(Apologies in advance for the long post – I just needed to lay out some thoughts on the topics mentioned)
In the world of networks, there is constant ebb and flow as to where the intelligence is located at – are the terminals (devices) at the very edge of the network supposed to be smart or dumb? Where is the intelligence and, perhaps more important, where is the control?
The Internet: from network-centric to device-centric .. and to network-centric again?
On the (fixed) Internet, the prevailing setup been that the network is dumb and most intelligence is in the devices, usually PCs. It hasn’t always been like that though, as even I remember the times of the VT and X terminals when the devices we used were really, really dumb without any storage or anything – just acting as a window to the server.
But then came the PCs and almost all the activity shifted to the devices themselves. As they gained storage and processing power, they began to take on many functions to themselves and even with connected applications like e-mail, the servers became just the messengers. With the proliferation of PCs at homes and intermittent or non-existent connectivity with the outside world, this made perfect sense. Now, however, with near-ubiquitous and always-on broadband becoming the norm in the developed countries, the trend is again towards the servers or the “network”. With Google leading the way, the browser is becoming our interface to the intelligent applications that all really reside on the server-side.
Whether this trend reverses itself again in another 15-20 years when the PCs of that time have equal storage capacity to the Google server-farm today, remains to be seen. In any case, Sun’s and Scott McNealy’s long-time vision of “the network is the computer” is again a big step closer to being reality than it was just 5 years ago.
On the business side, the Internet was so new by the time the balance started to shift towards the PCs that most ISPs started with the business model of being mainly a connectivity provider – essentially a dumb pipe. Though there has been continuing debate for years now as to how sustainable this model is, they seem to be doing just fine.
However, all is not well. The network is still relatively “dumb” in the sense that it enforces little or no control on the data sent and received by the devices. Though some operators are (in vain, I might add) trying to block and filter P2P traffic etc, relatively little real-time layer 7 filtering goes on in the networks – and none in the backbone. This combined with the advanced state of the PCs and their poor security thanks to unmentioned operating systems creates some serious security issues, of which botnets are one of the most visible ones. With no or primitive “rebel control” on the network side and only presumed control on the device side, large hordes of compromised PCs can for the most parts wreak havoc and spew spam at will.
Mobile: from network-centric to device-centric? But how far?
On the mobile networks side, the entire business of the mobile operators has been based on smart (or at least tightly controlled) networks. The devices have been quite dumb and pretty much all the services have required extensive support from the network to work. Now things are changing.
The mere prospect of mobile operators potentially becoming just dumb pipes owes itself to the vastly improved capabilities of the mobile phones. The devices are getting more and more capable and intelligent by the day, to the point where calling mobile phones “multimedia computers” no longer sounds quite as stupid as it did a couple of years ago when the term was coined. With the latest terminals one can even make phone calls without any special support from the network itself; all you need is IP connectivity.
Hence the mobile operators’ relatively recent fear of being delegated to merely the providers of basic connectivity – bit pipes. If they become providers of nothing more than IP connectivity, they are facing a massive overhaul of their traditional business models. This, even though it may turn out to be inevitable, is of course something not all operators are exactly looking forward to. This reluctance is curious in another context; the true core competence of most mobile operators is precisely in running the connectivity network.
Converging networks, diverging policies?
As the inevitable convergence of the networks (fixed and mobile) continues its march, it’s worth considering which policies will be adopted for the combined network, the makings of the future Internet. Is it the generally laissez-faire model of the Internet as it is now? Or the somewhat stricter network control model that’s being attempted by some? Or the strict mobile-operator-style model?
The starting point of the policies are worlds apart, but as we’ve seen, both sides are moving towards each other. I would argue that this is necessary, because a converged network with diverged policies is just not possible. While the fixed side may not really want – or arguably need – more control, this is the direction it’s being pushed to. However, the relatively huge processing power and capabilities of the terminals on the fixed side as well as the bandwidth available may successfully fight the control that’s being imposed on them through encrypted traffic, overlay networks and the likes.
