WTF is DIDO

[senectus]

Looks like a serious case of futuristic BS to me..

It looks like to make it work you need every user to have their own AP, and to top it off every AP needs to be a damned supercomputer to figure out the bandwidth.

[Renai]

We get one of these "wireless breakthroughs" every few years, but I am yet to see one of them really start to deliver in any real-world sense. The gigabit wireless chip unveiled by the CSIRO is another example ...

[HubertCumberdale]

I read a bit on this the other day. Far too vague for me to form a detailed opinion on this. No doubt the anti-NBN crusaders and the coalition will claim it as the great magical StarTrek technology they've been desperately hoping for to discredit the NBN even before getting the facts but I doubt even if what the inventors claim is true will we see anything practical over the build time of the NBN.

[Jeremy Kime]

It does look suspiciously like the telecommunications equivalent of perpetual motion. As with anything, seeing is believing.

[looktall]

It does look suspiciously like the telecommunications equivalent of perpetual motion.

never miss an opportunity to quote the simpsons.

[Renai]

This has to be one of the most bullshit media releases I have ever seen about the NBN (with relation to DIDO). Can't believe politicians are able to publish this rubbish ... I mean seriously -- this is a technology nowhere near commercialisation!

http://www.sophiemirabella.com.au/Me...TERNATIVE.aspx

LABOR’S BETA SOLUTION – DENY THE ALTERNATIVE

03-August-2011

Remember those rabid beta-tape aficionados of the late 70s and early 80’s?

You know, the ones who swore they had found the one true technology and held firm to their allegiance as the video library shelves became chock-o-block full of VHS tapes and the beta tapes were relegated to a dark, dingy corner out the back before disappearing altogether.

“Beta’s better!” they would cry in frustration. And technically, it could be argued they were right. Problem was, consumers voted VHS with their wallets. And beta, despite it’s small band of loyalists, died as a mainstream technology.

Just 30 years later even VHS has gone the way of the dodo and their successor DVDs are rapidly on the way to obsolescence, with hard drive and flash disk technology offering the compact convenience consumers want.

This Government kind of reminds me of the beta ideologues.

Their commitment to spending $36 billion of your money over the next decade to roll out fibre-to-the-premises technology with the National Broadband Network flies in the face of logic.

Labor’s claim that the NBN will “future proof” Australia seems naïve in the extreme – does Senator Conroy really believe that this cumbersome cabling to homes will be the technology that consumers want in 20, 30 or 50 years? It’s unlikely to be what consumers want in a couple of years – much less the 10 it is expected to take to roll it out.

It doesn’t even seem to be what consumers want now. The two NBN test sites in Australia have been connected to some 6,000 premises, but currently the NBN has only 41 active customers. Seriously. A 0.68% take-up rate - hardly a case-study in consumer demand.

But, just like those beta ideologues, Labor is sticking to its guns. But, honestly, what good is a product – whatever its merits – if no-one is buying?

With perhaps this in mind, Labor is paying Telstra an extra $12 billion or so to rip up all their copper network so folks will have no alternative but the NBN. Even the beta extremists didn’t advocate the destruction of the all super 8 films!

Like Labor’s other big policy dud – the carbon tax – it all seems a little pointless. And expensive. And with negligible benefits.

I don’t claim to be an expert. But American Steve Perlman is. He developed Apple’s Quick-time media player software (something those original beta-heads would have been impressed by) and he’s now announced a new wireless technology – DIDO – that has the potential for internet download speeds to be 1,000 times what they are currently while not suffering the current performance degradation caused by the number of wireless users.

Importantly, it’s wireless technology – the sort that allows people to access the internet through their smartphones or laptops wherever they are. The sort consumers want.

True – DIDO is not available right now. But neither is the NBN – except to those test areas where people just aren’t joining up.

Does anyone really believe, given the massive leaps in internet technology we’ve seen in recent years, that this new improved wireless will not eventuate? In fact, it will probably be sooner rather than later.

Question is, how many billions of dollars of taxpayers money will be wasted digging up streets and laying new cables before this Government realises that the NBN is an expensive waste and consumers are voting with their wallets for more convenient, appropriate and affordable technology?


