bit-tech.net

Researchers develop terabit wireless antennas

Researchers develop terabit wireless antennas

Future wireless devices could transfer data at a rate of 1Tb/s or more, thanks to theoretical work carried out at Georgia Tech on graphene-based antennas.

Researchers at the Georgia Institute of Technology have designed a graphene-based antenna that could potentially allow for data transfer rates as high as a terabit per second over a metre range.

The system, developed by a team from the Broadband Wireless Networking Laboratory at Georgia Tech led by director Ian Akyildiz, uses graphene - sheets of carbon just one atom thick arrayed in a honeycomb structure - to create narrow strips between 10 and 100 nanometres wide and one micrometer long, forming terahertz frequency antennas. Electrons oscillating on the surface of each graphene strip, known as plasmonic waves, interact with electromagnetic waves at the terahertz frequency in order to receive or transmit a signal.

According to the team's calculations, such a terahertz radio system could transfer data at a rate of one terabit per second - roughly 2,330 times faster than 802.11n Wi-Fi. While that is only sustainable at a range of a metre or less, as a close-range data-transfer tool the team's graphene-based antenna could prove extremely useful indeed: drop the range to a handful of centimetres and the data transfer rate could reach as high as 100 terabits per second.

Such a system could be a boon for external peripherals that rely on the transfer of large quantities of data. A high-definition video camera, for example, could dump all its footage in under a second just by being placed near a laptop or desktop equipped with the team's antenna, or a smartphone quickly download rented or purchased films for on-the-go viewing.

Sadly, as is often the case with such 'breakthroughs' involving the wonder-material graphene, the technology is far from a commercial reality just yet. '[The team's work] points out and provides a set of classical calculations on estimates of sizes and performance: it points out that there is something worthwhile here' explained Phaedon Avouris, IBM research fellow and graphene expert, in an interview on the subject with MIT Technology Review. It doesn’t solve the whole problem, but points out an opportunity.'

The team's work, which has up to this point been purely theoretical, will need proving with a prototype device - something Akyildiz claims is due for unveiling before the end of the year - and then the antenna will need to be mated to other high-performance hardware in order to reach anywhere near the terabit speeds promised. However, with the research due to appear in the IEEE Journal of Selected Areas in Communication in the coming year, those are problems that are likely to get many eyes eager to help make the next big breakthrough in high-speed wireless communications possible.

14 Comments

Discuss in the forums Reply
SpAceman 5th March 2013, 11:40 Quote
Getting an IC to send/receive the data is going to be a nightmare.
Expect silly amounts of buffering.
Bede 5th March 2013, 13:23 Quote
+1, hard drives of all varieties are still the major bottleneck here - this may be able to transfer at multiple terabits per second, but if there's no memory capable of reading or writing at that speeed this is pretty irrelevant.
law99 5th March 2013, 13:28 Quote
It's a bit like saying the wheels on my car are rated to 200mph.

Yes, that is as much as the tyres can take... but the car won't go past 88mph without setting fire to the ground and causing a whole heap of additional confusion. Got pretty weird between me and my Mum also.
CraigWatson 5th March 2013, 13:35 Quote
Quote:
Originally Posted by law99
Yes, that is as much as the tyres can take... but the car won't go past 88mph without setting fire to the ground and causing a whole heap of additional confusion. Got pretty weird between me and my Mum also.

lol! +rep for the reference ;)
Gradius 5th March 2013, 14:54 Quote
Cancer anyone? I don't like wireless one bit, much less the mobile phones (even more cancer).
Gradius 5th March 2013, 14:56 Quote
Hence, IBM already put information than 1THz = kills and mutate your DNA !
Blackshark 5th March 2013, 17:26 Quote
As the transfer (theoretical) rate increases, the efficiency decreases. Hence our new 802.11AC gear is only 20 to 30% better than the best N gear - but should be 2 to 300% better - at least if I listen to the marketing teams at every single manufacturer.

So by the time we get up to Terabit connection - it will likely only be 100% better than 802.11N gear (3x3). Ill keep that thanks.
Nexxo 5th March 2013, 18:01 Quote
The only drawback is that if you walk through the beam you spontaneously combust.
azazel1024 5th March 2013, 19:18 Quote
The heck with main storage being a limitation here, memory interface with be a limitation here. There aren't all that many video cards running GDDR5 that can push beyond 1Tbps data rates, let alone 100+Tbps.

Your average computer only has around 25GB/sec of main memory bandwidth...roughly 1/5th what this thing is "supposedly capable of".

I think the other issue you are going to have is this is likely to be ridiculously LoS, not simply uber short range. So basically you are likely to have to line this up pretty precisely AND be at super short range. I guess I could see it as a physical connector, but with no actual electrical contact (IE something kind of like a mag safe connector, but instead of electrical contacts, antennas in the "connector").

I mean, I guess I can see how it could be super useful...but I also fail to see how fiberoptics aren't potentially as fast or faster (hint, they are, they run at higher frequencies than terahertz radiation) and you basically need a physical connector with this "invention".

