AI Lund Lunch seminar: Neuromorphic technology ? Rethinking the computer for AI

Publicerad: 2025-12-15

Topic: Neuromorphic technology ? Rethinking the computer for AI

When: 11 March, 2026 12.00 to 13.00 CET

Where: Online - link by registration 

Speakers: 

• Mattias Borg, Electromagnetics and Nanoelectronics, Lund University
• Baktash Behmanesh, Integrated Sensors and Adaptive Technology for Sustainable Products and Manufacturing, Lund University

Abstract

Artificial Intelligence (AI) has revolutionized countless domains, but its growing energy footprint poses a critical challenge. The primary bottleneck lies in data movement - shuttling information between memory and processors consumes far more energy than computation itself. In contrast, the human brain achieves remarkable efficiency by computing where the data resides. Neuromorphic technology takes inspiration from this principle, rethinking the computer architecture to suit the heavy data loads of AI. By leveraging materials with memory properties, such as memristors, we can enable analog, in-memory computation that drastically reduces energy costs. 

This talk explores how neuromorphic hardware can disrupt conventional AI processing and why success will depend on a co-design approach, requiring interdisciplinary teams integrating expertise in materials science, circuits and system architecture, and algorithm development. 

We will outline our current effort (NeuTec) which will bring together Swedish neuromorphic researchers at all technology levels, bridging disciplines and forming common visions, with the ultimate goal for Sweden to lead in a new era of sustainable, brain-inspired computing. 

Speaker bios

Mattias Borg is Senior Lecturer in Nanoelectronics at Lund University, with previous research positions within the semiconductor industry, including IBM and Ericsson. His research focuses on the materials and device aspects of neuromorphic technology, as well as three-dimensional integration of emerging technologies with Si CMOS. 

Baktash Behmanesh received his PhD in 2017 from Sharif University of Technology, Tehran, Iran. In 2018, he joined the EIT department of Lund University, where he is currently an Associate Senior Lecturer focusing on analog/RF Integrated Circuit design. His research interests include RF and millimeter-wave front-end design, design of frequency synthesis circuits for wireless communication systems, and neuromorphic circuit design.

Registration

To participate is free of charge. Sign up at ai.lu.se/2026-03-11/registration and we send you an access link to the zoom platform.

Thesis defense: 3D Integration Technology and Near-Memory Computing for Edge AI

Publicerad: 2025-11-27

Arturo Prieto defends his thesis "3D Integration Technology and Near-Memory Computing for Edge AI".

Link to thesis in LU Research Portal.

Zoom link.
Zoom ID: 66847927874.

Higher performance through increased technology integration has focused on scaling transistor dimensions. However, the manufacturing process is increasingly expensive and faces technical challenges in the development of new breakthroughs. Evaluation of the third dimension has emerged as a promising alternative to scaling, which enables stacking of semiconductor components with 3D interconnections. Different technologies present different integration strategies, where 3D sequential integration (3DSI) enables small pitch for 3D contacts, allowing for high-integration circuits. A library of standard cells has been designed and characterized according to 3DSI, enhancing the high-integration capabilities of the technology for digital designs. This library compiles the required predefined logic cells that can be used in the design of a digital integrated circuit (IC).

The design of ICs as a foundation for edge AI is focused on enhancing memory and computing resources to improve the processing capabilities of such platforms. Computing architectures are traditionally based on the concept of von Neumann architecture, which distinguishes computing and memory units as two independent entities. However, near-memory computing (NMC) is presented as a viable alternative to the von Neumann architecture that brings computation closer to memory. NMC is non-intrusive to the conventional low-level structure of SRAM and enhances memory bandwidth for hardware acceleration. The integration of accelerators into resource-constrained platforms has been evaluated, expanding the functionality with custom hardware tailored for computation-intensive AI workloads. Furthermore, flexibility has been achieved by providing modularity to the design architecture.

The proposed architectures are evaluated by programs that highlight the performance of integrated AI hardware accelerators into edge devices, emphasizing the importance of software and hardware co-design. The contributions of this thesis focus on 3DSI technology circuit design and NMC architectures evaluating performance, energy and area efficiency.

Thesis defense: Near-Memory Computing Architectures for Scalable Edge AI Applications

Publicerad: 2025-11-10

Masoud Nouripayam defends his thesis "Near-Memory Computing Architectures for Scalable Edge AI Applications."

Zoom link.
Zoom ID: 69455644174

Link to the thesis i LU Research Portal.