END

[HubertCumberdale]

Really? VHS vs Beta? This is the comparisons they are making now? I love the "DENY THE ALTERNATIVE" in the title, guess Sophie didn't even realise at this stage that alternative isn't even in an alpha stage let alone a beta stage (lol) meanwhile fibre is real, available and we know it works. oh she also must have missed what Conroy had to say about it:

I hope that this wireless breakthrough is developed, because wireless and fibre to the home are complementary — people will want both

So much for denying the alternative.

[Dean Harding]

I always thought the DIDO guy was more famous as being the guy behind OnLive. But then, that's a gaming thing and games are a tool of the devil. So we'll instead go back to 1985 when he was working at Apple to find something else he worked on...

[deteego]

Here is a good article about it

Has OnLive’s Steve Perlman Discovered Holy Grail of Wireless? | Epicenter*| Wired.com

Apparently its possible not because its "magic" or because it "disobeys" the laws of physics, but merely because it bypasses them.

Shannons law doesn't really apply anyways, wireless speeds with MIMO have already "broken" the speeds of shannons law. Of course this has nothing to do with the law being broken, its just that people are smart and come up with a solution where the law doesn't apply

This has to be one of the most bullshit media releases I have ever seen about the NBN (with relation to DIDO). Can't believe politicians are able to publish this rubbish ... I mean seriously -- this is a technology nowhere near commercialisation!

LABOR

She is right, there is no such thing as future proof, anyone that knows anything about history of technology would realize that every technology, has been, and will be, superseded by something else

[NightKhaos]

I actually spent time reading the patent and came to the conclusion that it is MIMO on steroids, which isn't a bad thing, it's just after reading the patent it was almost as if "Well that was an obvious inference of the technology".

Personally I'm excited to see technologies like these being developed. But I still think we need a solid fixed line infrastructure to support it.

Wireless in a lot of ways reminds me of PGMs ("Smart Missiles"). I realise that is a curious comparison but bare with me a moment.

By making smarter and smarter munitions it will be possible to use these weapons in far more desirable situations. Imagine a munition that can direct the explosion away from a certain angle for danger close strikes, or a munition that can fly itself into an bunker and find a structural weak point to bring the whole structure down. We're close to having technology now that can do this.

However, this comes at a cost. An AIM-9 Sidewinder has a unit cost of $85,000. So what is the military doing research into? "Velocitas Eradico", I am speed and I destroy. 7 pounds of ferrous material fired to 2400 meters per second. Or better yet, two tonnes of re-enforced concrete dropped from a plane.

This DIDO technology is awesome, it will allow us to do great things in communications, and will bring about a new era of mobility, but like PGMs, this technology will come at a cost. And ultimately, as the technology gets more and more complex, laying a cable in the ground is going to look a hell of a lot more attractive in a lot of circumstances.

I do however see strong indications that there is going to be a period of rapid development into wireless technology, and exploring the realms of possibly. I just hope that in that time we don't neglect our fixed line infrastructure, regardless of what happens politically.

[Jeremy Kime]

This has to be one of the most bullshit media releases I have ever seen about the NBN (with relation to DIDO). Can't believe politicians are able to publish this rubbish ... I mean seriously -- this is a technology nowhere near commercialisation!

Agreed, that article is full of fail. This quote in particular demonstrates the author's apparent confusion:

"True – DIDO is not available right now. But neither is the NBN – except to those test areas where people just aren’t joining up."

DIDO is a technology, the NBN is an infrastructure project. The technology the NBN is being built on (GPON, LTE etc) is well and truly "available" and has been for a long time. DIDO is not "available" because it hasn't even been proven to work yet, much less been proven to be commercially viable.

If the author doesn't understand something as fundamental as this, there is little hope for the rest of the article.

[deteego]

Agreed, that article is full of fail. This quote in particular demonstrates the author's apparent confusion:

"True – DIDO is not available right now. But neither is the NBN – except to those test areas where people just aren’t joining up."

DIDO is a technology, the NBN is an infrastructure project. The technology the NBN is being built on (GPON, LTE etc) is well and truly "available" and has been for a long time. DIDO is not "available" because it hasn't even been proven to work yet, much less been proven to be commercially viable.

If the author doesn't understand something as fundamental as this, there is little hope for the rest of the article.

