The major US cablecos have been setting the pace in incorporating wireless into their service strategies, and until now Comcast has been the frontrunner. But its actual commercial launch only came this year, and was followed quickly by an alliance with Charter to cooperate on mobile services, bringing the two largest cable operators together in an axis which could be a threat, down the line, to the mobile carriers. Now Charter has provided more detail on its own plans to support wireless and converged services, and eventually a quad play to challenge Verizon and AT&T.
Although the US cablecos dabble periodically with acquiring spectrum (Comcast secured airwaves in the recent 600 MHz auction), they have scarcely deployed licensed cellular networks because of the cost and the huge headstart commanded by the MNOs. Even when the major cablecos partnered with Sprint to form a combined spectrum holding company, which was designed to underpin a joint national 4G network, the venture fell apart amid conflicting agendas and the sheer power of the big two telcos.
That power looked set to be consolidated further as Verizon and AT&T snapped up many smaller players, but in fact, the competitive landscape has shifted again and there is a better chance for new entrants – as long as they are armed with deep pockets and strong customer reach – to chip away at the power of the leaders. Verizon and AT&T have been taking higher risks recently in their M&A strategies, looking to move up the value chain by acquiring content and media activities. They realize that mobile dominance may no longer be just about who can buy and control the most spectrum, whether via auctions or acquisitions.
That has been the mainstay of mobile operator power throughout the life of the cellular industry, but now there are new chances for non-spectrum holders to offer mobile services using shared spectrum and localized RANs and packet cores. This is the chance Comcast and Charter have in their sights (and they may, of course, add a cellular acquisition, of T-Mobile or Sprint, or both, to expand their scale in future). But even without such a deal, they can challenge the MNOs at last, not in terms of national coverage, but by deploying wireless where it can best enhance their model and harness their existing assets of dense wireline networks, strong content partnerships and high levels of consumer awareness.
A network of LTE or 5G small cells, augmented by WiFi, running in unlicensed spectrum (perhaps with MulteFire), and with its own packet core, could be built specifically to support a certain area (such as a city downtown) or a certain service (for a particular vertical market or IoT application). That would enable the cableco to invest in build-out only where that could be targeted at a proven revenue generator, while reducing its reliance on an MVNO deal with a national MNO – though that would still be necessary to allow its users to roam.
All these trends are reflected in Charter’s new filing with the FCC, which sets out its wireless strategy. Like Comcast, it started with WiFi homespots and hotspots, as a low cost way to allow its cable subscribers to access content and the Internet wherever they went. Charter’s major shareholder is Liberty Global, which is pursuing similar WiFi strategies in Europe via various subsidiaries. But on both sides of the Atlantic, cablecos want to add cellular to the mix, to support services that require the predictable performance and improved security of LTE – but they do not want to rely entirely on a conventional MVNO deal, in which they have little control over how the network behaves.
In Belgium, Liberty went as far as acquiring an MNO, and of course, Charter has been tipped as a possible bidder for TMO or Sprint in the US. Another route is to emulate famous disruptors such as Free Mobile in France, whose cableco parent built a network at very low cost. This was done primarily by supporting WiFi-first services for consumers, to reduce the need for cellular capacity; and by harnessing its existing broadband lines to add small cells to its WiFi homespot base, and so create an inside-out dense wireless network. That cost base enabled it to undercut existing mobile services and spark a price war.
Some of those tactics were seen in Comcast’s initial mobile launches under its Xfinity brand. Though less dramatic than Free’s, they undercut other services for its own cable subscribers and some key target demographics, and relied heavily on the cableco’s huge WiFi network to reduce reliance on the cellular MVNO with Verizon. Charter is likely to take a similar approach when it activates its own Verizon MVNO, but the two cable majors are certainly looking for more control than those deals bring them, both individually and via their new alliance.
Comcast CEO Brian Roberts has previously said that a mixture of WiFi hotspots and homespots, the cableco’s MVNO deal with Verizon, and its own cellular roll-out, would allow Comcast to launch mobile services without the “kind of investment” that would normally be involved.
So while the WiFi-first MVNO strategy will be the first stage in Charter’s inside-out approach, when it launches commercially next year, it will soon expand into “our own mobile core infrastructure by deploying licensed LTE small cells,” it said in its filing. That will significantly augment the MVNO with its own networks for targeted user groups, as outlined above.
“Charter is actively testing licensed small cell technologies which will put us on the path to providing next generation super-fast, high capacity broadband to our customers,” the company wrote on its website. “We will combine these 4G LTE and 5G technologies with our existing advanced high speed network to profoundly enhance connectivity forms of our subscribers.”
Charter’s wireless chief Craig Cowden recently held a meeting with FCC chairman Ajit Pai, and other FCC officials, to discuss its plans and the outcomes were detailed in its filing. In particular, it described its progress towards greater network self-determination, and from localized indoor networks to full outdoor services.
