News

PhD Offer: Low-power Software-Defined Baseband Processor for Flexibility and Security

Published 5/14/2020

Keywords : Hardware architecture, IoT, Low-power, Cybersecurity

Abstract


The concept of ubiquitous system will have a real application and an effective deployment through the
paradigms of Internet of Things (IoT) or Cyber-Physical Systems (CPS) for the industry of the future. Nevertheless
securing IoT devices is a challenging task which will be in the future one of the key elements allowing to
make the difference in a competitive market and especially to build services acceptable to users for which data
protection corresponds to a fundamental issue. The lifecycle of a device is also an important concern regarding
updates & sustainability.
In this context, we propose to address the challenge of cybersecurity specifically at the level of the baseband
processor which is part of the communication unit of an IoT device. Indeed, the threat model is changing
due to the rise of Software-Defined Radio (SDR), the appearance of low-cost high-performance platforms and
associated software that offer cybercriminals a much easier access to the lower layers of communication networks.
In addition, the security flaws in algorithms or protocols at the level of communications have and will have a
strong economic impact as for example recently with the update of Orange network security for the GSM part
(flaw GSM protocol A5/1) [1]. It is a problem to be addressed from conception and which takes all the more
important if forecasts of millions objects that should operate on battery for long time (more than 10 years) are
realized.
The proposed project aims to integrate the cybersecurity threat from the design phase of architectures
running software-defined waveforms used in IoT devices with low resources. For this, the objectives are flexibility
and the integration of software updates mechanisms while considering low-power consumption and security
(e.g. authentication and confidentiality) constraints. It is about developing a flexible, reconfigurable, durable
and secured architecture under very low-power consumption constraints (lower than mW) in order to meet the
challenges of future connected devices. The proposed work will use the architecture of the RISC-V processor to
rely on a very active community, contributing to the development of an open solution and above all to allow an
audit.

 

Context
IoT devices are growing and they have an important attack surface. One of the potential entry point concerns
their communication capabilities. Indeed the threat model regarding digital communication is changing [2] due
to the rise of Software-Defined Radio, the development of low-cost high-performance platforms and associated
software that allow cybercriminals a much easier access to the lower layers of communication networks. Thus,
security breaches on communication-level algorithms or protocols have and will have an impact. Similarly,
updating the LoRaWAN standard to improve the security (upgrade in v1.1) shows that the ability to upgrade
protocols is mandatory. This is a problem that must be addressed from the design point of view. And the
problem becomes more important with the forecasts of massive IoT where devices should work on battery for
long periods of time (greater than 10 years).
Software-Defined Radio is a concept that could offer flexibility. These last years SDR has been studied and
applied where ultimately hardware architectures and resources were not necessarily a strong constraint. In the
context of constrained IoT, the execution of a software-defined waveform becomes more challenging. This is why
currently the communication part of a communicating device is generally separated and linked by an I2C or SPI
bus to the SoC executing the application part. However, recently an industrial solution integrating the radio part
within a SoC with the Texas Instrument CC1352R goes in the direction of the execution of several waveforms
on the same architecture (Cortex M0). Moreover, recently [3] proposed extending the RISC-V instruction set
for the LoRaWAN protocol. Chen et al. [4] propose a specific architecture allowing software-defined waveform
execution for IoT.
Other solutions as the work done by Roux et al. [5, 6] where authors share the observation of the need
for protection for the IoT base their solution on spectrum-level logs to extract behaviors or see attempts
of connections exploiting potential vulnerabilities on radio protocols. Their approach is based on adding an
Intrusion Detection System (IDS) at the network core level which uses machine learning techniques to classify
flows using PHY layer metrics from probes positioned in the network. Their IDS is original because it processes
information at the physical layer which makes it independent of layers encountered in the IoT. However, in
their work they apply this concept by adding probes and a dedicated Intrusion Detection System node which
adds complexity to the deployment. Indeed, their working hypothesis is that due to the strong constraints in
connected devices it is not possible to add security/protection mechanisms into a device.
Our approach aims to remove this constraint and add these features of security/protection at each node to
make surveillance and protection more robust. To address this lock we think the flexibility of an IoT SDR will
allow the addition of software code for supervision and protection inside a node.

 

Challenges
As part of this work, two issues will be addressed. The first issue deals with the definition of a hardware
architecture based on RISC-V processor for applying a software-defined waveform in a context of very low-power
consumption in order to meet the challenges of IoT in terms of energy efficiency. The second issue deals with
integration into the design of protection mechanisms in the baseband processor on the one hand to protect the
system and the data handled and on the other hand to guarantee a secure update of a software waveform. These
include guaranteeing versioning, authentication, confidentiality and integrity of data. For this second challenge
we can in particular rely on some previous work carried out in Michael Grand’s PhD thesis [7, 8].
From these two issues to be treated jointly, we identified two main scientific challenges. The first challenge
is linked to the definition of a low-power architecture allowing to run several waveforms in the Sub-Ghz bands
for the IoT. The second scientific challenge is linked to the security mechanisms that should be implemented
in the baseband processor. It is about imagining the software mechanisms and hardware to be integrated with
processor and memory in order to guarantee security properties.
According to the answers given by this research, the development prospects for a supervision within a
baseband processor could be considered in order to expand the threats model to the detection and protection
of radio attacks aiming at exploiting protocols vulnerabilities or with strategies targeting the battery lifetime.
The idea would be to propose inside a communication node a solution inspired from an Intrusion Detection
System. Finally, it would also target to implement security techniques exploiting the physical layer (PHY secrecy,
fingerprinting, etc.)

