Click here to go to our main radio page
Click here to go to a page on full duplex transmit/receive (new for June 2020)
New for January 2020! This is just the start of a STAR page, if this is your field of expertise please send us some content and/or references, like Gavin did.... As is the case on all Microwaves101 pages, please call us out if you see any incorrect information here.
Here we are talking about transmitting and receiving communications signals at the same frequency... sometimes called SF-STAR (single-frequency simultaneous transmit and receive). Note that it is not a huge trick to do this if TX and RX are separated in frequency. When you hear the term 'half duplex", it refers to a system that can transmit and receive, but not at the same time. When you hear the term "full duplex", the system is capable of simultaneous TX and RX operation.
There may be some contemporary work in STAR for radar, but that would be like a weird cousin that you try to put on the edge of a family photograph so you can crop him out later. Note that the original HAWK missile system of Cuban missile crisis fame features STAR in an FM-CW radar format. Click the video below and advance to 60 seconds to see the TX/RX antennas side-by-side.
Homing all-the-way killer (HAWK) missile system
There are plenty of non-microwave examples of STAR systems. That yard light that goes on after dark? It has a "receiver" to sense light, that must be isolated from the unit's self-lighting, otherwise the feedback can cause the light to twinkle, or even pulse on and off. A bull horn receives and transmits sound, the microphone is directional, and so is the speaker (in the opposite direction).
Half-duplex radio communication systems' history
The Walkie talkie (really, it was called the "Handie-talkie") was invented in 1941 by Motorola, and is the first example of a half-duplex radio communication system with shared antenna. The user had to push the talk button to transmit a signal, and if both users transmitted at the same time no one but a third listener would hear anything. Thus, the user had to get used to saying "over" when done transmitting. "Over and out" meant we are done with the whole conversation. Interestingly, the first hand-held radios use frequency modulation, invented by Edwin Armstrong. Armstrong's efforts to get broadcast radio to use FM eventually caused him to commit suicide in 1954, but at least he got to see his baby widely adopted by the US Army.
By the 1960s, the advent of cheap transistors made hand-held radios (half duplex) available in toys. Below are two commercials for these products, listen for the word "over" to indicate that transmit is being switched off.
In the Dick Tracy model, you had to have some wires go from the belt-mounted unit to the speaker/microphone you wore on your wrist.
Dick Tracy two-way radio, circa 1961
In the Star Trek communicator, at last you had a complete hand-held unit. This dates to perhaps 1974, five years after Star Trek went off the air.
Scott this is Kevin, my Bike is broken... can you help me? Over... Circa 1974
Back to the subject at hand. The real problem is, how do you prevent the transmit signal from jamming the receive channel? Depending on system parameters, you might need 80 dB isolation between transmit and receive channels. There is no single magic wand that will do this, you have to break it down. Also, you need to consider how much bandwidth your system needs. Solutions that work at a single frequency will fall off over a wide bandwidth. Here's a thesis on a wideband STAR system
https://etd.ohiolink.edu/!etd.send_file?accession=osu1499802814937457&disposition=inline
Here's a public domain paper that provides a good overview of STAR, and a nice list of references for you to chase down. And it describes how a 18- degree rat-race coupler can be used with a patch antenna to get over 40 dB TX/RX isolation.
http://www.jpier.org/PIERC/pierc78/03.17051205.pdf
Circulator
A circulator between TX, RX and antenna can easily provide 20 dB isolation, but isolation will decrease depending on the return loss of the antenna.
Polarization
If you transmit in one polarization and receive in a different one, you can get maybe 20 dB isolation. Typically, you would use circular polarizations, with left hand in one direction and right-hand in the other.
Separate antennas
Yes, this is could be considered a form of cheating, but what better way to get isolation than distance? Go back and look at the HAWK video at the beginning of this page to see how your grandparents made a STAR radar back in the 1960s.
Y0u can also improve isolation by adding absorptive shielding between TX and RX antennas.
Analog cancellation
The idea here is to sample off a coherent copy of the transmit signal and subtract it from the received signal. This is an analog circuit, and is in front of the LNA so loss really matters. Maybe you can get 30-40 dB reduction here.
Kumu Networks (kumunetworks.com) claims they can do analog self-interference cancellation at the receiver of 55dB using their analog adjustable finite impulse response (FIR) filter chip. All you eat is the loss of a coupler. Thanks to Chris!
Digital processing
In a phased array you could try to use one sub-array for transmit and another for receive, like MIT did in ths paper:
https://archive.ll.mit.edu/publications/technotes/TechNote_STAR-array.pdf
Phased array
A phased array may be naturally beneficial for STAR, as the power per element is reduced and the use of multiple LNAs can be used to increase the linearity of the system
Voice, or data?
A final note from Chris on where this is all heading:
5G will all be orthogonal frequency division multiplexing (OFDM) modulations for multiple-access on a basestation just like existing LTE, and I believe will have completely separate up and down bands to simplify the non-basestation hardware.
It’s not that we don’t care about voice anymore… is that voice is data now. In fact, did you know that audio from a phone call is not even transmitted as a representation of the audio waves at the microphone? It’s not like they just have some ADC feeding the transmitter with a bitstream. To compress the data more for transmission, phones process the microphone input and represent the sound as a combination of tonal “syllables” which are then encoded into binary data. Each syllable in the phones library is a sort of audio sound that humans can make. This allows human speech to be represented by the fewest number of bits. It’s also why listening to music and dogs barking over a phone sounds so terrible. We also have “HD” audio now where they expand the syllable library to improve fidelity. Faster connection? Less compression!