To enhance aerial image detection in complex environments characterized by multiple small targets and mutual occlusion, we propose an aerial target detection algorithm based on an improved version of YOLOv5 in this paper. Firstly, we employ an improved Mosaic algorithm to address redundant boundaries arising from varying image scales and to augment the training sample size, thereby enhancing detection accuracy. Secondly, we integrate the constructed hybrid attention module into the backbone network to enhance the model’s capability in extracting pertinent feature information. Subsequently, we incorporate feature fusion layer 7 and P2 fusion into the neck network, leading to a notable enhancement in the model’s capability to detect small targets. Finally, we replace the original PAN + FPN network structure with the optimized BiFPN (Bidirectional Feature Pyramid Network) to enable the model to preserve deeper semantic information, thereby enhancing detection capabilities for dense objects. Experimental results indicate a substantial improvement in both the detection accuracy and speed of the enhanced algorithm compared to its original version. It is noteworthy that the enhanced algorithm exhibits a markedly improved detection performance for aerial images, particularly under real-time conditions.

With the advent of smart cities, the significance of the Internet of Things (IoT) is gaining greater prominence. At the same time, the safety early warning system in the IoT has a significant impact on real-time monitoring and the response to potential risks. Despite the advancements made in edge-assisted IoT deployments, several challenges and constraints persist. Given the potential threat to life posed by safety-related messages, ensuring the authenticity of messages in the edge-assisted IoT safety warning system is crucial. However, considering the identity privacy of devices participating in the edge-assisted Internet of Things system, directly verifying the identity of the sending device is undesirable. To address this issue, in this work, we design a linkable group signature scheme that allows devices to anonymously send safety-related messages to edge nodes, defending against Sybil attacks while ensuring the traceability of malicious device identities. Then, we present a high-efficiency conditional privacy-preserving authentication (CPPA) scheme based on the designed group signatures for the safety warning system in edge-assisted IoT. This scheme effectively protects device identity privacy while providing a reliable authentication mechanism to ensure the credibility and traceability of alert messages. The proposed scheme contributes to the field of safety warning systems in the context of edge-assisted IoT, providing a robust solution for privacy preservation and authentication.

Malicious certificate authorities (CAs) pose a significant threat to the security of the Internet of Things (IoT). Existing CA auditing schemes primarily rely on passive detection and public data collection, lacking real-time and comprehensive monitoring. In this paper, we propose a novel double-authentication preventing signature scheme based on an SM2 algorithm, referred to as Dap-SM2. We further enhance its functionality by introducing Self-C2AD, a CA auditing mechanism with self-enforcement. In our proposed mechanism, any malicious CA that generates two certificates with different descriptions (such as public key and basic information) for the same IoT device will immediately lose its private key. To ensure the security of our proposed scheme, we provide a detailed security analysis of Dap-SM2. The analysis demonstrates that our Self-C2AD mechanism meets the necessary security requirements, offering robust protection against malicious CAs. Additionally, we conduct an efficiency evaluation and present experimental data to illustrate the promising potential of our construction for future IoT applications. By introducing the Dap-SM2 scheme and the Self-C2AD mechanism, we address the critical issue of malicious CAs in the IoT domain. Our approach provides real-time and comprehensive auditing capabilities, surpassing the limitations of existing schemes. The security analysis confirms the effectiveness of Dap-SM2, while the efficiency evaluation and experimental data demonstrate its suitability for practical IoT applications. In summary, our work presents a novel solution to combat the threat of malicious CAs in the IoT context. The Dap-SM2 scheme, coupled with the Self-C2AD mechanism, offers enhanced security and real-time auditing capabilities. The security analysis validates the robustness of our approach, while the efficiency evaluation and experimental data showcase its potential for future IoT deployments.

With the development of autonomous driving technology, the accuracy requirements for electronic maps are becoming increasingly high.And high-precision maps, as a key carrier of intelligent IoT System in the field of transportation, has become a necessity for autonomous driving. The drawing of high-precision maps cannot be separated from the data collection of convoys, but there are security issues, such as pseudo mobility attacks and low-data privacy in this process. Therefore, a hybrid authentication architecture for map collection vehicles and base stations (BSs) based on alliance blockchain is proposed. This architecture has significant effectiveness in resisting mobile pseudo BS attacks, achieving authentication of access point vehicles by mobile BSs. At the same time, in order to effectively adapt to high-precision map scenarios, the proposed key distribution strategy and high-precision map update method are both lightweight. Last but not least, we analyze and evaluate the performance of ESMU through rigorous theoretical analysis and extensive experimental verification. The results demonstrate its high efficiency and strong robustness.

The Space-Air-Ground Information Network (SAGIN) provides extensive coverage, enabling global connectivity across a diverse array of sensors, devices, and objects. These devices generate large amounts of data that require advanced analytics and decision making using artificial intelligence techniques. However, traditional deep learning approaches encounter drawbacks, primarily, the requirement to transmit substantial volumes of raw data to central servers, which raises concerns about user privacy breaches during transmission. Federated learning (FL) has emerged as a viable solution to these challenges, addressing both data volume and privacy issues effectively. Nonetheless, the deployment of FL faces its own set of obstacles, notably the excessive delay and energy consumption caused by the vast number of devices and fluctuating channel conditions. In this paper, by considering the heterogeneity of devices and the instability of the network state, the delay and energy consumption models of each round of federated training are established. Subsequently, we introduce a strategic node selection approach aimed at minimizing training costs. Building upon this, we propose an innovative, empirically driven Double Deep Q Network (DDQN)-based algorithm called low-cost node selection in federated learning (LCNSFL). The LCNSFL algorithm can assist edge servers in selecting the optimal set of devices to participate in federated training before the start of each round, based on the collected system state information. This paper culminates with a simulation-based comparison, showcasing the superior performance of LCNSFL against existing algorithms, thus underscoring its efficacy in practical applications.