Utilizing complex-valued neural networks (CVNNs) in wireless communication tasks has received growing attention for their ability to provide natural and effective representation of complex-valued signals and data. However, existing studies typically employ complex-valued versions of simple neural network architectures. Not only they merely scratch the surface of the extensive range of modern deep learning techniques, theoretical understanding of the superior performance of CVNNs is missing. To this end, this paper aims to fill both the theoretical and practice gap of employing CVNNs in wireless communications. In particular, we provide a comprehensive description on the various operations in CVNNs and theoretically prove that when the output dimension is large, the CVNN requires fewer layers than its real-valued counterpart to achieve a given approximation error of a continuous complex-valued function. Furthermore, to advance CVNNs in the field of wireless communications, this paper focuses on the transformer model, which represents a more sophisticated deep learning architecture and has been shown to have excellent performance in wireless communications but only in its real-valued form. In this aspect, we propose a fundamental paradigm of complex-valued transformers for wireless communications, including the complex-valued embedding module, encoding module, decoding module, and output projection module. Leveraging this structure, we develop customized complex-valued transformers for three representative applications in wireless communications: channel estimation, user activity detection, and joint design of pilot, feedback quantization, and precoder. These applications utilize transformers with varying levels of sophistication and span a variety of tasks, ranging from regression to classification, supervised to unsupervised learning, and specific module design to end-to-end design. Experimental results verify the theoretical advantage and effectiveness of the complex-valued transformers for the above three applications compared to other traditional real-valued neural network-based methods.

DreamCreek is a system for grading the capacity of lithium-ion battery cells. An electric vehicle (EV) battery pack consists of thousands of lithium-ion cells; these cells must have a balanced capacity, measured by a formation and grading phase. This phase is costly, taking 14-20+ hours. DreamCreek aims to optimize this process, by collecting data from partial charge. Using the data, it determines the capacity of lithium-ion cells by employing both machine learning prediction and logic deduction, to reduce the usage of energy and increase the production. We will demonstrate how DreamCreek works with a guided tour, and show how it reduces the time of the formation and grading phase to 4 hours, with an error rate in the range of [0.06 %, 1%].

This article provides a systematic overview of the basics of talkative power conversion (TPC). TPC is an emerging technique for simultaneous information and power transmission, in which data modulation is integrated into a switched-mode power converter. The data sequence is embedded in the ripple voltage, which is superimposing the output voltage of the converter. In contrast to conventional power line communication (PLC), TPC can be used universally, not only in grid applications. Aspects of power electronics (PE) and digital communication are presented in a structured form, including new perspectives such as multiple-input multiple-output (MIMO) techniques applied to TPC, adaptive modulation and channel coding, and advanced receiver design with adaptive channel and load estimation. The new aspects aim to mitigate the inherent shortcomings of TPC.

Deep learning-based Autonomous Driving (AD) perception models often exhibit poor generalization due to data heterogeneity in an ever domain-shifting environment. While Federated Learning (FL) could improve the generalization of an AD model (known as FedAD system), conventional models often struggle with under-fitting as the amount of accumulated training data progressively increases. To address this issue, instead of conventional small models, employing Large Vision Models (LVMs) in FedAD is a viable option for better learning of representations from a vast volume of data. However, implementing LVMs in FedAD introduces three challenges:(I)the extremely high communication overheads associated with transmitting LVMs between participating vehicles and a central server;(II)lack of computing resource to deploy LVMs on each vehicle;(III)the performance drop due to LVM focusing on shared features but overlooking local vehicle characteristics. To overcome these challenges, we propose pFedLVM, a LVM-Driven, Latent Feature-Based Personalized Federated Learning framework. In this approach, the LVM is deployed only on central server, which effectively alleviates the computational burden on individual vehicles. Furthermore, the exchange between central server and vehicles are the learned features rather than the LVM parameters, which significantly reduces communication overhead. In addition, we utilize both shared features from all participating vehicles and individual characteristics from each vehicle to establish a personalized learning mechanism. This enables each vehicle’s model to learn features from others while preserving its personalized characteristics, thereby outperforming globally shared models trained in general FL. As a demonstration of the proposed pFedLVM, this paper focuses on the semantic segmentation (SSeg) task. Extensive experiments demonstrate that pFedLVM outperforms the existing state-of-the-art approach by 18.47%, 25.60%, 51.03% and 14.19% in terms of mIoU, mF1, mPrecision and mRecall, respectively.

Talkative power conversion (TPC), which embeds information into the switching ripple to achieve the simultaneous transmission of power and data without installation of additional hardware, can reduce the additional communication infrastructure in cyber-physical systems of smart grid and renewable power plants. The TPC in DC/DC converters has been the subject of numerous investigations, but comparatively few studies have focused on the DC/AC converters. In this paper, the space-vector pulse width modulation (SVPWM)-based TPC techniques including both variable zero vector width modulation (VZVWM) and variable zero vector position modulation (VZVPM) for three-phase DC/AC converters are proposed. The DC/AC converters can transmit information to the receiver, crossing the LCL filter and power cable, with a good power quality. The sliding FFT method is employed for demodulation, and the bit rate is analyzed. Finally, simulations and experiments prove the viability of the SVPWM-based three-phase DC/AC talkative power converters with a bit rate of 100 bps and 1-blt length time delay.

Talkative power conversion (TPC), which can simultaneously transfer the energy and data by the power electronics converters and avoid using extra communication infrastructures, is an emerging technique for the digital power and digitalization of power system. In this paper, the TPC technology is used in the dual-active bridge (DAB) converter, in which the switching frequency and common phase are used as the degrees of control freedom respectively to modulate the data and power simultaneously. The transient performances of the DAB converters modulated by the PWM-frequency shift keying (PWM-FSK), which is a simple and effective technique for the TPC. The constraints on carrier frequency are derived, and the influence of data communication on power conversion is analyzed. The demodulation strategy based on the fast Fourier transform (FFT) algorithm is used to acquire the sent data. The simulation and experimental results clearly validate the effectiveness and correctness of the theoretical analysis.

A variety of machine learning approaches have been applied to synthetic aperture radar (SAR) automatic target recognition. The performances of these approaches rely strongly on the quality and quantity of training data. In real-world applications, however, it is challenging to obtain sufficient data suitable for these approaches. To alleviate this problem, a novel deep generative model for SAR image generation is proposed, which is an extension of Wasserstein autoencoder. The network structure and reconstruction loss function of the model have been improved according to the characteristics of SAR images. The experimental results demonstrate that our model is superior to other classical generative models in SAR image generation. The generated images can be directly used as training samples, thereby extending the training data set and improving the recognition accuracy.