its putative new owner be first-to-market with mainstream 11ad/WiFi/cellular combinations for devices or small cells.
Wilocity shipped its first commercial silicon in December 2012, and says it has shipped more than a million chipsets to PC manufacturers, with Dell the most high profile customer. In February, the start-up announced its first smartphone product, the Wil6300 chipset , which delivers speeds of up to 4.6Gbps at low power (200mW-300mW).
Wilocity has so far raised about $105m from venture capital funds such as Benchmark, Sequoia and Tallwood, as well as an impressive range of strategic investors, including Qualcomm itself and Marvell. Perhaps most interestingly, in November it announced that Cisco had taken a minority stake of undisclosed size, and that it was collaborating with the networking leader on proof-of-concept trials for small cells that integrated 3G/LTE with 11ad, particularly for enterprise applications such as video communications.
At the time, Cisco’s Bob Friday, CTO of the enterprise networking group, stated: “11ad and LTE will start to merge next year’. Cisco is investing heavily in small cells via its acquisitions of Ubiquisys and Intucell, and its Wi-Fi activities. The mutual interest in Wilocity is another signal of its increasingly strategic cooperation with Qualcomm in this area. It will be using Qualcomm’s small cell silicon and an expanded order from Cisco would be a good reason for the San Diego firm to sharpen its focus on integrated cellular/WiFi cell platforms, especially for the enterprise.
Tal Tamir, Wilocity’s co-founder and CEO, sees triband products enabling hotspots of very localized capacity within broader enterprise networks, perhaps to serve meeting rooms; to improve wireless coverage and speed in public areas such as stadiums; and to offload data from wireline or cellular networks for specialized, localized functions such as medical imaging.
There is also considerable discussion of WiGig as a low cost option for small cell backhaul, because of its high capacity and relatively uncluttered frequencies (for now at least). One of the critical debates in the small cell community will be whether the 60GHz and higher ‘millimeter wave’ bands will be best harnessed for backhaul or for access itself. This issue will be even more central as discussions begin about what might constitute the next generation of wireless standards, informally known as 5G and likely to hit commercial reality around 2020. There is little consensus, and many of the early R&D ‘contributions’ are really vendors posturing for influence. But most agree that the 5G network will be based around ever-smaller cells delivering ever-higher capacity, in a way that will be optimized for IoT devices, not just consumers. All that will mean a closer attention to the high frequency spectrum, for access as well as backhaul or device interconnection.
Therefore, Qualcomm’s interest will go beyond immediate market opportunities and it will look to harness Wilocity’s more generic expertise in high frequency radio technologies. It would be particularly welcome if it could start to drive a new air interface, though many believe this should not be the focus of 5G work. However, neither WiGig nor its 60GHz rivals like WirelessHD were really designed for mobile access, while LTE is not ideal for high frequencies.
Lauri Oksanen, VP of research and technology at Nokia, said in a recent interview that LTE was “not best suited” for millimeter wave and a new technology may be needed to make the best of this spectrum, perhaps via ‘nanosites’, optimized for the ultra-low latencies required by the IoT. Even if LTE or 802.11 could be evolved to suit this scenario, major breakthroughs will be required in antenna technology to turn the vision to workable reality.
Last month, a research conference was held in New York specifically to talk about millimeter wave cellular networks, focusing on access rather than backhaul or in-room transmission. A keynote speaker was Ted Rappaport, serial cellular entrepreneur and New York University professor, who is carrying out R&D to use millimeter wave spectrum as a dynamic, on-demand resource for mobile devices. For instance, the narrowbeam signals could bounce off buildings and other objects to a far greater extent than in current systems, adding ad hoc capacity and supporting what Qualcomm call’s ‘opportunistic’ small cells. Qualcomm has demonstrated such cells, which only activate if a user is in reach, together with millimeter wave backhaul relays, in its San Diego test network. Some start-ups are taking even more radical views of how mobile cells will work in future, including the much publicized ‘P-Cell’ from Artemis, which creates a very tiny cell of capacity on-demand, wherever a user appears.
Another important area of turning high frequencies into suitable mobile access spectrum lies in antenna design. Most cellular frequencies have wavelengths a few inches long, with antennas to match, but millimeter waves are far shorter, and so enable new phased array antenna designs, which can be very precisely aimed and rapidly reoriented. The phased array changes direction based on the charge applied to a given element of the array so that multiple signals can be steered to one device, boosting speed, and the beam can ‘follow’ the moving user. Phased array technology is only viable for mobile communications when coupled with a powerful processor to keep reconfiguring the antennas so that they are still directed at one another ‘ an area where Qualcomm, Intel and others are conducting R&D, and where one firm, Silicon Image, has already created a commercial product. This phased array antenna was designed for use with another 60GHz technology, WirelessHD.
Silicon Image, which acquired 60GHz CMOS pioneer Sibeam in 2011, works with WiGig as well as the less prominent WirelessHD, and ‘ no surprise ‘ is also a close partner of Qualcomm’s in several initiatives, including the 60GHz standards themselves, and the larger firm’s AllJoyn platform for internet of things connectivity.
Assuming Qualcomm closes on Wilocity, it will not only indicate deepening interest in next generation WiFi and in high frequency mobile technology. It will also be a counter to Intel, which has been a massive promoter of WiGig and sees the standard as another chance to gain power in carrier networks. Intel has the opposite agenda from Qualcomm’s ‘ to center broadband wireless networks on unlicensed spectrum, rather than the licensed bands whose platforms are alien to the chip giant. It failed in its bid to push WiMAX as an alternative 4G platform to LTE, and one driven by the PC and IT sector rather than the 3GPP community. But in helping to create the WiFi explosion, it has won greater carrier network influence as 802.11 becomes part of the mobile platform.
And this could be magnified if WiGig, or another unlicensed technology for high frequencies, becomes a cornerstone of ‘5G’, since Intel has been an early contributor and supporter ‘ and these networks have heavier reliance on processing power, the area where Intel excels far more than it does in radio.
Intel can clearly see WiGig moving to other high frequency bands beyond 60GHz, licensed or unlicensed, and itself seizing leadership of high frequency mobile platforms. It has an executive in charge of this very program – Ali Sadri, now senior director of millimeter wave standards in Intel’s mobile and wireless division, and also the leader of the WiGig Alliance. In an interview with EETimes last year, Sadri said that WiGig ‘will play a very serious role as an augmentation of the cellular infrastructure”, as a complement to LTE ‘in the same way that WiGig augments WiFi”.
While access will be important, the more immediate opportunity for WiGig in the cellular system need not wait for 5G ‘ Sadri is leading a project to demonstrate WiGig as a short range, high speed backhaul link for 3G/4G small cells. His group is also investigating whether other high frequencies could be harnessed for mobile access as well as their current uses in backhaul.
But Intel’s R&D indicates that the 28GHz and 39GHz backhaul bands could also support gigabit-plus access links over distances of at least 200 meters. “Our target is to start working on new media access controller and physical layer chips in 2016-2017, and we are also shooting for a project the industry can launch in 2020,’ Sadri said.
However 5G ends up looking by then, it is sure to throw a brighter spotlight on high frequencies and very small cells. It is it not inevitable that the same giants will dominate this generation, but we can be sure the Intel-Qualcomm battle will be part of the picture, and that both will be snapping up the cleverest start-ups in the years ahead.