Rapport Final

M5HESTIA

mmW Multi-user Massive MIMO Hybrid Equipments for Sounding, Transmissions and HW ImplementAtion

(2016-2019)

Name of the coordinating person: Maryline HELARD

Date of the report: 17th November, 2019

e-mail: maryline.helard@insa-rennes.fr

 

CominLabs fundings:

  • 2 PhD students + 2 post Docs (22mm + 18mm)
  • Equipments, prototyping, travels: 120 K€

 

Participant organisation name Participant

short name

Contact Person
IETR, UMR CNRS 6164, Rennes, www.ietr.fr IETR Prof. Maryline HELARD

maryline.helard@insa-rennes.fr

Lab-STICC UMR CNRS 6285, Plouzané,

www.lab-sticc.fr

Lab-STICC

(Telecom Bretagne)

Dr. François GALLEE

francois.gallee@telecom-bretagne.eu

Orange Labs, Cesson Sévigné,

www.orange.com/fr/Innovation

Orange Christian GALLARD

christian.gallard@orange.com

IRT (Institut de Recherche Technologique) b<>com, Rennes, www.b-com.com/ b<>com Dr. Rodolphe LEGOUABLE

rodolphe.legouable@b-com.com

 

Abstract

M5HESTIA project aims at designing advanced M-MIMO antennas, characterizing / modelling the outdoor mmW channel and proposing innovative algorithms in order to demonstrate, a full Multi-user massive-MIMO hardware (HW) platform operating in the 60-GHz band.

1. Context of the studies

At the beginning of the project, one of the 5G challenges was the use of new high frequencies with large bandwidths. The use of millimeter waves (mmWave) was expected for indoor but also, and it was new, for outdoor transmissions in the context of small cell deployment. Since the beginning of the project, the 26GHz and 72 GHz bands have been defined.

Another challenge of 5G was the improvement of the spectral efficiency by the increase of the number of antennas at the access point or at the Base Station level with the use of massive MIMO systems.

1.1.  Why studying hybrid massive MIMO for mmWave transmissions?

In the one hand, severe path losses of the propagation channel at high frequencies can be compensated by the gain that can provide a great number of antennas. On the other hand, the smaller the wavelength, the smaller the physical dimension of a larger antenna array making possible the realization of massive MIMO systems. In addition, in theoretical massive MIMO studies, it is commonly assumed that all treatments are digital, requiring as many RF chains as transmit antennas. Nevertheless, increasing the number of antenna RF chains at high frequencies lead to a lot of problems such as the complexity in terms of algorithms and HW integration, the system consumption, calibration…

Hybrid beamforming, with digital and/or analogue beamformings, applied to systems where the number of RF chains is less than the number of antenna can be a solution to some of the implementation problems.

1.2.  Identified challenges were for M5EHSTIA

a) Improve the link budget

  • The high path losses in millimetre waves impose to use directive antenna with very high gains to ensure a sufficient SNR (Signal to Noise Ratio)
  • Channel sounding antennas and transmission antennas have to be designed
  • HW implementation is very critical: calibration, interference, integration losses…

b) Increase of the number of antennas in hybrid MIMO system

  • Less number of RF chains, several antennas linked to one RF chain
  • Optimisation: trade-off between the number of antennas and the number of RF chains  and of digital and analogue treatments

c) Channel measurements and channel modelling

  • Need of channel models for theoretical studies
    • Development of a deterministic channel simulator well suited for 5G mmW scenarios
    • Blockage effect, Los dominant channels
  • Need of new channel sounders for massive MIMO
    • In both elevation and azimuth, for both indoor and outdoor communications
    • Exploitation of indoor and outdoor data from Aalto University
  •  Need of tracking position and orientation algorithm studies

d) Develop  proof of concept by designing one HW platform

  • To implement new enablers
  • To integrate new services
  • To validate the use of new frequencies
  • To make experiments and demonstration
1.3.   Organization of the M5HESTIA project in 5 work-packages
  • WP1: Project Managment and Dissemination
  • WP2: Antenna Design and RF architectures
  • WP3: Propagation and Channel Modelling
  • WP4: Baseband Signal Processing
  • WP5: Baseband Implementation & Platform Integration

2. Consortium

2.1.  Multi-disciplinary project

The research carried out in the framework of the M5HESTIA project is multi-disciplinary. It ranges from (1) microwave theory, electromagnetism and antenna design, (2) propagation, to (3) digital communication domains. Due to numerous technical and scientific challenges to deal with, efficient collaborations between experts of these three domains led to a real optimisation of the global system.

2.2.  Academics and industrial partners

The consortium of M5HESTIA project gathers two academic partners members of CominLabs (IETR and Lab-STICC) and two industrial partners (Orange and IRT b<>com). This constitutes an original project frame for CominLabs, with a close public-private partnership and a very strong involvement of the two industrials.

