If you wanted to build a home network, you wouldn’t want to start from where we are today. The average home, depending upon how rapidly it updates its devices, has a CD player, a DVD player, a TV set or two, or three, a WiFi network, some coaxial cables, a twisted pair coming into the home from the telephone network, Ethernet connections off your home gateway, at least one MP3 player, around 4 mobile phones, which in turn connect to at least two, probably three cellular network signals each (GSM, Edge, CDMA, UMTS). You have devices which talk to one another using serial ports, parallel ports, or talk across Bluetooth, through USB, using S-Video, component video, composite video, DVI, Firewire, HDMI, HomePNA, G.Hn, Wireless HD and Homeplug. And have signals modulated in DSL, QAM, VDSL, ADSL2+, and tuned services adhering to DVB-S, DOCSIS, DVB-C and DVB-T to name but a few.
Homes have devices which are powered by batteries, trying to co-exist with devices which have mains supply and some which use a ton of electricity (CRT TV set). We have devices which store on disk, on analog tape, use lasers to burn holes in substrates (DVD and CD) and multiple sizes, formats and software to drive screens. And to add to all of these, digital content is ripe for piracy, so spends most of its time encrypted, specifically so that other devices cannot play it ‘ which is the whole point of a home network. Commercial content comes from radios, satellite dishes, hybrid fiber coax, UHF antennas and out of boxes containing DVD players, as well as over the internet.
It’s a mess. The digital home is only part digital, mostly incompatible and to many, incomprehensible.
We’re not sure why anyone wants to fix where we are today, instead of throwing it all away and starting again, but effectively a bunch of companies which control the bulk of all that technology (around 200 of them) formed the DLNA back in 2003 and set about making sense of it. In our view they haven’t got very far, and they have plenty of distance to go.
The core idea behind DLNA is to turn all of this into one big happy network. Neat eh? But it’s not that straightforward, hence the 7 year delay ‘ so far.
For a start the DLNA needed to limit communication to the most popular formats, and then convert to and from these formats automatically in a manner that works first time, every time. So while we all think Wi-Fi is cool and easy to work with, because we’re used to it, as a distribution vehicle for in-home video, it sucks. So by settling on it, it shuts out most of the ‘good stuff’, the HD commercial video assets that most US and European homes pay so much for, simply because when WiFi has no MIMO or any ability to get a reply from a client device and resend packets, its loses about 10% to 15% of what it sends first time. Which is why DLNA has had to also include MoCA, designed for shifting compressed HD signals around a home on Coax, and Ethernet, which is the output from most broadband pipes, along with WiFi. MoCA might be the market leader, but this is a very fragmented market.
Already DLNA it has a split personality, because built into DLNA is the idea of isochronous and asynchronous data, data which is attached to a real time clock, and has to be somewhere on time, and data that can get where it’s going as and when. It doesn’t matter if the button which turns on your TV has a 200 millisecond delay, but if the next frame in an MPEG video file arrives late, your TV image goes wonky. So there are multiple pathways, which operate at different levels of Quality of Service (promises to get data there on time) built into multiple network types. Some of these networks, such as WiFi began life without any QoS mechanisms, so they have had to be retrofitted.
There are conversions to and from larger screens, from different aspect ratios, different resolutions, from different codecs and different audio file types and these transitions have to be built into silicon and happen swiftly and silently, without fuss or failure.
Even then, the job itself is not done when files can move from one device to another, and find their correct format. Take the simple activity of trying to view a TV program on the TV in your bedroom, complete with keeping your exact bookmark of where you left watching it in the living room. Even if you are using a multi-room DVR, in effect a single hard drive version of that content, you need to take control of a remote device from your remote control in your bedroom. The view of what’s on the DVR storage needs to be shown to the TV screen in the bedroom, and to do that it may need to go through a huge number of authentications to ensure this second TV is a valid member of the household and not some pirate device, and then swap encryption keys with the storage device, and begin deciphering the TV data packets.
All of this has to happen every time, without fail, or else who will be to blame? The TV manufacturer in the bedroom, the remote control designer, the home networking backbone, or the DVR and its software? No-one would know. So all the devices would be suspect immediately.
