Augmented reality (AR) systems are increasingly deployed in tactical environments, but their reliance on seamless human-computer interaction makes them vulnerable to cognitive attacks that manipulate a user's perception and severely compromise user decisionmaking. To address this challenge, we introduce the Perception Graph, a novel model designed to reason about human perception within these systems. Our model operates by first mimicking the human process of interpreting key information from an MR environment and then representing the outcomes using a semantically meaningful structure. We demonstrate how the model can compute a quantitative score that reflects the level of perception distortion, providing a robust and measurable method for detecting and analyzing the effects of such cognitive attacks.

Cognitive attacks in mixed reality (MR), e.g., latency perturbations that induce frame-time jitter, can divert visual attention and degrade task performance. We study 2D gaze prediction under such disturbances and propose a time-aware sequence model that handles irregular sampling by supplying elapsed times Δt between observations and conditions on sparse event/object context available at prediction time via learned token embeddings. Using time-based windows, we evaluate within-user and cross-user temporal generalization on MR recordings spanning multiple attack intensities. Results indicate accurate, time-robust gaze regression under latency perturbations, supporting adaptive MR interfaces in adversarial settings.

In location-based services(LBSs), it is promising for users to crowdsource and share their Point-of-Interest(PoI) information with each other in a common cache to reduce query frequency and preserve location privacy. Yet most studies on multi-user privacy preservation overlook the opportunity of leveraging their service flexibility. This paper is the first to study multiple users’ strategic cooperation against an adversary's optimal inference attack, by leveraging mutual service flexibility. We formulate the multi-user privacy cooperation against the adversary as a max-min adversarial game and solve it in a linear program. Unlike the vast literature, even if a user finds the cached information useful, we prove it beneficial to still query the platform to further confuse the adversary. As the linear program's computational complexity still increases superlinearly with the number of users’ possible locations, we propose a binary obfuscation scheme in two opposite spatial directions to achieve guaranteed performance with only constant complexity. Perhaps surprisingly, a user with a greater service flexibility should query with a less obfuscated location to add confusion. Finally, we provide guidance on the optimal query sequence among LBS users. Simulation results show that our crowdsourced privacy protection scheme greatly improves users’ privacy as compared with existing approaches.

In federated learning (FL), clients cooperatively train a global model without revealing their raw data but gradients or parameters, while the local information can still be disclosed from local outputs transmitted to the parameter server. With such privacy concerns, a client may overly add artificial noise to his local updates to compromise the global model training, and we prove the selfish noise adding leads to an infinite price of anarchy (PoA). This paper proposes a novel pricing mechanism to regulate privacy-sensitive clients without verifying their parameter updates, unlike existing privacy mechanisms that assume the server's full knowledge of added noise. Without knowing the ground truth, our mechanism reaches the social optimum to best balance the global training error and privacy loss, according to the difference between a client's updated parameter and all clients' average parameter. We also improve the FL convergence bound by refining the aggregation rule at the server to account for different clients' noise variances. Moreover, we extend our pricing scheme to fit incomplete information of clients' privacy sensitivities, ensuring their truthful type reporting and the system's ex-ante budget balance. Simulations show that our pricing scheme greatly improves the system performance especially when clients have diverse privacy sensitivities.

In location-based services (LBSs), it is promising for multiple users to cache and share their Point-of-Interest (PoI) information with each other to reduce overall query frequency and preserve location privacy. Yet most studies on multi-user privacy preservation overlook the opportunity of leveraging service flexibility, where many users are flexible and may add obfuscation to individual LBS query. This paper is the first to study how multiple users cooperate to query with obfuscation against the adversary’s optimal inference attack, by leveraging their mutual service flexibility. Unlike the literature, even if a user already finds the shared PoI information useful, we prove it beneficial for him to further query with obfuscated location to confuse the adversary. To save the computational complexity of the max-min adversarial game problem and derive the closed-form solution, we also propose a binary approximate solution, which is proved to guarantee good privacy performance for an average user. Perhaps surprisingly, the user with greater service flexibility should choose to query the LBS with less misreported location, to maximally confuse the adversary. Finally, we numerically compare our optimal and approximate solutions with the existing approaches to show our effective privacy improvement.

When using location-based services (LBSs), a user obtains points-of-interest (PoI) information by providing the LBS platform with his current geo-location. Such a search leads to potential privacy leakage if an adversary has access to his geo-data. Traditionalk-anonymity mechanisms instruct a user to bear the overhead to report his current location together withk- 1 dummy locations to confuse the adversary, which only work well given a large numberk. Aware of the common practices that a user is actually flexible in service requirement (e.g., as long as the searched PoIs are within his walking distance), we propose a novel approach to help the user gain location privacy from service flexibility for the challenging case of a small numberk. By analyzing the strategic interaction between the user and the adversary in a dynamic Bayesian game, we prove that the user’s equilibrium strategy depends on the adversary’s capability of accessing geo-data. Takek= 2 for example, we find that if the adversary is not likely to access both geo-data, the user should report the two dummy locations at two different directions of his real location, and otherwise at the same direction. Perhaps surprisingly, the user may benefit from the adversary’s access to more geo-data. Furthermore, we extend the game-theoretic approach for multi-query and arbitrary user location distributions. Numerical results show that our approach obviously outperformsk-anonymity mechanisms especially under a small numberk.

When using location-based services (LBSs), a user obtains points-of-interest $(\text{P}\text{o}\text{I})$ information by providing the LBS platform with his current geo-location. Such a search also leads to potential privacy leakage if an adversary has access to his geo-data. Traditional k-anonymity mechanisms instruct a user to bear the overhead to report his current location together with k-1 dummy locations to confuse the adversary, which only work well given a large number k. Aware of the common practices that a user is actually flexible in service requirement (e.g., as long as the searched PoIs are within his walking distance), we propose a novel approach to help the user gain location privacy from service flexibility for the case of a small number k. By analyzing the strategic interaction between the user and the adversary in a Bayesian game, we prove that the user with service flexibility should never report his real location for searching PoIs nearby. Instead, he should jointly use all k dummy locations to confuse the adversary’s inference of his real location. Take $k=2$ for example, we manage to show that if the adversary is not likely to access both dummy geo-data, the user should report the two dummy locations at two opposite directions of his real location, and otherwise at the same direction. Perhaps surprisingly, our approach may enable the user to benefit from the adversary’s access to more geo-data. Finally, extensive simulations using some real data show that our mechanism obviously outperforms k anonymity mechanism especially under a small number k.

With the rapid development of sensor technology and wireless network technology, wireless sensor network (WSN) has been widely applied in many resource-constrained environments and application scenarios. As there are a large number of sensor nodes in WSN, node failures are inevitable and have a significant impact on task execution. In this paper, considering the vulnerability, unreliability, and dynamic characteristics of sensor nodes, node failures are classified into two categories including unrecoverable failures and recoverable failures. Then, the traditional description of the interaction results is extended to the trinomial distribution. According to the Bayesian cognitive model, the global trust degree is aggregated by both direct and indirect interaction records, and a novel trust model based on node recovery technique for WSNs is proposed to reduce the probability of failure for task execution. Simulation results show that compared with existing trust models, our proposed TMBNRT (trust model based on node recovery technique) algorithm can effectively meet the security and the reliability requirements of WSN.