It wrote: “The wireless component of Charter’s network is transitioning from a nomadic WiFi network to one that that supports full mobility by incorporating WiFi with multiple 4G and 5G access technologies to deliver a seamless connectivity experience. In navigating this transition, Charter is emphasizing an ‘inside-out’ strategy, focusing first on wireless solutions inside the home and office, and then eventually expanding outdoors.”
One key enabler for such strategies will be the combination of small cells – which can be deployed as ‘sub-nets’ with their own localized controller and virtualized packet core – with cellular technologies that can be run in shared or unlicensed spectrum. Early examples like LTE-Unlicensed and LTE-LAA (Licensed Assisted Access) require an anchor network in licensed spectrum, which does not preclude the subnet model, since most cablecos will still want to support national mobility via an MVNO agreement.
However, this would require the MNO to support an MVNO partner with ambitions to drain away some of the national player’s profits by targeting subnets at high value user groups. The cablecos will have more freedom to pursue their business models with technologies that do not require a licensed-band host, notably the emerging MulteFire system, which is initially being implemented for 5 GHz and the general access (unlicensed) portion of the US’s CBRS band in 3.5 GHz. In future, other such shared spectrum options are likely to emerge in the US and elsewhere, as well as unlicensed 5G technologies (Qualcomm, which spearheaded LTE-U and MulteFire, is heading up a 3GPP group focused on this, although the standards will not appear until Release 16).
Charter has said it sees three main use cases in the shared access portion of the band – neutral host; point-to-point and point-to-multipoint fixed wireless services; and private LTE networks for Industrial IoT and other applications.
The CBRS scheme, which has three levels of access including unlicensed, is currently being in a consultation period. It is very important to the subnet model, but there are fears that the system may be compromised because of opposition from various quarters – T-Mobile is leading the argument that 3.5 GHz should be assigned to 5G not 4G, to bring the US into line with other major markets including China; and there are many disputes over the licensed portion and whether that should be allocated using traditional long licenses or shorter term, lighter terms and conditions to encourage innovation.
Charter, in its FCC meeting, petitioned the regulator to expedite the release of the spectrum for both unlicensed and licensed use, and in the latter area, to keep licences geographically no bigger than a county.
“Large licence sizes would limit access of the band to the country’s largest wireless carriers and would likely result in deployment only to the most densely populated areas,” Charter argued in its filing. “Smaller licence sizes will enable new entrants, like Charter, to tailor their investment and deployment plans by leveraging existing infrastructure … By contrast, increasing the geographic licensing size of PALs to PEAs (which are too large to match up with Charter’s network) would preclude new entrants such as Charter from investing in the 3.5 GHz Band due to their large geographic size.”
This argument goes to the heart of Charter’s (and Comcast’s) multi-staged approach – start with WiFi, then add cellular in unlicensed subnets, then open more options by using licensed spectrum, but preferably without having to invest many millions in national franchises. Charter said in its filing that it could build a licensed spectrum small cell network economically by harnessing its existing hybrid fiber-coaxial (HFC) network to provide backhaul, power and sites (the three major impediments to large-scale densification, especially in the US and especially outdoors).
This is a clear appeal to the FCC’s desire to open up more competition and service innovation in the progress towards 5G, the IoT and a proliferation of wireless-first enterprises. Charter is arguing that, if it gains access to desirable spectrum, it will jumpstart the process of densification and bring much-needed capacity and coverage to US cities and enterprises, because it has the site infrastructure already in place. By contrast, Sprint, AT&T and others have struggled to achieve their density targets because of these logistical issues. The FCC has been trying to push through rules which would streamline the process and reduce city authorities’ power to obstruct or delay small cell deployments, but these are being opposed by many municipalities – while the cablecos are sitting on readymade infrastructure which could unblock the bottleneck, Charter asserts.
“We are going to be leaders in wireless,” Craig Cowden, SVP of wireless technology at the company, recently told the SCTE Cable-Tec Expo. He also talked about Charter’s ambitions to run its own core network, not just on a discrete local basis, but to manage all its wireless services from a virtualized platform. Initially, it will use Verizon’s EPC, IMS infrastructure, SIM cards and roaming relationships but it is already evaluating EPC vendors and testing its own virtual EPC, Cowden said.
Charter is already using virtual EPCs to manage its two mobility trials in Tampa, Florida and Charlotte, North Carolina. “In terms of our full-scale EPC infrastructure, it will be a virtualized infrastructure I’m sure, but the actual timing of that is really more timed to this migration from a reseller MVNO to a full MVNO,” Cowden said.
The Tampa and Charlotte tests focus on mobile services, mainly with outdoor small cells. He said the company will use four vendors (unnamed) per market, and expects to operate a total of 200 sites. The tests will start later this month, and will focus in particular on mobile hand-off.