DIDO actually has been shown to work, its just a question of getting it into practical and commercial reality

There is a lot of 'politics' around DIDO, because for example it would make current mobile telecommunications topology redundant, there are no 'base towers' in DIDO, and the model is distributed (thats what it stands for) instead of centralized with telcos (as it currently is)

This DIDO technology is awesome, it will allow us to do great things in communications, and will bring about a new era of mobility, but like PGMs, this technology will come at a cost. And ultimately, as the technology gets more and more complex, laying a cable in the ground is going to look a hell of a lot more attractive in a lot of circumstances.

There isn't anything actually complicated about DIDO, instead of having a few massive base stations you have mini base situations located everywhere, and realistically they would be inbuilt into routers

[Jeremy Kime]

DIDO actually has been shown to work

Shown to whom?

[NightKhaos]

There isn't anything actually complicated about DIDO, instead of having a few massive base stations you have mini base situations located everywhere, and realistically they would be inbuilt into routers

I'm sorry, but have you actually read the patent in question? The signal processing requirements of the system are non-trivial. I doubt very much this technology can scale very well with the technology "built into routers" because they would not be powerful enough to handle the calculations required in order to manage hundreds of clients on the channel they are operating.

[Gwyntaglaw]

So, we've had TDMA and CDMA, now this is a particularly intricate implementation of SDMA (Space Division Multiple Access).

The claims are impressive. But there is no independent verification of those claims, and no way of separating hype from sober fact.

And a few passages give me pause straight away:

This paper describes how DIDO is dramatically different than conventional wireless technology, how DIDO works, what we have running so far, and the mind-blowing applications DIDO makes possible.

It's the segue into PR-talk ("mind-blowing applications") that rings alarm bells.

And then there's the abrupt leap to spruiking Steve Perlman's own OnLive gaming service.

I'm left to register my deep skepticism. Hope, and interest, but skepticism nonetheless.

[stryqx]

I actually spent time reading the patent and came to the conclusion that it is MIMO on steroids, which isn't a bad thing, it's just after reading the patent it was almost as if "Well that was an obvious inference of the technology".

That's exactly what I thought after first reading the patent, but it's the wrong conclusion. DIDO uses the same principles as MIMO, but complements this with transmit filtering to overcome interference issues along with the APs connected to a data centre to calculate the required transmit encodings for X transmitters to talk to Y receivers.

The first commercial implementation of DIDO will be for OnLive. The issue will be subsequent implementations, as the critical part of how DIDO operates according to the lodged patents will be the software and algorithms used in the data centre. The APs and transceivers appear to purposely be kept as low cost as possible, pushing as much of the smarts as possible back to the data centre.

Oh, and Shannon's Law looks pretty safe. At the end of the day DIDO is providing a separate channel to each endpoint, courtesy of the transmit filtering to make it interference proof.

[Renai]

I was discussing this whole technology with a couple of mates yesterday. My opinion is that while it seems to have some legitimacy, it will take at least another half-decade to make its way to the technology roadmaps of the major equipment vendors (God knows they have had a hard enough time deciding on things like 802.11n and LTE). However, my mate reckons this is a fundamentally game-changing technology which could really change everything.

It will be interesting to see how things develop and if the technology gets anywhere in practice ... some technologies just don't seem to get off the ground -- for example the CSIRO's gigabit Wi-Fi:

Multi Gigabit Millimetre Wave Wireless

[deteego]

I'm sorry, but have you actually read the patent in question? The signal processing requirements of the system are non-trivial. I doubt very much this technology can scale very well with the technology "built into routers" because they would not be powerful enough to handle the calculations required in order to manage hundreds of clients on the channel they are operating.

Routers deal with such processing requirements currently, they are basically a minicomputer

If its an issue they simply put a better processor in the router, and processors aren't exactly expensive these days

[NightKhaos]

Routers deal with such processing requirements currently, they are basically a minicomputer

If its an issue they simply put a better processor in the router, and processors aren't exactly expensive these days

Really? You're going to talk to someone specialising in embedded systems and then throw the line "just put a better processor in it" at him?

Note the post above from stryqx, who also read the patient in detail. Note that he says you need to calculate the appropriate calculations in a data-centre. This is because they are non-trivial. A router with an embedded processor is not powerful enough to do this kind of work. You will need a cluster.

Current routers only have to deal with this problem over a small number of receivers and transmitters. Generally less than 10. For a DIDO network its hundreds to thousands of transmitters and receivers. Even with a dedicated decoder you run very quickly into Moore's law.

