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.

A millimeter-wave (mm-wave) dual-strip coupled dual-patch antenna is proposed, which is achieved by high density interconnect antenna-in-package process. The dual patches and dual strips are closely placed on the same layer and produce two resonances with broadside radiation and x-direction polarization. Compared with the state-of-the-art single-layer patch antennas, the proposed antenna can achieve both compact radiator size and wide bandwidth, which is useful to form mm-wave array. Furthermore, the antenna radiator and the feedline are separated by the antenna ground to decrease the interaction from radio frequency components in the active system. Characteristic mode analysis method is utilized to analyze the antenna element to reveal the modal behavior of the proposed antenna. For demonstration, a prototype of 2 × 4 array is fabricated, which has the antenna element size of 0.36 × 0.37 × 0.05 λ ₀ ³ (λ ₀ is the operating wavelength at 26 GHz in free space), the 10 dB impedance matching bandwidth of 20.6% (22.7-27.9 GHz), and the peak gain of 14.2 dBi.

A millimeter-wave (mm-Wave) quasi-Yagi antenna utilizing a pair of back-to-back U-shaped full-wavelength dipoles and a folded shorted stub-loaded half-wavelength strip resonator is proposed. The high impedance of the U-shaped full-wavelength dipole overcomes the mismatching problem when it is close to the reflector, enabling compact size in the endfire direction. Three reflection zeros can be obtained from the full-wavelength dipole, the half-wavelength strip resonator, and the equivalent inductor and capacitor derived from the folded shorted stub. The three reflection zeros and appropriate impedance make the proposed antenna have wide bandwidth. For demonstration, a prototype of 1 × 4 array is designed with the antenna element planar size (without ground) of 0.57 λ ₀ × 0.18 λ ₀ (λ ₀ is the wavelength in the free space at center frequency), the 10-dB impedance matching bandwidth of 49.7% (24.2 - 40.2 GHz), and the average gain of 9.12 dBi. The proposed compact wideband antenna array can be a good candidate for 5G mm-Wave wireless application.

A millimeter-wave (mm-wave) substrate integrated waveguide (SIW)-fed endfire antenna is proposed by using double-faced multiple dielectric patch (DFMDP) resonator, named SIW-fed endfire DFMDP antenna. The DFMDP resonator consists of DFMDPs and the extend substrate of the SIW, where DFMDPs are a number of dielectric patches on both surfaces of the extend substrate of the SIW. The endfire radiation is supported by the DFMDP resonator as well as the open-ended SIW. Multiple TE ^y _{n 11} modes and low quality factor can be found in the DFMDP resonator to obtain wideband operation. Four antennas with different number of DFMDPs are studied and compared. The antenna using 3 × 3 DFMDPs is preferred considering both the antenna performance and process capability. A prototype operating at 33.5 GHz is designed, analyzed, and demonstrated. The measured results exhibit the 10 dB impedance matching bandwidth of 19.19 GHz (23.97-43.16 GHz) or 57.14% and the average gain of 9.45 dBi.

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.