I mean, I can see a handful of applications where you litterally cannot interface directly, but need to have at least a few cm gap. However, as a replacement for wires...it won't be as you'll still be able to push more through a wire probably both in theory and in fact (higher frequencies possible, easier to multichannel it by bundling multiple wires or fibers). And of course the obvious that it only works at roughly a meter or less...and likely you are going to need a surface antenna on receiver and transmitter and likely can't have ANYTHING inbetween as it likely won't be able to penetrate even a piece of paper (maybe glass or another substance that is opaque to terahertz radiation).
azazel1024 5th March 2013, 19:34 Quote
Quote:
Originally Posted by Blackshark
As the transfer (theoretical) rate increases, the efficiency decreases. Hence our new 802.11AC gear is only 20 to 30% better than the best N gear - but should be 2 to 300% better - at least if I listen to the marketing teams at every single manufacturer.

So by the time we get up to Terabit connection - it will likely only be 100% better than 802.11N gear (3x3). Ill keep that thanks.

Well 802.11ac gear is 5Ghz stuff and the best 802.11n 3:3 40Mhz stuff I have seen can only push 180-200Mbps on an ideal day, downhill with a tail wind (the best 2.4Ghz stuff 3:3 40Mhz I've seen something north of 200Mbps, around 240 is the best I have seen...again, downhill with a tail wind). The early 802.11ac 80Mhz stuff I have seen (claimed 1300Mbps speeds) seems to hit roughly 340-360Mbps at best, which is about an 80% increase in speed compared to the best 802.11n 5Ghz routers I have seen with only a claimed increase in speed of 288%.

So, your comparison isn't completely out, but the gains are somewhat better than what'd you mentioned...and this is also only the first generation of 802.11ac (draft). I'd suspect, even without some of the "exotic" 4:4 and 160Mhz setups future 802.11ac will bring, we will likely see somewhat better routers and clients come down the pike even for just regulary 1300Mbps 802.11ac stuff. I wouldn't expect miracles, but it might push 400Mbps in the end.

Anyway, you get roughly 1/3rd of the claimed increase in performance (claimed, 288%, actual about 80%).

Still and all, I'd take over 300Mbps in halfway decent conditions!!! My current 802.11n 2:2 40Mhz router and laptop Wifi card can only hit about 140Mbps actual data rates downhill with a tail wind and I consider that pretty decent.

I just wish they'd bring some of the improvements of 802.11ac to the 2.4Ghz realm. Sure, you can't do 80Mhz in 2.4Ghz (there isn't enough spectrum), but there isn't really anything stopping you from running 4:4 spatial streaming or doing 256QAM encoding (802.11n is 64QAM). Just 256QAM gets a theoretical bump up from 150Mbps per stream to 200Mbps per stream. I don't know how well that bit tracks on real vs theoretical speeds...but I'll take anything I can get in 2.4Ghz.

My current spread is a 1 acre property and I need at least 2APs to cover the thing (well, with better APs/Antennas. Currently my 2 cover the house really well, but outdoors the coverage extends to most of my front yard and only about 50% of my rear yard). That is in 2.4Ghz. 802.11ac is supposed to be pretty good with range...but I bet it could be better doing 2.4Ghz and some of the stuff from the 802.11ac specs.

Though...maybe I am mistaken though. Maybe 802.11ac APs and Wifi cards CAN do some of the 802.11ac trickery in 2.4Ghz (though obviously limited to 40Mhz). Bah. Whatever.
abezors 5th March 2013, 21:40 Quote
Less hating on graphene please! It's turning out to be *the* material of the future. The amount of new and existing roles it can improve or invent are growing every month.
ch424 6th March 2013, 00:27 Quote
Quote:
Originally Posted by azazel1024
The heck with main storage being a limitation here, memory interface with be a limitation here. There aren't all that many video cards running GDDR5 that can push beyond 1Tbps data rates, let alone 100+Tbps.

Your average computer only has around 25GB/sec of main memory bandwidth...roughly 1/5th what this thing is "supposedly capable of".

... have you never heard of Moore's law?

Most modern CPUs push far more than 1Tbit/sec around. Once this technology actually reaches production quality, there'll be no problem saturating it.
Glix 6th March 2013, 01:41 Quote
My USR 5462 still works. :>

My biggest gripe is signal fluctuation and crappy aerials. :( What's the point of all of the bandwidth if it drops to 10kb/s because of a dead zone or aerial cable breakage.
SlowMotionSuicide 6th March 2013, 09:25 Quote
Quote:
Originally Posted by azazel1024
Wireless transfer bonanza.

I'd be perfectly happy for a wireless technology that wouldn't crap out on me once or twice a day. :(

As it stands my wireless network coverage and signal strength is excellent everywhere in the household, except where I decide to place the laptop. And the router requires hard reset at least once a day.
Log in

You are not logged in, please login with your forum account below. If you don't already have an account please register to start contributing.



Discuss in the forums