The integration of the Mobile Crowdsensing (MCS) and Industrial Internet of Things (IIoT) brings enormous volumes of data that generate significant commercial value. However, the data contain a wealth of sensitive information about devices' environmental situation and collective activities, which draws a flock of adversaries and poses an unprecedented security challenge. Furthermore, sensing gadgets deployed in the IIoT device are usually resource‐constrained and often do not have adequate 3C resources (i.e. communication, computing, caching) to run sophisticated privacy‐preserving methods, making them easier targets for attacks in data sharing. Therefore, a risk‐adaptive privacy protection scheme R‐D ^P for MCS‐enabled IIoT gadgets is proposed, which comprises a closed‐loop risk‐awareness process and an adaptive privacy protection method D ^P (a dissemination process with perturbation). The closed‐loop process dynamic awareness of risks and threats in MCS task feeds appropriate privacy protection advice to the decision‐makers for the task. In addition, D ^P was designed as a lightweight and risk‐adaptive privacy protection method to meet the operational needs of 3C resource‐constrained gadgets. The analysis and evaluation show that R‐D ^P provides satisfactory privacy protection while the availability of statistical features reaches more than 96%, and the time complexity is only O (1) for sensing gadgets.

Local differential privacy (LDP), a robust privacy-protection model, is widely adopted in the Industrial Internet of Things (IIoT) due to its lightweight, decentralized, and scalable. However, its perturbation-based privacy-protection mechanism hinders distinguishing between any two data, thereby facilitating LDP poisoning attacks. The exposed physical-layer vulnerabilities and resource-constrained prevalent at the IIoT edge not only facilitate such attacks but also render existing LDP poisoning defenses, all of which are deployed at the edge and rely on ample resources, impractical. This work proposes a LDP poisoning defense for IIoT in the resource-rich aggregator. We first reveal key poisoning attack modes occurring within the LDP-utilized IIoT data-collection process, detailing how IIoT vulnerabilities enable attacks, and then formulate a general attack model and derive the poisoned data’s indistinguishability. This work subsequently analyzes the poisoning impacts on aggregated data based on industrial process correlation, revealing the distortion of statistical query result’ temporal similarity and the resulting disruption of interattribute correlation, and uncovering the intriguing paradox that adversaries’ attempts to stabilize their poisoning actions for stealth are difficult to maintain. Given these findings, we propose PoisonCatcher, a solution for identifying poisoned data, which includes time-series detectors based on temporal similarity, attribute correlation, and pattern stability metrics to detect poisoned attributes, and a latent-bias feature miner for identifying poisons. Experiments on two representative real-world datasets indicate that PoisonCatcher successfully identifies poisoned data, demonstrating robust identification capabilities with F2 scores above 90.7% under various attack settings.

In the Internet of Vehicles, vehicular crowdsensing is crucial for alleviating traffic congestion and ensuring the safety of autonomous driving. However, practical vehicular crowdsensing processes face dual challenges of skewed spatial distributions of vehicles and inadequate data quality guidance. These issues cause sensing redundancy in high-participation areas (HPAs) and coverage deficits in low-participation areas (LPAs), while also leading to unstable data quality. Given that participants’ decisions are profoundly influenced by bounded rationality and psychological preferences, this paper proposes a collaborative incentive mechanism integrating behavioral economics and psychology (BEP-IM) to drive sustained spatial coverage and proactive sensing shaping. First, to mitigate coverage deficits in LPA, a reference-dependent two-sided selection and bidding strategy (RD-TSB) is designed to guide participants toward LPA via a reference-driven utility evaluation. Concurrently, a loss-aversion-based sustained incentive strategy (LA-RPI) is introduced to enhance their sustained participation within LPAs by amplifying loss perception. Furthermore, to overcome weak data quality constraints, an operant conditioning-based proactive sensing shaping strategy (OC-SFQ) is constructed, utilizing a closed-loop mechanism of relative improvement, variable-ratio reinforcement, and association updating to drive participants to output high-quality data. Simulation results demonstrate that the proposed mechanism effectively increases participation frequency in LPAs and optimizes sensing data quality.