A PR problem at the mobile operators – or worse
At the core of this discussion lies the separation of services and access, a setup that has been working nicely on the Internet for a long time now. And consumers still need the services and the mobile operators could provide them.So why would the mobile operators want to fight the split? Because when consumers go look for the best services for their devices, they are very unlikely to look at the ones offered by their mobile operator, that’s why. With practically infinite choice available, providing sufficient recommendation and search tools are available, the chances of the operator offering the best services are vanishingly small. Of course, being at the bottom of all customer satisfaction surveys doesn’t help the mobile operators. Having innovation roll-out times measured more in decades than months doesn’t do much good either.
The option of getting similar services from elsewhere exposes the operator services for what they really are; often mediocre at best and overpriced dysfunctional rip-offs at worst. In essence it’s a grave that mobile operators have been digging for themselves for a long time and something that they are now being pushed into.
Before proclaiming that mobile networks should work in identical fashion to the fixed Internet, there are a couple of small problems to consider. One of them is whether we want to repeat the botnet problems of the fixed Internet on the mobile side? The answer to that is that we do not, because the mobile side wouldn’t survive it. While botnets are, on the large scale of things, a nuisance in the Internet, the backbone pipes are thick enough to survive the occasional attack from even hundreds of thousands of machines. Mobile networks don’t have this luxury – a single device sending or receiving data at capacity has a very detrimental effect on the efficiency of the network in its proximity. Take a thousand compromised devices and you’re quickly looking at a network meltdown, so network control of the devices is to some degree a justified position.
But to what degree? And should we adopt similar measures on the fixed side?
Who do you trust?
In my opinion we’re looking at an inevitable split between access and service providers. I realize that to most people familiar with the industry, this is a “Well, *duh*” statement. For the consumers, however, it raises tantalizing options – as well as one very important question: who do you trust? This is something the advanced users have to deal with now and the masses have the option of dealing with in the near future. The question needs to be answered because to offer personal services (which mobile services often are), you need to trust the service provider.
Do you trust the network provider, even if it doesn’t trust you? Do you trust a multinational corporation based in some foreign country? Besides service experience and quality issue, this is a PR and an image question and a very personal one at that. And that makes the choice irrational at times. Why so? For example, I am fine with Google holding and managing selected portions of my e-mail but I’ll make damn sure Sonera doesn’t get any of that. This is so even though the former is potentially a much more likely candidate to hand out my information to government-affiliated organizations and is generally a greater threat to my privacy, however you want to define that. Sonera? Well, I just don’t like them. (Though I’m not as naïve as to believe that organizations like CIA or the NSA wouldn’t already have everything they need, including my e-mails if that’s their whim of the day)
In an attempt to draw some kind of conclusions from this tirade, I’ll give my personal predictions:
- In countries that fulfill the two key premises (advanced mobile networks and advanced mobile devices), the power balance will inevitably shift away from the mobile operators. This is because other parties can offer more tempting services than the operators and something operators need to deal with.
- The relegation of mobile operators to pure bit-pipes will take many years (if it ever even materializes), but losing even smallish customer segments is enough to drive further introspection in the operator community. In the medium term, the majority of subscribers will still want only cheap voice and SMS and, which is what operators can often deliver if they want.
- Data ARPU will continue to increase – the increase is supplemental at first but will start eating into voice ARPU in the longer run. In the longer run, the data ARPU is going to come down also.
- At some point, some operator will figure out that “embracing the Internet” doesn’t mean adding fancy boxes to their networks or even making deals with selected partners. It means fully opening their networks to and from the outside world.
- Of the key mobile operator assets, existing customer billing relationship is coming under fire from the (free) Internet service. Location is still theirs to lose, but the soon-to-be-ubiquitous GPS receivers in handsets don’t bode well for that either. In essence, most foundations of the operator business are under fire and will continue to be so.
Then again, I might be totally wrong. That’s been known to happen. And YMMV in any case.
