Two-way Radio Breakthrough To Double Wi-Fi Speeds

An anonymous reader writes "Scientists at Stanford University have built a radio that can transmit and receive at the same time on the same frequency. The breakthrough could lead to a twofold increase in performance for home wireless networks and end that annoying habit of pilots finishing every sentence with 'over.'" But you can still do it if you like. I'm not judging.

Comments

Innovative

[Lumpy]

Doing this On the same frequency is remarkable. but the gains they are claiming can be had right now by using TWO frequencies. Transmit on channel 1 receive on channel 12.. the other end does the opposite. the thing is, 90% of Ethernet traffic is not bi directional. it's packetized so their claims of DOUBLE will not be realized. when you set up a network connection from half duplex to full duplex you do not see a double in speed, just a double in capacity.

Re:Innovative

[Phreakiture]

Well, dial-up modems don't matter much at this time (save for some corner cases), but I'll take it on anyway . . .

The phone line does, actually, have 112kb/s of bandwidth, but it is divided by the telephone network to go in opposite directions. 56k in, 56k out. At the trunk level, they actually travel over separate wire pairs (i.e. if you were to get a DS0 or a T-1 or higher, you have a transmit pair and a receive pair).

As for the notion that modems do this trick already, it is completely true. There are two main differences between a modem and a radio, though. First is that the modem can reasonably expect that, under normal conditions, the signal level of what it is receiving from the other end will not change much, and that its required transmit power will not change at all. Second is that the signalling going on in a modem is all at particularly low frequencies (4kHz and down) versus those going on via wireless which will be between one and 10 orders of magnitude higher in frequency, which is a tad more difficult to operate on.

Let me take that last point and expand on it a little. It is completely reasonable to take a modulated signal of a few kHz up to maybe a few tens of MHz, sample it digitally, push it through a DSP, slap some math on it, and get some sort of accurate filtering to take place. In dealing with higher frequencies, this is far more difficult, and achieving this, I believe, is the breakthrough.

Re:Innovative

[MattskEE]

GPS and CDMA use something completely different. Spread spectrum techniques like GPS and CDMA take a signal with (for example) 1MHz bandwidth and spread that data over a 100MHz bandwidth. Now up to 100 people employing this technique can transmit over that 100MHz bandwidth simultaneously, but there is no gain in throughput because it's the same in the end as those 100 users transmitting in a 1MHz bandwidth with user 1 at 1.000GHz, user 2 at 1.001GHz, and so on. The benefit of spread spectrum is that it's hard to segregate each radio into such a small bandwidth without interfering with adjacent users. It could not be used for full duplex single frequency radio because the transmitted signal would still swamp out the received signal, unless it were combined with isolation/nulling techniques like these Stanford guys are using.

The research page for the work in this article is here: http://sing.stanford.edu/fullduplex/
They are using multiple techniques to selectively null out the transmit signal at the receiver. Their main novelty is spatial nulling of the antenna. Two antennas transmitting the same signal will have points in space where the signals destructively interfere and cancel. If they are spaced by an odd number of half wavelengths then this includes the entire line between the two antennas, so this is where the receive antenna is placed. Then they use existing analog and digital techniques to further cancel out the component of the transmitter which appear at the receiver.

Although the techniques for this are well known the trick is getting it to actually work effectively, because you need to achieve very high isolation from your own transmitter to receiver in order to avoid the transmitter effectively jamming the receiver. Their antenna nulling is apparently what gave them that extra isolation they needed.

Re:Innovative

[philip.levis]

Hey, it's just a news article. Here's the more technical stuff: http://sing.stanford.edu/fullduplex/ Short answer is the fact that the two transmit antennas are at different distances means they need a power difference in order to match amplitude at the receive antenna. This in turn limits the depths of nulls at distance.

Re:Innovative

[commodore6502]

>>>Phone modem speeds weren't limited to 56kbps by technology
>>>...was arbitrarily limited by the FCC

Completely and totally false. Digital phone lines have 8000 samples per second at 7 bits. That yields 56000 bits per second maximum. (Analog lines are limited to 33800 bps/3429 baud.) So it's a technological limitation.

The FCC imposed a *power limit* due to reports of crosstalk between lines. The power limit reduces the max speed to 53,300.

Actual information

[Zurk]

How this actually works :
The Challenge in Achieving Full-Duplex

The problem that has historically prevented full-duplex is that, when a node transmits, its own signal is millions of times stronger than other signals it might hear: the node is trying to hear a whisper while shouting. The challenge is canceling the node's own transmitted signal (shout) from what it receives (whisper). Existing approaches, such as digital cancellation and noise cancellation circtuis, can cancel some of the transmitted signal, reducing its strength, but not enough to make a node able to receive.

Antenna Cancellation

Our design uses two transmit antennas one receive antenna per node. The transmit antennas send the same data and the receive antenna is placed such that there is destructive interference from the two transmit antennas, thus reducing self-interference. Offsetting the two transmit signals by half of the wavelength causes them to cancel each other, creating a null position where the transmitted signal is much, much weaker.

Combining antenna cancellation with cancellation through a noise cancellation circuit gives ~50dB reduction in self-interference before the RF signal is demodulated and sampled to the digital domain. Digital cancellation removes the residual interference.

