Safe and efficient vehicular networks in modern Intelligent Transportation Systems (ITS) rely heavily on ultra-reliable and low-latency communication (URLLC) as a fundamental requirement. To address this, researchers propose a hybrid deep learning–based Vehicle-to-Everything (V2X) framework that integrates Convolutional Neural Networks (CNN) with Long Short-Term Memory (LSTM) networks to enhance real-time decision-making. The framework initially collects heterogeneous V2X data from Vehicle-to-Vehicle (V2V), Vehicle-to-Infrastructure (V2I), Vehicle-to-Pedestrian (V2P), and Vehicle-to-Network (V2N) sources, including GPS coordinates, vehicle speed, Received Signal Strength Indicator (RSSI), and channel state information. Data preprocessing employs normalization techniques such as Min–Max scaling, sliding window methods, dimensionality reduction, and time-series transformations to generate model-ready inputs. CNN components extract spatial features such as road layouts, vehicle density, and interference patterns, while LSTM modules capture temporal dynamics including signal fluctuations, handover delays, and congestion evolution. The proposed framework was evaluated using a dataset of 10,000 V2X communication samples collected at a sampling frequency of 10 Hz to capture dynamic vehicular network conditions. By fusing spatial and temporal representations, the framework predicts optimal communication strategies, enabling adaptive routing, dynamic power control, and stable link selection. This hybrid approach improves computational efficiency and predictive accuracy, supporting autonomous driving, enhanced traffic management, real-time safety, and congestion forecasting in next-generation V2X networks.
Rockfalls are one of the most significant geohazards that endanger transportation infrastructures in the mountainous region of Kurdistan Region of Iraq, particularly along the Barzan Highway within Duhok City. This study introduces a high-resolution, multi-parametric Rockfall Susceptibility Assessment (RSA) applied to the delineation of hazardous zones along this important road transportation network. The AHP-based framework significantly enhances the accuracy and decision reliability of rockfall susceptibility assessment by incorporating expert-driven weighting of ten conditioning factors. Via Geographic Information System (GIS) based AHP, the study incorporates ten conditioning factors obtained from sophisticated remote sensing data. A sub-meter Digital Surface Model (DSM) was built by mobile LiDAR Scanning (Stonex X120Go) to characterize the micro-relief and to upscale structural discontinuities detection. The Land Use/Land Cover (LULC) was also classified with a sequence CNN model using the Sentinel-2 satellite data, with an overall accuracy of 97.40%. The AHP analysis, verified by expert opinions with a Consistency [Formula: see text], revealed that slope (37.29%) and structural discontinuities (11.71%) were the most significant contributors to instability. The resulting susceptibility map shows that 26.31% of the corridor is classified into High to Very High susceptibility zones, predominantly along the steep limestone cuts of the western ridge. A longitudinal profile analysis, supported by Theis solution, establishes a clear relation between man-made excavation and hazard maxima, with the maximum susceptibility index (0.40) observed in vertical cut sections. These findings highlight the efficacy of integrating deep learning and mobile LiDAR within the AHP framework, offering local authorities an accurate, station-level zonation map for targeted mitigation efforts.
Machine learning models play a crucial role in predictive analytics, particularly in cloud-based environments where vast amounts of performance data must be efficiently processed and analysed. This study measures the usage patterns of cloud Central Processing Unit (CPU) utilization, which contains Virtual Machines (VMs) usage pattern, and presents a comprehensive evaluation of multiple regression models — Linear Regression, Decision Tree (TR), Gradient Boosting Machine (GBM), and Multi-Layer Perceptron (MLP) Regressor — to determine their effectiveness in modelling cloud resource utilisation patterns. The performance of each model was assessed based on two key error metrics: Mean Squared Error (MSE) and Mean Absolute Error (MAE). Our findings reveal that Linear Regression performed the worst, with high error rates, indicating its inability to handle non-linearity in the data. Decision Tree (TR) improved upon Linear Regression but tended to overfit, leading to moderate accuracy. Gradient Boosting (GBM) further enhanced performance, striking a balance between accuracy and computational efficiency. However, the MLP Regressor significantly outperformed all models, achieving an MSE of 0.000137 and an MAE of 0.00640, demonstrating exceptional predictive accuracy and generalisation capability. While neural network-based models like MLP require greater computational resources, they exhibit superior ability in capturing complex relationships within the dataset. The results of this study suggest that MLP Regressor is the most effective model for cloud resource prediction, given its minimal error metrics. Future research should explore hyperparameter tuning, hybrid modelling approaches, and the integration of deep learning techniques to further enhance predictive performance while optimising computational costs.