Clearly, the control of backhaul and sites infrastructure could be critical to the cablecos’ ability to start encroaching on the MNOs’ territory, at least in selected markets. Also at the Cable-Tec Expo, there was discussion of how the HFC network could be used to support densification – whether for a cableco directly or to assist an MNO partner (a quid pro quo – MVNO deal for small cell backhaul and sites – would certainly improve the wireless business case for cablecos which are not ready to build their own network like the big two).
Critics of the cable strategy have argued that most of the cable lines would be sub-optimal for small cell backhaul, let alone fronthauling Cloud-RANs. Modern fiber is clearly the preference, and that has led telcos like Verizon to invest in acquisitions and partnerships that increase their access to dense fiber. But CableLabs, the R&D arm of the US cable industry, believes there is plenty of life in established DOCSIS HFC networks yet.
The organization has teamed up with Cisco to examine how to optimize DOCSIS-based HFC for wireless services. The main aim is to enhance latency in order to use HFC infrastructure as backhaul for 4G and eventually 5G cellular traffic, and to provide a lower cost backhaul network than fiber-only.
In some areas, some cablecos do have dense fiber to support their own FttX deployments and small cells roll-outs. But as John Chapman, CTO of Cisco’s cable access business unit, told the conference, the termination point for that fiber (the cable optical node) is rarely well-equipped for small cells, either because it is in the wrong location, or lacks the necessary power or real estate to host the active cell itself. In that case, he argued that the coax network is better suited to enable small cells, and includes modems and gateways that more easily support the location and real estate requirements of wireless.
The main downside is latency, hence the new project with CableLabs. Chapman said a DOCSIS-connected small cell today would have to deal with the combined latency of transmission to a small cell that is then connected to both a cable modem and a cable modem termination system (CMTS).
Chapman believes that this could be circumvented by ‘pipelining’ a request from the LTE device directly to the CMTS using a new technique called a bandwidth report (BWR). The BWR message would request bandwidth before it was actually needed to reduce the time to grant that bandwidth to the device.
Cisco and CableLabs presented their technical paper on the topic at Cable-Tec Expo and Cisco ran a demonstration on its stand, though both admit that challenges remain, including how to map QoS measurements between DOCSIS and cellular systems. However, they believe those will be overcome in the coming year, which would enable cablecos to embed small cells with their cable modems, lowering the cost of inside-out cellular to WiFi levels (Comcast is pointing the way with its work to embed the LPWAN technology, LoRa, into its home gateways.)
This would be another way for cablecos to support wireless where it matters – indoors and in high value outdoor areas such as downtowns – with a completely different cost base from that of an MNO. They could apply the WiFi economics, which they have already harnessed for WiFi-first, to cellular inside-out strategies, leaving the most challenging and expensive (and potentially unprofitable) parts of a roll-out – the wide area national network – to the MNOs. At that point, the huge investment in nationwide licensed spectrum and real estate may flip from being an unchallengeable asset for the MNO, to being a costly burden.
Comcast and Charter team on mobile services:
In May, Comcast and Charter, the two largest US cablecos, made a pact about wireless activities, which might could make them more likely to join forces to buy their own mobile operator in future – and anyway, will accelerate the progress of cablecos’ mobile and quad play services, towards being real challengers for the established operators.
The two companies agreed to cooperate on common platforms and device procurement to accelerate their entry into the cellular market. This deal will be exclusive for one year.
Both have an MVNO agreement with Verizon (Charter’s courtesy of its acquisition of Time Warner Cable and BrightHouse), and Comcast has announced plans for mobile services – Xfinity Mobile – which could undercut the MNOs in many target subscriber bases, harnessing its residential lines and extensive WiFi to reduce its overall data delivery costs.
The cooperation is designed to “accelerate and enhance each company’s ability to participate in the national wireless marketplace”. Building on existing technology and roaming deals in the CableWiFi initiative, the two operators said they will now focus on creating common operating platforms for wireless; technical standards development and harmonization; device logistics; and emerging wireless technology platforms.
“By working with the team at Comcast, we can not only speed Charter’s entry into the marketplace, it will also enable us to provide more competition and drive costs down for consumers at a similar national scale as current wireless operators,” said Charter CEO Tom Rutledge, in a statement at the announcement.
The breadth of the cooperation will create something more influential than just a deployment partnership, with the potential to create a new wireless R&D powerhouse. The companies have both been involved in testing technologies in 5G high frequency spectrum and they could conceivably pool their efforts, and even their joint holdings of 214,000 fiber route miles, which “could provide a critical backbone that would enable the densification anticipated for 5G services”, as Walter Piecyk of analyst firm BTIG put it to the Wall Street Journal.
However the alliance pans out, there will be knock-on effects on the established MNOs in terms of greater competition, and a reduced ability to buy up a cableco in the near future.
Analysts at New Street Research believe the closer cooperation between the two large cablecos will hasten a Sprint/TMO tie-up, writing in a client note: “The emergence of cable as a serious new entrant significantly improves the odds of a Sprint / T-Mobile merger being approved. This announcement on its own may not be enough to convince regulators that we are in a five-carrier market, but it helps.”