For more information :
http://sing.stanford.edu/fullduplex/
The actual paper (PDF) :
http://sing.stanford.edu/pubs/mobicom10-duplex.pdf

Re:"Over"?

[BitZtream]

and the fact that you've shut up, seem to work well enough. :)

No it doesn't, I hope you don't fly anymore you're making a careless mistake about something that should have been taught to you before you took the ground exam.

The purpose of an End of Transmission marker is so that everyone listening has confirmation that they received ALL of your transmission as intended. So if for some reason my transmission is cut off and it seems like just silence you as a listener know it was cut off because you didn't hear an End of Transmission marker and you can request that it be repeated.

If you think 'silence' is the way to tell, you don't need to be anywhere in an aircraft except the passenger cabin, there are a ridiculous number of airline accidents that result because of just this sort of stupidity, a fine example is a Pan-Am flight which started a takeoff roll after knowing they didn't get a full transmission from the tower ... which told them to hold until another aircraft which had to taxis back up the runway itself could clear it. About 500 people died that day because some idiot thought silence was good enough and ignored procedure which would have been to ask for a repeat. Half way down the run way, as their 747 approached rotation speed, out of the fog appeared another jumbo jet, turning off the runway right in front of them. All because they knowingly didn't follow preceedure and ask for a repeat when the cockpit voice recorder clearly shows them noticing, pointing out, and ignoring the missing End of Transmission marker.

The click when you release the mic is there because idiots like you couldnt' follow procedure so they took it out of your hands in order for everyone else that actually has a clue to be safer. Either way, your lack of understanding of why procedure is the way it is gives me a very disturbing feeling.

Limitations

[bheilig]

The signals will only perfectly cancel when they are separated by a distance that is exactly one half the wavelength. Assuming you separate the two transmit antennas by this distance at the carrier frequency, then there will be a limitation on the available bandwidth. This is because the further you get away from center frequency, and away from the ideal antenna spacing, the less destructive interference you will have (and the more your transmit signal will leak into your receive signal). So you will double your capacity for only narrowband channels.

The pdf gives actual numbers. I just wanted to point out that there is a limitation on bandwidth.

You might also think, "If I know what I'm transmitting, why can't I just subtract it from what I receive?" This has to do with the dynamic range of the receiver, which is a function of the number of bits in your analog to digital converter. You must attenuate your received signal so that you don't saturate your converter. Have you ever turned the volume up so loud that you begin to hear distortion? It's the same thing.

So you are receiving this loud unwanted transmit signal, and this soft receive signal. You must lower the volume so that you are not distorting the highest signal. This lowers the volume on the desired signal as well. You can lower it so much that your analog to digital converter is not able to differentiate between a 1 and a 0 anymore.

I think if you could have an A2D with enough bits that you didn't care if you received the transmitted signal, then you could just carefully subtract out the unwanted transmit signal. Maybe I should patent that? Meh. I'm probably wrong.

Re:"Over"?

[natehoy]

Because in any case, the pilot has finished his/her message when you hear the mic click. Surely you don't think the conversation is going to continue?

More importantly, the pilot and controller speak to each other in very precisely defined and very concise language. It's pretty obvious when one of them is done yakking, the mic click is a convenience, like the "over" used to be before all radios had mic clicks.

A typical initial approach might go something like this:

"Bangor approach, Cessna five-two-five-Lima-Charlie, 12 miles west, descending 5000 with information Sierra, full stop."

This tells the controller that:

1. You are intending to make an announcement to the controller at Bangor Center in charge of approaches (in case you fucked up your frequency, they can correct you quickly and get you on the right frequency).
2. You are a Cessna, US-registered, with tail number N525LC.
3. You are 12 miles to the west of the airport, at 5000 feet, and descending.
4. You have listened to their weather/conditions report recently, which is their update "S" (Sierra), and the letter is updated whenever the information is updated (usually once an hour). That means you already know the wind speed, altimeter settings, and preferred runway, and have adjusted all of your instrumentation and expectations appropriately.
5. You are requesting approach vectors for the currently-active runway (which you already know) and you intend to land there (full stop, as opposed to a touch-and-go or a practice approach but not a landing).

The controller will respond with something like this:

"Cessna Five-Lima-Charlie, Information Sierra current, enter 45 left downwind for runway one-eight-zero, report midfield"

This means:

1. The controller has acknowledged your presence, confirmed that you have the latest weather, and picked an abbreviation for your tail number that does not conflict with any other aircraft currently operating in his airspace. That will be your designation for the duration of your talk with this controller.
2. The controller wants you to enter the pattern at a 45-degree angle on the upwind side of the runway and call you again when you are properly established in a left downwind and abeam the middle of the runway.
3. There is no known traffic on that side of the field that will conflict with your entry, because the controller didn't mention any.

The conversation will proceed, with both the pilot and controller keeping radio use to the absolute minimum necessary to communicate what they need to say. If the frequency is really quiet, they might exchange a few jokes or snide remarks, but "over" is usually in the domain of CB radio, old timers who used to deal with really crappy radios, and bad movies.

Interruptions to what a pilot or controller is saying are obvious because of the way the language is constructed. This is done on purpose. If you say "Bangor approach, Cessna three-five..." then stop talking, you're going to hear a controller say something like "Unknown Cessna starting in three-five, please repeat, message not received." in just a very small handful of seconds.