Features

Your phone as a wireless carrier bag - a plea for sanity and a single, common control channel.

It’s a sad fact that the concept of lemmings hurling themselves over a cliff is apocryphal. Disney invented it and popularised it in the 1950’s film “White Wonderland”. If Disney hadn’t then someone would need to be going out and reinventing the concept, as there’s nothing more relevant to the wave of hype and hysteria regarding mobile wireless.

I believe in wireless for mobile devices. But I also believe that if it’s going to be successful it’s going to need to arrive in a fairly orderly manner, where the current plethora of wireless standards will eventually be whittled down into a smaller number of solutions. The mantra of “anything and everything will fit” may have some validity for a PC platform, but once we move outside that to truly mobile devices it’s a dangerous fallacy. The cautious should look at the raft of radio standards that are being promoted and ask themselves the basic question – why do so many of them start with the syllable “WHY?”  

At Ezurio we’re involved with short range wireless connectivity – in particular with Bluetooth, but increasingly with Wi-Fi and ZigBee and we’re aware of the fact that each is a totally separate entity. The recent announcement of co- operative work between Bluetooth and UWB shows a welcome move towards a shared understanding, but most of the time integration and co-operation remain dirty words. Once again, the question is “WHY?”   

What is a handset?  

Once upon a time we all knew what a mobile phone was. It was a device that let us talk to others without having our location restricted by a scant metre of curly cable.

As the critical mass of users grew we, the users, discovered a new use that none of the network providers had foreseen – we invented text messaging, probably the most expensive method of sending information that has ever been developed. Using it to send the complete works of Shakespeare would cost around $5,000, which is considerably more than amazon.com will charge you for a nice, leather bound edition. The next revolution was as mundane and annoying as downloadable ring tones, which grew out from a free Christmas novelty generated by Nokia’s engineers into a multi-million dollar industry. The revenues involved in these two areas are such that everyone within the industry is desperately trying to work out what the next money spinner will be. As they don’t know what it will be, they’re taking the approach of cramming every conceivable piece of new technology into a handset, oblivious to the fact that the two great revenue streams of SMS and ringtones were serendipitous developments that grew out of the user community, not the brains of the mobile marketing departments.  

But there’s another stream of handset development working alongside the content delivery models. It’s driven partly by the PC industry, who believe that massive bandwidth, delivered to any device, anywhere in the world is a staple of human existence in the twenty-first century. As well as the Blackberry addicts, conditioned in a Pavlovian manner to email on the move, the advocates of this theory cite the iPod phenomenon, which is a clever piece of marketing that has so far persuaded almost 20 million people that it’s cool to carry a hard drive around with them, as long as it’s embedded in a piece of white plastic. Put the two together and we get the concept that has been described as the iPhone, where a 2GByte hard drive is squeezed into the mobile phone to give us music, voice and a failing battery wherever we go. The corollary to this picture is that to get data onto the drive at a reasonable rate needs something faster than GPRS (which at current rates would cost me $3,000 to fill up my hard drive in download charges alone, before I even think about paying for the music). Hence the interest in putting every other conceivable radio into a handset in order to allow instant music and video on the move.      

The Phone as a Wireless Carrier Bag  

In the beginning your phone had a single radio. As time progressed that GSM radio gained complexity, coping with two different frequencies, then three and now often four. Thanks to some clever design in terms of silicon integration that still means only one physical radio implementation. Along the way it gained a second radio in the form of Bluetooth, which is likely to be there for the duration, as it’s the only radio that is low power and which also supports voice. So as long as we want headsets and car kits, Bluetooth is here to stay. Next came a new radio for UMTS. So we’re looking at a current top end generation of handsets that contain three radios, with the majority of mid range handsets containing at least two.  

That’s manageable. Not least because as far as the user is concerned there’s only one radio that they have to set up, which is the Bluetooth radio. The GSM radio is so deeply embedded that it just works – all the user has to do is to dial a number. At this point it’s worth pointing out that setting up a Bluetooth connection is at the limit of most users’ competence. And equally at the opposing limit of most phone designers’ understanding of a user interface. There is a link here, and it’s one to hold in reserve for later in this exposition.  

This is where we move into the concept of the mobile phone as a carrier bag for radios. Because the advocates of wireless want to start adding more radios to the everyday handset.  

