Hungarian Algorithm Research Articles

A phased resource allocation algorithm for full‐duplex device‐to‐device communications underlaying cellular networks

AbstractDevice‐to‐device (D2D) and full duplex (FD) technologies demonstrate immense potentials to improve the spectral efficiency of wireless networks. However, their synergistic combination to reap added benefits is proved to be technically challenging. This letter investigates the throughput‐maximization resource allocation problem for FD D2D communications underlaying cellular networks. The problem is decoupled into two subproblems, namely power allocation and channel assignment, which are solved by using a novel quadratic transform based scheme and the Kuhn–Munkres algorithm, respectively. Simulation results show that the proposed algorithm outperforms existing ones in terms of overall system throughput.

Don’t get lost in the crowd: Graph convolutional network for online animal tracking in dense groups

Tracking the behaviour of animals in group-housed situations is a critical area of study for precision livestock farming, but it poses challenges in diverse and crowded environments. Existing methods often struggle with false positives and false negatives due to the complexities of these scenarios. In this study, we propose a robust computer vision algorithm for long-term (>10mins) animal tracking, with a primary focus on group-housed pigs. Our method addresses the limitations of current approaches by effectively handling errors from the detector in challenging environments. Our approach integrates Graph Convolutional Networks.(GCNs) with deep learning-based object detection techniques. We represent the data as a graph structure, with nodes corresponding to multiple animal detections over several frames. Edges connect thedetections that do not appear in the same frame. The model's objective is to perform edge classification, where each edge is associated with a scalar representing the probability of being the same object. To enhance the robustness of edge classification, we combine the classifier withan ensemble predictor, enabling accurate tracking of animal identitiesfor extended periods. Notably, our graph convolutional model achievesaccurate re identification of non-consecutive animal detections inreal-time without the need for trajectory prediction or ID association. We evaluate our method by comparing its performance against othertracking methods, including DeepSORT—a state-of-the-art animal trackingmethod that relies on a Kalman filter, a deep appearance descriptor, and the Hungarian algorithm for trajectory prediction and ID association.Our comparison demonstrates the advantages of our method, as it (1)improves the IDF1 score, indicating enhanced ID association accuracy, by 1.72%; (2) mitigates 93% of errors resulting from ID-switch anddeviation; (3) extends the tracking duration of each animal byapproximately 66% in challenging conditions; and (4) achieves aprocessing speed that is 1000 times faster than DeepSORT when operatingat 22fps with the detector. The potential applications of our method extend to various aspects of livestock management. By accurately tracking individual animals, we can monitor behaviours such as feeding, drinking, and aggressive interactions. This information can lead to improvements in animal welfare, resource allocation, and breeding strategies. Follow-ups of this research are available at: https://gitlab.kuleuven.be/m3-biores/public/m3pig.

HVD-Net: A Hybrid Vehicle Detection Network for Vision-Based Vehicle Tracking and Speed Estimation

Visual surveillance of on-road vehicles has emerged as a vibrant research topic within the fields of computer vision and Intelligent Transportation Systems (ITSs) to mitigate traffic problems. Road accidents result in traffic congestion, vehicle damage, injuries, and fatalities. The likelihood, severity, and fatality of automobile collisions are all increased by speeding. Autonomous Speed Limit Violation Detection (SLVD) mechanism offers the best way to estimate speed with high precision using Intelligent Transportation Systems (ITSs) technology. Vision-based Vehicle Speed Monitoring (VSM) pipeline is proposed in this work, which includes mechanisms for vehicle detection, tracking, and speed measurement. VSM pipeline is based on a three-tier architecture, exploiting a single RSU-camera. In the first tier, a real-time CNN-based Hybrid Vehicle Detection Network (HVD-Net) is designed for vehicle detection. The HVD-Net utilizes multi-level and multi-scale features to minimize the impact of vehicle scale variation and maximize detection accuracy. Secondly, a Simple Online Real-time Tracker (SORT) along with HVD-Net is adopted to track vehicles' trajectories. The tracker utilizes a Kalman filter for vehicle state estimation and a Hungarian algorithm to solve the multi-vehicle association problem. Finally, a mechanism is presented to estimate the vehicle speed in the motion plane to mitigate speed overestimation. Empirical evaluations on three datasets demonstrate that the proposed vehicle detection, tracking, and speed estimation schemes perform better when compared with the relevant and state-of-the-art schemes. The estimated pipeline vehicle speed is achieved with accuracy of 87.242%.

