PhD defence: Contributions to Securing Software Updates in IoT

Published: 2022-09-26

Thesis title: Contributions to Securing Software Updates in IoT

Author: Pegah Nikbaht Bideh, Department of Electrical and Information Technology, Lund university

Faculty opponent:  Professor Valtteri Niemi - University of Helsinki, Finland.

Location: E:1406 E-huset, Ole Römers väg 3, LTH, Lund University, Lund.

Streamed at: https://lu-se.zoom.us/s/61131664328

Thesis at LU Research Portal

Abstract

The Internet of Things (IoT) is a large network of connected devices. In IoT, devices can communicate with each other or back-end systems to transfer data or perform assigned tasks. Communication protocols used in IoT depend on target applications but usually require low bandwidth. On the other hand, IoT devices are constrained, having limited resources, including memory, power, and computational resources. Considering these limitations in IoT environments, it is difficult to implement best security practices. Consequently, network attacks can threaten devices or the data they transfer. Thus it is crucial to react quickly to emerging vulnerabilities.
These vulnerabilities should be mitigated by firmware updates or other necessary updates securely. Since IoT devices usually connect to the network wirelessly, such updates can be performed Over-The-Air (OTA). This dissertation presents contributions to enable secure OTA software updates in IoT.

In order to perform secure updates, vulnerabilities must first be identified and assessed. In this dissertation, first, we present our contribution to designing a maturity model for vulnerability handling. Next, we analyze and compare common communication protocols and security practices regarding energy consumption. Finally, we describe our designed lightweight protocol for OTA updates targeting constrained IoT devices.

IoT devices and back-end systems often use incompatible protocols that are unable to interoperate securely. This dissertation also includes our contribution to designing a secure protocol translator for IoT. This translation is performed inside a Trusted Execution Environment (TEE) with TLS interception.

This dissertation also contains our contribution to key management and key distribution in IoT networks. In performing secure software updates, the IoT devices can be grouped since the updates target a large number of devices. Thus, prior to deploying updates, a group key needs to be established among group members.
In this dissertation, we present our designed secure group key establishment scheme.
Symmetric key cryptography can help to save IoT device resources at the cost of increased key management complexity. This trade-off can be improved by integrating IoT networks with cloud computing and Software Defined Networking (SDN).
In this dissertation, we use SDN in cloud networks to provision symmetric keys efficiently and securely.
These pieces together help software developers and maintainers identify vulnerabilities, provision secret keys, and perform lightweight secure OTA updates. Furthermore, they help devices and systems with incompatible protocols to be able to interoperate.

PhD defence: Massive MIMO for Dependable Communication

Published: 2022-09-26

Thesis title: Massive MIMO for Dependable Communication

Author: Sara Willhammar, Department of Electrical and Information Technology, Lund university

Faculty opponent:  Prof. Jean-Paul Linnartz - Signify (Philips Lighting) and Eindhoven University of Technology, Netherlands.

Location: E:1406 E-huset, Ole Römers väg 3, LTH, Lund University, Lund.

