Your browser is not supported. Please update it.

18 June 2015

Qualcomm MuLTEfire sets LTE free from licensed spectrum

Qualcomm’s detractors like to portray the chip giant as a religious fanatic around 3GPP, but sometimes it behaves more like the Vicar of Bray, who pragmatically changed his religious affiliations in order to hang on to his parish. After years of sniffing at WiFi and WiMAX, the company acquired Airgo and Flarion (both pioneers of IEEE technologies) in 2006, followed by Atheros in 2011 and Wilocity in 2014, becoming a major powerhouse in WiFi. Even if, in its dream world, 3GPP platforms – and indeed, CDMA – would have remained unchallenged, in reality Qualcomm knows when to accept that another technology has turned into an opportunity too big to be missed.

Its latest apparent volte-face comes in the controversial area of LTE for unlicensed spectrum. Having pushed heavily to use LTE in the 5 GHz band only to provide supplemental capacity for a licensed-spectrum anchor network, it has now announced MuLTEfire, which enables LTE to run standalone in 5 GHz – potentially making it a WiFi-like weapon for non-licence owners, such as cable operators, to create competitors to the cellcos’ services.

Qualcomm was one of the most active early promoters of LTE technology for unlicensed bands, which is now being standardised by the 3GPP. The supplemental downlink idea meant that LTE operators would be able to access the additional capacity of the 5 GHz band in a more integrated way than relying on WiFi offload.

This is the focus of the LTE-LAA (Licensed Assisted Access) and LTE-U versions of the technology – the former implements ‘listen before talk’ to avoid interference and will be finalized in Release 13. In countries like the US and Korea, where LBT is not mandated, the existing LTE-U variant could be used, subject to regulatory approval.

However, along with fears of interference with WiFi (still unresolved), this approach led to accusations that the 3GPP community not only wanted to hijack the valuable 5 GHz band in order to weaken WiFi, but was determined to keep LTE only usable for operators with expensive licences and macro networks. There has been plenty of speculation that it might even ‘forget’ about the companion platform, LTE-LWA, which would allow the 4G standard to run alone in 5 GHz.

But Qualcomm, whose interests seemed to lie entirely in protecting its traditional customers and operators, has suddenly lent its weight to an LTE free-for-all. There is some logic to this – a more open LTE platform could broaden the customer base for the networks, and therefore its chips, and tip the balance of power against WiFi by providing a viable alternative for ISPs in 5 GHz (particularly if the 3GPP vendors’ claims of their technology’s efficiency advantages are proven).

In particular, it could deliver to cablecos, ISPs and other non-traditional wireless providers the undoubted advantages of 3GPP technologies – nothing to do with data rates or even spectral efficiency, but actually the way they authenticate a user, and support fully mobile hand-off and roaming, in an invisible way. Qualcomm says MuLTEfire would deliver all those to a wide range of providers, while being as easily deployable as WiFi.

Introduced in a blog post by Qualcomm technical marketing manager Matt Branda (and also in its filing with the FCC as part of the LTE-U comment process), MuLTEfire claims to provide “LTE-like performance benefits to more deployment scenarios with WiFi-like simplicity – a leaner, self-contained network architecture that is suitable for neutral deployments where any deployment can service any device.” Qualcomm also said MuLTEfire would “create expanded opportunities for small cell deployments, especially in hyper-dense environments and indoor locations”, by providing complementary capacity and coverage options for MNOs.

This perhaps indicates how far WiFi has progressed towards being a carrier-grade platform, provoking defensive moves by 3GPP vendors. The Wireless Broadband Alliance and WiFi Alliance, as well as many WiFi vendors, have developed technologies to support cellular-like hand-off and authentication, including the Passpoint and Next Generation Hotspot standards and the WBA’s roaming frameworks. These are only at the beginning of their adoption curve, and have not yet achieved the same user experience as cellular, but they are only a few steps away.

One way to keep WiFi at bay is to offer non-MNO providers LTE as an alternative, even if that means arming the challengers to Qualcomm’s traditional cellco allies. If MuLTEfire comes to market (it is still a concept at this stage) – or indeed, LTE-LWA offerings – it would lure providers with the promise of filling the gaps which still exist in WiFi. Even if WiFi has caught up in terms of authentication, LTE will still score when it comes to high-mobility, high-QoS services, notably voice and video. “The ultimate goal of MuLTEfire is to ensure the best possible user experience for wireless access to the Internet or when making video/voice calls, especially in hyper-dense environments,” Branda wrote.

MuLTEfire will use the signals and channelization of the LTE radio link and incorporate some of Qualcomm’s evolving technologies, such as UltraSON, geared to dense small cell environments, whether LTE or WiFi. However, the chip giant cannot fully overcome its disdain for unlicensed spectrum, noting in its FCC filing that LTE-U and LTE-LAA would still be more robust than MuLTEfire or LWA, because they carry out all the control signaling, mobility, authentication and so on in the licensed spectrum of the anchor network.

