Mitsubishi Electric (France)

Phase discrepancy is a major source of error in the measurements of core losses, so reactive cancellation techniques are used to improve the measurement accuracy. To do so, an air core transformer is introduced in the measurement setups. In this work, an alternative methodology for correction of the phase discrepancy is presented, which makes use of the air core transformer introduced in previous techniques. Unlike reactive cancellation techniques, whose objective is to reduce the reactive voltage in the core under test, the proposed method directly addresses the phase discrepancy based on the fundamental assumption that there are no core losses on an air core transformer. The proposed methodology does not rely on Fast Fourier Transform algorithms, which minimizes the required computational resources for the post-processing of the measured waveforms.

ABSTRACT Bio‐based materials with high elongation are essential for applications requiring flexibility and durability, such as packaging, wearable sensors, or conformable biomedical films. Biobased films were produced by solvent casting from suspensions of Hydroxypropyl methylcellulose (HPMC) and TEMPO‐oxidized cellulose nanofibrils (CNFt). The rheological properties were investigated through flow curves and strain sweep measurements. Even with a small addition of nanocellulose, 6 wt%, the resulting blend exhibited strong shear‐thinning behavior characteristics of CNFt. CNFt contributes to structuring the suspension, even if the percolation threshold is not reached, proving reinforcing interactions between HPMC and CNFt. The films ensued from these suspensions were characterized by tensile testing. Increasing the nanocellulose content in the deformable HPMC films enhanced mechanical strength. However, further increasing the CNFt content resulted in very brittle films. A screening of potential plasticizers to enhance elongation property was then made. Incorporating 20 wt% of polyurethane diol increases strain at break by 58%, while 20 wt% of glycerol led to a 107% increase in comparison with the pristine HPMC/CNFt films. By adjusting the HPMC/CNFt ratio and adding an appropriate plasticizer, the mechanical properties of the films can be tailored to specific needs, leading to the production of transparent, homogeneous self‐standing films with improved mechanical deformation.

This paper proposes a Bayesian prediction framework for the dynamic interference, caused by the movement of the interfering device. The framework relies on a hidden Markov model where the hidden state is the device mobility and the observable is the measured interference. To efficiently predict the interference, we consider jointly exploiting the mobility correlation and the radio correlation in a probabilistic manner. The benefits in doing so is to keep all available information about the estimation meanwhile to have the flexible to handle the algorithm output. The numerical evaluations shows that the proposed approach outperforms the conventional algorithm such as the Auto-Regressive prediction.

This work presents a new method for enhancing communication efficiency in stochastic Federated Learning that trains over-parameterized random networks. In this setting, a binary mask is optimized instead of the model weights, which are kept fixed. The mask characterizes a sparse sub-network that is able to generalize as good as a smaller target network. Importantly, sparse binary masks are exchanged rather than the floating point weights in traditional federated learning, reducing communication cost to at most 1 bit per parameter (Bpp). We show that previous state of the art stochastic methods fail to find sparse networks that can reduce the communication and storage overhead using consistent loss objectives. To address this, we propose adding a regularization term to local objectives that acts as a proxy of the transmitted masks entropy, therefore encouraging sparser solutions by eliminating redundant features across sub-networks. Extensive empirical experiments demonstrate significant improvements in communication and memory efficiency of up to five magnitudes compared to the literature, with minimal performance degradation in validation accuracy in some instances.

Modern communication systems rely on accurate channel estimation to achieve efficient and reliable transmission of information. As the communication channel response is highly related to the user's location, one can use a neural network to map the user's spatial coordinates to the channel coefficients. However, these latter are rapidly varying as a function of the location, on the order of the wavelength. Classical neural architectures being biased towards learning low frequency functions (spectral bias), such mapping is therefore notably difficult to learn. In order to overcome this limitation, this paper presents a frugal, model-based network that separates the low frequency from the high frequency components of the target mapping function. This yields an hypernetwork architecture where the neural network only learns low frequency sparse coefficients in a dictionary of high frequency components. Simulation results show that the proposed neural network outperforms standard approaches on realistic synthetic data.

Constant Modulus Algorithm (CMA) in dual-polarization optical coherent transmission is revisited for optical access. We introduce a refined CMA, exhibit singularity-free operation up to 5 ps/km <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0.5</sup> polarization mode dispersion in 32GBaud-QPSK system and suggest new ideas.

Channel charting builds a map of the radio environment in an unsupervised way. The obtained chart locations can be seen as low-dimensional compressed versions of channel state information that can be used for a wide variety of applications, including beam prediction. In non-standalone or cell-free systems, chart locations computed at a given base station can be transmitted to several other base stations (possibly operating at different frequency bands) for them to predict which beams to use. This potentially yields a dramatic reduction of the overhead due to channel estimation or beam management, since only the base station performing charting requires channel state information, the others directly predicting the beam from the chart location. In this paper, advanced model-based neural network architectures are proposed for both channel charting and beam prediction. The proposed methods are assessed on realistic synthetic channels, yielding promising results.

Multicore processor architectures facilitate the integration of applications with different real-time and security requirements. We propose a mixed-criticality system relying on hardware-assisted virtualization that not only shares CPU and memory resources, but also an Ethernet Time-Sensitive Networking (TSN) interface, between applications, taking into account their Quality of Service (QoS) and security requirements. This paper shows how TSN network features can be efficiently exploited in industrial and automotive mixed-criticality systems.

Modern communication systems rely on accurate channel estimation to achieve efficient and reliable transmission of information. As the communication channel response is highly related to the user's location, one can use a neural network to map the user's spatial coordinates to the channel coefficients. However, these latter are rapidly varying as a function of the location, on the order of the wavelength. Classical neural architectures being biased towards learning low frequency functions (spectral bias), such mapping is therefore notably difficult to learn. In order to overcome this limitation, this paper presents a frugal, model-based network that separates the low frequency from the high frequency components of the target mapping function. This yields an hypernetwork architecture where the neural network only learns low frequency sparse coefficients in a dictionary of high frequency components. Simulation results show that the proposed neural network outperforms standard approaches on realistic synthetic data.