Is DIDO The Answer?

Everyone’s talking about pCell – the DIDO (distributed input distributed output)  technology which looks like the answer to the wireless industry’s prayers.

It would be a fine thing thing if every one got the full data rate of the wireless signal from their base station no matter how many other people were using the same base station.

That's what pCell does. Or that's what pCell's inventor, Steve Perlman, says it does.

"Each user is able to utilise the full Shannon limit of the channel," says Perlman. Modestly, Perlman says he believes the technology will "completely transform the world."

He's looking for partners to join his company Artemis Networks in effecting the transformation.

Now usually, when someone goes public with a demo to the press in the search for backers, it means they've already approached the obvious candidates and been turned down.

You'd think Perlman would have made the trips to San Diego, Irvine, Santa Clara, Espoo, Stockholm, Seoul et al.

So there's a presumption that pCell is not all it's cracked up to be.

As it's All way above my head I asked Rupert Baines, CMO of Real Wireless, formerly vp at picoChip and Mindset and now advisor to Accelleran.

And this is what he said:

"What is being demoed seems entirely within the capabilities of existing LTE technology.

Demo 1 shows a total throughput of 60 Mbps with 4 HD videos at 7.5 Mbps each, and 2 4K videos at 15 Mbps each. For a 10 MHz LTE system, the peak single cell throughput is about 100 Mbps (with 4 antennas at least for the base station). So, demo 1 is showing the system achieving about 60% of single cell capacity.

Demo 2 shows a total throughput of about 20 Mbps with 8 iphones playing HD videos (at ~ 2.5 Mbps). Comparing that to a 5 MHz LTE system, the peak single cell throughput is about 50 Mbps (with 4 antennas at least for the base station) . So, demo 2 is showing the system achieving about 40% of single cell capacity.

20MHz is 200Mbps so even easier

Admittedly, achieving close to advertised peak throughput of a cellular system itself is an achievement, but not impossible in a carefully orchestrated demo."

On Thevsubstantive claims for the underlying technology, Baines says:

"Mr Perlman is claiming 1000X advantage, with no new interference, that is a huge claim: Turbo codes (which genuinely were revolutionary, as once in a generation advantage) "only" gave 3dB."

"Most technologies are offering +20%, perhaps +50% - 2X as an extreme - but 1000X would be remarkable."

"Personal cell is the idea that you can make a cell arbitrarily small - each terminal has a signal set of waves that it alone receives, from a constellation of antennas," adds Baines, "you have lots of antennas then each user sees a different combination of signals. If you have enough antennas and and lots of very smart maths, then that is viable."

"But it is sort of part of the same idea as a lot of mainstream concepts: CoMP, massive MIMO, distributed MIMO etc. So the logic is good."

"On the other hand, that really needs user co-operation, and Mr Perlman says it is one way i.e. it upgrades the network but can still use normal, existing, phones," continues Baines, "if so, that would be remarkable: it would be very important, it doesn't need new devices, and it means it doesn't need standards. He could do it directly, without waiting or co-operating with 3GPP etc. That would be huge."

"It also means it could work with all technologies: LTE, WiFi, 5G etc. That would be huge too BUT I don’t understand how that can work...

How do you focus these beams without feedback path?
How does the network know where you are in order to create the required ensemble across different antennas?
How do UEs resolve all the different MIMO signals without some dedicated hardware?"

Baines' conclusion is "all in all I'm pretty dubious."

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>