Time-Efficient Joint UAV-BS Deployment and User Association Based on Machine Learning

In this paper, a time-efficient mechanism is proposed to solve the joint unmanned aerial vehicle (UAV) base station deployment and user/sensor association (UDUA) problem aiming at maximizing the downlink sum transmission throughput and reducing the time of on-line computations. In this work, two relevant sub-problems are decoupled from the joint UDUA problem: the first one is denoted as the user association sub-problem, for certain UAV base station (UAV-BS) positions, this sub-problem is settled for finding the strategy which matches aerial and ground nodes optimally. The second sub-problem is the positioning of UAV-BSs in order to obtain the best possible solution to the user association sub-problem from all possible positioning combinations for the UAV-BSs. In the proposed mechanism, the Kuhn-Munkres algorithm is used to solve the first sub-problem as an equivalent bipartite matching problem. For the UAV-BS deployment sub-problem, when the two user distributions own a high similarity, we theoretically prove that little performance decline will be introduced when the new user distribution’s optimal strategy is compared with choosing the optimal UAV-BS deployment strategy of stored user distributions. Based on our mathematical analyses, the similarity level between user distributions is well defined and becomes the key to solve the second sub-problem. According to experimental findings, the proposed UDUA mechanism, when compared to benchmark approaches, can provide near-optimum system performance with regard to average downlink total transmission throughput and failure rate with significantly decreased computing time.

Disease Detection and Identification of Rice Leaf Based on Improved Detection Transformer

In recent years, the domain of diagnosing plant afflictions has predominantly relied upon the utilization of deep learning techniques for classifying images of diseased specimens; however, these classification algorithms remain insufficient for instances where a single plant exhibits multiple ailments. Consequently, we view the region afflicted by the malady of rice leaves as a minuscule issue of target detection, and then avail ourselves of a computational approach to vision to identify the affected area. In this paper, we advance a proposal for a Dense Higher-Level Composition Feature Pyramid Network (DHLC-FPN) that is integrated into the Detection Transformer (DETR) algorithm, thereby proffering a novel Dense Higher-Level Composition Detection Transformer (DHLC-DETR) methodology which can effectively detect three diseases: sheath blight, rice blast, and flax spot. Initially, the proposed DHLC-FPN is utilized to supersede the backbone network of DETR through amalgamation with Res2Net, thus forming a feature extraction network. Res2Net then extracts five feature scales, which are coalesced through the deployment of high-density rank hybrid sampling by the DHLC-FPN architecture. The fused features, in concert with the location encoding, are then fed into the transformer to produce predictions of classes and prediction boxes. Lastly, the prediction classes and the prediction boxes are subjected to binary matching through the application of the Hungarian algorithm. On the IDADP datasets, the DHLC-DETR model, through the utilization of data enhancement, elevated mean Average Precision (mAP) by 17.3% in comparison to the DETR model. Additionally, mAP for small target detection was improved by 9.5%, and the magnitude of hyperparameters was reduced by 324.9 M. The empirical outcomes demonstrate that the optimized structure for feature extraction can significantly enhance the average detection accuracy and small target detection accuracy of the model, achieving an average accuracy of 97.44% on the IDADP rice disease dataset.

An Electric Vehicle Charging Method Considering Multiple Power Exchange Modes’ Coordination

To achieve sustainable environmental development, numerous countries and governments have been vigorously promoting the proliferation of electric vehicles (EVs) through a series of policy measures and economic subsidies. With the increasing number of EVs, multiple EV charging modes are being researched to satisfy owners’ requirements. In this paper, an EV charging method considering multiple power exchange modes’ coordination is proposed for meeting owners’ requirements with cost-effectiveness. In the proposed method, the battery swapping (BS) station, building-to-vehicle (B2V) station, and grid-to-vehicle (G2V) station are considered. In G2V stations, EVs charge from the power grid. In B2V stations, distributed renewable energy generation is considered as the energy provider. This study contemplates the use of photovoltaic power systems in B2V stations for the charging of EVs. In BS stations, the power exchange among batteries and the power grid is considered. The battery energy storage is utilized for reducing the battery degradation cost (BDC) and power cost. EVs are dispatched to the corresponding charging stations according to the electric price, BDC, and so on. In the dispatching process, the particle swarm optimization (PSO) algorithm and Hungarian algorithm are applied. Several case studies are presented to validate the effectiveness of the proposed method and the power matching between EVs and charging modes is discussed.

