As the 3GPP committees ramp up, I have seen a renewed push for “full duplex” communications. That’s a shorthand way of saying that the same spectrum is used for uplink and downlink, on the same antenna, at the same time.
Full duplex has been pursued by the R&D community for at least 28 years. I goofed around with this idea in 1998-2001, and at the time, I concluded that it was going nowhere in the mobile market. Using techniques including both analog and digital cancellation, we were able to achieve two-way communications using narrow CDMA channels.
But the problem was that any abrupt change to the radio frequency (RF) conditions threw our correction loop out of whack, and the algorithm required about 60 seconds to re-converge. I made the difficult decision to abandon the project because I could not see any way to improve that time by eight orders of magnitude. It seemed to me that fixed wireless communications might be able to use the concept, but mobile links could not.
How to duplex radio signals
There are multiple ways to duplex radio signals. We have frequency division duplex (FDD), where the same antennas are used for transmit and receive functions. Uplink and downlink are separated into separate frequency bands to avoid self-interference. Then, we changed to time division duplex (TDD), where the same band is used for uplink and downlink, with time slots allocated to each. The term “full duplex” indicates that the same antenna is used for uplink and downlink, in the same band, at the same time.
Today, computing power is much better than my old 2001 prototype. I’m sure that a modern correction loop can converge much faster. Qualcomm recently demonstrated a network in FR3 frequency bands that I’m sure is much better than my old experiments. But I would like to point out that the Qualcomm demo is not actually full duplex. Instead, I would call it “spatial duplex”.
Recent demos and the ongoing 3GPP discussion are based on separate antennas for the uplink and downlink functions. Using a separate array for each link direction adds at least 40-50 dB of isolation, making it possible to solve the digital cancellation challenges.
But here’s the problem: Even at 6-8 GHz, this approach makes the antenna bigger.
Every inch of space on a tower costs money in monthly rent. Bigger antennas can cost thousands of dollars for the operators every year, so the 3GPP committees should think carefully about adopting this approach. Will the additional capacity justify the additional expense?
Spatial duplex could be a very useful technology to boost capacity in FR3 links. Imagine a 400 MHz wide channel at 7-8 GHz, with a 256T or higher order Giga-MIMO antenna array.
The spatial duplexing impact
With spatial duplexing in one or two 100 MHz sub-bands, this 6G anodeB could handle a tremendous amount of capacity — around 20x the capacity of a typical 5G channel, and even more in the critical uplink.
On the other hand, spatial duplexing will have very limited use in lower bands, because the additional size of the antenna arrays will be economically infeasible. So this is one 6G technology that is unlikely to migrate down to existing frequency bands.
I will be covering the capacity vs. cost question in an upcoming report, because I believe that cost and business factors should be fully considered before the 3GPP committees adopt a new technology. Please contact me if you have input on the business aspects of full duplex and spatial duplex communications.
Joe Madden is principal analyst at Mobile Experts, a network of market and technology experts that analyzes wireless markets. Disclaimer: Nokia is a client of Mobile Experts.
Opinions from industry experts, analysts or our editorial staff do not represent the opinions of Fierce Network.