German researchers are claiming a world record, using a 237.5 GHz carrier and photonic mixing to achieve a 100 Gbps wireless link. Don't throw away that WiFi kit just yet, however: while the reach, at 20 metres, is good enough to cover most household applications, the setup is a little bit exotic at this point in the system's …
This will turn into a solution for certain problems, e.g. "microwave links", however once you lay down a fibre, you can get 100 GBits with off the shelf technology _today_.... per wavelength. And you can do around 100 wavelengths over such a fibre today.
So while it is a solution for certain problems, and certainly can evolve into a stopgap solution, it still doesn't keep us from finally getting off our collective asses and laying a dedicated pair of fibres to every household.
Re: Well yes
Agreed with everything you say BUT fibre requires the opening and closing of the roads and pavements - which is a long and costly process.. especially when it has to be repeated every 10 or 15 years..
Re: Well yes
FTTN is optimal in the long run. By being supplied by a dedicated physical medium It avoids many issues experienced by wireless and it's shared medium, which is in simple terms a range of frequencies beamed into the air. Interference from electrical and mechanical equipment and not to mention the ever present and growing issue of interference from competing wireless networks sharing the same frequencies or even the issue of more users on the same access point and wireless channels = more latency and less throughput...
Wireless, even though in a lab can produce incredible throughput, will suffer the "shared medium" issue in real life deployments. Something FTTN will forever be unaffected by.
Concerning Existing WiFi implementations in populated areas, like a busy street in bangkok you will find over 30 competing and overlapping wifi signals in the 2.4GHz band, as population densities grows the clutter ever increases and newly opened frequencies like the 5GHz band will soon become crowded, especially if using max bandwidth multiple channels.
Opening and closing of roads
Yes, it's a costly process, but it can be done.
Where I live, the power co. had decided to move the supply cables from masts to ducts in the ground. Since they had the diggers out and the roads open anyway, they put down empty tubes as well, and everyone in the village got a letter asking "Sir, in preparation for FTTH, should we lay down a tube from the road to the foundation of your house free of charge?". Sure, go right ahead. So they did.
Later, the power cos established a joint venture service provider, and to all the households where they had laid down tubes to the foundation, they now offered free installation of the fibre and the terminal gear if you subscribed to the service. Which I had anticipated, so I got 30/30 Mbit/s internet (or better - it's usually at 37/34) on top of a bunch of TV channels and telephone. And I saved on the overall monthly fee.
Sure, the fibres have a life span of 30 years, but they are in tubes, so no need to dig up the roads and pavements when the time comes.
Re: Well yes
1. Considering our current wiring was laid down starting in 1900, chances are good that dedicated fibres will last for another 50 years.
2. There is a new and exciting invention called the "tube". It's a hollow enclosure for cables which makes it easy to replace them without opening and closing any roads.
The only practical problem is joining fibres. That's harder than joining copper, but I'm sure there are decent plug solutions which would work on a decent scale.
@JosefMartin (Re: Well yes)
Pretty sure you won't find anyone else in your neighbourhood transmitting at 237.5 GHz. Not yet, anyway, given what is required to generate the signal. It would make sense and be quite simple for the government to restrict use of that frequency to node-to-home connections.
Re: Well yes
The link is *very* directional line-of-sight with a beam width of a few degrees, like a flashlight -- it's intended for a small number of backhaul point-to-point links, not broadcast radio with many potentially interfering transmitters.
Stating that FTTx is optimum in the long run is obvious, until you're in a small village where 1km of road needs digging up to reach a few subscribers. In this case a small dish at each end would be a lot cheaper.
Another possible application is links between the large number of microcell basestations needed for widespread 4G/5G access in cities. Digging up city streets to install the fibre needed to link these -- because many more basestations will be needed than at present to get the network capacity -- is also a lot more expensive than point-to-point radio links between rooftop dishes.
“a narrow-band terahertz carrier is photonically generated by mixing comb lines of a mode-locked laser in a uni-travelling-carrier photodiode. The uni-travelling-carrier photodiode output is then radiated over a beam-focusing antenna. The signal is received by a millimetre-wave monolithic integrated circuit comprising novel terahertz mixers and amplifiers.”
A whole paragraph of awesome!
I've had an accident.....
Far off from scrapping the need for copper/fibre links. But it looks like it could be the start.
What about laaaaaaaaaaaaaaaaaaaaaat'ncy?
This technology will be great in niches. It won't beat fibre for throughput, reliability, or durability. Could come in handy for instant or hard to cable point-to-point application though.
This sort of technology is more or less line of sight, subject to rain attenuation, chugs relatively high power relative to fibre, but the real usage lid is interference between systems. There's only one radio spectrum available and carving it up spatially becomes increasingly expensive with more users.
- Leaked screenshots show next Windows kernel to be a perfect 10
- Product round-up Coming clean: Ten cordless vacuum cleaners
- Something for the Weekend, Sir? I need a password to BRAKE? What? No! STOP! Aaaargh!
- Episode 13 BOFH: WHERE did this 'fax-enabled' printer UPGRADE come from?
- Vulture at the Wheel Ford's B-Max: Fiesta-based runaround that goes THUNK