[deteego]

Really? You're going to talk to someone specialising in embedded systems and then throw the line "just put a better processor in it" at him?

Note the post above from stryqx, who also read the patient in detail. Note that he says you need to calculate the appropriate calculations in a data-centre. This is because they are non-trivial. A router with an embedded processor is not powerful enough to do this kind of work. You will need a cluster.

Current routers only have to deal with this problem over a small number of receivers and transmitters. Generally less than 10. For a DIDO network its hundreds to thousands of transmitters and receivers. Even with a dedicated decoder you run very quickly into Moore's law.

And currently routers are embedded with underclocked broadcom chips, they can easily handle the calculations required. Your making it sound like calculating the maths for wireless is a much larger bottleneck then it already is

[NightKhaos]

That's exactly what I thought after first reading the patent, but it's the wrong conclusion. DIDO uses the same principles as MIMO, but complements this with transmit filtering to overcome interference issues along with the APs connected to a data centre to calculate the required transmit encodings for X transmitters to talk to Y receivers.

The first commercial implementation of DIDO will be for OnLive. The issue will be subsequent implementations, as the critical part of how DIDO operates according to the lodged patents will be the software and algorithms used in the data centre. The APs and transceivers appear to purposely be kept as low cost as possible, pushing as much of the smarts as possible back to the data centre.

Oh, and Shannon's Law looks pretty safe. At the end of the day DIDO is providing a separate channel to each endpoint, courtesy of the transmit filtering to make it interference proof.

And currently routers are embedded with underclocked broadcom chips, they can easily handle the calculations required. Your making it sound like calculating the maths for wireless is a much larger bottleneck then it already is

Go read stryqx's, post please. You clearly haven't. As stated: DIDO will be backed by a cluster in a data-centre. The calculations for the system cannot, and will not, be done in an embedded device.

[deteego]

Go read stryqx's, post please. You clearly haven't. As stated: DIDO will be backed by a cluster in a data-centre. The calculations for the system cannot, and will not, be done in an embedded device.

I wasn't talking about the cluster in the data center at all....

[NightKhaos]

I wasn't talking about the cluster in the data center at all....

You said:

There isn't anything actually complicated about DIDO, instead of having a few massive base stations you have mini base situations located everywhere, and realistically they would be inbuilt into routers

Although the routers will be cheap and simple the signal processing requirements for DIDO are such that there will have to be a lot of work offloaded into a cluster at a data-centre. The entire work of the system will not be done by the routers.

If you neglected to mention the required cluster, I apologise, but the fact the calculations cannot be run in on the routers.

[Gwyntaglaw]

It appears that just-retired Senator Helen Coonan has drunk the DIDO Kool-Aid - see her valedictory address, in which she not only swallows the DIDO press release claims, but also refers to fibre as "sub-optimal":

Recent reports of new experimental wireless technology, for example, if proven, may well allow access to the internet up to 1000 times faster than is possible now on conventional wireless networks, without the drawback of degraded speeds with multiple users of the network. I hasten to add this is experimental technology and yes, wireless applications do need access to fibre back haul to connect to the network, but I believe the lessons to be learnt from rapid technological change are these:

There are enormous hazards inherent in picking one dominant technology, Fibre to the Home for a new ubiquitous network, when all the risk is borne by taxpayers who will likely be left with a sub-optimal network when something more efficient comes along, as it surely will over a ten year roll out.

Secondly tax payers are right to wonder whether an investment of some $50-80 billion is worth the money when there is no guarantee that a spend even of that magnitude, will future proof Australia’s telecommunications infrastructure needs for the future.

[NightKhaos]

It appears that just-retired Senator Helen Coonan has drunk the DIDO Kool-Aid - see her valedictory address, in which she not only swallows the DIDO press release claims, but also refers to fibre as "sub-optimal":

It clearly shows a lack of understanding of how DIDO actually works. The fact that it requires a cluster data-center doing matrix solving all day and everyday shows that the system is probably not going to work out all that cheap compared to GPON fibre.

I'm starting to think a better solution would be to have active nodes, but still GPON fibre (or direct fibre), but that's a minor consideration. Telco's can already afford to run direct fibre to their towers, and let's face it, with the requirements for a (centralised) cluster for everything to work with, DIDO is still going to be operated by a greedy Telco.