Streamed at: https://lu-se.zoom.us/s/61738905353

Thesis at LU Research Portal

Abstract

Cellular communication is constantly evolving; currently 5G systems are being deployed and research towards 6G is ongoing. Three use cases have been discussed as enhanced mobile broadband (eMBB), massive machine-type communication (mMTC), and ultra-reliable low-latency communication (URLLC). To fulfill the requirements of these use cases, new technologies are needed and one enabler is massive multiple-input multiple-output (MIMO). By increasing the number of antennas at the base station side, data rates can be increased, more users can be served simultaneously, and there is a potential to improve reliability. In addition, it is possible to achieve better coverage, improved energy efficiency, and low-complex user devices. The performance of any wireless system is limited by the underlying channels. Massive MIMO channels have shown several beneficial properties: the array gain stemming from the combining of the signals from the many antennas, improved user separation due to favourable propagation -- where the user channels become pair-wise orthogonal -- and the channel hardening effect, where the variations of channel gain decreases as the number of antennas increases. Previous theoretical works have commonly assumed independent and identically distributed (i.i.d.) complex Gaussian channels. However, in the first studies on massive MIMO channels, it was shown that common outdoor and indoor environments are not that rich in scattering, but that the channels are rather spatially correlated. To enable the above use cases, investigations are needed for the targeted environments. This thesis focuses on the benefits of deploying massive MIMO systems to achieve dependable communication in a number of scenarios related to the use cases. The first main area is the study of an industrial environment and aims at characterizing and modeling massive MIMO channels to assess the possibility of achieving the requirements of URLLC in a factory context. For example, a unique fully distributed array is deployed with the aim to further exploit spatial diversity. The other main area concerns massive MIMO at sub-GHz, a previously unexplored area. The channel characteristics when deploying a physically very large array for IoT networks are explored. To conclude, massive MIMO can indeed bring great advantages when trying to achieve dependable communication. Although channels in regular indoor environments are not i.i.d. complex Gaussian, the model can be justified in rich scattering industrial environments. Due to massive MIMO, the small-scale fading effects are reduced and when deploying a distributed array also the large-scale fading effects are reduced. In the Internet-of-Things (IoT) scenario, the channel is not as rich scattering. In this use case one can benefit from the array gain to extend coverage and improved energy efficiency, and diversity is gained due to the physically large array.

PhD defence: Sparse Codes on Graphs with Convolutional Code Constraints

Published: 2022-09-26

Thesis title: Sparse Codes on Graphs with Convolutional Code Constraints

Author: Muhammad Umar Farooq, Department of Electrical and Information Technology, Lund university

Faculty opponent:  Professor Laurent Schmalen, Karlsruhe Institute of Technology, Germany.

Location: E:1406 E-huset, Ole Römers väg 3, LTH, Lund University, Lund.

Streamed at: https://lu-se.zoom.us/s/68250582310

Thesis at LU Research Portal

Abstract

Modern coding theory is based on the foundation of the sparse codes on graphs, such as the low-density parity-check (LDPC) codes, and the turbo-like codes (TCs) with component convolutional codes. The success of the LDPC codes and the TCs lies in their ability to perform low-complexity iterative message passing decoding procedures. The iterative message passing decoders that exchange messages probabilities, or beliefs, within the code graph are known as the belief propagation (BP) decoders. The BP decoders are sub-optimal, whereas maximum-a-posteriori (MAP) decoders for these codes are computationally infeasible. These codes can be optimized for their BP decoding performance, which improves their error rate performance in the waterfall region at the cost of a performance loss in the error-floor region. On the contrary, optimizing these codes for the MAP performance results in an improved performance in the error-floor region at the expense of a degraded performance in the waterfall region.

In practice, the BP decoding performance of the LDPC codes and the TCs, in the asymptotic block length regime, is determined by computing their BP decoding thresholds from the density evolution (DE), or the extrinsic information transfer (EXIT) chart techniques. The MAP decoding thresholds can be obtained with an application of the area theorem to the BP decoder performance. The graphs of the LDPC codes and the TCs are optimized for the BP, and the MAP decoding performance by using their decoding thresholds. For very large block lengths, spatially coupled (SC) versions of LDPC codes, and the TCs-which are optimized for MAP decoding performance- were shown to achieve excellent BP decoding performance in both the waterfall and the error-floor region, thanks to the threshold saturation. However, the BP decoding performance of these spatially coupled codes suffer from a high error-floor at a moderate to short code block length.

BP and MAP decoding thresholds of TCs on a binary erasure channel (BEC) have previously been investigated via the DE analysis. The capacity achieving SC-TCs were determined, where the underlying TCs were optimized for the BP and MAP performance. The TCs ensembles, parallel concatenated codes, serially concatenated codes, braided convolutional codes, and hybrid concatenated codes, with varying component convolutional codes strengths were considered in these investigations.

This thesis focuses on investigating the BP decoding performance of SC-TCs- which were earlier investigated for the BEC- on an additive white Gaussian noise (AWGN) channel. Furthermore, the problem of high error-floor of SC-TCs for short to moderate block lengths is investigated and addressed with a design of an optimized convlutional permutor in constructing the SC-TCs. Finally, the connection of the TCs and the LDPC codes is explored by introducing a family of convolutional codes (CC) based generalized LDPC codes (GLDPCs). These research areas are summarized under the following three topics.

