ค้นหาผลงานวิชาการ

Fall detection for elderly care has gained substantial research attention, with Wi-Fi Channel State Information (CSI) emerging as a promising non-intrusive sensing modality that addresses privacy concerns associated with camera-based systems. However, the adoption of deep learning in this domain requires models that balance high accuracy with computational efficiency, enabling deployment on resource-constrained devices. This study presents a comparative analysis of four lightweight Convolutional Neural Network (CNN) architectures-MobileNetV2, MobileNetV3, EfficientNet-B0, and ShuffleNetV2-that has been presented for Wi-Fi CSI-based fall detection. The models were evaluated using classification metrics, including accuracy, loss, precision, recall, and F1-score, as well as efficiency metrics such as model size, parameter count, FLOPs, and inference time. Experimental results showed that the highest classification accuracy of 98.61% was achieved by EfficientNet-B0, along with the lowest loss and best F1-score (0.986), while a near-fastest inference time of 8.53 ms was maintained. This combination of predictive performance and computational efficiency has been demonstrated to highlight strong potential for practical, real-time fall detection applications. Meanwhile, ShuffleNetV2 achieved a comparable accuracy of 98.47% and an $F 1$-score of 0.985, with the fastest inference time of $\mathbf{7. 1 9}$ ms and a small model size of only 5.21 MB. These results indicate that ShuffleNetV2 provides an excellent trade-off between accuracy and computational cost, making it a highly suitable candidate for deployment on real-time edge devices.

Increasing awareness of sustainability issues, together with ever‐changing technological disruption, has led enterprises, especially in emerging markets, to invest more in digital technologies to enhance their sustainable competitiveness in the market. Relying on 163 firm‐year observations covering the years 2019–2023 from Indonesia, Malaysia, and Thailand, we explore the effects of digital transformation (DT) on the corporate sustainability performance. Our findings support resource‐based view theory , suggesting the adoption of digital technology by firms primarily drives their sustainability performance, particularly in environmental and social domains. To ensure the robustness of our main findings, we perform several robust tests, including the 2SLS IV method, panel fixed‐effect regression, and quantile regression. The practical implications of our findings highlight the importance for both listed companies and regulators in developing corporate and national digital strategies aimed at improving the corporate sustainability performance of the country.

Nowadays, leading manufacturers of enterprise-grade networking devices offer the dynamic ARP inspection (DAI) feature in their Ethernet Switches to detect and prevent ARP cache poisoning attacks from malicious hosts. However, MikroTik Ethernet switches do not yet support this feature. Within MikroTik-based networks, three potential approaches exist to prevent ARP cache poisoning attacks, each with drawbacks. This paper proposes an innovative approach called Gateway-controlled ARP (GCA) to prevent ARP cache poisoning attacks on a router-on-a-stick (RoaS) network using MikroTik networking devices, where a single router performs inter-VLAN routing through one physical interface. With this approach, all Ethernet switches are configured to forward ARP messages from hosts directly to the router for inspection and handling. A RouterOS script based on the GCA approach was implemented and executed on the router to handle all incoming ARP requests from any host in all VLANs, ensuring all hosts receive legitimate ARP responses from the router. This approach can effectively prevent spoofed ARP packets sent by malicious attackers. This approach was tested and evaluated on an actual RoaS network, focusing on processing time, CPU Load, and response time. The evaluation results show that the approach effectively prevents ARP cache poisoning attacks.