Perhaps this is why the two most conspicuous non-members of DLNA are Apple and Google ‘ both with set top offerings, and both offering a hybrid, anti-pay TV, pro-internet, way of managing video content. Apple takes care of all this complexity one way ‘ it designs everything end to end, and Google takes care of it another ‘ it expects IP traffic at every step of the journey to run on the Android OS, with a full return path for interactivity. Google has spent most of its life as a search engine on a PC and is only now starting to make moves onto operating software for handsets and set tops. It doesn’t need all that complexity.
This is why it is easy to think of DLNA, the digital home networking stack, being like a big house of cards, an 80 to 100 standard stack that has taken 7 years to build, which may simply fall down any second. The DLNA has co-opted these standards, not made them, and got manufacturers to create implementations of each standard which work together ‘ it’s no mean feat. But still all of this effort relies on the whim of content companies.
At least that’s pretty much the way that content companies view it, which is why one of the reasons that all of this has moved on so slowly.
Initially DLNA thought, ‘Okay let’s get the plumbing right and by the time we’ve done that, the content companies might have made decisions about how they want to go.’ Which is why all of its early examples were made up of video which the consumer owned ‘ so on its white papers on the site it talks about as use case where a consumer wants to share a short video of her daughter taken on her mobile phone and drags you through what you would have to do in a pre-DLNA world, compared to a DLNA enabled world. And it’s a lot.
But now 7 years on, DLNA has achieved all that and got thousands of devices to become DLNA certified, today 2,000 TV models have DLNA on them, and yet its vision isn’t complete ‘ not yet, until it can embrace encrypted commercial content and every other kind of device.
We recently had a briefing on DLNA from Dr. Stephen Palm, also Technical Director at Broadcom, who now sits on the main Board of DLNA. It became clear from talking to Palm just how much control the content businesses have. They are happy right now with protected DLNA streaming, but still have a long way to go before they embrace two further key concepts ‘ storage on a second device and software only based protection services.
First a history lesson.
The DLNA was formed in June 2003 as the Digital Home Working Group, and at the time we said it was one of the most significant organizations and one of the most powerful technology groupings to address the digital home.
Today there is not much more of a digital home than there was back in 2003, and all of the efforts of the original 17 companies and the 28 which Palm lists as Promoter members today, are only just coming home to roost. Sure there’s a bit more broadband, a few more pay TV homes and a lot more home WiFi, but in the main, so far this organization has had close to zero impact.
Since its formation, it has been present at IPTV and Cellular shows around the US and Europe ‘ usually in the form of a tiny DLNA stand, which was almost a figure of fun, tucked away in a corner, on hand to recruit new members. A DLNA speaker would always whine on about the individual standards stack and their particular implementations, like they meant something, always a technician and always lacking in market relevance, and always preaching to the converted. This giant was effectively sleeping.
Talking to Palm, we can see that this has changed. The DLNA has a new Board, with the chair now taken by Nokia, Vice Chair by Microsoft and new directors from Broadcom, Intel, Panasonic, Philips, Samsung, and Sony and Awox, a small French DLNA software stack provider.
What’s interesting about the formation of this group is that it was put together so that it could NOT be hijacked. The original 17 Promoters had votes, and contributor members, of which there are now 200, do not. We assume this is still true of the 27 current Promoters although Palm tries to de-emphasize the control that the Promoters have and suggest that it’s a club of 200 where everyone sorts out their differences.
Today the 27 promoters include all the board members but also Access, the Japanese CE software supplier, CableLabs, Cisco, Comcast, DirecTV, Dolby, DTS Digital Entertainment (US audio specialist), Ericsson, HP, Huawei, LG, Motorola, Pioneer, Qualcomm, Rovi, Sharp, Technicolor, Toshiba and Verizon.
From the original members, out of Promoter status has gone Fujitsu, Gateway, IBM, Intel, JVC Kenwood, Lenovo, NEC, Philips and STMicroelectronics, as well as Texas Instruments, which joined after the formation and then fell back into the ranks. Most of the companies that you would expect to be in there, are in there as contributors, with a few exceptions.
There are no content companies. They clearly do not see it as their proper business to go poking around in the underlying plumbing within the core electronics, even though they want absolute control over the environments which they give permission to carry their content.
So although Sony is represented, not Sony Pictures. Neither is there Fox, Paramount, NBC Universal or Warner Brothers ‘ at least not directly. They have affiliates businesses involved ‘ Comcast will soon be able to speak for NBC, Time Warner Cable is only recently separated from Time Warner, so we’re sure they know in detail what’s going on.
