The detection of surface defects on printed circuit board (PCB) plays a vital role. However, current defect detection methods face significant challenges, such as frequently misidentifying non-defective areas as defects, low defect recognition capability, and diminished accuracy in identifying minor defects. To address these challenges, a novel defect detection network based on the YOLOv7-tiny network framework is proposed (SMC-YOLO). In this paper, we have improved the (Spatial Pyramid Pooling, Cross Stage Partial Channel) SPPCSPC module by adding three additional convolutional layers and introducing an extra feature channel below the max-pooling layer. These enhancements help in capturing fine-grained features and preserving crucial information, which is essential for detecting tiny defects on PCB. Additionally, we have modified the connection method between the max-pooling layers, changing from a parallel to a serial connection to reduce information loss. On this basis, incorporating the multi-head self-attention (MHSA) mechanism at the SPPCSPC module output enhances the network’s ability to focus on critical features across various scales, effectively capturing and utilizing detailed information, thereby improving the detection performance for small defects. Furthermore, the introduction of the content-aware reassembly of features (CARAFE) lightweight upsampling operator restores fine-grained details like defect edges and textures. It also preserves contextual information such as spatial relationships and background consistency, resulting in improved upsampling outcomes. Finally, The effective integration of the normalized wasserstein distance (NWD) loss function and the efficient IoU (EIoU) loss function further enhances the positioning accuracy and convergence efficiency of the SMC-YOLO defect detection algorithm. In comparison to the original YOLOv7-tiny network, the proposed network showcases notable advancements in detection performance. The experimental results demonstrate that the proposed network achieves a detection speed of 114.94 FPS while maintaining the mAP value of 97.4%. It is evident that the SMC-YOLO network effectively detects surface defects in PCB manufacturing and holds promising potential for implementation on embedded systems.

The dielectric resonator antenna (DRA) with wide bandwidth, compact size and low profile is considered as an attractive candidate for 5G wireless communications. However, most of the reported DRAs either have bulky volumes or have limited bandwidths. In this paper, a kind of compact and low-profile DRA with extended bandwidth is proposed. By observing the E-field distribution differences between the two target modes and modifying the dielectric characteristics (such as material and dimension) of the designated region where the E-field is very weak for fundamental TE ₁₁₁ mode but quite strong for high-order TE ₁₃₁ mode, the frequency of TE ₁₃₁ mode can be significantly affected and shifted down to merge with that of TE ₁₁₁ mode which is less affected. As a result, extended operating bandwidth can be obtained. This technique also benefits from not having to enlarge the planar size of the high permittivity DR, therefore making the antenna compact enough for 5G beam-scanning and low profile applications. For demonstration, linear- and circular- polarization antenna prototypes were designed and measured. The LP antenna has a −10 dB impedance bandwidth of 17.3% and a peak gain of 7.1 dBi while the CP antenna achieves a 3-dB axial ratio (AR) bandwidth of 12% and a peak gain of 6.6 dBic. Both the antennas have compact volumes of no larger than $0.35{\lambda } _{0} \times 0.35{\lambda } _{0} \times 0.11{\lambda } _{0}$ . Based on the LP and CP antenna elements, two $1\times5$ antenna arrays with the capability of beam scanning are designed and simulated. Wide scanning angles of ±45° and ±40° can be obtained for the LP and CP arrays, respectively.

A hybrid method of the volume integral equation based overlapping domain decomposition method and the FFT-based method (VIE-ODDM-FFT) using nested uniform Cartesian grid is proposed for the analysis of electrically large inhomogeneous dielectric objects whose usual fast algorithm models are too large for the user’s computer to accommodate. The integral equation is first built in the entire volumetric domain, and then the entire volumetric domain is partitioned into some non-overlapping geometric subdomains. The SWG basis functions are used to expand the equivalent electric flux density in the entire domain. Then several concepts such as basic subdomain, extended region and complementary region are introduced. The subproblem in an extended region is quickly solved by using the FFT-based method with DILU preconditioner, and the action of the complementary region of an extended region on this extended region itself is quickly calculated by means of FFT. In this scheme, FFT acceleration is achieved by fitting Green’s function (FG) onto the nodes of one or more nested uniform Cartesian grid. The convergence of the outer-iterative scheme of the proposed VIE-ODDM is investigated theoretically and numerically, and demonstrated to be very good, and hence, a very large problem can be effectively solved in an ordinary computer. Some numerical examples are provided to demonstrate the correctness, robustness and efficiency of the proposed method.

In this paper, a novel grouping scheme of the basis functions within the framework of the FFT-based method is proposed for creating a block-sparse structure for the near-matrix, and then FFT-based method with near-matrix compression is established for the efficient analysis of multiscale problems. For a multiscale problem, if a FFT-based method is required to maintain both higher efficiency and higher accuracy at calculating the far-interactions, then the near-matrix will be inevitably very large. Compared with the traditional FFT-based method, the proposed method can significantly reduce the near-matrix memory requirement without increasing the time spent. Several numerical examples are provided to demonstrate the correctness and the efficiency of the proposed method.

In Wireless Sensor Networks (WSN), there are three conflicting important issues, including extensive coverage, low node redundancy, and energy consumption for secondary deployment. To balance network coverage and longevity, a WSN coverage model with three objectives is developed. A novel multi‐strategy gold rush optimizer is proposed for the coverage optimization problem in WSN. First, the fusion search method becomes extensive and flexible by combining the migration strategy with the Levy flight mechanism and joining a follower group. Then, the dynamic opposite learning strategy is used to increase population diversity in the panning stage. In addition, a multi‐player collaborative strategy is utilized to improve search ability of the algorithm in the collaboration stage. Finally, the random differential mutation method is used to try to assist the algorithm in escaping local optima. The simulation results show that this algorithm compared with other algorithms performs better on all three objectives considered.

This letter presents a novel low-profile patch antenna with shared aperture and two-dimensional (2-D) scanning capability. Compared with the state-of-the-art designs, the proposed antenna offers low cost and easy integration due to its simple structure. The millimeter-wave (MMW) stacked patch antenna array is embedded in the weak electric-field (E-field) region of the anti-phase TM20 mode and TM12 mode of the microwave (MW) dual-patch antenna. This design interleaves two antenna types within the same aperture, achieving natural isolation. Slots on the MW patch modify the current path of the anti-phase TM20 mode and the TM12 mode, enabling dual-mode operation with low profile. The stacked patches provide wideband operation in MMW frequency band. Additionally, periodic slots on the MW patch facilitate MMW beam scanning, achieving ±45o beam scanning at both E- and H-planes. For demonstration, a prototype of the proposed MW and MMW shared aperture patch antenna is fabricated and measured, which has a total size of 0.38λL × 0.48λL × 0.022λL (λL is the operating wavelength at 4.6 GHz in free space). Measurements show a 10-dB impedance matching bandwidth of 6.6% and 27.1%, and a peak gain of 7.2 dBi and 15.5 dBi in MW and MMW bands, respectively.