Unmanned aerial vehicles (UAVs) equipped with visible light communication (VLC) systems are envisioned to simultaneously provide secure data transmission and nighttime illumination. However, when buildings obstruct the optical links, both connectivity and lighting are disrupted, which may severely compromise system reliability and safety. This paper investigates an artificial noise-based physical layer security (PLS) scheme for a VLC-enabled UAV communication system in a multiuser environment with potential eavesdroppers, while explicitly incorporating awareness of shadowed area caused by blockage and enabling the UAV to autonomously adjust its trajectory to proactively avoid such shadow coverage to the ground users. We formulate a joint optimization problem of user association, power allocation, and UAV trajectory design to maximize the average secrecy rate of the system, while taking into account illumination requirements, shadowing effects, and UAV mobility. To tackle this mixed-integer and non-convex optimization problem, we decompose it into three subproblems and transform them into tractable convex forms. Furthermore, we also develop an iterative algorithm by leveraging successive convex approximation techniques under a block coordinate descent framework to efficiently obtain a suboptimal solution. Simulation results demonstrate that the proposed scheme can achieve fast convergence and improve the average secrecy rate at least by 51.1% compared with conventional schemes. Moreover, the algorithm still exhibits robustness and efficacy in exploiting the spatial-temporal trade-offs under severe eavesdropping threats and shadowing with diverse user geometries, highlighting its practicality for secure nighttime urban VLC-UAV communication.

Reasoning is a fundamental ability towards artificial general intelligence (AGI). Fueled by the success of deep learning, the neural logic machines models (NLMs) have introduced novel neural-symbolic structures and demonstrate great performance and generalization on reasoning and decision-making tasks. However, the original training approaches of the NLMs are still far from perfect, the models would repeat similar mistakes during the training process which leads to sub-optimal performance. To mitigate this issue, we present a novel framework named Failure Reflection Guided Regularizer (FRGR). FRGR first dynamically identifies and summarizes the root cause if the model repeats similar mistakes during training. Then it penalizes the model if it makes similar mistakes in future training iterations. In this way, the model is expected to avoid repeating errors of similar root causes and converge faster to a better-performed optimum. Experimental results on multiple relational reasoning and decision-making tasks demonstrate the effectiveness of FRGR in improving performance, generalization, training efficiency, and data efficiency.

Modern software systems heavily rely on external libraries developed by third-parties to ensure efficient development. However, frequent library upgrades can lead to compatibility issues between the libraries and their client systems. In this paper, we introduce Compsuite, a dataset that includes 123 real-world Java client-library pairs where upgrading the library causes an incompatibility issue in the corresponding client. Each incompatibility issue in Compsuite is associated with a test case authored by the developers, which can be used to reproduce the issue. The dataset also provides a command-line interface that simplifies the execution and validation of each issue. With this infrastructure, users can perform an inspection of any incompatibility issue with the push of a button, or reproduce an issue step-by-step for a more detailed investigation. We make Compsuite publicly available to promote open science. We believe that various software analysis techniques, such as compatibility checking, debugging, and regression test selection, can benefit from Compsuite. The demonstration video of Compsuite is available at https://www.youtube.com/watch?v=7DQGsGs_65s.

Modern software systems are complex, and they heavily rely on external libraries developed by different teams and organizations. Such systems suffer from higher instability due to incompatibility issues caused by library upgrades. In this article, we address the problem by investigating the impact of a library upgrade on the behaviors of its clients. We developed CompCheck , an automated upgrade compatibility checking framework that generates incompatibility-revealing tests based on previous examples. CompCheck first establishes an offline knowledge base of incompatibility issues by mining from open source projects and their upgrades. It then discovers incompatibilities for a specific client project, by searching for similar library usages in the knowledge base and generating tests to reveal the problems. We evaluated CompCheck on 202 call sites of 37 open source projects and the results show that CompCheck successfully revealed incompatibility issues on 76 call sites, 72.7% and 94.9% more than two existing techniques, confirming CompCheck ’s applicability and effectiveness.