In the first topic, we have computed the iterative decoding thresholds of SC-TCs on the AWGN channel via the Monte Carlo density evolution (MC-DE) methods. The MC-DE methods are time consuming, which has motivated us to introduce an efficient alternative that predicts the AWGN thresholds of SC-TCs with the knowledge of their BEC thresholds. The results show that the estimated thresholds via the MC-DE method and the predicted thresholds are very close for the capacity achieving randomly punctured SC-TCs. For the high rate uncoupled TCs, which are obtained by randomly puncturing their mother code, the predicted thresholds are improved by incorporating the estimated AWGN threshold of the mother code ensemble into the threshold prediction method.

In the second topic, we have introduced the design of a single block-wise periodic time-varying convolutional permutor to construct the SC-TCs. The convolutional permutor is designed by applying the unwrapping procedure to an optimized block permutor, which optimize the bit error rate (BER) performance of a TC in an error-floor region. We showed that a convolutional permutor obtained via the unwrapping procedure inherits the properties of its parent permutor. Due to this reason, the BER performance of block-wise periodically time-varying convolutional permutor based SC-SCCs does not suffer from a high error-floor problem at short block lengths, which was demonstrated through the simulation results.

In the third topic, we have introduced the families of regular and irregular CC-GLDPCs. The CC-GLDPCs enabled us to connect TCs and LDPC codes in terms of their graph structures. The BEC thresholds and the minimum distance properties of the regular CC-GLDPCs were compared to the regular LDPC codes. Furthermore, we performed an exhaustive grid search using the BEC thresholds of the class of CC-GLDCPs, and determined the design configurations of optimized CC-GLDPCs. The results suggest that, for regular graphs, it is possible to find a sparser CC-GLDPCs than the LDPC codes at the expense of a slightly negligible loss in the performance. Furthermore, the BP optimized CC-GLDPC is observed to have a better BP and MAP thresholds than the turbo codes on the BEC.

Intelligent assistants and tools - ELLIIT tech talk 3

Published: 2022-10-31

You can watch this and more ELLIIT tech talks at Youtube.

Intelligent assistants and tools

Release date: 15 November 2022

Spoken language: Swedish and English

Subtitles: Swedish and English

Content

Mapping and Positioning With Images, Sound and Radio

With the help of information from satellites, radio waves, images and sound, we can soon let robots catch a hair in the air, but how does it work? How can we really figure out where things are? Professor Kalle Åström explains how different positioning techniques work, what we can use them for and the mathematics behind.

Kalle Åström, Mathematical Imaging Group, Lund University

Navigation and radio channel estimation? - a tale of a doctoral thesis

Professor Bo Bernhardsson explains radio-based navigation. We dive into the background of a doctoral thesis by the former PhD student Anders Mannesson. What does it mean to do a PhD, how does a thesis come about, what does it contain, and what was the problem that Anders solved?

Bo Bernhardsson, Professor, Automatik Control, Lund University

Conversation about intelligent assistants and tools

• Bo Bernhardsson, Professor, Automatik Control, Lund University
• Kalle Åström, Mathematical Imaging Group, Lund University
• Charlotta Falvin, Board Professional
• Johan Wester, host and moderator 

Access

You can watch this and more ELLIIT tech talks at Youtube.

Read about ELLIIT tech talk at elliit.se

AI, machine learning and large scale processing - ELLIIT tech talk 2

Published: 2022-10-31

You can watch this and more ELLIIT tech talks at Youtube.