Users' grouping algorithm for fairness improvement of NOMA‐based multi‐beams satellite networks intended for 5G

AbstractNon‐Orthogonal Multiple‐Access (NOMA)‐based Multibeam Satellite Networks (MBSNs) are intended for 5G implementation, and 5G networks set the requirement for both high system capacity as well as high system reliability to all users, which is achieved via high system's fairness. The NOMA network requires a user grouping algorithm (UGA), and most existing UGAs for 2‐users NOMA‐MBSNs, favor the system's capacity maximization with little regard to the system's fairness. Thus, this paper proposes a UGA that maximizes the fairness of 2‐users NOMA‐MBSNs. The algorithm seeks to group 2M users into M‐pairs satisfying three requirements, (i) a minimum channel‐gain margin; (ii) high channel‐correlation coefficients; and (iii) there is high fairness amongst the resulting pairs, in term of their respective channel‐correlation coefficients, to increase the users' fairness of the system. To achieve this, the 2M users are split into two sets of M‐far and M‐near users based on users' channel gains for pairing far and near users with high channel‐correlation coefficients. It will turn it into a maximization Bipartite Matching problem (BPMP) of channel‐correlation coefficients. The BPMP is addressed by means of the Hungarian Method (HM) to ensure high fairness amongst resulting pairs regarding their channel‐correlation coefficients. The proposed UGA demonstrates superior fairness compared to existing UGAs.

On the Variance of Interval Type-2 Fuzzy Sets

This paper presents some theoretical considerations about the variance of a fuzzy set and introduces a covering–based algorithm for computing the bounds of the variance of an interval Type–2 fuzzy set. The proposed algorithm obtains the bounds of all possible variances of an interval Type–2 fuzzy set unlike other algorithms which are based on the center of centroid of an interval Type–2 fuzzy set to obtain a relative variance which is proven to be always bigger than the absolute variance. A comparison to the relative variance of an interval Type–2 fuzzy set is given through different examples where the proposed algorithm shows better precision than the KM algorithm.

Optimization of Packaging Carton Production at PT X with the Hungarian Method and Multi Attribute Utility Theory

Household shopping activities in Indonesia have increased by 51% percent since the pandemic period (BPS, 2022). The increase affects household ingredients in packaging. PT X produces various types of packaging for household needs with various size specifications through the make to order system has experienced an overload of production quantity resulting in a delay of 36.19%. PT X needs to conduct problem analysis and production planning to improve ontime delivery. Problem analysis is carried out by direct observation methods and management interviews. Production planning is carried out by dividing and prioritizing tasks using the Hungarian optimization method and Multi Attribute Utility Theory

INFORMATION-LOGICAL MODEL OF EDUCATION OPTIMIZATION IN REMOTE MODE

The educational optimization process is widely researched in the theoretical aspect. Analyzing existing sources made it possible to highlight the research issues: the need to create an optimizing distance studying model, in which work with weaker students becomes possible both within the educational process and in individual or group independent work. The study aims to develop an information-logical model for optimizing distance studying. This model should provide for the learning process organization in such a way as to strengthen the weaknesses of students, reveal their potential and focus on the comprehensive development of knowledge and skills. The task formalization is carried out using the Hungarian algorithm and Boolean variables. The main work limitation is the operation with integers. Discreteness manifests itself already at the modelling stages in many problems, for example, when working with Boolean variables. An example with the most straightforward information model construction using the logical functions "true/false" with the transition to a chain of matrices is given. It demonstrates the studying optimization algorithm and presents an expanded information-logical model. The presented model was preliminary tested in one of the academic groups of the S. Seifullin Kazakh Agro Technical Research University. Students' knowledge inspections were carried out at the approbation beginning. Then a student group working on a student project was divided into subgroups according to the algorithm. The knowledge inspection showed an 11.3% improvement in results at the work's end. Further research on this topic may consist of expanding the presented model's capabilities and developing appropriate modules for knowledge control and algorithmization of related tasks.

Autoencoder-based efficient resource allocation in device-to-device communication

Device-to-device (D2D) communication offloads the base station (BS) traffic, enhances spectral efficiency, and provides efficient data transmission with low latency between devices in close proximity. However, there are many issues in D2D communication, such as mode selection, device discovery, security, privacy, resource allocation, and interference management. At the time of the resource allocation, there is an issue of interference and eavesdropping by the malicious user on D2D pairs. Here, an eavesdropper can decrypt the data exchanged between the D2D user devices. This paper focuses on the efficient resource allocation in D2D communication that improves sum rate, computational complexity, and information secrecy capacity by reusing cellular user’s (CU) resources. Here, we propose an efficient resource allocation scheme incorporating the artificial intelligence (AI) algorithm-based autoencoder that overcomes information ambiguity when there is the same weight in the data rate matrix of the Hungarian algorithm. First, we identify cellular and D2D users’ communication data rates and generate a data rate matrix using the channel gain, received signal, and signal-to-interference-plus-noise ratio (SINR) values. The generated data rate matrix is passed as an input into an autoencoder. An autoencoder performs the dimensionality reduction using latent space and reconstructs the optimal cost matrix. The AI model’s result shows that the autoencoder achieves 83% training and 90% validation accuracy using the mean squared error (MSE) loss function. Then, we determine the optimal cost using the Hungarian algorithm. As we aim to maximize the sum rate and reduce the computation cost, our simulation result shows that the resource allocation performs efficiently using the proposed scheme.