First in line is Wi-Fi, otherwise known as 802.11b or 802.11g, which operates at 2.4GHz – the same frequency as Bluetooth, but which, for technical reasons, needs a separate chip and antenna. Quite why we need Wi-Fi in a phone is still open to debate, but the most general explanation is that it will allow VoIP calls to be made when you’re in range of a Wi-Fi hotspot, thus destroying the revenue of the mobile operator. It’s not the only way to reduce call costs within particular areas. It can be done without a new radio, by being clever with the SIM. For more on that approach have a look at www.coffeetelecom.com. Although Wi-Fi is already going into handsets, the application still has to prove it’s worthwhile. Those with a long memory will recall the less than successful CT2 network, which had a surprisingly similar business plan.  

On to the next: NFC or Near Field Communication. NFC is a short range radio which is being proposed for secure payments as well as a method of setting up other radios. On the latter score its proponents claim it’s necessary because other radios are too difficult to set up. They don’t seem to see the irony in proposing another one to solve the problem.  

Back to the Wi-Fi camp again, who will admit that the issue with putting 802.11g into a phone is that the average hotspot only has the capacity to support a single digit number of concurrent VoIP calls. However they have a solution called 802.11a (also known as Wi-Fi), which will solve the problem by putting in another radio that works at 5.1GHz. Which gets us to three new radios in our carrier bag phone.  

However, Wi-Fi is losing its novelty in the PC sponsored radio world, where all of the smart money is now getting directed to Ultra Wide Band, or UWB. UWB promises even faster data rates with reduced power consumption, and is obviously such a good thing that everyone wants to do it their own way. As a result the standards group has split into two competing camps, both hoping that market acceptance will turn them into a de facto standard and hence become king of short range radio.  

If you thought we were there on new radios those last few words – short range radio - gave the game away. There’s a gulf between the wide area coverage of the GSM and UMTS networks which let the cellphone communicate at distances of tens of kilometres, and the short range radios that struggle to get past the tens of metres. The PC industry keeps coming back to the need for high speed bandwidth any time, any where, any place. Although the mobile industry is moving ahead with higher speed networks, using techniques such as HSDPA, the PC industry appears to lack faith in the capacity or timeliness of the traditional networks to deliver. Instead it is promoting its own vision of the future – WiMax.

WiMax came into existence as a Metropolitan Area Network, capable of providing T1-type speeds over a medium range wireless link. It’s evolving from the original IEEE 802.16 standard to a new, lower power mobile version called 802.16e, which is slated as the next best thing (where have we heard that before) for our future iPhones, allowing us instant email with really bloated attachments as well as music videos on the move. And surprise, surprise – it’s another radio to add to our already overloaded carrier bag of a phone. Moreover, unlike the unlicensed short range radios, WiMax uses a licensed band, which means that it needs network operators, infrastructure and more subscription fees.  

The optimist would suggest that these are all the radios we’ll ever need in a handset. The pragmatist will point out that two years ago we’d only ever envisaged three – GSM, Bluetooth and UMTS, and if we’re up to eight just twenty four months later, there will be those clamouring for twelve a year from now. Our future phone looks like it’s going to be a very busy piece of gear.   

The advantages of adding radios  

This is the paragraph where it all starts to get difficult for the proponents. Any network operator or content provider should be asking what opportunities these new radios bring. By looking at the radio in isolation it’s possible to devise some interesting new services and business plans, but take them together and the future looks very different.  

What does adding five more radios bring to the handset? It can be summed up in three bullet points:  

Greater drain on the battery 

More use of pcb real estate 

More impenetrable user interfaces

The big issue comes from the fact that each of these new radios is a totally separate entity. Each one will need additional chips and antennae to be added to the main board of the mobile device. With the current state of technology that accounts for between 1 and 2 sq cms for each new radio. Five new radios is equivalent to around 50% of the board area of a small handset.  

That independence of radios carries another burden. Each one will need to power up every few seconds to find out if there is anything for it to do. Every time it does it takes a bit of power from the battery. Supporting multiple radios means a much larger battery, and there’s no new battery technology on the horizon that is going to keep up with this increased power requirement.  

All of the new radio standards have taken the approach of trying to reinvent the wheel in how they connect and how they talk to applications, so that means a set of new user interfaces for each radio. That’s where we bring back that thought about how difficult it is for users to set up a Bluetooth radio.

 Bluetooth is probably three to five years ahead of any other radio standard in defining how to make the wireless connection simple. Adding each new radio adds a level of user complexity which will result in them giving up on all of them. Recently UWB made a far sighted decision to work with Bluetooth as an upper application layer. That will help to hide it from the customer, which is a good move in an otherwise self-serving technology area.   

The Economics  

As well as taking up space, power and user tolerance, multiple radios add another feature – cost. Each new radio will add between $10 and $20 to the overall manufacturing cost of the device. Adding five new radios will push up the cost of the handset by around $50. That will only happen if someone is prepared to pay for it. The question is, who will?  