AI, machine learning and large scale processing

• Introduction to the theme AI, machine learning and large scale processing
  Fredrik Tufvesson, Johan Wester

• Privacy-Preserving Machine Learning
  The use of AI and machine learning in different areas increase, but how can we ensure that we have ethical, reliable and privacy-preserving AI? Fredik Heintz guides us to how to avoid bias, get sufficient data for training and ensure that our privacy is preserved even when we use AI together with health data.
  Fredrik Heintz, Professor of Computer Science

• Digital twins in health care
  The use of AI and machine learning in different areas increase, but how can we ensure that we have ethical, reliable and privacy-preserving AI? Fredik Heintz guides us to how to avoid bias, get sufficient data for training and ensure that our privacy is preserved even when we use AI together with health data.
  Claes Lundström, Sectra; Research Director, Sectra
  Gunnar Cedersund, Associate Professor, Biomedical Engineering

• Conversation about large scale data processing
  Johan Wester, Fredrik Heintz, Claes Lundström and Jonas Björk
  Jonas Björk, Professor of Epidemiology, Lund University

Host: Johan Wester

Access

You can watch this and more ELLIIT tech talks at Youtube.

Read about ELLIIT at elliit.se

Natural and Artificial Cognition ? 1, 2, many

Published: 2022-09-06

Recordings: https://www.ai.lu.se/2022-10-27/

Cognition means sensing, processing, and sharing information, leading up to observable behaviour. Both natural and artificial cognition are extremely relevant topics today: The large-scale automation of society and advent of increasingly complex autonomous (robotic) systems with cognitive capabilities, sometimes called the fourth industrial revolution, will cause comprehensive changes to human society and the lives of individuals.

The changes create challenges and consequences that needs to be understood for individual and collective cognition in the natural and artificial worlds. Taking advantage of the opportunities presented by the new technologies and automation and developing novel and appropriate technology for a sustainable future demands concerted effort and wide investigation of both natural and artificial cognition.

Research about human cognition in the natural world can be used to make robots and AI behave in ways that are familiar and nonintrusive to us. Cognitive modelling and robotic systems can serve as tools to validate theories in natural cognition. Groups of agents -- people, robots, and animals ? in novel constellations create behavioural interactions that in turn will influence existing collective behaviour.

Lund University has research with different, yet mutually relevant perspectives, methods, and goals connected to the areas of natural and artificial cognition. 

On 27 October the newly appointed profile area, Natural and Artificial Cognition: 1,2, many, invites to a workshop that discusses this research. The workshop will address how aspects of human, animal, or artificial cognition, through communication and other forms of interaction, influence behaviour in the individual, dyadic, or collective context.

When: 27 October 09.30 ? 15.30 (CET)

Where: E:A, E-huset, John Ericssons väg 2/Ole Römers väg 3, Lund, Sweden

Registration: Free to attend but you need to sign up at ai.lu.se.

Programme

9.30 Registration, fika & mingle

10.00 Session 1

Introduction
Kalle Åström, Mathematical Imaging Group, Coordinator AI Lund, Lund University

Neuronal mechanisms underlying the selection of actions
Per Petersson, Integrative Neurophysiology, Lund University

From Human Skills to Robot Skills
Volker Krueger, Computer Science and RobotLab LTH, Lund University

Navigational decisions - from insect neural circuits to nanophotonic hardware
Stanley Heinze, Functional zoology, NanoLund, Lunds University

Modelling Cognition in Robot
Christian Balkenius, Cognitive Science, Lund University

Dangerous decisions in mosquitos
Marcus Stensmyr, Animal Physiology, Lund University

12.00 Lunch and mingle

13.15 Session 2

Humanoid robots in social situations
Ingar Brinck, Theoretical Philosophy, Lund University

Language is heard and seen. About speech and gestures
Marianne Gullberg, General Linguistics and Humanities Lab, Lund University

Assessment of mental health with language analysed by AI
Sverker Sikström, Psychology, Lund University

Cognition in musical practice

• Creative strategies and instrumental interaction in artistic knowledge building, Karin Johansson, Malmö Academy of Music, Lund University
• Musician and instrument as a cognitive system, Markus Tullberg, Malmö Academy of Music, Lund University

On Large-Scale Coordination
Anders Rantzer, Automatic Control, Lund University

CANCELED: Political communication in a polarised digital democracy
Hanna Bäck, Political Science, Lund University
Replaced by online lunch seminar 23 November 2022 - open for registration

15.00 Fika & mingle

15.30 End of workshop

Organisation and contact

• Hanna Bäck, Professor, Department of Political Science
• Marie Dacke, Professor, Department of Biology
• Sverker Sikström, Professor, Department of Psychology
• Ingar Brinck, Professor, Theoretical Philosophy
• Kalle Åström, Professor, Mathematics (Faculty of Engineering)
• Emma Boberg, AI Lund Communication
• Jonas Wisbrant, AI Lund Communication

ELLIIT tech talks 1: Industry 4.0

Published: 2022-10-18

You can watch this and more ELLIIT tech talks at Youtube.