Cloud resource allocation is a process of analyzing the Virtual Machines (VMs) availability and allocating it for running tasks. However, the longer queues in the network increased the delay in allocating the resources. So, a significant cloud Resource Allocation (RA) framework is proposed using Conhattan K-Means (C-K Means) and Aranmoid-Ridge Bilateral Long Short-Term Memory (AR-BiLSTM). Initially, the cloud users are registered and log in with their required tasks. During registration, a Service Level Agreement (SLA) is created betwixt the cloud server and user. Based on the SLA and tasks, a digital signature is created. Then, the tasks are clustered by using C-K Means; afterward, the clustered tasks are prioritized. Next, the workloads of the VMs are predicted by preprocessing the data and then extracting features from it. Then, optimal features are selected and given to AR-BiLSTM. Then, the features of the prioritized tasks and VMs are extracted and analyzed to allocate the suitable VM to the particular tasks. Here, the created signature is verified to ensure the user authentication for allocating resources. The analysis results proved the superiority of the proposed framework in allocating cloud resources by utilizing the maximum resources of 0.95.
Constant workspace algorithms use a constant number of words in addition to the read-only input to the algorithm. In this paper, we devise algorithms to efficiently compute relative hulls in the plane using a constant workspace. Specifically, we devise algorithms for the following three problems: (i) Given two simple polygons [Formula: see text] and [Formula: see text] with [Formula: see text], compute a simple polygon [Formula: see text] with a perimeter of minimum length such that [Formula: see text]. (ii) Given two simple polygons [Formula: see text] and [Formula: see text] such that [Formula: see text] does not intersect the relative interior of [Formula: see text] but it does intersect the relative interior of the convex hull of [Formula: see text], compute a weakly simple polygon [Formula: see text] with a perimeter of minimum length such that [Formula: see text], the convex hull of [Formula: see text] contains [Formula: see text], and [Formula: see text] does not intersect the relative interior of [Formula: see text]. (iii) Given a set [Formula: see text] of points located in a simple polygon [Formula: see text], compute a weakly simple polygon [Formula: see text] with a perimeter of minimum length such that [Formula: see text] and [Formula: see text] contains all the points in [Formula: see text]. To our knowledge, no prior work devised algorithms to compute relative hulls using a constant workspace, and this work is the first such attempt.
We study the problem of comparing a pair of geometric networks that may not be similarly defined, i.e., when they do not have one-to-one correspondences between their nodes and edges. Our motivating application is to compare power distribution networks of a region. Due to the lack of openly available power network datasets, researchers synthesize realistic networks resembling their actual counterparts. But the synthetic digital twins may vary significantly from one another and from actual networks due to varying underlying assumptions and approaches. Hence the user wants to evaluate the quality of networks in terms of their structural similarity to actual power networks. But the lack of correspondence between the networks renders most standard approaches, e.g., subgraph isomorphism and edit distance, unsuitable. We propose an approach based on the multiscale flat norm, a notion of distance between objects defined in the field of geometric measure theory, to compute the distance between a pair of planar geometric networks. Using a triangulation of the domain containing the input networks, the flat norm distance between two networks at a given scale can be computed by solving a linear program. In addition, this computation automatically identifies the 2D regions (patches) that capture where the two networks are different. We demonstrate our approach on a set of actual power networks from a county in the USA. Our approach can be extended to validate synthetic networks created for multiple infrastructures such as transportation, communication, water, and gas networks.
