CsiNet, a deep neural network framework, utilizes an autoencoder to efficiently transmit downlink channel state information (CSI) in the feedback link, which reduces the cost of feedback, and significantly improves the quality of the reconstruction. However, the model with massive parameters incurs a lot of storage space and high computational complexity, which is impractical for low‐cost and low‐power edge devices. In this work, a lightweight CsiNet based SNIP (Single‐shot Network Pruning) is implemented, which prunes the model using the gradient information of the first training epoch, eliminating both pretraining and the complex pruning schedule. Numerical simulation results show that, under the same compression rates, the method can achieve a similar or even better reconstruction effect and more effectively reduce computational complexity, compared to traditional lightweight methods.

We propose a lightweight CsiNet model based on SNIP. SNIP‐based CsiNet model is more suitable for practical engineering applications, especially resource‐constrained mobile edge devices.image

Current natural gas load forecasting encounters with the conundrum of unsatisfying accuracy and interpretability. To address the challenge, a multi‐variate forecasting method is proposed, which contains three phases: First, an integrate history‐climate‐holiday factor set is established to provide multi‐perspective for a more explainable forecast; Second, factor fusion interaction between features and instances is carried out based on hierarchical contrastive learning, which contributes to inter‐intra factors potential relationships exploration. Third, a multivariate forecasting model named ResRNN is trained using fused target dataset. Due to its innovation in structure and loss, forecasting accuracy is further improved. Finally, the authors’ method's superiority is confirmed by several groups of comparative experiments and results demonstrate that it outperforms mainstream methods.

This letter proposes a multivariate natural gas load forecasting method, the core of which lies in integrate history‐climate‐holiday factors fusion and multivariate forecasting model named ResRNN. Rather than over‐reliance on historical gas load series, the authors’ method introduces and fuses essential influencing factors for target dataset enrichment by performing contrastive learning in a hierarchical way over feature‐wise and instance‐wise. To forecast gas load accurately based on the fused target dataset, ResRNN is proposed to innovatively utilize residual connections for more direct temporal trends capture and interpretability improvement.image

In this letter, a design of virtual guarded SiPMs fabricated in a standard 0.35μm standard complementary metal oxide semiconductor (CMOS) process is introduced. The performance of these virtual guarded cells (VGC) is compared to that of conventional cells with real guard rings, referred to as physical guarded cells (PGC). Specifically, the photon detection efficiency (PDE) of both types of SiPMs is evaluated. For this, PDE calculations are conducted at different overvoltages (OVs) and the PDE is compared depending on the wavelength before conducting measurements. The results demonstrate that the VGC SiPM outperforms the PGC SiPM, exhibiting a true PDE of (22.5 ± 0.5)%, which is significantly higher than the PDE of (10.9 ± 0.3)% obtained for the PGC SiPM. The superior PDE of the VGC SiPM is attributed to a larger active or photosensitive area due to the virtual guard rings and a thinner n‐layer in the photosensitive region.

In this letter, we propose a design for virtual guarded silicon photomultipliers (SiPMs) manufactured using a standard 0.35 m CMOS process. We compare their performance to that of conventional SiPMs with physical guard rings. Our results show that the virtual guarded SiPMs achieve a higher photon detection efficiency (PDE) of (22.5 ± 0.5)% compared to (10.9 ± 0.3)% for the physical guarded SiPMs.image

The article shows that transient chaos phenomena can be observed in a generalized memristor Chua's circuit where a nonlinear resistor is introduced to better model the real memristor behaviour. The flux‐charge analysis method is used to explain the origin of transient chaos, that is attributed to the drift of the index of the memristor circuit invariant manifolds caused by the charge flowing into the nonlinear resistor.

Transient chaos phenomena are shown in a Chua's circuit with a real memristor device. The origin of transient chaos is explained via the drift of invariant manifolds caused by the charge in a nonlinear resistor.image

This letter presents an improved Self‐Interference Cancellation (SIC) receiver with Demodulation Reference Signal (DMRS)‐based power estimation for In‐band Full‐duplex (IBFD) transmission. The performance of the SIC scheme is highly dependent on the power of transmit signal. The DMRS‐based power estimation can obtain accurate power. Numerical results show that SIC with DMRS‐based power estimation improves the performance of an IBFD system.

This manuscript presents a novel approach to enhance self‐interference cancellation (Self‐IC) in In‐band Full‐duplex (IBFD) transmission. By utilizing Demodulation reference signal (DMRS)‐based power estimation, the proposed technique accurately estimates the power of the self‐interference signal, leading to improved Self‐IC quality and higher data throughput in IBFD communication systems. Experimental results demonstrate the effectiveness of the DMRS‐based power estimation approach, highlighting its significance in achieving high‐frequency efficiency.image