One of the hardest things to do is to reverse the habit of a lifetime. The habit for the mobile industry has been looking ever upwards to the next higher band in the spectrum in order to deliver ever higher data speeds.

1G started its life around 450 MHz in some countries and then 900 MHz became the norm. 2G began its life at 900 MHz and was extended to 1800 MHz. 3G was focussed on 2100 MHz. The ever higher spectrum band story was less clear cut with 4G with a schism between going up to 2600 MHz (for high data speeds) and down to 800 MHz (to improve coverage). But the hype was all about higher data speeds. In fact each new mobile generation was sold on the basis of offering high data speeds - with the time to down-load a DVD almost becoming the way to judge how good the next mobile technology generation was.

The quest for ever higher data speeds in higher spectrum bands started to go off the rails in the formative period of 5G. The prevailing consensus view of vendors was that 5G was going to deliver 1 GB/s at 28 GHz. (26,000 Mhz).

It turned out to be a costly error for those who followed that path. I was among the first to say that 28 GHz was going to be a mobile coverage disaster. The higher the frequency used the shorter the distance the mobile signal can travel. All the gaps left without mobile signals and only be filled by adding new mobile towers in the gaps. That needs new investment and results in higher running costs. Everyone had taken their eyes off the ball on just how costly mobile coverage at 28 GHz would turn out to be.

As luck would have it, a world radio conference in 2015 did not allow its mobile use in Europe. In 2016 I was sent over to Brussels to find a solution. I recommended pivoting downwards to mid-band 3.4-3.8 GHz with 700 MHz delivering national coverage of much reduced data speeds and 26 GHz providing huge capacity relief for traffic hot spots. The three spectrum bands were called "5G pioneer bands".

The 3600 MHz band was a compromise in seeking an optimal larger bandwidth/larger coverage combination. It caught the mobile world’s imagination as the 5G “sweet spot”. Ofcom played a leading role, having the RSPG chair at the time, in delivering this re-think across Europe in record time (and some might say “just in time”).

We then come to the arrival of the first reflections on 6G around five years ago. Right on cue the old thinking kicked in. Many research centres were suggesting the next logical step would be terahertz. What planet were these academics on? There is no point in mutli-Gbit/second mobile data speeds if the cost of coverage is so astronomic that no mobile operator would ever deploy it.

So if it is not to be terahertz and not to be mmWave spectrum for 6G – what are the choices left. The large system vendors see 7 GHz (7000 MHz) as the next opportunity…but that is still on the road to ever higher data speeds...and shorter distances the mobile signal will travel.

This is where everyone is looking in the wrong direction.

Higher data speeds are far less important to users today as video delivery has moved from down-loading to video streaming. The rate of increase in mobile users watching streamed video on their smartphones is slowing. We need to think differently about the future. What has never been achieved so far is to make mobile connectivity "highly reliable" across the enture country. That is what consumers have always wanted and the true measure of "excellence" of a "mobile" service. It should be policy maker's next goal. It would be greatly facilitated by a new spectrum band fairly low in the spectrum. Why a new band is so valuable is that it could be layered over an existing congested mobile infrastructure to provide "universal reliability" as a premium service for a relatively modest investment.

The UK finds itself in the fortunate position of planning the eventual close-down of digital terrestrial TV services that uses spectrum at 600 MHz. That is ideal for national coverage. At least half of that spectrum could come free within the next 5-7 years.

So we have a market need for more reliable mobile connectivity, the right spectrum to deliver it on coming on stream and the time to fashion the technology to pick up all the other issues where improvements are possible, such as lower latency and lower jitter.

A 6G is likely to be the generation that breaks old habits - we don't need even faster mobile data speeds. Instead it could be the route by which very high priority AI traffic can be delivered with a certainty that today's mobile networks were never designed to deliver.