Gian Marti
Ph.D. Student, Electrical Engineering
ETH Zurich
Integrated Information Processing Group
Email: marti at iis.ee.ethz.ch
I am a final-year doctoral student in electrical engineering and information technology at ETH Zurich. My advisor is Prof. Christoph Studer, and I am part of the Integrated Information Processing (IIP) Group. In my doctoral studies, I work on multi-antenna (MIMO) methods for jammer mitigation in wireless communications. Before I joined the IIP Group, I was with the Signal and Information Processing Laboratory (ISI) under Prof. Hans-Andrea Loeliger, where I worked on blind image deblurring and on inverse problems with discrete constraints. Prior to that, I had written my Master's thesis (also at the ISI, but under the guidance of Prof. Amos Lapidoth) in the area of Shannon theory.
I have obtained my BSc and MSc degrees in electrical engineering and information technology from ETH Zurich in 2016 and 2019, respectively, and I have held internships at ABB, Kistler, and NVIDIA. For the duration of my MSc, I was awarded a scholarship under ETH's Excellence Scholarship and Opportunity Programme. I am a recipient of the ETH Medal. In 2023 and 2024, I achieved second place at the Student Paper Contest of the Asilomar Conference on Signals, Systems, and Computers.
My research interests are at the intersection of fundamental theory, well-principled algorithms, and practical systems. In my doctoral research, I have been able to touch on all of these areas.
Selected Publications
Spatial filtering based on multiple-input multiple-output (MIMO) processing is a powerful method for jammer mitigation. In principle, a MIMO receiver can null the interference of a single-antenna jammer at the cost of only one degree of freedom - if the number of receive antennas is large, communication performance is barely affected. In this paper, we show that the potential for MIMO jammer mitigation based on the digital outputs of finite-resolution analog-to-digital converters (ADCs) is fundamentally worse: Strong jammers will either cause the ADCs to saturate (when the ADCs' quantization range is small) or drown legitimate communication signals in quantization noise (when the ADCs' quantization range is large). We provide a fundamental bound on the mutual information between the quantized receive signal and the legitimate transmit signal. Our bound shows that, for any fixed ADC resolution, the mutual information tends to zero as the jammer power tends to infinity. Our bound also confirms the intuition that for every 6.02 dB increase in jamming power, the ADC resolution must be increased by 1 bit in order to prevent the mutual information from vanishing.
In multiple-input multiple-output (MIMO) wireless systems with frequency-flat channels, a single-antenna jammer causes receive interference that is confined to a one-dimensional subspace. Such a jammer can thus be nulled using linear spatial filtering at the cost of one degree of freedom. Frequency-selective channels are often transformed into multiple frequency-flat subcarriers with orthogonal frequency-division multiplexing (OFDM). We show that when a single-antenna jammer violates the OFDM protocol by not sending a cyclic prefix, the interference received on each subcarrier by a multi-antenna receiver is, in general, not confined to a subspace of dimension one (as a single-antenna jammer in a frequency-flat scenario would be), but of dimension L, where L is the jammer's number of channel taps. In MIMO-OFDM systems, a single-antenna jammer can therefore resemble an L-antenna jammer. Simulations corroborate our theoretical results. These findings imply that mitigating jammers with large delay spread through linear spatial filtering is infeasible. We discuss some (im)possibilities for the way forward.
MIMO processing enables jammer mitigation through spatial filtering, provided that the receiver knows the spatial signature of the jammer interference. Estimating this signature is easy for barrage jammers that transmit continuously and with static signature, but difficult for more sophisticated jammers: Smart jammers may deliberately suspend transmission when the receiver tries to estimate their spatial signature, they may use time-varying beamforming to continuously change their spatial signature, or they may stay mostly silent and jam only specific instants (e.g., transmission of control signals). To deal with such smart jammers, we propose MASH, the first method that indiscriminately mitigates all types of jammers: Assume that the transmitter and receiver share a common secret. Based on this secret, the transmitter embeds (with a linear time-domain transform) its signal in a secret subspace of a higher-dimensional space. The receiver applies a reciprocal linear transform to the receive signal, which (i) raises the legitimate transmit signal from its secret subspace and (ii) provably transforms any jammer into a barrage jammer, which makes estimation and mitigation via MIMO processing straightforward. We show the efficacy of MASH for data transmission in the massive multi-user MIMO uplink.
Wireless systems must be resilient to jamming attacks. Existing mitigation methods based on multi-antenna processing require knowledge of the jammer's transmit characteristics that may be difficult to acquire, especially for smart jammers that evade mitigation by transmitting only at specific instants. We propose a novel method to mitigate smart jamming attacks on the massive multi-user multiple-input multiple-output (MU-MIMO) uplink which does not require the jammer to be active at any specific instant. By formulating an optimization problem that unifies jammer estimation and mitigation, channel estimation, and data detection, we exploit that a jammer cannot change its subspace within a coherence interval. Theoretical results for our problem formulation show that its solution is guaranteed to recover the users' data symbols under certain conditions. We develop two efficient iterative algorithms for approximately solving the proposed problem formulation: MAED, a parameter-free algorithm which uses forward-backward splitting with a box symbol prior, and SO-MAED, which replaces the prior of MAED with soft-output symbol estimates that exploit the discrete transmit constellation and which uses deep unfolding to optimize algorithm parameters. We use simulations to demonstrate that the proposed algorithms effectively mitigate a wide range of smart jammers without a priori knowledge about the attack type.