As technologies in unlicensed spectrum, notably Wi-Fi, become an integral part of a cellco’s network, the logic of separating the radios which work in licensed or unlicensed bands starts to break down. In a future world where spectrum is more dynamically allocated and where different service providers can share bands, wireless technologies will start to spread across all kinds of frequencies, according to application rather than operator ownership. So Globalstar and Amazon are trialling a flavor of Wi-Fi in licensed airwaves in the US, while Qualcomm has made the highest profile bid so far to run LTE in the unlicensed bands.
Qualcomm has historically fought hard to preserve the cellular status quo, where operators had full control of their spectrum and ran technologies (preferably CDMA-based) which were optimized for their services and airwaves. It was therefore hostile to Wi-Fi and WiMAX as carrier tools, although it soon recognized that keeping Wi-Fi ‘ with all its baggage of disruptive models and free spectrum ‘ out of the mobile broadband picture was a losing battle. The mass consumer uptake of Wi-Fi, its place as the default home broadband technology, and then the cellcos’ urgent need for new spectrum capacity of almost any kind, all drove the king of CDMA to acquire first Airgo, in late 2006, and then its larger rival Atheros in early 2011.
The latter was Qualcomm’s most important and strategic acquisition of recent times, setting it up to enter home, vehicle and other markets and to broaden its reach beyond CDMA and LTE, as well as bolstering its IPR by buying both the early creators of advanced MIMO-based Wi-Fi. However, the attachment to licensed cellular technologies, in which Qualcomm has such sales and IPR dominance, dies hard. This was seen at the firm’s recent analyst day, when CEO Paul Jacobs talked about the interest in extending LTE-Advanced to unlicensed bands, in order to create a unified platform for HetNet, and a superior user experience to Wi-Fi.
He was quick to insist that there was no agenda to ‘step on Wi-Fi’, which is now a major business for Qualcomm, but he said the two technologies could coexist more intimately than they currently do, to minimize interference and network complexity and improve data experience. But while he emphasized that the two technologies could be ‘better neighbors’, it was clear where the heart of this CDMA engineer-turned-CEO really lies. He told Electronics Weekly: “We paid for that very sophisticated LTE receiver in the devices. It’s a better receiver than the receiver that’s been used for Wi-Fi. That means you get better capacity and better coverage out of an LTE unlicensed deployment.”
He added that, as 4G cells get smaller, they are becoming more like Wi-Fi access points in design, cost, deployment procedures and behavior. He commented: “You see a lot of companies that are looking at you getting access to spectrum and they’ll be able to use both licensed and unlicensed bands with the small cell technology.” Indeed, Qualcomm has already proposed this disruptive approach in the expanding area of ‘homespots’ (home Wi-Fi access points in which some capacity is left unsecured for community use, which are becoming an important way for broadband operators to expand Wi-Fi availability at low cost). In May, the chip giant said it could adapt the idea of the homespot to the cellular world, showing off a reference design for a femtocell which was small enough to integrate into consumer electronics devices such as set-top boxes and consoles, which would be used for sharing bandwidth with nearby mobile users.
CTO Matt Grob presented a model called ‘neighborhood small cells’, which effectively opens up the residential femtocell for shared access, and incorporates it into other home products rather than requiring a separate box. He said: “A little bit of that signal leaks outside and starts to provide little pieces of coverage outdoors. And if enough of it is happening indoors, then pretty soon you end up with pretty good outdoor coverage.’ His vision is to use Qualcomm’s UltraSON self optimizing network technology to combine signals from many cells in a neighbourhood, pooling them into a single network and coordinating the signals to mitigate interference.
This could be another piece in the jigsaw of technologies and frequencies with which carriers are assembling their mobile capacity picture, but Qualcomm’s bid to expand its favored technologies into licensed bands will also focus on emerging new bands, such as 3.5GHz, whose short range and high capacity suit them to small cells ‘ and which some regulators will open up on an unlicensed or lightly licensed basis.
At the recent analyst day, Rasmus Hellberg, Qualcomm’s senior director of technical marketing, also pushed the ‘unlicensed LTE’ idea, focusing this time on the existing 5GHz Wi-Fi band. He claimed many features of LTE in 5GHz that Wi-Fi cannot deliver ‘ and if it can convince enough of the ecosystem (and the regulators), it would be seizing the upper hand from competitors such as Broadcom, which are strong in Wi-Fi and less so in cellular, though it could also be accused of attacking its own Atheros business.
Like Jacobs, Hellberg was keen to avoid the implication that Qualcomm was trying to weaken Wi-Fi and promote the technology where it has better market and technology leadership. Instead he described peaceful coexistence of LTE with its inferior cousin, with the former better managed, better coordinated and faster, supporting lower power consumption, and with its beams able to penetrate furthe, than Wi-Fi’s. He argued that, in a network combining LTE in licensed and unlicensed spectrum, fewer nodes would be needed to deliver the same capacity than in a licensed LTE/Wi-Fi offload scenario ‘ or the same nodes would deliver more capacity.
There are many obstacles in the way of Qualcomm’s LTE nirvana, not least regulatory approvals. It would also require cellular small cells to come down in price by around 40% to match the falling price curve of Wi-Fi chips. And it would need to add some of Wi-Fi’s recent advances, such as new MIMO capabilities and dynamic channel selection, to unlicensed LTE.
But Qualcomm has been thinking for years about how LTE could be extended into new spectrum bands, some previously thought unusable or untouchable, in order to meet carriers’ demand for data capacity and grow demand for its preferred technology. It pioneered SDL (Supplemental Downlink) and carrier aggregation support when it sold its unpaired 700MHz spectrum (previously used for MediaFLO TV) to AT&T, helping the carrier harness that unpromising slice of airwaves to boost its LTE network. Back in 2008, it bought the L-Band spectrum in the UK, also usable for SDL and to extend LTE into new territories.
Going a step further, into unlicensed spectrum, would be a logical next move in this long term thinking ‘ even if it brings the wrath of the mighty Wi-Fi community, which has seen off many technologies before and infiltrated the licensed mobile world very effectively, down upon it. That has never deterred Qualcomm, of course, or it would never have pitched CDMA against the might of GSM-focused vendors, operators and regulators all those years ago. And its timing is right to start a discussion of whether, in the HetNet, it will be better to use a single technology where possible to simplify roll-out and devices. This must not turn into an attempt to set LTE against Wi-Fi, since both have a role and an established ecosystem, while both have disadvantages too. It can be argued that Wi-Fi is a WLAN system which has been shoehorned into the mobile sector for which LTE was optimized, but that does not prevent carriers being drawn to its massive ecosystem and its low costs.
It has always been obvious that at some point there must be a war between the companies with a vested interest and lead in cellular technologies, against the rising tide of vendors who have grown big on Wi-Fi. The 160MHz width of unlicensed channels compared with auctioned 20MHz in cellular is enough to make any spectrum holder cry ‘not-fair’. But at the same time the world’s regulators see how the cozy world of cellular makes spectral assets squeeze vast quantities of cash out of beleaguered consumers who they have vowed to protect ‘ and these same regulators created Wi-Fi and see that is it a force for good (cheapness) in services. They may well take strides to prevent the arrival of paid spectrum working alongside unlicensed spectrum in the same system.