To reach the very ambitious goals of the M5HESTIA project, a closed collaboration has been set up with a complementary project funded by IRT b<>com (internal project at IRT) and also dedicated to mmW transmissions; this project is entitled 5M (Mm-Waves Multi-User Massive MIMO).

Therefore, the two projects M5HESTIA and 5M run in parallel to reach common and shared objectives: i) comparison between different MIMO processing techniques, by using the propagation channel model provided by the M5HESTIA project and ii) integration in the HW platform developed by IRT of the M-MIMO antenna array developed in the framework of M5HESTIA (see Figure 1).

The academic and industrial partners had common objectives:

  • Comparison of MU massive-MIMO hybrid techniques
  • Development of antennas: for channel sounding and for transmissions
  • Measurement campaigns and channel modelling
  • Integration of M5HESTIA antennas and RF front-ends in the b<>com platform

The main roles of industrial partners:

  • Orange was mainly involved in the characterisation of the antennas and the channel sounder as well as in the measurement campaigns to be carried out in the framework of another CominLabs Project.
  • IRT was mainly involved in the proof of concept integrating antennas and RF front-end developed by M5HESTIA as well as in the parallel IRT 5M project.

Figure 1: Working frame and interactions between M5HESTIA and 5M projects. Both projects deal with Multi-User M-MIMO mmW communications

Positioning of the M5HESTIA project: related international and national activities

To the knowledge of the partners of M5HESTIA project, this proposal was quite innovative and original as it mixed together higher spectrum bands (60 GHz), massive MIMO technology, multiple users access in an outdoor environment and prototyping.

To the best project members’ knowledge, M5HESTIA was one of the rare projects dealing with channel characterization and Massive MIMO techniques in mmWave spectrum.

A list of  projects dedicated to mmWave or massive MIMO was also provided (available in pdf format, on https://project.inria.fr/m5hstia/state-of-the-art/): M5HESTIA_Other_projects

3. State of the art at the beginning of the project (2016)

A detailed state of the art is available on https://project.inria.fr/m5hstia/state-of-the-art/

At the beginning of the project:

  • Mostly MIMO systems below 6 GHz (5G system)
  • Very few measurments, no channel model for massive MIMO
  • Mu-MIMO in mmWave for indoor systems: IEEE 802.11ad (WiGig) – 4.6 Gbps max
  • WRC’19: first frequency bands identification 57 – 64 GHz as potential candidate

In a nutshell and by WP:

  • WP2: antenna for channel sounding, massive MIMO antenna
    • NOKIA mmWave system (2015) at 73,5 GHz
      • No spatial multiplexing, one single RF port
      • Mutibeam capability in one direction  at the same time
    • NIST 60GHz sounder with fixed beams (2015)
  • WP3: channel models in mmWave
    • Few channel measurements and modelling with large antenna arrays
    • No channel model for Massive MIMO
  • WP4: massive-MIMO for mmWave
    • Massive MIMO studies: mainly theoretical, do not take into account number of RF chains limitation and the specificities of mmWave propagation
    • Emergence of the Hybrid Beamforming Concept
  • WP5: existing platforms
    • Mostly M-MIMO below 6 GHz (5G system)
    • M-MIMO in mm waves for indoor systems : IEEE 802.11ad (WiGig) – 4.6 Gbps max

A more detailed “State Of the Art” was  provided at the beginning of the project (available in pdf format, on https://project.inria.fr/m5hstia/state-of-the-art/): SoA_2015

 

4.  State of the art at the end of the project (nov. 2019) and M5HESTIA results

Current state of the art:

  • 5G proposes 26 GHz (over up to 400 MHz) for outdoor transmission (mgNodeB)
  • Hybrid M-MIMO system up to 256 antenna array (with lens) -> Identical direction as M5HESTIA
4.1.   Antennas for channel sounding and transmission and RF architectures

Current state of the art:

  • Tokyo Institute of Technology 2019: Chip 802.11.ay without beamforming
  • No complete front-end RF with simultaneous beams

Provided by M5HESTIA: antennas for channel sounding at 60GHz, antenna for M-MIMO (8 RF chains in parallel) [Pham19a,19b,19c, Guer19] 

a) Development of a modular approach for lens-based antennas

Easy duplication for other sounding and transmission antennas and other frequencies

b) Providing a 60 GHz MIMO channel sounding antenna

Used by Orange and IMT-A for measurement campaigns

c) Design of different systems

A wide scan angle transmit array antenna

Different multibeam 60 GHz systems

  • for channel sounding with analogue or digital beamforming
  • -for wireless communication: until 8 IQ links and 2 users (used by IRT)

d) Results in accordance with:

  • 11 ad norms (beamsteering)
  • 11 ay standardization (multi-user multibeams)

4.2. Propagation, channel modelling and measurements

Current state of the art:

  • Still few channel measurements and modelling with large antenna arrays
  • No model for massive MIMO

Provided by M5HESTIA: improvement of the existing channel modelling (PyLayer) and contribution to mobility scenarios [Shah19, Bald19] 

a) Analysis and use of the measurements campaigns from Altoo University

  • Better modeling of small scatterers in the outdoor channel
  • Integration in our PyLayers channel model

b) Theoretical contributions for location and tracking in 5G systems

  • A new joint heuristic beam selection and user position and orientation tracking approach
4.3. Baseband signal processing

Current state of the art:

  • Mixed analogue/digital or hybrid architectures widely assumed for communicating systems based on large-scale antenna arrays, especially for mmWaves.
  • Various theoretical results for Massive MIMO systems relying on the today’s common assumption of (sparse) ray and cluster based channel representations in mmWaves
  • Confirmation of the interest for a two-stage precoding/beamforming strategy: both analog and digital precodings (5G)

Provided by M5HESTIA: New results in terms of theoretical performance and algorithms for M-MIMO systems taking into account mmWave propagation characteristics and relying on hybrid precoding and/or index modulation  [Sheh17, 18a, 18b, 18c, 19a, 19b, 19c, Mokh19]

a) Theoretical performance of large-scale MIMO systems

  • with spatially clustered and ray-based channel representations (from WP3)
  • with spare and LOS dominated channels
  • Under Los dominant channel: performance of hybrid beamforming with number of RF chains equals to the number of users are the same than for a full digital beamforming system

b) Hybrid beamforming

  • New efficient and low complexity beamforming strategies proposed
    • Analogue Los beamsteering + digital interference cancellation (ZF)
      • Simple and optimal solution in pure LoS scenarios
      • Remains robust in sparse channels
    • Simple and efficient : ZF + EGT-based beamforming
      • Remains robust if some paths are blocked or shadowed

 

c) Application to spatial modulations

  • Design of new index modulation strategies and related theoretical results
4.4.   mmWave M-MIMO plateform

Current state of the art:

  • 5G proposes 26 GHz (over up to 400 MHz) for outdoor transmission (gNodeB)
  • Hybrid M-MIMO system up to 256 antenna array (with lens) -> Identical direction as M5HESTIA

Provided by M5HESTIA: integration of the antennas in the b<>com platform for beamsteering allowing 2 parallel and different beams in a 2 GHz bandwidth leading to high data rate transmission (more than 5 Gbps per spatial stream)

Integration of the WP2 antennas in the b<>com platform

  • Beamsteering allowing 2 parallel beams in a 2 GHz bandwidth
  • Real time video services transmission
  • Validation of the analog and digital parts of the powerful b<>com platform
  • Platform that can be used to test new digital enablers in real environments
  • High throughput (5 Gbps), OFDM and LDPC, 2 GHz bandwidth, 256-QAM, indoor/outdoor transmis-sions, beam scanning/tracking

 

5.      Future possible works and use of results

5.1.   Antenna and RF front-end architecture

The modular structure allows for almost easy extension to different types and number of antennas. The lens antenna provided interesting solution for high gain antenna array.

The developments at 60 GHz could be easily be extended to other frequencies (around 26 or 72 GHz).

The antenna would be integrated by Orange in their channel sounder and are integrated by IRT b<>com for a multi-user 60 GHz transmission.

They would also be used by LabSTICC for their channel sounder and by IETR in a further evolution of its National Instrument massive MIMO platform for mmWave communications.

5.2.        Channel sounding

A full and dynamic channel sounder based on Software Defined Radio equipment will be developed in the framework of a one year “Action –Innovation” funded be CominLabs entitled mmWSounder in order to:

  • to integrate antennas provided by M5HESTIA into a dynamic channel sounder
  • to carry out  additional channel measurements (indoor and outdoor)
  • to go on improving channel modelling
  • to check theoretical simulation results with “measured channels”

Same partners are involved.

5.3.  New algorithms

Some simple proposed algorithms will be checked with realistic “measured channels” in the framework of new mmWSounder project.