Efficient Distributed Inference of Deep Neural Networks via Restructuring and Pruning

In this paper, we consider the parallel implementation of an already-trained deep model on multiple processing nodes (a.k.a. workers). Specifically, we investigate as to how a deep model should be divided into several parallel sub-models, each of which is executed efficiently by a worker. Since latency due to synchronization and data transfer among workers negatively impacts the performance of the parallel implementation, it is desirable to have minimum interdependency among parallel sub-models. To achieve this goal, we propose to rearrange the neurons in the neural network, partition them (without changing the general topology of the neural network), and modify the weights such that the interdependency among sub-models is minimized under the computations and communications constraints of the workers while minimizing its impact on the performance of the model. We propose RePurpose, a layer-wise model restructuring and pruning technique that guarantees the performance of the overall parallelized model. To efficiently apply RePurpose, we propose an approach based on L0 optimization and the Munkres assignment algorithm. We show that, compared to the existing methods, RePurpose significantly improves the efficiency of the distributed inference via parallel implementation, both in terms of communication and computational complexity.

Partial offloading in device-to-device-assisted MEC network: A utility optimization approach

In order to meet the growing demand for faster and more efficient communication and computing in internet of things networks, the device-to-device (D2D)-assisted multi-access edge computing (MEC) network has emerged as a promising solution. However, most existing literature overlooks the monetary cost associated with computing and communication resources when it comes to computation offloading. Motivated by this, we first formulate a novel utility function which is defined as the tradeoff between the benefit of the delay reduction via partial offloading and the cost of the purchased computing and communication resources. Then, we propose the partial offloading-based utility optimization (POUO) algorithm to solve the utility optimization problem. Afterward, we determine the resource allocation policy and offloading ratios while taking into account the constraints of the utility optimization problem. In addition, the offloading pairing decision in the device-to-edge server (D2E) offloading mode is modeled as a knapsack problem and solved using a greedy algorithm. For D2D offloading mode, we transform the offloading pairing decision into an assignment problem and apply the Hungarian algorithm to solve it. Finally, we validate and evaluate the proposed POUO algorithm across a range of system parameters. Simulation results show that the POUO algorithm improves the completed tasks ratio, the total utility, and the average delay.

Hybrid approach of Hungarian method to find optimal solution for solving Fuzzy Transportation Problem using Hexagonal Fuzzy numbers

This research article explains about a hybrid approach of Hungarian method to solve Hexagonal Fuzzy Transportation Problem. Step by step algorithm is given in this article. A numerical illustration is given to verify the new hybrid algorithm. Within short duration, optimal solution is derived easily.

Discrete optimizations’ problems of deliveries of heterogeneous products

The article touches upon issues related to the tasks of the national project «Smart City». The paper analyzes the problem of supplying a conditional consumer with heterogeneous products in accordance with his demand for deliveries from several suppliers in a situation with fixed surcharges, in addition to the cost of purchasing each unit of production. For the situation under study, a model of a reduced transport type with a discontinuous piecewise linear objective function of minimized total costs and with a system of linear constraints is constructed. A method for finding the optimal solution one of the many such solutions, based on the ideas of the Hungarian algorithm is proposed, the justification of which is given on the basis of the corresponding lemma. A refinement of the method is presented in the presence of some restrictions related to suppliers. The polynomial complexity of the method, i.e. the problem is quickly solvable, and the significant limitation of the applicability of the method within the framework of the model are noted. Further possible research directions of a stochastic or fuzzy nature are indicated.

Identification and Tracking of Vehicles between Multiple Cameras on Bridges Using a YOLOv4 and OSNet-Based Method