Artificial intelligence (AI) and machine learning (ML) are rapidly permeating nearly every aspect of modern life, from personal devices and autonomous systems to industrial automation and environmental monitoring. The growing demand for intelligence at the network edge is reshaping how computing hardware is conceived and built. Edge AI platforms are expected to deliver high throughput within tight energy and area budgets, operate reliably at low voltages, and adapt to diverse workloads, while data movement between processors and memory continues to dominate system cost. These trends position memory-centric computing as a compelling alternative to conventional architectures. By tightly coupling local memory, approximate processing, and near-memory execution, this work advances the development of compact, high-throughput, and energy-efficient AI hardware. 
 

Publicerad: 2025-10-23

AI & Digitalization Breakfast Seminar: Certifiably Optimal Anisotropic Rotation Averaging

Publicerad: 2025-09-29

When: Wednesday October 15, 2025 at 09:00-10:00

Where: Control Seminar Room M:3170-73

Register: for free breakfast at the seminar, register no later than October 10th by using this link.

Abstract: Rotation averaging is a key subproblem in structure from motion. Semidefinite programming relaxations are often used to solve the problem, and there are theoretical results analyzing difficulty and optimality. However, previous methos focus on the isotropic setting, where the intrinsic uncertainties in the measurements are not fully incorporated into the resulting optimization task. Recent empirical results suggest that moving to an anisotropic framework, where these uncertainties are explicitly included, can result in an improvement of solution quality. However, global optimization for rotation averaging has remained a challenge in this scenario. In this talk we show how anisotropic costs can be incorporated in rotation averaging. We also demonstrate how existing solvers, designed for isotropic situations, fail in the anisotropic setting. Finally, we propose a stronger relaxation and empirically show that it recovers global optima on a number of tested datasets.

Speaker bio: Carl Olsson is a Professor at the Computer Vision and Machine Learning division at Centre for Mathematical Sciences, Lund University. He obtained his PhD at Lund University in 2009, and has since then been in Lund and at Chalmers. His research addresses large scale optimization methods with applications in computer vision. A particular interest is in mapping and navigation problems such as structure from motion. His research aims at improving reliability of algorithms by developing methods that provide globally optimal inference, independent of initialization.

The Breakfast Seminar Series is sponsored by the LTH Profile Area AI & Digitalization.

Thesis defense: Ashkan Sheikhi

Publicerad: 2025-09-04

Title of thesis: Scaling massive MIMO with imperfect transceivers.

Link to thesis in Lund University Research Portal

Zoom link.
Zoom ID: 67836255687.

The number of users and the information transmitted over wireless networks have been growing constantly during the last decades. Nowadays, the pace of this growth is extremely sharp because of the new applications which heavily rely on wireless networks to meet users' demands. Wireless networks infrastructures are constantly developing to meet these demands. Massive multiple-input multiple-output (MIMO) and large intelligent surface (LIS) are two of the main technologies which are the key-enablers for the current and future wireless networks. The performance gains achieved from these system are mainly due to the large number of deployed transceiver chains, which enables serving more users by exploiting spatial domain multiplexing to meet the higher service requirements. The possibility to scale up these systems is a necessity to constantly meet the network demands. Deploying massive MIMO and LIS systems with non-ideal hardware components is of great importance to make the scalability of these systems feasible. In theory, the performance of these systems can grow unboundedly by scaling up the number of transceiver chains. However, assuming ideal hardware components for the transceivers is not realistic from a practical point of view, since the number of transceiver chains are in the order of hundreds to thousands, and the deployment cost, processing complexity, and power consumption can limit the scaling of such systems. 

This work presents an analysis of hardware quality, complexity, power consumption, versus performance of wireless communication systems, with a particular focus on massive MIMO and LIS architectures. We derive closed-form scaling laws that relate analogue front ends power consumption to key system and environmental parameters, such as bandwidth, signal-to-noise-plus-distortion-ratio (SNDR), and fading conditions, enabling informed decisions for low-power design. For massive MIMO systems, we explore both traditional and machine learning-based digital pre-distortion (DPD) strategies. In particular, we propose optimization of per-antenna DPD sizes under hardware constraints and adaptive neural DPD allocation strategies based on channel conditions, demonstrating substantial capacity improvements and system cost reductions. We further analyze the effects of non-ideal receiver chains on LIS, and propose efficient antenna and panel selection schemes to sustain LIS performance with fewer number of transceiver chains. Finally, we propose an over-the-air (OTA) method to jointly perform DPD and reciprocity calibration in massive MIMO and LIS systems, mitigating transmitter non-linearity and non-reciprocity without dedicated hardware or iterative algorithms. Collectively, these contributions provide new insights and tools for the design of energy- and cost-efficient wireless systems that remain robust under realistic hardware constraints.

