COMMENT: Why DIDO does make a difference to NBN futures

Posted on: Sunday, 31st July 2011

CDMA was invented in 1942 but wasn’t used for military applications until the 1960s and for commercial applications until the 1990s. OFDM was conceived in the 1960s but, as with CDMA, didn’t hit the big time until decades later.

The next big paradigm in wireless communications announced last week—DIDO (distributed input distributed output) is potentially such a game changer that it will likely find its way into commercial applications in a considerably more compressed time period.

You might have seen some reporting of DIDO in the weekend popular press. Essentially it is an innovation that bends the convention of Shannon’s Law, which dictates that end bandwidth performance is always limited by the number of users sharing the resource. DIDO, created by the people who brought you Apple Quicktime, turns that on its head: effectively it would allow all users to get the full Shannon’s Law limit (aggregate capacity) of the wireless channel without suffering from any interference or having to share that channel capacity with other users. It does this through data centre-defined processing which determines the specific waveform requirement for each end user device and transmits accordingly. The end result is fibre-level 100Mbs/1Gig speeds.

In essence, if DIDO lives up to expectations, it will enable wireless networks to match residential fibre networks head on, with the benefit of the installation and mobility premiums that accrue to the economics of wireless communications. Indeed, DIDO’s proponents claim it will be cheaper to implement than today’s TDMA, CDMA and OFDM-wireless networks—something that stands in sharp contrast to copper-to-fibre migration.

The usual coterie of fibre apologists have been quick to come out and breezily concede that DIDO just might be a paradigm-changer but that, oh no, don’t believe the NBN Cassandras who will tell you that this just might be of relevance to current public policy debates. I would respond: don’t believe them.

For a start, let’s look at the point they concede and then proceed to brush over. We have constantly been told that we need expensive and pervasive fibre networks that fail to pass any “willingness to pay” test. Why? Because the “physics” of wireless cannot be changed and can never provide for ever growing demands for interactive high-speed bandwidth.

Barely 18 months after these statements were made with expensive certitude, this logic has been turned on its head. And don’t assume that all wireless innovation stopped on Friday, 29 July 2011. No one quite imagined the potential for TDMA, CDMA and OFDM when they were first conceived. We could potentially be on the threshold of something truly exciting here: the combination of the sheer horsepower of fibre-style high speeds with the alluring economics of mobility and ubiquity.

The same people who ask for giant public investments in civil works and networks supporting fibre (with all the uncertainties of demand) also clearly underestimate the power of capitalism to commercialise R&D for a market representing virtually all of humanity and for which there is a demonstrated demand. The very essence of innovation lies in its ability to subvert received wisdom about what can be achieved.

The second point being made by certain fibre apologists is that it will be many years before DIDO can be commercialised and, thus, it will have little impact on the economics of today’s network deployments. This is undoubtedly true for fixed network investments on a modest scale such as FTTN or DOCSIS 3 which typically can return costs in a number of years.

But the more expensive FTTH networks being built across the world entail payback periods stretching out to two to three decades. If one considers the enormous technological progress that has taken place in wireless over the past 20 years—which has seen narrowband AMPS with mere voice functionality evolve to LTE networks supporting 1Gbps at the cell, it would seem desperately naïve to assume that FTTH will have it all its own way in the coming 20 years. What’s more, the slow historical development of CDMA and OFDM was clearly constrained by processing and battery constraints that no longer manifest.

Indeed, DIDO is an exciting and refreshing reminder of the power of technological innovation and market economics in an industry which has become way too bogged down by the dreary constraints of political science and statism over recent times.

 Grahame Lynch