# Full Duplex Radio Communication On a Single Frequency

I am an ELECTRONICS final year student and I am working on a project, "Full Duplex Radio On a Single Frequency". So is it possible to design such a project. If so then give me the idea about it, how can i design such a circuit, what do i need for it (mircrontroller, raspberry pi etc)? I have some idea about it but i don't know how to implement it.

If you have a single frequency with a single antenna, the best you can do is half duplex on the RF side -- you can't transmit and receive at the same time on the same frequency. If you have two antennas sufficiently separated (on both ends), you might be able to use signal diversity and geometry to separate signals; this would work better on higher frequencies (>5GHz?)

However, an easier way to do this would be to do timed half duplex (the two ends take turns transmitting), and digitize, compress, and packetize audio so that the audio is full duplex, even though the RF is only half duplex. TDMA?

• Yep, time-share is the only option. This could actually be done with analog equipment, similar to Hedy Lamar's version of spread spectrum (timed frequency jumps on both ends). With high enough switching rate and appropriate filtering, the operators wouldn't hear the switching... – Zeiss Ikon Nov 22 '19 at 13:04
• This is actually wrong. You can do same-frequency full-duplex communications – that's not easy, but it's established technology! In essence, you use an actual circulator, or "build one in software" by calculating the effect of your transmission and subtracting it from the received signal. That works, and can be shown to be arbitrarily good compared to a self-interference-free channel in both directions. – Marcus Müller Nov 24 '19 at 17:22
• That sounds messy and difficult! Also, this would only work if the received signal and transmitted signal were at sufficiently similar power levels. A cursory search shows circulatory isolation seems to be <20db. The best SDR might give you 100db. Could work on a transmission line, but I think it'd be limited over the air. Interesting point though! – user10489 Nov 25 '19 at 1:55
• @user10489 it's messy, but not overly difficult. Self-cancellation is a pretty mature problem. You're overestimating how limited this is on the air – friend of mine whipped up an over-the-air demonstrator in little more than a day to win a transmission rate contest at a conference. This stuff works, and isn't limited to lab settings! And: it's important for frequency duplex systems, too, which only in theory don't have the problem of TX crosstalking into RX. Modern transceivers do this all the time! – Marcus Müller Nov 26 '19 at 10:41
• Don't forget that you might accept that your doable data rate take a hit – as long as that hit is below 50% (which would be the hit it took if you did, say, time-sharing or frequency division multiplex). And that can be pretty doable; sure, for example, in a tradeoff system, your weaker, i.e. RX signal gets a worse quantization, sure, your TX signal can't use the maximum TX power, but if that's still better in total than just doing time sharing, then you win. – Marcus Müller Nov 26 '19 at 10:51

Until the last 20 or so years, this would have been extremely difficult to accomplish in the analog domain. With advances in processing power applied to DSP however, Spatial Multiplexing is a reality -- that is, the ability to send different signals on the same channel simultaneously. It is done by having separate, and multiple TX/RX radio chains -- each carrying different information -- separated by some distance causing different "paths" through reflection, refraction, etc of each link... processing is able to distinguish one stream of data from another through characterizing each path-channel, yet they occupy the same frequency. It is not entirely an intuitive concept. Lots of math.

Most of us have already experienced something similar (though not exactly the same) with MIMO (multiple in multiple out) technology that came with the 802.11n standard for WiFi. Those routers you may have with multiple antennas (like 4, or 6, or 8....) are using a form of MIMO. This significantly increases channel capacity, and hence throughput on a single frequency -- so you can happily watch Netflix on your laptop.

Here is an example of a project done by Stanford on employing this technique on Full Duplex / Same channel. Its a bit of a slog to read, but you're a student -- so you might actually like it :)

https://web.stanford.edu/~skatti/pubs/nsdi14-mimo.pdf

Also, here is a brief tutorial on MIMO -- implemented for WiFi -- discussing some of the basics around spatial multiplexing. Cool stuff.

https://www.electronics-notes.com/articles/antennas-propagation/mimo/what-is-mimo-multiple-input-multiple-output-wireless-technology.php

• This is exactly what I was referring to with "signal diversity" or antenna diversity. However, the comments on my question indicate there is yet another way.... – user10489 Nov 28 '19 at 18:18
• Transmit and Receive diversity are different than MIMO and are actually a form of SISO. MIMO is sending different information on the same channel simultaneously, whereas diversity is transmitting or receiving the same information on multiple transmitters and receivers. Path diversity is definitely the root to both however. – W4QA Nov 28 '19 at 21:44