Throughout most of the world handsets are sold on a subsidy model. Although some users will buy a new handset outright, increasingly the phone is used as a bribe, either to prevent or encourage churn, depending on whether the user moves network or not. A few years ago, if I wanted a 2 Megapixel digital camera I’d have to go out and pay several hundred dollars for one. Now, I phone up my network operator and say I’m moving to another network and hey presto, I get one for free. Whereas I used to keep my digital camera for three or four years, I can now pull that trick and get a new one every twelve months.  

Where’s the logic. A good camera phone costs the network around $50 more. This year there will be over 100 million camera phone sold, which will cost the networks $5 billion. I don’t believe that they are going to make that much from MMS messages. The question is whether giving me a new phone increases my ARPU, or whether they’re just buying my custom. Can you see the Emperor’s new clothes?  

At least with cameras they can try to convince themselves that as soon as the issues are sorted out we will embrace MMS and the revenues will flow in. We’ll see. Remember that neither SMS nor ringtones were developed by the mobile industry. Whereas WAP and MMS were.  

But what happens with all of the new radios that are being planned? It’s yet another $50 or more of cost, and the business models that are being proposed mostly REMOVE revenue from the network operators by promoting VoIP calls and encouraging downloads from local internet access or home PCs. Why should a network operator subsidise them? Add together the cost of five radios and a camera phone and the networks are giving away $10 billion a year in technology that generates no revenue for them. It’s a path that makes the 3G licence spend look justifiable.   

The Good News  

There is one potential positive feature, which will at least slow down the rush for multiple radios. It’s the national regulatory requirements for unlicensed radio transmission. Amidst all of the hype surrounding these radio standards, there is very little debate as to whether they are legal. The three radios already in the bulk of handsets – GSM, UMTS and Bluetooth all worked hard to ensure that the frequencies they use are globally approved, or that radios were designed to cope with the different national requirements. That’s a task that has yet to be done for most of the other radios. UWB is currently only legal in the US, and there only under an FCC waiver. In the UK, Ofcom has recently released a research paper indicating that there are significant interference issues to be solved before UWB can be deployed. An earlier study suggested that using the FCC frequency mask would lead to UWB failing to generate revenue until 2018. WiMax is also likely to face delays in approval for general mobile use in much of the world. And even Wi-Fi faces different power requirements as you move from country to country within Europe, which will be exacerbated with the MIMO technology within 802.11n.  

This may not be an issue for PC vendors, who have the space to fit radios using mini-PCI cards which can be populated onto a motherboard at production, or even at final distribution. Handheld devices are built differently – everything is soldered to the main board – hardware is rarely customised to an individual market. The approvals costs for mobile devices mean that manufacturers want global platforms. And until the regulatory standards allow devices to ship globally, multiple radios will be confined to a few high end handsets.  

With luck this will give the industry time to look at how to work together to simplify the way radio is presented to the user. I don’t believe we need so many different radios. What we do need is a common method to interface with applications, to display information to the user, and preferably a single radio that can control the other radios, rather than letting all operate autonomously. The recent announcement of collaboration between Bluetooth and UWB points the way forward. It would be good to see this approach encompass other standards so that by the time we have multiple radios, there is a standard implementation. All it needs is agreement to use one low power radio as the common signalling and control channel, along with a single, integrated protocol stack. That saves cost, user interface complexity, power and money. But it means standards groups will need to start to work together for a common good.   

The Future – Software Defined Radio?  

Of course, at some point in the future technology will be capable of producing a software defined radio that can work across all bands and protocols. It is a vision that will solve many of the problems, but will equally open up a new can of worms, as the efficient spectral usage that can come from Software Defined Radio is totally at odds with today’s regulatory regimes, which manage spectrum by the MHz. Once SDR becomes possible we should consider rewriting the rulebook. The conspiracy theorists out there may argue that this is what companies such as Intel are already surreptitiously doing – backing new radio standards each year and forcing a pace of change onto regulatory authorities such that they buckle and acquiesce, leaving the future spectrum open to all comers. The question is whether getting mugged improves your quality of life in the short to medium term.  

Only time will tell whether we will look back and see this as the period where spectrum regulation started to crumble. If that is the ulterior motive, it should give cause for those building business plans upon the new radios to pause and consider. We are throwing technology at mobile customers faster than they or the infrastructure can absorb it. And yet those pushing these future technologies still talk of business plans.    

Refs:  Ofcom research on UWB benefits and interference  

Visit www.ezurio.com for more white papers and background articles about wireless technologies.