The Internet of Vehicles (IoV) provides an effective solution for alleviating traffic congestion and enhancing road safety. However, shared traffic messages in IoV may deviate from on-road conditions due to self-interest protection or insufficient sensor performance. Therefore, evaluating the trustworthiness of shared messages is essential for vehicles to make informed decisions. To this end, a truth-oriented trust model for shared traffic message is proposed, which is inspired by human trust establishment mechanisms (HS-TEMs). Firstly, we quantify the integrated trust value (I-VT) of the message sender by fusing self-experience-based vehicle trust (SEB-VT) and peer-recommendation-based vehicle trust (PRB-VT). In SEB-VT, a sample-size-dependent smoothing factor dynamically trades off prior information and empirical evidence, reducing instability under small-sample conditions. In PRB-VT, we employ link analysis to compute the reference degree of recommendation information, which mitigates biases arising from subjective cognitive limitations. Secondly, with the I-VT of vehicles, we calculate event trust (ET) by differentiating message attitudes and quantifying their relative influence, which effectively reduces the impact of individual bias on the final judgment. The simulation results show that HS-TEM can accurately and fairly evaluate the credibility of messages, which helps vehicles make informed decisions.

Local Differential Privacy (LDP) is a widely adopted privacy-protection model in the Internet of Things (IoT) due to its lightweight, decentralized, and scalable nature. However, it is vulnerable to poisoning attacks, and existing defenses either incur prohibitive resource overheads or rely on domain-specific prior knowledge, limiting their practical deployment. To address these limitations, we propose PEEL, a Poisoning-Exposing Encoding theoretical framework for LDP, which departs from resource- or prior-dependent countermeasures and instead leverages the inherent structural consistency of LDP-perturbed data. As a non-intrusive post-processing module, PEEL amplifies stealthy poisoning effects by re-encoding LDP-perturbed data via sparsification, normalization, and low-rank projection, thereby revealing both output and rule poisoning attacks through structural inconsistencies in the reconstructed space. Theoretical analysis proves that PEEL, integrated with LDP, retains unbiasedness and statistical accuracy, while being robust to expose both output and rule poisoning attacks. Moreover, evaluation results show that LDP-integrated PEEL not only outperforms four state-of-the-art defenses in terms of poisoning exposure accuracy but also significantly reduces client-side computational costs, making it highly suitable for large-scale IoT deployments.

Local Differential Privacy (LDP) has emerged as a dominant privacy paradigm in the Internet of Things (IoT) due to its lightweight, decentralized and scalable nature. However, the indistinguishability guarantee inherent to LDP mechanisms inadvertently masks adversarial manipulation, rendering poisoned data difficult to differentiate from legitimate perturbations. Existing poisoning defense mechanisms either incur prohibitive overheads incompatible with resource-constrained IoT devices or rely on rigid domain priors. To address this, we propose PEEL, a Poisoning-Exposing Encoding mechanism for LDP, which reformulates poisoning exposure as a geometric structural verification problem within the encoding subspace. Departing from probabilistic statistical filtering, PEEL exploits the rank disparity between legitimate sparse perturbations and arbitrary adversarial injections. By projecting LDP outputs onto a low-rank structural support, the mechanism forces poisoning attempts to manifest as deterministic reconstruction residuals, thereby exposing structure-violating output and rule poisoning attacks. Theoretical analysis confirms that PEEL, as a post-randomization encoding layer, preserves the unbiasedness and statistical accuracy of the underlying LDP protocols, while deterministically exposing structure-violating output and rule poisoning attacks. Extensive evaluations across six protocols and two real-world datasets confirm that PEEL reduces client-side overheads by orders of magnitude compared to cryptographic baselines, while maintaining F1 scores exceeding 99.8% against structure-violating output and rule poisoning attacks.