In this paper, we design a friendly jammer selection scheme for the social Internet of Things (IoT). A typical social IoT is composed of a cellular network with underlaying Device-to-Device (D2D) communications. In our scheme, we consider signal characteristics over a physical layer and social attribute information of an application layer simultaneously. Using signal characteristics, one of the D2D gadgets is selected as a friendly jammer to improve the secrecy performance of a cellular device. In return, the selected D2D gadget is allowed to reuse spectrum resources of the cellular device. Using social relationship, we analyze and quantify the social intimacy degree among the nodes in IoT to design an adaptive communication time threshold. Applying an artificial intelligence forecasting model, we further forecast and update the intimacy degree, and then screen and filter potential devices to effectively reduce the detection and calculation costs. Finally, we propose an optimal scheme to integrate the virtual social relationship with actual communication systems. To select the optimal D2D gadget as a friendly jammer, we apply Kuhn-Munkres (KM) algorithm to solve the maximization problem of social intimacy and cooperative jamming. Comprehensive numerical results are presented to validate the performance of our scheme.

In smart grids, real-time electricity data uploaded by smart meters may be analyzed by an attacker with other data analytics methods, which may expose users’ privacy. To ensure user privacy, differential privacy methods are often used to process data. However, these methods reduce the accuracy of the data results obtained by the center and lead to unavailability of the data. In this paper, we address this problem and propose a distributed differential privacy protection scheme. Two methods of data noise addition and data perturbation are fused and used in the protection scheme. Data accuracy is improved by optimizing the noise generation method. To address the problem of quantitatively balancing the users’ privacy needs with the central analytics needs, this paper describes the needs of both through mathematical definitions, i.e., data accuracy and data privacy, and proposes a privacy budget that balances data accuracy and privacy. The performance of the proposed scheme is evaluated using the typical power data, which proves the excellent performance.

In the realm of edge computing, the analog aggregation based Federated Learning Over the Air (FLOA) emerges as a promising technology, offering heightened communication efficiency and privacy provisions. This approach involves concurrent transmission and aggregation, where edge devices (workers) collectively upload their local updates to a Parameter Server (PS) through shared time-frequency resources. The PS then obtains averaged updates only but not the individual local ones, reducing latency and communication costs. However, this simultaneous process exposes FLOA to vulnerabilities, particularly Byzantine attacks. Addressing this concern, we introduce an innovative framework utilizing Unmanned Aerial Vehicles (UAVs) to assist in a heterogeneous FLOA. This framework mitigates the impact of Byzantine attacks while preserving the advantages of over-the-air computation for efficient federated learning. The UAVs engage in over-the-air computation, collecting gradients from local workers and aggregating them. Subsequently, the UAVs transmit these aggregated gradients to the PS in different time slots. Robust aggregation techniques are applied at the PS to combine updates from UAVs. To enhance the robustness of over-the-air transmissions from workers to UAVs, we propose a power control policy. The resilience of the proposed framework to attacks is demonstrated through the derived expected convergence rate, validated by experiments on real datasets.