When an August 3GPP meeting on 6G standards concluded that 6G would initially be based on the OFDM waveform that underpinned 4G and 5G, adding that further studies would continue into certain further enhancements if required, the overall direction might have looked like bad news for companies that have long been proposing alternative waveforms for 6G.
One of those companies is Cohere Technologies, which has been proposing the OTFS (Orthogonal Time Frequency Space) waveform as a candidate since before 5G was standardised. While the industry has not settled on OTFS as either a 5G or (yet) a 6G technology, the company has made some commercial progress adapting the fundamentals of OTFS for use as a spectrum optimisation tool on top of any OFDM-based base station.
CEO Ray Dolan says that this work will be deployed at large carriers “next year.”
Enter the Pulsone
Now the company is announcing the commercial launch of a new concept that it is calling Pulsone – so-called because it integrates the “pulse” of a radar with the tone-based technology of OFDM.
First mentioned to TMN by Dolan in this video, Pulsone is the name Cohere is giving to its commercialisation of the Zak-OTFS Waveform as a foundation of Integrated Sensing and Communications Networks, and for 5G and 6G Non Terrestrial Networks (NTNs). (Zak in OTFS refers Zak transform: the conversion of delay doppler information to the time domain that is sent over the channel.)
Neural receiver trial with Nvidia platform
Along with Virginia Tech and Duke University, the company is also set to demonstrate a demo of the waveform with a neural receiver that works in Delay-Doppler channel environments. The demo, which will take place at Nvidia’s GTC Conference and uses Nvidia’s Jetson GPU platform, will show live video and a simultaneous radar view of the wireless channel. While this receiver marks a new step for Cohere’s away from the Intel FlexRAN platform, the neural receiver is different in another way from other neural receivers ; it doesn’t operate using offline training.
Anton Monk, SVP of Strategy, Cohere Technologies, said, “One of the biggest concerns that the industry has with these very cool neural receivers is that you train them in some environment and that’s great, but what happens if something changes? This particular receiver is low complexity: it can work for OFDM and OTFS using the same structure and it converges within a single symbol time.”
“The other thing that is new is that for integrated sensing one of the biggest questions is how much overhead you are adding to do the sensing? And what we’re showing is a data driven, data directed, sensing capability that has no additional overhead, no capacity overhead because of the stability of OTFS. Operating in the sensing domain is very stable and so you’re able to do sensing using the data in the network, the inherent data. So those are two unique things that we’re showing, as well as OTFS in a real Nvidia platform.”
Dolan added the company would have working prototypes of the technology in the first half of 2026.
We think there’s significant potential, huge interest from the defence sectors and other public safety and national security sectors
OTFS and 6G
So is there still a path forward for Cohere’s OTFS? Dolan and Monk pointed to a few positive factors.
One is the large amount of US government dollars that are being thrown at defence applications for new sensing technologies, including the putative Golden Dome missile defence system. Aside from that grand ambition, military uses of ISAC could be for uses such as drone swarm detection, or for deploying undetectable comms. And it won’t hurt here that CohereTechnologies is a US company.
“We think there’s significant potential, huge interest from the defence sectors and other public safety and national security sectors,” Dolan said.
Then there’s the development of ISAC as a 6G use case, with potential uses in areas such as industrial automation and automated driving, which is an example of ISAC’s capability for situational awareness, although Dolan concedes that business models in the mobile operator space are still to be proved.
Monk said, “I believe that national security interests can drive adoption by operators globally in a way that URLLC and industrial IoT could not in the past. When national security interests come first and foremost, as they are now, then I think Government plays a bigger role than they have in the past. They haven’t really been able to influence standards at all in the past, but you can see it now. I don’t think there’s a clear answer here, but things are changing with respect to ISAC.”
Another positive is that, according to Cohere, Zak-OTFS is a much better candidate for building Integrated Sensing and Communications (ISAC) than OFDM is.
Monk said, “In sensing we’re concerned with metrics like range resolution, number of targets, and there’s theoretical research that shows that OTFS can detect four times the number of targets and four times better resolution. That’s significant.
“That may not have been that important in the past when radars are looking for one missile or an airplane but what’s changed in the defense space is the threat of drone swarms, and in the Public Safety Safety space, detecting many numbers of pedestrians, many numbers of cars. So it’s not a traditional radar and these new sorts of metrics become really important. We know we’re convinced, but we have to prove it with real hardware. And that’s the announcement – that we’re building this in collaboration with partners.”
Structurally, Dolan and Monk add that OTFS can be implemented on top of OFDM, as a phased approach.
If, in fact, our Zak-OTFS has significant benefits relative to an OFDM outcome – and we think it gives much higher resolution, much lower complexity – we’re going to show that by demonstration, not by argument. We’re going to do it very soon.”
“If the industry chooses to stay in an existing derivative of OFDM, we can form that using our Pulse zone as well,” Dolan said. “We have modes where we can operate over OFDM in a very simple way. For the full-fledged Zak-OFTS Pulsone, we can implement the waveform on top of the existing 5G paradigm; meaning the same scheduler that’s used for 5G can be used for OTFS. We’re not saying that the receiver or the transmitter is going to be the same, but we have ways to sit on top of the standard and prove advanced capabilities and much higher performance, not just for the waveform and communications, but also for sensing.”
Monk explained further, “OTFS is a mathematical generalisation of existing 3GPP waveforms OFDM and TDMA. It represents those sort of two extremes – pulses which are time domain and tones or phases which are frequency domain. So in addition to its more general ability to work in complex channels it can collapse to either of those two domains; TDMA or OFDM. We showed that through a very simple pre-coder we could generate either of those two, providing a phased approach for operators to adopt it [OTFS]”
That said, Dolan said that Cohere is going to bring its story to the standards bodies, insisting on the benefits of the waveform both for communications such as NTN, and for ISAC.
“We’re going to try to make a persuasive case. There are folks that are so frustrated that 5G didn’t do anything, and yet they spent hundreds of billions of dollars on capex. So now they are saying, ‘I don’t want anything that’s going to cost money anymore. Just innovate for free.’
“So well OK, let’s see how much of this ISAC the industry can do with OFDM. In fact, if they crush it with an OFDM waveform and ISAC, then we were wrong and we’ll just render that with OTFS. But if, in fact, our Zak-OTFS has significant benefits relative to an OFDM outcome – and we think it gives much higher resolution, much lower complexity – we’re going to show that by demonstration, not by argument. We’re going to do it very soon.”
