There are leaks this week, the earliest we can find picked up by Light Reading, which say Time Warner Cable is playing with a Microsoft Mediaroom pilot, for full IPTV over cable.
It is not un-anticipated. Microsoft once in 2003 sold 5 million licenses to Comcast for its Microsoft TV Foundation Edition – software and a new interface on cable set tops. Initially Comcast used the Microsoft Programming Guide in the Seattle area on a Motorola set top and Time Warner Cable went on to use it in Beaumont. Both trials were later discontinued.
But Microsoft has to come back for another go at cable at some point. It is just a matter of when AT&T and Deutsche Telekom and others ‘ its Telco customers – were fully committed to their IPTV installations and when the benefits of IP were obvious enough for Cablecos to mean that the same technical approach that Mediaroom uses might also benefit cable.
Cable system have varying amounts of protected bandwidth in the hybrid Fiber Coax systems, but usually topping out at around 860 MHz of spectrum, with which to make channels for sending TV and for providing broadband. This spectrum can be applied in a variety of ways and right now it is analogous to how analog cable did the job ‘ with each TV channel making its way all the way to every single set top, and then a tuner resolving just one of the TV channels for viewing (or two if you have two tuners).
Switched digital video (SDV) uses a different approach and Time Warner Cable is known to have experimented widely with this, and has a number of towns using it. SDV works more like an IP Multicast using IGMP (Internet Group Management Protocol), which is the basis of IPTV. We say ‘more’ like it, because it’s not the same. But there are requests that are issued to subscriber to a TV channel, just like Multicast join requests, and the TV channels are distributed only to a local switch and not all the way to each home. These requests pull the channel that each home wishes to view, into the cable loop. Once it is there, if someone else wants to view this channel their equipment can be told where it is, so there can never be more channels going down the access loop than those being watched. Logically it is very similar to the wireless multicast systems like MBMS.
Since most homes are not watching TV at the same time, and many of them watch popular channels, this re-harvests huge amounts of spectrum on what is usually a 1 Gbps coax loop, which might have anything up to 2,000 homes on it. This can then be used for VoD, more HD channels and broadband using high speeds systems, often with two channels bonded together, such as in DOCSIS 3.0.
Even this requires a return path built into the system, to carry these channel subscribe requests, and unless a system has DOCSIS already built in, proprietary return paths are hacked out locally, in a non-standard way. Better still the Tru2way (OCAP) set tops rely on a standardizes return path so if Tru2way is installed this can be done in a standard way, although they are very expensive.
What’s attractive to cable companies is that the apparent, inherent advantages of their cabling architecture are mostly lost in their current architecture, due to the nature of their historical choice of technology. People from cable always go on about having a better underlying cabling infrastructure than telcos, but mostly it’s not true. Telcos are now on a path to reach 300 Mbps down a phone line, where their loops are short enough, and compete quite happily with cable into the distant future. But that’s only while cable relies on edge QAM and puts every TV channel it has, including HD channels, the full length of their infrastructure. In which case, you may ask, why are more not changing immediately?
And that’s the problem. Each cable loop has to change, lock stock and barrel and that includes the customer premises equipment (set top in this case) which is always the most expensive part of any network usually around 50% of the entire cost.
What cable needs is one of three things. 1) An extremely cheap, and yet fully SDV compatible set top, so that ripping out tons and tons of set tops doesn’t cost the earth 2) Another transition strategy that allows SDV customers to co-exist with normal cable customers or 3) a similar leap to full IP compatibility, which bypasses SDV and goes straight to full IPTV, but at the same time doesn’t lead to anything more than a concerted Capex spend which investors can live with.
And this is why Time Warner Cable would go to Microsoft. At Mediaroom it has taken an IPTV set top design, which used to cost $700 to make and driven its price down below $100, through promised volumes and it’s financial muscle. Not many companies could have done that. It also now has a team that has been educated by Verizon and AT&T in how to understand large scale IP networks. All of its major mistakes and software re-writes and years of waiting for IPTV at Mediaroom are behind it. It has a core, hard earned, understanding of how to get this job done. At least you would thinks so.
For our money the company best placed to build this for cable is Cisco. It has the bulking of the high value router business, as well owning the remnants of Scientific Atlanta, a business thought to be in the slow lane, within Cisco, but which is never broken out in quarterly results, so it’s hard to be sure. Scientific Atlanta gives Cisco about a 25% market share in US cable set top boxes up against the dominant Motorola, with more like a 50% share. Cisco would bring with it a strong consumer electronics division, which includes every variant of wi-fi that might be needed to send programming around a home.
So while there are definitely other companies who MUST go after this new opportunity, it is Mediaroom which is supposedly being adapted to the Time Warner Cable platform in a trial in Los Angeles, still at the planning stages. It is expected that the trial, if it goes ahead, will go live later this year.
The US cable operators are beset on every side right now. They have the authorities unsure if it is fair to allow them to operate TV Everywhere style internet strategies; there are streaming TV services planned by both Apple and Google and others, and both have the power to eat into their subscriber rates and drive up churn, not to mention the existing Netflix online movie service which is showing astronomical growth. Telcos have finally got to grips with selling almost as many TV homes as they lose in fixed line telephone homes to the cable operators, and none of the cable operators yet have a well rolled out cellular strategy. They have a clear strategy in partnership with Sprint and then Clearwire, but it’s not something that they can sell right now, or at least not to many people.
A move to pure IPTV will give them standard ways of adding new interactive services in the home, something that Tru2way was supposed to bring, but has failed at due to the high costs of its deployment.
The problem in the supplier community is that those companies with the best knowledge of the IPTV industry, outside of Microsoft, are mostly smaller businesses. Sure Motorola bought a passel of IPTV providers, led by set top company Kreatel and is now shipping high numbers of IPTV set tops, but most companies in this eco-system would not have the resources to bang on the doors of the larger US cable operators and wait the long lead times which they typically have, for business. What cable has done in the past is have its collective R&D group, Cablelabs, work out its transition strategies to newer technologies and then create standards that multiple providers can supply against. That actually might be what this trial is about, simply using Mediaroom to provide middleware and server power to a trial designed to define the future cable-IP transition.
Certainly companies like Arris already steeped in cable and, which acquired set top UI company Digeo last year, and BigBand Networks, would have to make adjustments into a new cable IP eco-system. Bigband already preaches IPTV from its web site and has been instrumental in installing SDV at some Time Warner Cable sites. Motorola we feel would find the adjustment painful and it would likely further damage its cable set top leadership in the US. Most encoder suppliers swing both ways and could work with either IPTV or cable technologies. But if Mediaroom gets its claws into this business, then DRM companies which have grown up around IPTV, will find it hard to get a foothold with perfectly appropriate technologies, due to the fact that Microsoft already has its own DRM technology, an adapted Windows Media DRM with an on-chip set top component. That would then eat into Motorola market control, which it exerts through its cable conditional access system.
Either way this is the starting gun to begin courting the cable community, and expect plenty of announcements in the coming months to address this issue.
Light Reading referred to internal projects at Time Warner Cable and Comcast code named ‘Longfellow’ and ‘Excalibur’ which are designed to work out a full IP transition strategy. Perhaps these plans will come to reality earlier than everyone has so far anticipated.