High Efficiency wiReless CMOS transceiver boosted by artificial intelligencE for 6G bandS and beyond

Project Information

HERMES

Grant agreement ID: 964246

DOI

10.3030/964246

EC signature date 21 December 2020

Start date 1 September 2021

End date 31 August 2026

Funded under

• EXCELLENT SCIENCE - Future and Emerging Technologies (FET)

Total cost

€ 3 271 457,50

EU contribution

€ 3 271 457,50

Coordinated by

UNIVERSITE DE BORDEAUX
France

Periodic Reporting for period 2 - HERMES (High Efficiency wiReless CMOS transceiver boosted by artificial intelligencE for 6G bandS and beyond)

Reporting period: 2022-09-01 to 2024-08-31

Summary of the context and overall objectives of the project

The future of wireless communications will go beyond connecting people or things to connecting smart robots or unmanned vehicles without human intervention. Only sub-THz frequencies can offer the spectrum to achieve unprecedented communication throughput. Yet current technologies suffer from barriers that prevent mass-market operations, such as high cost, limited bandwidth or a power-hungry hardware not compatible with autonomous systems. HERMES proposes the fusion of Artificial Intelligence (AI) and deep sub-micron CMOS technology to open a new generation of wireless transceivers. The HERMES project focuses on sub-THz frequencies between 140 and 160GHz, and will break new ground in the conversion of information from digital to THz using European CMOS technology to develop a highly integrated transceiver. For the first time, HERMES will deliver to the telecommunications industry a disruptive way of designing transceivers, with impact on production of billions of units that can be implemented in any autonomous system to communicate. This new wireless link will be a springboard for an innovation leap in the robotics and the security industry. Our ambitious and risky approach goes significantly beyond the SoA: we will demonstrate that the power of AI developed by computer sciences research and associated with an original electronics signal processing technique can push CMOS technology to release outstanding performances. The project will produce a chipset of a low-cost radio that exchanges tens of Gbps and will test it in use cases of unmanned vehicles.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

WP1 has demonstrated that adding learning algorithms combined to Walsh signal processing techniques were able to perform jointly spectrum sensing, and system non-idealities mitigation. Specifically, an autoencoder for end-to-end deep learning communication modelling predicted the symbols and the Walsh coefficients by taking into account the transmitter, channel and receiver altogether. It reduced the power consumption by a ratio from 3 to 5. The signal processing pre-distortion techniques exhibited a power consumption reduction by a ratio of 10 and a 10-time faster convergence for similar performances compared to advanced DPD.

WP2 designed and measured the converters based on Walsh signal processing in the context of Carrier Aggregation communications. It achieved several GHz of bandwidth (5GHz for the digital-to-analog and 2.5GHz for the analog-to-digital) with SoA results with 0.34pJ/bit and tackling the Boris Murmann’s SoA.

WP3 designed and measured radiofrequency front-end components (mixer, frequency synthesis, power amplifier, low noise amplifier). Topologies have been chosen, schematics layout have been drawn and simulated and circuit have been fabricated. All the circuits are functional and operate according to the specification at sub-THz frequencies.

WP4 has monitored the measurements methodologies and facilities, and initiated the co-integration roadmap of the building blocks to design the radiofrequency front-end into a single die.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

Telecommunications: HERMES’s first outcome will be chipsets for the telecommunication industry. The solution demonstrated that the low-cost technology, here 28nm CMOS FDSOI from STMicroelectronics, was able to reach sub-THz frequencies, to increase the energy efficiency by a ratio of 10 and give access to GHz- large bandwidths instantaneously. The targeted frequency >100 GHz will provide strong technology for 6G and engineering support foundations for use into complex and evolvable System of Systems (SoS). The market gets more and more interested in the need of circuit to carry out the connectivity from device to backend systems, such as the UxVs sector for video and data backhauling with secured and reliable short-range communication features.

Robotics: Smart robots market is now emerging with a need of smart connectivity such integrated sensing and communication (ISAC). Robots will require to learn from their environment and experience, collaborate, and learn from the behavior of humans. The chipset demonstrated as a connectivity solution to help smart robots exchange the tens of Gbps of data rate to collaborate. A potential extension would be to apply HERMES to ISAC for robotics.

Security: the solution brought a joint hardware-software solution for cognitive radio, broadband communications, signal intelligence, dynamic filtering and especially spectrum sensing. IDE will demonstrate the use of the solution for enhanced detection capabilities in the civilian security. We expect this demonstration to bring interest of large industrial companies in investing in this new kind of sophisticated solutions. It will give room for more efficient and performant solutions with the use of integrated AI.

HERMES Logo