FUll Duplex Active Cancellation for wireless communication and co-exisTence

Informations projet

FUDACT

N° de convention de subvention: 656940

DOI

10.3030/656940

Projet clôturé

Date de signature de la CE 1 Avril 2015

Date de début 1 Juillet 2015

Date de fin 30 Juin 2018

Financé au titre de

• EXCELLENT SCIENCE - Marie Skłodowska-Curie Actions

Coût total

€ 240 172,20

Contribution de l’UE

€ 240 172,20

Coordonné par

UNIVERZITET U NOVOM SADU FAKULTET TEHNICKIH NAUKA
Serbia

Periodic Reporting for period 2 - FUDACT (FUll Duplex Active Cancellation for wireless communication and co-exisTence)

Période du rapport: 2017-07-01 au 2018-06-30

Résumé du contexte et des objectifs généraux du projet

The development of fifth-generation (5G) wireless networks is expected to provide technologies to support 1000x increase in data rates. To support the vision of hundreds of billions of connected devices after year 2020, many of the new technologies need to be developed. 5G is expected to utilize many more spectral bands, several of them at millimeter-wave frequencies, and much denser deployment of wireless infrastructure. As a result, co-existence between systems will be the key problem to be addressed at the physical level. In “Full Duplex Active Cancellation for wireless communication and co-exisTence” (FUDACT) we explore a set of new co-existence technologies culminating in a new radio transmission paradigm to improve spectrum efficiency. The FUDACT goal is to support two Europe 2020 flagship initiatives: Digital agenda for Europe by addressing some of the H2020 Future Internet challenges and Innovation Union by paving the way to innovation-friendly environment.

Due to demand for significant increase in network capacity over the next 10 years small cells such as pico and femto cells, that bring network closer to the user, play significant role. By supporting extreme massive network densification, they enable spectral resources to be reused in space by virtue of their smaller footprints. Additionally, to support a growing number of bands in an area and cost efficient manner there is a strong request for frequency-flexible receivers with less or no external filters. However, in submicron CMOS processes absence of significant RF filtering in the presence of the strong interferer will cause receiver compression and desensitize receiver due to reciprocal mixing. Some receiver topologies, like mixer-first receiver, have natural resiliency to strong out-of-band blockers as they first convert signal to the current domain thus avoiding issues coming from the low power supply and limited voltage headroom. In FUDACT, mixer-first receiver was designed to fulfil femtocell requirements for 5G communication networks in STMicroelectronics 28 nm UTBB FD-SOI technology. To achieve state-of-the art performance in this deep-submicron technology, powerful technique-known as body biasing- have been used. Within FUDACT a new approach to analog design optimization with body biasing has been proposed that not only considerably reduces the circuit active area and complexity but also optimize and recovers the transistor/circuit performance in the most robust process corners.

To accomplish FUDACT goals we will be pursuing the following closely related objectives: (1) Determine the technology limits of active cancellation of self-interference in wireless transceivers (2) Demonstrate an architecture capable of cancelling a +23dBm transmitter and a 0dBm of narrowband blocker (3) Investigate mixer-first receiver architectures capable of tolerating higher blocker signal levels (4) Apply the techniques (1-3) in the design of a full duplex future-oriented wireless communication network.

Travail effectué depuis le début du projet jusqu’à la fin de la période considérée dans le rapport et principaux résultats atteints jusqu’à présent

First, thorough investigation of the full-duplex current state-of-the art published solutions was performed followed by work of full-duplex transceiver system level and circuit level specifications. The project is based on a new, 28nm FDSOI technology that is not in volume production yet. As a result, the technology parameters change as well as the corresponding design kits. Therefore, researcher put additional effort into automatizing characterization of the 28nm FD SOI technology, generating Matlab code for direct data access for specific biasing points and control parameters. Also, researcher has written Matlab scripts for using that data base for performance optimization and better understanding of trade-offs between technology and system/circuit topology. Researcher has used body biasing technique, available in 28nm UTBB FD-SOI, for fast and efficient change of power and performance characteristics . The main goal was to use this technique to recover the whole design space of the mixer-first receiver in the most robust corners (SSA, SFA, FSA, FFA). New generation receivers, like mixer-first receiver, introduce previously not encountered demands for some circuit blocks. In the traditional receiver architecture base-band amplifier is at the end of the receiver chain and needs to satisfy receiver linearity requirement. However, in mixer-first receiver base-band amplifier is just after down-conversion mixer where due to mixer switches bidirectional behavior it has to satisfy requirements related to receiver impedance matching, noise and linearity. The design goal was to keep approximately same power consumption and the receiver design Figures of Merit (Noise Figure, S11, 3rd order Harmonic Rejection, 5th order Harmonic Rejection, etc.) within +/- 10% range. The paper that presents novel analog optimization methodology that led to significant decrease of the active devices area and tight control of the design parameters will be submitted to IEEE TCAS-II journal. Paper explaining new approach to class AB base-band amplifier design with body biasing, able to significantly reduce circuit area and complexity and recover performance in even the most robust corners, was presented at S3S 2017 conference. Project results related to the body-biasing technique and analog circuits device optimization were presented at 26th International Workshop on Post-Binary ULSI Systems 2017, EWDTS 2017 and ISSE 2018. The new approach to rapid analog and digital integrated systems design was presented during hands-on “Integrated Systems Design” Workshop/ Bootcamp.

Progrès au-delà de l’état des connaissances et impact potentiel prévu (y compris l’impact socio-économique et les conséquences sociétales plus larges du projet jusqu’à présent)

FUDACT utilizes advances of the state of the art 28nm UTBB FD-SOI CMOS technology and support the European flagship in More than Moore race through unique collaboration between STMicroelectronics, BWRC, FTS and the researcher. Knowledge transfer from the world-class facility such as Berkeley Wireless Research Center to Europe (both to academia and industry) and finished mixer-first receiver design will support growth of the European design ecosystem around FD-SOI. The performance optimization methodology proposed by the researcher introduces a new paradigm- design optimization should not start from typical process conditions like in CMOS technology thus bringing additional benefits besides tight parameter control- circuits area reduction. The novel analog design optimization procedure is already used as a part of 28nm UTBB FD-SOI STMicroelectronics portfolio of successful applications. The planned dissemination activities during and beyond FUDACT will try to provide competitive edge to the next generation engineers either by organizing bootcamps or sharing design methodology and supporting Matlab scripts at Zenodo and mirjanavidenovicmisic.com site.

my-die.png