Publicerad: 2025-07-03

Thesis defense: Exploiting multiple antennas in maritime radio channels.

Publicerad: 2025-06-09

Zoom link

Link to thesis in LUCRIS

Reliable maritime communication systems are essential for both safety-critical operations and emerging applications such as autonomous shipping, remote pilotage, and drone-assisted search and rescue. These scenarios demand ultra-reliable, low-latency wireless connectivity, where communication outages are unacceptable. Equally important is the need for dependable positioning systems. In situations where GNSS (Global Navigation Satellite System) signals are unreliable, a robust backup positioning solution is essential. Recent wireless technology trends, such as massive multiple-input multiple-output (MIMO) in fifth-generation (5G) and distributed MIMO in future sixth-generation (6G) networks, involve the deployment of large-scale antenna arrays to enhance reliability and capacity. Inspired by these developments, this thesis investigates the use of multiple antennas in maritime radio channels to improve both communication reliability and positioning capability. The investigation focuses on sea surface fading mitigation and ranging, supported by both theoretical analysis and real-world measurements. A high-performance wideband distributed massive MIMO channel sounder operating in the 5 GHz band was developed to support this research. The sounder is also well-suited for broader 6G research, including distributed MIMO and joint communication and sensing. The thesis demonstrates that deploying multiple antennas, particularly in vertical configurations, yields significant advantages for maritime wireless systems. Through a combination of analytical modeling and empirical measurements, it shows that vertical antenna arrays can effectively mitigate deep fading caused by sea surface reflections. Closed-form expressions based on the two-ray model are derived to identify optimal antenna spacing and array configurations and to provide practical design insights. Experimental validation in open-sea environments confirms that even arrays of only three elements can enhance link reliability by up to 15 dB. Furthermore, the two-ray model is extended to enable GNSS-independent ranging between vessels and base stations on land without requiring time synchronization, achieving sub-10-meter accuracy with eight vertically distributed antennas, offering a viable independent positioning method.

Thesis defense: Code-based Cryptography: Attacking and Constructing Cryptographic Systems

Publicerad: 2025-05-26

Zoom link.

Link to Thesis in LU Research Portal.

We analyze several lightweight code-based cryptosystems in the first three works, ranging from stream ciphers to wPRFs and authentication protocols. We investigate the design weaknesses that allow us to launch attacks using various techniques. In particular, we analyze a novel LPN-based stream cipher called Firekite, a wPRFs construction, and an HB-like authentication protocol named LCMQ. Using diverse techniques in conjunction with information-set decoding algorithms (ISD), our studies improve previous results (if any) and impose stronger security parameters for said constructions. Then, we draw connections between lattice-solving algorithms and traditional syndrome decoding algorithms with our new proposal: a sieving-style ISD algorithm. Our algorithm offers a novel time-memory trade-off in solving relevant code-based parameters. In the low error-weight regime, the sieving-style ISD can use memory more efficiently without losing its competitiveness in computational performance. Thus, we introduce a valuable and practical alternative to cryptanalysis. The last two papers look at the novel RSDP problem from a new perspective - the Oracle model, analogous to the LWE or LPN problems. We construct an HB-like authentication protocol, replacing the LPN problem with the (Oracle)
RSDP problem, showing its remarkable adaptiveness to the most secure designs. In practice, RSDP structures allow incredibly efficient operations, rivaling those of LPN. Moreover, RSDP also achieves high-security guarantees with modest parameters, yielding significant superiority regarding communication cost. Finally, we expand the cryptanalysis of the RSDP problem, especially when many RSDP samples are allowed with a BKW-style solver. We analyze the concrete complexity of RSDP in new regimes outside of CROSS parameters. Hence, our work is a useful calibrating tool for similar RSDP-based cryptosystems in the future.

Thesis defense: Moving towards cognitive radio access networks: transforming MIMO complexities into opportunities

Publicerad: 2025-05-20

Dino Pjani? defends his thesis.

Zoom link.

Link to thesis in LU Research Portal.