5.4.  Proof of concept and plateform

Further developments in 5M

  • Testing new MU-MIMO precoding schemes
  • Virtual Reality use-case wireless transmission

Figure 2: HW demonstration of M5HESTIA

6.      What about the future of the Phd Students and post-docs?

Concerning antennas and RF architectures:

  • 1 PhD student (Pham Tuyen) co-direction IETR- LabSTICC
    • PhD defense: 17/12/2019
    • After his PhD: not yet defined
  • 1 post-doc SAD ( Jialai Weng) at IMT Atlantique (22 months)
  • After his post-doc: Lytid, Paris in mmWave development

Concerning propagation

  • 1 post-doc SAD (Arash Shahmansoori ) at IETR
    • After his post-doc: Mitsubishi MERCE

Concerning signal processing

  • 1 PhD student (Mohamed Shehata) co-direction IETR- LabSTICC
  • Phd defense: 8 /11/2019
  • After his Phd: Mitsubishi MERCE (since the 1st of November)

7.   Publications

In the framework of the project, we presented papers in 9 conferences (9 CICL: 2 EuCAP, 1 EuMW, 3 PIMRC, 1 VTC, 2 ICT), in 1 summer school and we published 1 papers (2 RICL: 1 IEEE Signal Processing Letters).

Two paper journals are under reviewing (1 IEEE Trans. On Propagation and 1 IEEE Wireless Communication Letter) and one in preparation for submisson to IEEE Tr. on Wireless Communication.

A session dedicated to M5HESTIA was organized the 8th November 2019 during an International Workshop on “mm-Wave and THz communications for beyond 5G applications” co-organized by IETR and Orange. https://wave-and-thz-com-66.webself.net/home

The publications from academics are listed below for each work package.

WP2: antenna design and RF architecture

  • Pham, J. Weng, K. Pham, F. Gallée, R.Sauleau, V-Band Beam-Switching Transmitarray Antenna for 5G MIMO Channel Sounding, 13th European Conference on Antennas and Propagation (EuCAP), Krakow, Poland, 2019
  • Guerrero, F. Gallée, C. Kärnfelt, Novel 60 GHz DRA topology adapted to the LTCC technology process,13th European Conference on Antennas and Propagation (EuCAP), Krakow, Poland, 2019
  • Pham, F. Gallée, R. Sauleau, A Beam-Steering Transmitarray Antenna for 5G MIMO Channel Sounding in V-band, EuMW 2019, Paris, France
  • Pham, K. T. Pham, F. Gallée, and R. Sauleau, A Wide-Angle Beam-Steering Bifocal Transmitarray Antenna for 5G Channel Sounding at V-band, IEEE Transactions on Antennas and Propagation (submitted)

WP3: channel modelling

  • Shahmansoori, B. Uguen, G. Destino, G. Seco-Granados, H. Wymeersch, “Tracking Position and Orientation through Millimeter ave Lens MIMO in 5G Systems, “IEEE Signal Processing Letters, Vol 26, N°8, August 2019
  • Balde, B. Uguen ; A. Karttunen and K. Haneda, Identification of Wave Scatterers in an Urban MicroCellular Environment at 32 GHz”, PIMRC 2018, Bologna, 2018

WP4: baseband signal processing

  • Shehata, M. Crussière, M. Hélard, P. Pajusco, Hybrid Analog and Digital Precoding in Millimeter Wave Massive MIMO Systems with Realistic Hardware and Channel Constraints, 2017 IEEE SPS Summer School on Signal Processing for 5G Wireless Access, Gothenburg, Sweden: 2017
  • Shehata, M. Hélard, M. Crussière, A. Rozé, C. Langlais, Angular Based Beamforming and Power Allocation Framework in a Multi-User Millimeter-Wave Massive MIMO System, 87th Vehicular Technology Conference: VTC2018-Spring: 2018
  • Shehata, M. Crussiere, M. Hélard, Hybrid Beamforming for Multi-User MISO Channels with Equal Gain Transmission: A Robust and Spectral Efficient Approach, ICT 2018 | 25th International Conference on Telecommunications: 2018
  • Shehata, M. Crussiere, M. Hélard, P. Pajusco, Leakage Based Users Selection for Hybrid Beamforming in MillimeterWave MIMO, 2018 IEEE 29th Annual International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC): 2018
  • Shehata, A. Mokh, M. Crussière, M. Hélard, P. Pajusco, On The Equivalence between Hybrid and Full Digital Beamforming in mmWave communications, IEEE ICT 2019, 26th Hanoi, Vietnam
  • Shehata, M. Crussière, M. Hélard, On the Theoretical Limits of Beam Steering in mmWave Massive MIMO Channels,  IEEE 30th Annual International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC) : 2019
  • Mokh, M. Shehata, M. Crussière, M. Di Renzo, M. Hélard, “Hybrid Beamforming for Receive Spatial Modulation”, To be submitted to IEEE Tr. on Wireless Communication
  • Shehata, A. Mokh, M. Crussière, M. Hélard, “Analytical Performance of Hybrid Beam Index Modulation”, Submitted at IEEE Wireless Communication Letters

Comments are closed.