Interestingly there is no Apple or Google, and we think that for Google this is an oversight and for Apple it’s a political statement ‘ almost the position that Apple only needs to talk to Apple devices.
Ericsson, Nokia and Technicolor and Awox are the only European businesses represented and the only pay TV operators among the promoters are Comcast, AT&T, Verizon and DirecTV – the largest US cable co, largest DTH company and the two largest telcos.
What you get the sense of is that this is the start of the operational phase of DLNA. Before the techies were in charge, now the businessmen have stepped in, though Palm in particular clearly has an immensely strong technical grounding.
But right at the beginning there was something missing, and we asked the initial chairman in 2003, Scot Smyers from Sony, many times about DRM and many times he said, ‘We are aware of the problems posed by DRM,’ and that was about that.
It turns out that the issue of DRM falls to DTCP, and the recent arrival of DirecTV came about directly because of this.
DTCP is used in DLNA in its Internet Protocol version called DTCP- IP. The original protection was called the 4Cs then later the 5Cs and was a piece of research between IBM, Intel and two Japanese CE firms, but it is now owned by Hitachi, Intel, Matsushita, Sony, and Toshiba. It is a method for setting up a protected encrypted connection between two devices, even when their first ever communication has to be presumed to be eavesdropped.
It uses something called an Elliptic Curve Digital Signature for authentication and the Elliptic Curve Diffie-Hellman algorithm for key exchange. These rely on the AES encryption and Public Key Infrastructure and session keys, once established, are 128-bits long and use the AES base cipher to encrypt content.
The idea behind DTCP is that two devices ‘ server and client say ‘ in a DLNA network, share a secret over a non-secret channel that may be intercepted.
They already have a public and private key system of their own and they manage to swap public keys or clues to their public keys, so that the rest of the key can be worked out. Each device in your home can manage this negotiation, and as long as one of them has access to the content, it can reach the device. Their original keys are initially allocated when they are shipped.
It was first implemented in the setting of the 1394 Firewire serial bus standard and in fact DLNA can be thought of a little like taking the aims of Firewire and extending them throughout the home and updating them from their 1995 beginnings, when that standard was set. The idea behind Firewire was to replace and extend the SCSI (Small computer systems interface) bus into the consumer electronics world and it was virtually where the Plug and Play concept came from, of plugging something in while it was playing and yet it still working without a reboot.
The DTCP standard has branched out to cover a variety of media types and now supports USB, IP, WiFi and Bluetooth. But it was during its 1394 days that DTCP developed a reputation as being highly reliable in protecting content from illegal re-use.
When asked about it Palm said ‘Oh you are talking about providing a means to enjoy commercial content on a DLNA device. Converting from conditional access encryption and re-encrypting with DTCP-IP is the right way to go about doing multi-room DVR and all other forms of encrypted commercial, in-home, distribution,’ and he gives the example of DirecTV, which has only recently joined the DLNA but which has built an impressive multi-room DVR in precisely this way.
In DLNA speak there are a number of device types ‘ servers, renderers and controllers and players as well as uploaders and downloaders. Servers acquire, record, store and protect content, and source content for other devices. Renderers playback content from servers after being configured by a controller, players can do it on their own. Uploaders copy content to servers while downloaders copy content from servers. There is also a special class of mobile media server and mobile media player, and networks can be limited to certain functions too.
So when Conditional Access content is played out inside a server, instead of just storing the content, it can re-encrypt the content using DTCP-IP and send it over purely as a stream, to a rendering device.
But in all cases right now, a Pay TV operator is only allowed to let its content go to a DLNA certified player and in order to do that, it must be a device which ships with a DTCP-IP base key, hidden within it somewhere in hardware. So right now letting content play to a Tablet would require for a version of the tablet to be DLNA certified, and for that it must have a key hidden in some part of its hardware which is protected from prying. So the iPad cannot be connected to a DLNA device to enable content playout unless Apple adds new hardware.
Recently at IBC, we saw set top maker ADB actually doing this, and it assured us that it was perfectly possible to do this in a software application. ‘DLNA is interested in running software certification ‘ it’s not difficult to offer DLNA in software and there are 7 or 8 apps out there, but they are currently uncertified. That’s because there is no software certification program yet in place, although we are working towards one,’ said Palm.
