The D-Band May Be the Path Forward for 6G ResearchMay 23, 2020 by Nicholas St. John
For now, 5G is on the main stage. But behind the curtains, 6G might not be as far away as we think.
While the rest of the world is getting their hands on 5G technology, researchers are laying the groundwork for 6G. This will surpass the current capabilities of 5G with more than 10 Gbps data rates and 1-millisecond latency. It may also connect to 100 times more devices than 4G LTE and offer 1,000 times more the bandwidth along with other improvements, noted by Thales in a review of 5G technology.
The 6G spectrum and KPI targets. Image used courtesy of 6G Research Visions
Now, we know what 6G needs to satisfy at a minimum, but how are we going to do it? To answer this question, researchers are looking into frequency ranges and the circuits that will be transmitting and receiving these signals.
The D-Band and 6G
According to a paper written by the electronics laboratory CEA-Leti, communications via waves in the millimeter band (mmWaves)—specifically the D-band—are the channels that 6G technology will travel. The D-band is just a specified set of frequency ranges for wireless communication. In a presentation on D-band, Nokia outlines these ranges:
- 130–134 GHz
- 141–148.5 GHz
- 151.5–155.5 GHz
- 155.5–158.5 GHz
- 158.5–164 GHz
- 167–174.7 GHz
These all total up to 31.7 GHz of frequency to be utilized for 6G—that is a lot of bandwidth! This is compared to the approximately 13 GHz of bandwidth that 5G has, according to Qualcomm. This upgrade may enable more connections per unit area. While these frequencies are open, physical barriers will have an effect on these waves. CEA-Leti suggests that high-gain antennas and circuitry are needed.
Circuitry Up to the D-Band Challenge
The University of Grenoble created a design with such a high-gain antenna. This antenna module utilizes some of the D-band, operating between 114 GHz and 138 GHz with a maximum of 25 dBi gain and a minimum gain of 22 dBi.
The circuit aims for ultrafast short-range communication with an integrated frequency multiplier. This multiplier is comprised of a series of self-mixer circuits and amplification stages. The self-mixers double the frequency of the output with respect to the input signal. Below is the block diagram for the antenna driver circuit and the circuit for an individual mixer stage.
Antenna driver block diagram and layout. Image used courtesy of Francesco Foglia Manzillo et. al
In regard to the total antenna driver circuit, the part labeled “input balun” is where the differential signal comes into the first mixer stage. The active part of the circuit (the transistor circuitry—the orange part of each block) for both the mixer and amplifier stages are the same sizes, but the output inductors and transformers (the green shape in the layout picture) are optimized in each stage for the frequencies they are dealing with.
So, we see that the inductors and transformers decrease in size as the frequency increases. Also, while the first amplification stage is a single amp, the corresponding stages are cascaded amplifiers for double the gain per stage. All of these amplifiers are identical, common source pseudo-differential stages.
NMOS Transistors Optimize Frequency Response
Now, we can see the mixer schematic implements only NMOS transistors. This is because CEA-Leti reported concern about the CMOS circuitry’s ability to respond to the high frequencies of the D-band. The mixer circuit implements two transistors for the differential input (M5 and M6), connected in a pull-down fashion to differential pairs of NMOS transistors (M1 and M2; M3 and M4).
Self-mixing stage schematic and layout of a coupling transformer. Image used courtesy of Francesco Foglia Manzillo et. al
M1 and M3 are connected to the one side of the output transformer, while the second is connected to M2 and M4.
Now, since M5 and M6 are controlled by differential signals, M1 and M3 conduct at opposite half cycles of the input wave. This means that for the first half cycle, M1 is closed and M3 is open, and during the second half cycle, M3 is closed and M1 is open.
This same thing is happening to M2 and M4. The output nodes of M1/M3 and M2/M4 combine for a differential signal and is twice the frequency of the input because the transistor combinations are on and off at different half cycles. The mixer then couples via the transformer to the following amplification stage.
All of the NMOS transistors in this circuit (mixer and amplifier) are low threshold transistors, optimizing the frequency response of the circuit.
How Will 6G Move Forward?
The D-band and novel circuit designs like the antenna circuit can be a good initial framework for 6G technology. While we have an idea of where 6G signals will travel, there is still a long way to go. We'll still need to use this band for longer range communication and consider the challenge of large-scale integration.
SIRADEL's ray-tracing tool has helped CEA-Leti develop an electronically steerable antenna to avoid outdoor objects that block THz wireless performance. Image used courtesy of CEA-Leti
The D-band has the space and capability to surpass 5G by a longshot, and new transistor circuit designs will allow such high gains for antenna transmission and receiving, which will allow us to communicate over high frequencies.
For now, 5G is on the main stage, but behind the curtains, 6G might not be as far away as we think.