Chinese People's Armed Police Force Engineering University

Abstract Most existing research on High Efficiency Video Coding (HEVC) video steganography is based on the premise that the receiver can acquire an unaltered bitstream. However, videos uploaded to social media platforms (e.g., YouTube, WeChat, TikTok) are prone to be re-encoded. This process can corrupt or entirely remove the information hidden in the original bitstream. To address this issue, this paper proposes a novel scheme that jointly leverages the frequency domain and the compressed domain. In the frequency domain, an invertible neural network is designed to conceal secret message, where the modifications are directly written into the YUV file. In the compressed domain, data is hidden by modifying Coding Unit (CU) partition during the HEVC encoding process. The impacts of recompression and the available embedding positions are presented through detailed statistics and analysis. Experimental results demonstrate that the proposed algorithm maintains high visual quality and security. Crucially, it ensures reliable and accurate extraction of the secret information even after the video has undergone recompression.

Internet of Things (IoT) facilitates intelligent interconnection and data exchange between devices. However, existing data aggregation schemes face challenges such as edge node disconnections, high computational overhead, and limited fault tolerance, which affect the reliability and efficiency of the system. To address these issues, this paper proposes a smart contract assisted and fault-tolerant data aggregation scheme without a trusted authority, named Cap. To realize it, we integrate blockchain technology and homomorphic encryption. Through smart contracts, we dynamically select and switch edge nodes to ensure that the data aggregation process continues even if some nodes go offline. Moreover, by leveraging homomorphic secret sharing, we effectively minimize communication overhead and ensure that users can exit the aggregation process without jeopardizing data integrity. This scheme provides an efficient and reliable solution for secure data aggregation in Iot environments and enhances the robustness and data security of the system. Performance evaluation show that, compared with existing schemes, Cap reduces computational and communication overhead by 30%.

This letter presents a wideband, high-gain, and low-profile dual-layer metasurface (MTS) circularly polarized (CP) antenna array for millimeter-wave (MMW) applications. A modal-split synergistic frequency tuning (MSFT) strategy guided by characteristic mode analysis (CMA) is proposed: by achieving controlled separation of orthogonal modes via H-shaped slots and utilizing the dual-layer MTS structure for synergistic tuning to broaden the modal bandwidth, this strategy fundamentally realizes the synchronized expansion of impedance and axial ratio (AR) bandwidths. Additionally, a novel dual-layer LC decoupling resonator (LCDR) network is introduced to efficiently suppress array mutual coupling by generating anti-phase compensating magnetic fields. Measured results of a 4 × 4 array prototype indicate that, with an ultra-low profile of only 0.13<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\lambda _{0}$</tex-math></inline-formula>, it achieves an impedance bandwidth of 39.20%, an AR bandwidth of 38.75%, and a peak gain of 20.45 dBic.

Background: In recent years, an accurate assessment of gun shooting performance is increasingly critical for Internet-of-Things (IoT)-based athlete training and shill monitoring. Where shooting actions involve complex wrist-gunarm coordination, rapid impulse-driven motions and shooterspecific execution patterns. Challenges: Existing Virtual Reality Shooting Performance Assessment System (VR-SPAS) fail to capture phase-specific coordination such a pre-short aiming stability, trigger pull smoothness and post-shot recoil recovery, even though VR-SPAS perform adequately in overall shooting outcomes. Proposed Model: To address these limitations, this research proposes an Action-Phase Gun Shooting Training System (AP-GSTS) for personalized training. Initially, multiaxis Inertial Measurement Units (IMU) signals are collected from wrist-mounted and gun-mounted sensors, which are synchronized with trigger events and segmented into pre-shot, firing as well as post-shot phases. Then, collected signals are preprocessed with Temporal Motion Normalization and Smoothing (TMNS) to preserve micro-instabilities. A Temporal Sequence Encoder (TSE) models phase-aware motion dynamics, which are compared to expert reference sequences to compute interpretable scores for stability, trigger execution and recoil control. Finally, an actionable feedback generator translates these scores into phase-specific recommendations for targeted skill improvement and adaptive training. Results: The experimental results demonstrates that the proposed AP-GSTS framework in terms of mean of 21.52%, Standard Deviation (SD) of 2.22 from 0-25 range, when compared to existing VRSPAS approach.

Health assessment is a critical component of prognostics and health management (PHM), and its accuracy and real-time performance directly determine the effectiveness of subsequent prediction, maintenance, and resource scheduling. In current health assessment practices, the weights of health indicators are typically calculated using a single weighting method, which often results in unreliable outcomes. To address this issue, this paper proposes a combined weighting method based on the geometric principle for the health assessment of launch-vehicle engines. This method primarily integrates the Analytic Hierarchy Process (AHP) and the Entropy Weight Method, taking both subjective experience and objective data into account. Finally, the rationality of the proposed method is validated through a case study.