Palm could give no time frame for software certification and we suspect there isn’t one and there is a reason for this. As much as pay TV operators WANT to offer cool Tablet content around the home, the content businesses – studios and broadcasters ‘ will not give permission yet to pay TV operators to work with software only security.
So this massively powerful organization can only really decide How to go about sharing content in the home, and it is organizations like the Motion Picture Association of America and all its members, which will decide When they are ready.
The studios have said streaming to physically secure devices is enough for now, let’s see how it goes, and that’s that. So if DLNA cannot go to a device without a hardware aid for security, how comes handsets are emerging which are certified. Quite simply, unlike Apple, handset makers are happy to go the extra mile and put in DLNA. Although for the most part, none of these have the capability of playing out TV programs.
You can search for DLNA certification on its website, and there are hundreds of mobile phones which are certified as M-DMS (mobile digital media server). In this category there are a few products which have made a noise about their DLNA credentials. Vodafone has come out with a DLNA certified MiFi style device which allows phones to share content through it and Samsung has launched both the Captivate and the Epic, a US WiMAX phone as DLNA enabled. A number of the LG Optimus smart phones have it, along with the Motorola Defy, the Samsung Galaxy Tab tablet, the Panasonic Lumix and the Nokia has at least 20 devices which are M-DMS enabled.
As far as we can tell, none of these have Mobile Digital Media Player functions, which would mean it is able to play digital content from other DLNA Certified devices, in other words none of them have the DTCP-IP capability for viewing re-encrypted content ‘ yet.
Our guess is that Motorola, when it comes up with a companion Tablet product for Verizon’s FiOS TV, which is expected in January, will be the first such device which enacts the type of security that is needed to view encrypted paid video. If we’re right it means that widescale use of such functions is a long way off. What we will initially see instead is control functions offered to tablets, not viewing functions.
‘The original model for handsets was simple – control applications would run on them, such as TV remote control. But as more storage got added to handsets and there were higher megapixel camera outputs, the use of content outside the home has become important,’ said Palm.
‘Also the addition of Wi-Fi on phones and the fact that it is rapidly becoming a standard for offload has created the small matter of Over The Top applications.’
So today after 7 years of DLNA work, we cannot even get a video safely from a Pay TV service to stream on a handset, never mind store on a handset for viewing in another home, although both are on the roadmap, which, at a guess might be delivered around 2017. But the fact that there is such a roadmap, makes this the true route to mobile TV in the same timeframe ‘ store programming on your DVR, copy it to your handset and view on the go ‘ but quite when Hollywood will give the go ahead on that is anyone’s guess and another 7 years is not out of the question.
‘It’s true that DLNA allows continuous delivery today, not delivery to storage,’ agreed Palm. ‘Technically there are problems to solve, and a lot needs to be done in that area, but basically the Content People want to see how streaming goes.’
What Palm points to is a future which content owners want, where every piece of content and therefore every DLNA server, becomes a store. DLNA has built something into DLNA called Content Directory Services. These are DLNA search tool which sit on every DLNA server to help advertise content that each device is authorized to play. Hollywood is now looking to these type of capabilities and instead of saying, ‘sorry you don’t have the right to play this content on that mobile,’ instead offering you a price to make a copy.
This has the potential to make DLNA exceptionally unpopular. Where once it was easy to simply ‘take’ a personal copy of content, it will become virtually impossible without a new class of pirating technology and after years of being under the cosh of piracy, don’t be too surprised if the idea of a rights locker starts to emerge to make good on this hard won DLNA technology.
We’ve written a lot about DECE and its rights locker ideas where you buy content once and then have the right to play it on many other types of devices, and can download those formats subsequently, and Palm says that the DECE technologists are trying hard to fit in with DLNA and make the two compatible.
It is likely we will move to a point where if you buy a copy of a piece of content, you have right to this resolution and all resolutions below this. So if you buy a mobile version, you will have to pay more for a DVD quality online version. On the other hand if you have a Blu-ray version of the content, then mobile copies might be authorized automatically. But not immediately if the content guys have any say about it. Remember studios lived in a world where you watched a film on the cinema, then watched in on a rented DVD, then saw it on pay per view, then watched it re-run on TV, then bought the director’s cut on DVD ‘ paying afresh every time you viewed it.
Technology is after all agnostic, like weapons, it depends on what companies choose to do with it.