We analyze hierachical relations between the minimum/maximum angle conditions, widely used in the interpolation theory and the finite element analysis, for tetrahedral partitions.
In response to the low efficiency of current environmental, social, and corporate governance rating prediction management methods in China, a stacked model theoretical framework integrating light gradient boosting machine feature filtering algorithm is proposed and applied to environmental, social, and corporate governance rating prediction management. By improving the feature importance of the output results of the stacked model base learner, the feature imbalance in the model is optimized. Then, a weighted strategy is designed to adjust the fitting performance of the model. The results showed that the improved stacked model increased accuracy by 8.3% and 4.98% respectively compared with the other two models in the self-built dataset. The fitting accuracy of the improved stacked model was 92.32%. The results indicated that the designed method optimized the input layer of the stacked model and enhanced the generalization ability. The weighted optimization strategy of the designed base learner can improve the model fitting effect and enhance the efficiency of stacked models in predicting and managing environmental, social, and corporate governance ratings. The proposed prediction method has positive application significance in environmental, social, and corporate governance rating prediction.
In this paper we consider the following [Formula: see text]-Central Path Problem (2CPP): Given a set of [Formula: see text] polygonal curves [Formula: see text] in the plane, find two curves [Formula: see text] and [Formula: see text], called [Formula: see text]-central paths, that best represent all curves in [Formula: see text]. Despite its theoretical interest and a wide range of practical applications, 2CPP has not been well studied. In this paper, we first establish criteria that [Formula: see text] and [Formula: see text] ought to meet in order for them to best represent [Formula: see text]. In particular, we require that there exists parametrizations [Formula: see text] and [Formula: see text] ([Formula: see text]) of [Formula: see text] and [Formula: see text] respectively such that the maximum distance from [Formula: see text] to curves in [Formula: see text] is minimized. Then an efficient algorithm is presented to solve 2CPP under certain realistic assumptions. Our algorithm constructs [Formula: see text] and [Formula: see text] in [Formula: see text] time, where [Formula: see text] is the total complexity of [Formula: see text] (i.e., the total number of vertices and edges), [Formula: see text] is the number of curves in [Formula: see text], and [Formula: see text] is the inverse Ackermann function. Our algorithm uses parametric search technique and is considerably faster than arrangement-related algorithms (i.e. [Formula: see text]) when [Formula: see text] as in most real applications.
In this paper we consider the problem of training a Support Vector Machine (SVM) online using a stream of data in random order. We provide a fast online training algorithm for general SVM on very large datasets. Based on the geometric interpretation of SVM known as the polytope distance, our algorithm uses a gradient descent procedure to solve the problem. With high probability our algorithm outputs an [Formula: see text]-approximation result in constant time and space, which is independent of the size of the dataset, where [Formula: see text]-approximation means that the separating margin of the classifier is almost optimal (with error [Formula: see text]), and the number of misclassified training points is very small (with error [Formula: see text]). Experimental results show that our algorithm outperforms most of existing online algorithms, especially in the space requirement aspect, while maintaining high accuracy.
We considered differential geometries of helical curves for ruled and developable surfaces in Euclidean 3-space E 3. We clarify the conditions for ruled surface to become stationary-axis ruled surface. Then, we concentrate on ruled surface generated by a line undergoing a one-parameter screw movement. Finally, we provide some modules to illustrate the main results.
This paper deals with the problem of reconstructing the path of a vehicle in an unknown environment consisting of planar structures using sound. Many systems in the literature do this by using a loudspeaker and microphones mounted on a vehicle. Symmetries in the environment lead to solution ambiguities for such systems. We propose to resolve this issue by placing the loudspeaker at a fixed location in the environment rather than on the vehicle. The question of whether this will remove ambiguities regardless of the environment geometry leads to a question about breaking symmetries that can be phrased in purely mathematical terms. We solve this question in the affirmative if the geometry is in dimension three or bigger, and give counterexamples in dimension two. Excluding the rare situations where the counterexamples arise, we also give an affirmative answer in dimension two. Our results lead to a simple path reconstruction algorithm for a vehicle carrying four microphones navigating within an environment in which a loudspeaker at a fixed position emits short bursts of sounds. This algorithm could be combined with other methods from the literature to construct a path tracking system for vehicles navigating within a potentially symmetric environment.