 

Scientific program
Firstly, we propose to study baseband processors for the use of protocols in the Sub-bands GHz available in
the scientific literature. This analysis will allow us to consider improvements and to propose techniques whose
main objective will be the trade-off between more flexibility and low-power consumption. An architectural
solution around the extension of the instruction set of a RISC-V processor for a software-defined waveform
dedicated to protocols in bands of Sub-GHz frequency is targeted. Regarding sustainability, which is part of a
problem of sobriety (energy and resource consumption which are also essential in the deployment of IoT), the
flexibility that a SDR radio can bring by allowing scalability of applications and extending the hardware life
cycle is an essential point that will be addressed.
Secondly, we will study how adding protection mechanisms to the design in order to obtain the secure
realization of these waveforms, intrusion detection and secure update mechanisms. There exists a fairly rich
literature in this area, one should study how to integrate these protection mechanisms into each IoT node with
resource constraints and very low-power consumption.
Although these two steps are presented sequentially here, the definition of an SDR architecture will integrate
the security dimension from the design phase in order to especially take into account potential security
vulnerabilities at the architectural level.

Candidat skills
— Master degree or equivalent.
— Key skills :
— architecture of processors
— VHDL, FPGA
— C, assembler
— Other skills (appreciated) :
— security for embedded systems
— network protocols, digital communication

 

Informations
— Supervisor : Guy GOGNIAT
— Co-supervisor : Philippe TANGUY
— Co-supervisor : Jean-Philippe DELAHAYE
— Laboratory : Lab-STICC (https://www.labsticc.fr/en/index/)
— Research team : MOCS (https://www.labsticc.fr/en/teams/m-10-mocs.htm)
— Location : Lorient
— Starting date : October 2020 (3 ans)
— Doctoral school : MathSTIC (https://ed-mathstic.u-bretagneloire.fr/)

 

Candidature

Email to Philippe TANGUY with :
— Motivation letter and full CV (student projects, ...)
— Complete academic records (MSc)

Deadline : as soon as possible


Contacts
GOGNIAT Guy
B guy.gogniat@univ-ubs.fr
T +33 (0)2 97 87 46 41
Professor (Professeur des universités)
TANGUY Philippe
B philippe.tanguy@univ-ubs.fr
T +33 (0)2 97 87 45 67
Associate professor (Maître de conférences)


Références
[1] Orange : une mise à jour de sécurité sur le réseau 2g bloquera certains téléphones. https://www.nextinpact.
com/news/105316-orange-mise-a-jour-securite-sur-reseau-2g-bloquera-certains-telephones.
htm. Accessed : 2020-29-01.
[2] Chaouki Kasmi. De la radio matérielle à la radio logicielle : impact sur l’étude de la sécurité des réseaux
sans l. page 17.
[3] Hela Belhadj Amor and Carolynn Bernier. Software-Hardware Co-Design of Multi-Standard Digital Baseband
Processor for IoT. In Design, Automation and Test in Europe (DATE), Florence, Italy, March 2019.
[4] Y. Chen, S. Lu, H. Kim, D. Blaauw, R. G. Dreslinski, and T. Mudge. A low power software-defined-radio
baseband processor for the internet of things. In 2016 IEEE International Symposium on High Performance
Computer Architecture (HPCA), pages 40–51, March 2016.
[5] J. Roux, É. Alata, G. Auriol, V. Nicomette, and M. Kâaniche. Toward an intrusion detection approach
for iot based on radio communications profiling. In 2017 13th European Dependable Computing Conference
(EDCC), pages 147–150, Sep. 2017.
[6] J. Roux, É. Alata, G. Auriol, M. Kaâniche, V. Nicomette, and R. Cayre. Radiot : Radio communications
intrusion detection for iot - a protocol independent approach. In 2018 IEEE 17th International Symposium
on Network Computing and Applications (NCA), pages 1–8, Nov 2018.
[7] Michael Grand, Lilian Bossuet, Bertrand Le Gal, Guy Gogniat, and Dominique Dallet. Design and implementation
of a multi-core crypto-processor for software defined radios. In Proceedings of the 7th International
Conference on Reconfigurable Computing : Architectures, Tools and Applications, ARC’11, page 29–40, Berlin,
Heidelberg, 2011. Springer-Verlag.
[8] M. Grand, L. Bossuet, G. Gogniat, B. L. Gal, J. Delahaye, and D. Dallet. A reconfigurable multi-core
cryptoprocessor for multi-channel communication systems. In 2011 IEEE International Symposium on
Parallel and Distributed Processing Workshops and Phd Forum, pages 204–211, May 2011

Display the whole heading

©2016-2020 Azimut - Website design & Interactive kiosks Legal information | Site map