An Improved Acoustic Diffusion Equation Model for Long-Channel Underground Spaces.

The acoustic diffusion equation model has been widely applied in various scenarios, but a larger prediction error exists when applied to underground spaces, showing a significantly lower characteristic of the sound pressure level in the later stage compared to field tests since underground spaces have a more closed acoustic environment. Therefore, we analyze the characteristics of underground spaces differentiating from aboveground spaces when applying the model and propose an improved model from the perspective of energy balance. The energy neglected in the calculation of the acoustic diffusion equation model is compensated in long channel underground spaces named "acoustic escape compensation". A simulation and two field experiments are conducted to verify the effectiveness of the proposed compensation strategy in long-channel underground spaces. The mean square error is used to evaluate the differences between the classical model and the improved model, which shows a numerical improvement of 1.3 in the underground field test. The results show that the improved model is more suitable for describing underground spaces. The proposed strategy provides an effective extension of the acoustic diffusion equation model to solve the problem of sound field prediction and management in underground spaces.

From Signal to Image: Enabling Fine-Grained Gesture Recognition with Commercial Wi-Fi Devices.

Gesture recognition acts as a key enabler for user-friendly human-computer interfaces (HCI). To bridge the human-computer barrier, numerous efforts have been devoted to designing accurate fine-grained gesture recognition systems. Recent advances in wireless sensing hold promise for a ubiquitous, non-invasive and low-cost system with existing Wi-Fi infrastructures. In this paper, we propose DeepNum, which enables fine-grained finger gesture recognition with only a pair of commercial Wi-Fi devices. The key insight of DeepNum is to incorporate the quintessence of deep learning-based image processing so as to better depict the influence induced by subtle finger movements. In particular, we make multiple efforts to transfer sensitive Channel State Information (CSI) into depth radio images, including antenna selection, gesture segmentation and image construction, followed by noisy image purification using high-dimensional relations. To fulfill the restrictive size requirements of deep learning model, we propose a novel region-selection method to constrain the image size and select qualified regions with dominant color and texture features. Finally, a 7-layer Convolutional Neural Network (CNN) and SoftMax function are adopted to achieve automatic feature extraction and accurate gesture classification. Experimental results demonstrate the excellent performance of DeepNum, which recognizes 10 finger gestures with overall accuracy of 98% in three typical indoor scenarios.

A novel finite-time average consensus protocol based on event-triggered nonlinear control strategy for multiagent systems.

We present a novel finite-time average consensus protocol based on event-triggered control strategy for multiagent systems. The system stability is proved. The lower bound of the interevent time is obtained to guarantee that there is no Zeno behavior. Moreover, the upper bound of the convergence time is obtained. The relationship between the convergence time and protocol parameter with initial state is analyzed. Lastly, simulations are conducted to verify the effectiveness of the results.