We study the problem of covering barrier points by mobile sensors. Each sensor is represented by a point in the plane with the same covering range [Formula: see text] so that any point within distance [Formula: see text] from the sensor can be covered by the sensor. Given a set [Formula: see text] of [Formula: see text] points (called “barrier points”) and a set [Formula: see text] of [Formula: see text] points (representing the “sensors”) in the plane, the problem is to move the sensors so that each barrier point is covered by at least one sensor and the maximum movement of all sensors is minimized. The problem is NP-hard. In this paper, we consider two line-constrained variations of the problem and present efficient algorithms that improve the previous work. In the first problem, all sensors are given on a line [Formula: see text] and are required to move on [Formula: see text] only while the barrier points can be anywhere in the plane. We propose an [Formula: see text] time algorithm for the problem. We also consider the weighted case where each sensor has a weight; we give an [Formula: see text] time algorithm for this case. In the second problem, all barrier points are on [Formula: see text] while all sensors are in the plane but are required to move onto [Formula: see text] to cover all barrier points. We also solve the weighted case in [Formula: see text] time.
International Journal of Computational Geometry & ApplicationsVol. 18, No. 01n02, pp. 1-2 (2008) Special Issue: Selected Papers from the 21st ACM Symposium on Computational Geometry (SoCG), 6–8 June 2005, Pisa, ItalyNo AccessGUEST EDITOR'S FOREWORDALON EFRATALON EFRATDepartment of Computer Science, The University of Arizona, Tucson, AZ 85721, USAhttps://doi.org/10.1142/S0218195908002519Cited by:0 Next AboutSectionsPDF/EPUB ToolsAdd to favoritesDownload CitationsTrack CitationsRecommend to Library ShareShare onFacebookTwitterLinked InRedditEmail Remember to check out the Most Cited Articles! Check out these titles in image analysis! FiguresReferencesRelatedDetails Recommended Vol. 18, No. 01n02 Metrics History PDF download
There exists a linear algorithm to decide whether a polyomino tessellates the plane by translation only. On the other hand, the problem of deciding whether a set of [Formula: see text] or more polyominoes can tile the plane by translation is undecidable. We narrow the gap between decidable and undecidable by showing that it remains undecidable for a set of [Formula: see text] polyominoes, which partially solves a conjecture posed by Ollinger.
We give some properties of semi-symmetric pseudo-Riemannian manifolds as an indecomposable irreducible Ricci pseudo-Riemannian manifold (i.e. the minimal polynomial of its Ricci operator is irreducible) is semi symmetric if and only if it is locally symmetric. We also show that any semi-symmetric pseudo-Riemannian manifold will be foliated. Moreover, if the metric is Lorentzian, the Ricci operator has only real eigenvalues and more precisely, on each leaf, it is diagonalizable with at most a single non zero eigenvalue or isotropic.
In this paper, we consider enumeration of geodesics on a polyhedron, where a geodesic means locally-shortest path between two points. Particularly, we consider the following preprocessing problem: given a point [Formula: see text] on a polyhedral surface and a positive real number [Formula: see text], to build a data structure that enables, for any point [Formula: see text] on the surface, to enumerate all geodesics from [Formula: see text] to [Formula: see text] whose length is less than [Formula: see text]. First, we present a naive algorithm by removing the trimming process from the MMP algorithm (1987). Next, we present an improved algorithm which is practically more efficient on a non-convex polyhedron, in terms of preprocessing time and memory consumption. Moreover, we introduce a single-pair geodesic graph to succinctly encode a result of geodesic query. Lastly, we compare these naive and improved algorithms by some computer experiments.