The world is full of clouds, and there will be more

Clouds - What are they and where are they? Why do we use clouds? How does a distributed cloud infrastructure look like in 5G and the upcoming 6G as examples? How do they affect energy consumption?

Maria Kihl, Professor, Broadband Communication, Department of Electrical and Information Technology

What are the pieces of the puzzle in smart industry?

What exactly is smart industry? What are the advantages, how does it work and what are the challenges? The choice of  computer between the cloud server and the smartwatch matters greatly for performance, energy consumption and privacy ? for both humans and machines. We discuss design decisions and priorities.

Charlotta Johnsson, Professor, Department of Automatic Control and  Rector Campus Helsingborg, Lund University

In the panel

Jonas Birgersson, IT entrepreneur

Host and moderator

Johan Wester, Skånemotor

Access

You can watch this and more ELLIIT tech talks at Youtube.

Read about ELLIIT at elliit.se

PhD defence: Systems with Massive Number of Antennas: Distributed Approaches

Published: 2022-08-26

Thesis title: Systems with Massive Number of Antennas

Author: Jesús Rodríguez Sánchez, Department of Electrical and Information Technology, Lund university

Faculty opponent: Associate Professor Francois Quitin, Belgium.

Location: E:1406 E-huset, Ole Römers väg 3, LTH, Lund University, Lund.

Streamed at: https://lu-se.zoom.us/s/67219695148

Thesis for download (PDF)

Abstract

As 5G is entering maturity, the research interest has shifted towards 6G, and specially the new use cases that the future telecommunication infrastructure needs to support. These new use cases encompass much higher requirements, specifically: higher communication data-rates, larger number of users, higher accuracy in localization, possibility to wirelessly charge devices, among others.

The radio access network (RAN) has already gone through an evolution on the path towards 5G. One of the main changes was a large increment of the number of antennas in the base-station. Some of them may even reach 100 elements, in what is commonly referred as Massive MIMO. New proposals for 6G RAN point in the direction of continuing this path of increasing the number of antennas, and locate them throughout a certain area of service. Different technologies have been proposed in this direction, such as: cell-free Massive MIMO, distributed MIMO, and large intelligent surface (LIS). In this thesis we focus on LIS, whose conducted theoretical studies promise the fulfillment of the aforementioned requirements.

While the theoretical capabilities of LIS have been conveniently analyzed, little has been done in terms of implementing this type of systems. When the number of antennas grow to hundreds or thousands, there are numerous challenges that need to be solved for a successful implementation. The most critical challenges are the interconnection data-rate and the computational complexity.

In the present thesis we introduce the implementation challenges, and show that centralized processing architectures are no longer adequate for this type of systems. We also present different distributed processing architectures and show the benefits of this type of schemes. This work aims at giving a system-design guideline that helps the system designer to make the right decisions when designing these type of systems. For that, we provide algorithms, performance analysis and comparisons, including first order evaluation of the interconnection data-rate, processing latency, memory and energy consumption. These numbers are based on models and available data in the literature. Exact values depend on the selected technology, and will be accurately determined after building and testing these type of systems.

The thesis concentrates mostly on the topic of communication, with additional exploration of other areas, such as localization. In case of localization, we benefit from the high spatial resolution of a very-large array that provides very rich channel state information (CSI). A CSI-based fingerprinting via neural network technique is selected for this case with promising results. As the communication and localization services are based on the acquisition of CSI, we foresee a common system architecture capable of supporting both cases. Further work in this direction is recommended, with the possibility of including other applications such as sensing.

The obtained results indicate that the implementation of these very-large array systems is feasible, but the challenges are numerous. The proposed solutions provide encouraging results that need to be verified with hardware implementations and real measurements.