The introduction of Multiple-Input Multiple-Output (MIMO) systems has dramatically transformed wireless communication systems, in particular in the Fifth Generation (5G) New Radio (NR) systems, fundamentally changing how signals are transmitted and received. MIMO technology deploys numerous antennas to transmit and receive multiple data streams simultaneously. The presence of obstructions and scatterers in wireless environments, varying in location, size, and shape, contributes to a high-dimensional feature space. As user devices move, the interaction of electromagnetic radio waves with surrounding objects and devices generates distinct patterns, called spatial fingerprints. By analyzing the behavior of the radio channel in real time through these spatial fingerprints and their temporal evolution, MIMO systems unlock significant opportunities for deeper insight into channel dynamics. These insights lay the groundwork for previously unforeseen functionalities in the Radio Access Network (RAN) domain of cellular networks, moving beyond the constraints of traditional approaches based on mathematical models and solutions.

Thesis defense: Principles and Solutions for Improved Availability and Code Vulnerability Detection

Publicerad: 2025-05-06

Syafiq Al Atiiq defends his thesis "Principles and Solutions for Improved Availability and Code Vulnerability Detection".

Link to thesis in LUCRIS.

Zoom link.
Zoom ID: 68034306533

His research explored the following question: how do we ensure our system remains available when facing attacks or failures?
He focused on three areas. First, he developed protocols for Internet of Things (IoT) devices, shielding these resource-limited devices from being overwhelmed by malicious traffic and help them recover after compromise. Second, he investigated security vulnerabilities in 5G networks, particularly in their AI-driven mobility prediction systems. He showed that even a small number of fake devices can significantly reduce prediction accuracy and identified strategies to make these systems more resilient. Finally, he explored how artificial intelligence can automatically detect vulnerabilities in software before they become security problems. His research revealed that different programming languages and vulnerability types perform variably.
 

LTH AI and Digitalization Breakfast Seminar: Sustainability perspectives towards distributed MIMO for 6G systems

Publicerad: 2025-04-28

For free breakfast at the seminar, please sign up Here no later than May 4th. 

Sustainability perspectives towards distributed MIMO for 6G systems

Speaker: Sara Willhammar, EIT

Abstract: This talk presents early results of an exploratory work and a high-level conceptual framework aiming at providing a structured way of how to qualitatively integrate sustainability perspectives in technology development. We analyze aspects of 6G systems in terms of environmental, social and economic sustainability, from the dimensions of ?sustainable 6G? and ?6G for sustainability?. This sustainability framework is then applied to distributed MIMO, one of the candidate technologies for 6G, to outline challenges and opportunities that should be considered in the development of the technology to be used for 6G systems. 

Speaker Bio: Sara Willhammar is a Postdoc at the Department of Electrical and Information Technology at Lund University, Sweden. She received her dual Ph.D. degree in Electrical Engineering from Lund University, Sweden and KU Leuven, Belgium in 2022, and her M.Sc. degree in Electrical Engineering from Lund University, Sweden, in 2017. Her research interests are within channel measurements, characterization and modeling for next generation wireless systems. She also has an interest in the broader field of wireless communication in terms of exploring sustainability perspectives in the development of such systems.

Thesis defence: On the Trustworthiness of Trusted Third Parties

Publicerad: 2025-02-25

Joakim Brorsson defends his thesis "On the Trustworthiness of Trusted Third Parties."

Zoom link: https://lu-se.zoom.us/j/67331586615
Zoom ID: 67331586615.

External link to thesis. 

Trusted Third Parties (TTPs) are a standard building block in secure systems and are used in many cryptographic protocols. A TTP is an entity which facilitates interaction between the users in a system. As the name suggests, TTPs are trusted, and can thus be relied on to perform critical parts of a protocol without need for external scrutiny. Due to this property, they can serve the role of making protocols easier to design, maintain and participate in.

Since TTPs are trusted, they often have access to some of the most sensitive information in a system; e.g., in backup systems, or when used for privileged roles such as certificate authorities. If a TTP is not as trustworthy as assumed, the consequences can be severe. This is often the case in practice, where multiple examples of failures among high-profile TTPs exist. For example, data leaks are common news and certificate authorities have been caught issuing incorrect certificates. This illustrates the problem that the trustworthiness of a TTP is often more of a convenient assumption than a system property motivated by properly investigated system mechanisms.

Thesis defence: Performance limits for microstrip patch antennas

Publicerad: 2025-02-12

Ben Nel defends his thesis:
Performance Limits for Microstrip Patch Antennas.

Zoom link: https://lu-se.zoom.us/j/63106127825
Zoom ID: 63106127825. 

External link to thesis:
https://portal.research.lu.se/en/publications/performance-limits-for-microstrip-patch-antennas