Qualcomm may be tempting new providers with LTE’s jewels in order to expand the whole addressable market, but there are huge question marks over whether LTE will be as attractive as it believes, as an option for providers using the 5 GHz band. Much will depend on adoption by the device ecosystem, which usually decides the fate of any technology. Qualcomm has considerable influence over handsets, of course, but increasingly the growth in device sales and usage is seen in non-cellphone products such as tablets, post-PC designs and wearables. To date, the cellular platforms dominated by Qualcomm have been ubiquitous in smartphones, and those have been the main driver of mobile data usage – but as that usage shifts to other form factors, the advantage swings to WiFi. After all, cellular has failed to achieve significant penetration of any other mainstream data gadgets, apart from handsets – cellular PCs and even dongles have remained niche compared to WiFi, and even tablets, which are basically stretched smartphones, have often been WiFi-only.
That, of course, is largely about cost – to the end user and the device maker. WiFi has thrived on a very open ecosystem which has encouraged far lower device chip and infrastructure prices than in the cellular world. If LTE-U were to be competitive with that, it could have a depressing effect on Qualcomm’s already vulnerable margins (and deter other LTE vendors from participating at all, stifling the ecosystem at birth).

However, the company has some experience, via Atheros, of balancing mass market WiFi chip sales with more expensive and rarefied offerings for high end enterprise and service provider kit, and so may hope to protect its LTE profits in the same way by targeting carrier-class deployments.

And of course, like everything Qualcomm does, MuLTEfire would give it an opportunity to harness its inhouse R&D to gain a headstart over rivals in a new market, to help define the way that market shapes up, and to collect IPR revenues. Running LTE in unlicensed spectrum will open an IPR can of worms, adding to pressure to adjust its licensing practices to be more like those of the IEEE platforms, but at least for a while, it would give Qualcomm a new base of potential royalty payments.

Meanwhile, no fewer than 36 companies have filed comments to the FCC as part of its process of evaluating LTE-Unlicensed. Most of the comments have acknowledged some benefits of LTE-U, such as spectral efficiency, high speeds and, for MNOs, a cost effective way to add capacity. But most have also expressed concerns over interference and potential weakening of operators’ existing WiFi investments.

  • AT&T echoed many reactions to the FCC’s original request for comment, which said the regulator was going beyond its remit in addressing specific technologies’ pros and cons. The carrier reiterated that the FCC must make any decision on LTE-U from a technology neutral standpoint, and said any interference issues must be resolved by standards bodies, not regulators.
  • Verizon and T-Mobile USA have been vocal supporters of LTE in 5 GHz and have promised to deploy it – the former’s filing suggested that date could be soon, since in the US, there is no need to wait for the listen before talk features of LAA. If the FCC approves, they could deploy existing LTE-U, for supplemental downlink only (LAA will also support uplink in future), almost at once.
  • Google, unsurprisingly, was more negative about LTE-U, though the search giant might well be expected to leap on a non-anchored implementation of 4G in the 5 GHz band, as another way to disrupt the established carrier business model and provide alternatives to the MNOs’ services. It noted the potential for interference with WiFi and commented: “The Commission should be vigilant in ensuring that deployments of LTE-U and LAA in unlicensed spectrum will not systematically exclude unlicensed-only technologies.”
  • Other WiFi supporters went a lot further, including Ruckus, which said pre-standard implementations, or LTE-U systems without listen before talk, would have “a potentially devastating effect on 802.11-based WiFi networks, especially for services such as voice and video over WiFi.”
  • And Cablevision, one of the most aggressive US cablecos in deploying WiFi as an alternative to cellular, wrote in a lengthy filing that “LTE-U/LAA proponents have distorted the historical, healthy dynamics of the unlicensed bands, using their FCC licences to unjustly and unreasonably exploit unlicensed spectrum, harm consumers and undermine an essential input for their competitors.” It demanded that the FCC “step up” to protect consumers and competition.
  • Ericsson used its filing to deny any prospect of pre-standard implementations coming to the US. It wrote that LTE-U specifications are already complete, since it is included in Release 12, not in R13 like LTE-LAA. R12 has multiple coexistence mechanisms for fair, shared usage with WiFi, it insisted, pointing to “extensive coordination between 3GPP and IEEE 802.11 on appropriate sharing characteristics to ensure coexistence between LTE-U/LAA and 802.11/WiFi.”
  • However, another filing, by Paul Nikolich, chair of the IEEE 802 LAN/MAN standards committee, claimed: “There has been no coordination between IEEE 802 and any standards body associated with LTE-U, because LTE-U was not developed by a standards body.” He wrote that he believed LTE-U to be a proprietary solution which does not use appropriate sharing mechanisms to ensure coexistence with IEEE 802.11 standards, and that there had been no coordination with the 3GPP beyond some presentations.
  • The WiFi Alliance said there was, at this stage, insufficient information about how LTE-U and LAA will coexist with WiFi and urged the FCC to continue monitoring developments to ensure there is sufficient attention to fair sharing. It also claimed LTE-U was a proprietary system developed by a few companies, because it used a technology (CSAT or carrier sensing adaptive transmission) which was not under consideration by 3GPP. “This lack of industry standard implementation of CSAT means that its impact on other users of shared spectrum will be variable and unpredictable,” it added.
  • On a similar theme, Broadcom wrote that LTE-U was unlikely to be a good neighbor to WiFi for various reasons – the coexistence algorithms are proprietary, there are variations in equipment, and coexistence is controlled at time of operation, among other reasons. It added to the claims that LTE-U is not a proper standard, since its LTE-U Forum is restricted in participation, and its sharing algorithms are proprietary. The company wrote: “The standards-setting process should evaluate coexistence with WiFi as it is currently deployed in the market: it should recognize that WiFi devices can operate using 20 to 160 MHz channels, incorporate MIMO technology, and use explicit transmit beamforming (TxBF).”