The openness of wireless networks enables flexible connections on devices, but also arises the security and privacy concerns on adversarial eavesdropping. By concealing the transmissions from eavesdropper, covert communication preserves the privacy of communications and has become a significant research topic. Whereas, existing works primarily focus on the covertness of communication links, neglecting the protection for the critical leader nodes response for information aggregation and decision making in information systems. To bridge this gap, we introduce a novel conceptcovert leader, whose identity remains concealed from all other nodes while performing leadership functions. This design inherently mitigates targeted attacks (e.g., eavesdropping, jamming, DoS) against leaders, enhancing both privacy and robustness of wireless network. Then, a distributed randomizedK-Covert Leader Election (KCLE) algorithm is designed, in whichKis a hyperparameter to depict the covertness of leader election and can be adjusted when KCLE algorithm is implemented in reality. Unlike cryptographic solutions, KCLE leverages physical-layer signal properties to elect a leader from at leastKcandidates with balanced energy-time tradeoffs, ensuring no leader identity leakage occurs during election. The algorithm operates without complex encryption, making it suitable for resource-constrained Internet-of-Things and ad hoc networks. Theoretical analysis and simulation results demonstrate KCLE’s correctness, covertness guarantees, and operational efficiency under SINR interference models.

This paper studies an efficient federated learning (FL) problem involving multiple edge-based clients with heterogeneous constrained resources. Compared with numerous training parameters, the computing and communication resources of clients in edge scenarios are usually insufficient for fast local training and real-time knowledge sharing. Besides, training on clients with heterogeneous resources may result in the straggler problem, which delays the convergence of FL. To address these issues, we proposeFed-RAA: aResource-AdaptiveAsynchronousFederated learning algorithm. Different from vanilla FL methods, where all parameters are trained by each participating client regardless of resource diversity, Fed-RAA adaptively allocates submodels of the global model to clients based on their computing and communication capabilities. Each client then individually trains its assigned submodel and asynchronously uploads the updated result. Theoretical analysis confirms the convergence of our approach. Additionally, an online greedy-based algorithm is designed for asynchronous submodel assignment in Fed-RAA, improving the convergence of Fed-RAA by optimal minimization on the training delay bound of submodels. Compared to state-of-the-art methods, our Fed-RAA algorithm reduces the time required to achieve the target accuracy by an average of$ 30.89\%$, demonstrating its superior efficiency on heterogeneous constrained computing and communication resources. To the best of our knowledge, this paper is the first resource-adaptive asynchronous method for submodel-based FL with guaranteed theoretical convergence.

Driven by Artificial Intelligence (AI) and Internet of Things (IoT) technologies, smart cities are becoming more sophisticated. In smart cities, intelligent applications such as traffic optimization, environmental monitoring and emergency response develop intelligent decisions with AI models. The core of smart decision-making lies in the intelligent analysis and processing of real-time data collected by IoT devices, which supports the intelligent and precise operation of cities. However, the credibility of data sources directly affects the quality of decision-making and the effectiveness of AI models. Issues such as tampering, falsification, or failure of data sources not only lead to biased decisions but may also jeopardize the stability and security of urban systems. Therefore, conducting trust evaluations on data collection terminals in intelligent scenarios is crucial for ensuring the reliable operation of smart cities. This paper proposes a multidimensional trust evaluation method for data collection in intelligent applications of smart cities. Moreover, this method can effectively detect sophisticated attacks, such as data poisoning and insider threats. By incorporating static, dynamic, and energy trust assessments, and applying the Technique for Order Preference by Similarity to an Ideal Solution (TOPSIS) to generate an integrated trust value, this approach ensures data credibility from the source and provides reliable data support for smart decision-making.

Due to the surge in mobile apps, online videos, and cloud gaming, central servers struggle with high traffic demands. Content delivery networks (CDNs) mitigate this by distributing content from edge caches, easing backbone network strain. Yet, current allocation algorithms, constrained by practical complexities and billing, often fail to optimally schedule traffic, causing server overload and increased costs. Our solution, CEQC-CDN, addresses these issues by considering DNS load balancing, regional hijacking, quality constraints, and server capacity limits. By incorporating these factors as linear constraints through binary decomposition and variable introduction, CEQC-CDN achieves global optimal traffic scheduling. Employing a greedy policy for monthly optimization, tests demonstrate an 8.57% reduction in CDN traffic costs versus conventional greedy algorithms.