The estimation of vehicle loads is a rising research hotspot in bridge structure health monitoring (SHM). Traditional methods, such as the bridge weight-in-motion system (BWIM), are widely used but they fail to record the locations of vehicles on the bridges. Computer vision-based approaches are promising ways for vehicle tracking on bridges. Nevertheless, keeping track of vehicles from the video frames of multiple cameras without an overlapped visual field poses a challenge for the tracking of vehicles across the whole bridge. In this study, a method that was You Only Look Once v4 (YOLOv4)- and Omni-Scale Net (OSNet)-based was proposed to realize vehicle detecting and tracking across multiple cameras. A modified IoU-based tracking method was proposed to track a vehicle in adjacent video frames from the same camera, which takes both the appearance of vehicles and overlapping rates between the vehicle bounding boxes into consideration. The Hungary algorithm was adopted to match vehicle photos in various videos. Moreover, a dataset with 25,080 images of 1727 vehicles for vehicle identification was established to train and evaluate four models. Field validation experiments based on videos from three surveillance cameras were conducted to validate the proposed method. Experimental results show that the proposed method has an accuracy of 97.7% in terms of vehicle tracking in the visual field of a single camera and over 92.5% in tracking across multiple cameras, which can contribute to the acquisition of the temporal–spatial distribution of vehicle loads on the whole bridge.

Alternate Approach (Penalty Approach) to Assignment Problem Solving and Comparison to Existing Approaches

Assignment problem is an important area in Operation Research and is also discussed in real physical world. In this paper an attempt has been made to solve the assignment problem using a new Method called the Penalty method. We discuss a numerical example by using the new Method and compare it with standard existing method which is the Hungarian method. We compare the optimal solution of the new Method and the Hungarian method. The new method is a simple procedure, easy to apply for solving assignment problem.

LBCC-Hung: A load balancing protocol for cloud computing based on Hungarian method

Cloud Computing can be defined as enabling computing resources to whoever want remotely upon demand. This necessarily requires using virtualization. A virtual machine (VM) is an emulation of a computer that runs in virtualization software. Tasks coming from different users are passed to the virtual machines to be processed. This paper studies the load balancing problem among various VMs; to ensure that the network resources are distributed in a fair way between various clients. This paper proposes solving the load balancing problem using a combinatorial optimization approach named the Hungarian method. The proposed protocol is named LBCC-Hung (Load Balancing Protocol for Cloud Computing Based on Hungarian Method). Using simulation, the performance of LBCC-Hung is measured and compared with two well-known methods; MIN-MIN and First Come First Serve FCFS methods. The simulation results proved that LBCC-Hung overperforms the others two protocols, in terms of both the Makespan and the throughput and the virtual machine utilization deviation.

Multi-UAV Aerial Mission Intelligent Assignment Algorithm Based on LightGBM Model

Abstract With the development of UAV technology, UAVs have the advantages of convenience, speed, and small size. However, when faced with a complex task environment, some traditional artificial intelligence algorithms will significantly increase computation time, and the algorithms that perform better in time performance often fail to adapt to the type and number of random tasks to be assigned. In this paper, we combine the Hungarian algorithm with improving the LightGBM model, simulate the implementation in various task scenarios, and verify that the improved LightGBM model completes the intelligent assignment of multiple UAVs without increasing time cost compared to the classical algorithm through comparative analysis. It is shown that the improved LightGBM model has an average time of 26.39s for 10 runs, which is higher than 16.91s for NSGA-II and 14.73s for PESA, but lower than 46.92s for SPEA2 and 50.01s for SPGA2. For a non-real-time assignment system, a better mission intelligence assignment at a certain time cost scheme is more in line with the system’s performance requirements. Therefore, the improved LightGBM model in this paper can obtain better solutions than the other four algorithms in solving the multi-objective multi-UAV task intelligence assignment problem, and the solution set coverage is larger than several other algorithms. This indicates that the algorithm studied in this paper is feasible in solving the multi-objective multi-UAV task intelligence assignment problem, and it also has some reference significance for solving the multi-objective problem and improving the coverage of the multi-objective solution set.

Vehicle detection and tracking using low-channel roadside LiDAR

Traffic object detection and tracking is one of the fundamental tasks in processing point cloud data to collect high-resolution vehicle trajectories using Roadside LiDAR sensors. Comprehensive, accurate, and complete traffic object tracking within the sensor’s scanning range still remains elusive, especially in complex traffic environments, mainly due to challenges like occlusion. To solve these issues, a novel vehicle detection and tracking method for low-channel roadside LiDAR in complex environments is proposed. Firstly, we use the L-shape fitting method to get a more accurate bounding box to extract the object features. Then, the decision tree with bagging algorithm is used to classify the traffic objects based on the selected features. Next, an improved Hungarian algorithm with the Kalman filter is used to predict the vehicle’s path considering the conditions of complete and partial occlusion. Finally, the effectiveness of the proposed framework is evaluated by comparing the roadside LiDAR data collected from four sites with the ground truth data obtained either from video-cameras or VeloView (software for point cloud data visualization). The result shows that the detection and tracking accuracy of the proposed method can reach up to 99.50% and 97%, respectively, which outperforms the state-of-the-art algorithms.