Network Flow Research Articles

Learning-augmented maximum flow

We propose a framework for speeding up maximum flow computation by using predictions. A prediction is a flow, i.e., an assignment of non-negative flow values to edges, which satisfies the flow conservation property, but does not necessarily respect the edge capacities of the actual instance (since these were unknown at the time of learning). We present an algorithm that, given an m-edge flow network and a predicted flow, computes a maximum flow in O(mη) time, where η is the ℓ1 error of the prediction, i.e., the sum over the edges of the absolute difference between the predicted and optimal flow values. Moreover, we prove that, given an oracle access to a distribution over flow networks, it is possible to efficiently PAC-learn a prediction minimizing the expected ℓ1 error over that distribution. Our results fit into the recent line of research on learning-augmented algorithms, which aims to improve over worst-case bounds of classical algorithms by using predictions, e.g., machine-learned from previous similar instances. So far, the main focus in this area was on improving competitive ratios for online problems. Following Dinitz et al. (2021) [6], our results are among the firsts to improve the running time of an offline problem.

Usage-aware representation learning for critical information identification in transportation networks

Extracting meaningful information from noisy high-dimensional data is attracting increasing attention as richer and higher resolution data is being collected and used for transportation system planning and management purposes. Discovering critical information via effective data representation learning not only helps reduce data dimension, it also enables a deeper understanding of the underlying properties of noisy data, which could then lead to better planning and operations decisions. In this study, we present a new perspective that, unlike most existing approaches in the general data science literature, the design of data representation should go beyond the data itself; it should incorporate an understanding of how the data is used in the domain-specific applications. We further argue that this design philosophy is particularly important for transportation data because of the high spatial correlations of transportation data brought by network interdependence. We propose a usage-aware representation learning framework by incorporating the information loss for downstream application into the data encoding-decoding process. The proposed approach is formulated as a Stiefel manifold optimization problem. The effectiveness of the proposed framework is demonstrated in two network applications: modeling transportation network flows and estimating network-level vehicular emissions. The performance of the learned representation from our approach is compared with existing approaches using multiple evaluation context, including data reconstruction quality, clustering, anomaly detection, and critical information identification, through case studies implemented in Sioux Falls, Boston, and San Jose networks. The good performance of our approach consistently observed in those experiments indicates the importance of incorporating the downstream data usage in the process of data representation learning.

Democratizing traffic control in smart cities

To improve the performance of systems, optimization has been the prevailing approach in the past. However, the approach faces challenges when multiple goals shall be simultaneously achieved. For illustration, we study a multi-agent system, where agents have a plurality of different, and mutually inconsistent goals. We then allow agents in the system to vote on which traffic signal controllers, which were trained on different goals using deep reinforcement learning, would control the intersection. Taking decisions based on suitable voting procedures turns out to lead to favorable solutions, which perform highly for several goals rather than optimally for one goal and poorly for others. This opens up new opportunities for the management or even self-governance of complex systems that require the consideration and achievement of multiple goals, such as many systems involving humans. Here, we present results for traffic flows in urban street networks, which suggest that “democratizing traffic” would be a promising alternative to centralized control of traffic flows.

Exploring global embodied carbon emissions transfer network—An analysis based on national responsibility

This paper explores the global embodied carbon emissions transfer network. To clarify the carbon emissions reduction responsibilities of each country globally, a complex interdependent network structure based on the 2016 World Multi-Regional Input-Output (MRIO) table and the transfer of production factors between industries is constructed. Using input-output and social network analyses, the structural features of the global carbon emissions flow network is uncovered, and then carbon inequality among countries from a multi-regional perspective is revealed. The research findings show that the scale of global embodied carbon transfer expands continuously. The direction of transfer exhibits asymmetry, regionalism, and complexity. Hungary, Romania, Slovenia, Iran, Ethiopia, and Iran are the main carbon-exporting countries, while the United States, Germany, the United Kingdom, Japan, Canada, Spain are the main carbon-importing countries. Furthermore, the block model analysis reveals that the carbon transfer and transmission effects between global regions tend to be centralized clusters, from “demand-driven” to “internalized supply-demand coexistence”. There is significant uneven exchange between wealth benefits and environmental burdens among blocks.

Impact of detour on traffic flow in branching Koch curve network with bottleneck

There are various routes with and without detours in city traffic network with a complex connectivity. Branching Koch curve fractal has such complex connectivity with singly, doubly, and multiply connecting links. We consider branching Koch curve as a city traffic network. We study the effect of detour routes (bypasses) on macroscopic traffic flow in branching Koch curve network with a bottleneck. The bottleneck effect depends on whether or not there are detour routes. We propose a mathematical dynamic model for the traffic flow with a bottleneck in the directed Koch curve network. The macroscopic dynamic equations of vehicular densities are derived in the Koch curve network. Vehicular densities and currents are obtained at all roads in the network by solving the density equations numerically. The fundamental diagrams (flow-density relations) are derived. The traffic currents (flows) depend highly on the position of the bottleneck due to the connectivity of the fractal network. It is shown where and how traffic congestion is induced by the bottleneck on the directed fractal network.

ActiveGuardian: An accurate and efficient algorithm for identifying active elephant flows in network traffic

Active elephant flows, which indicate the real-time data transmission status, are of primary interest in network management and various applications. However, existing network measurement efforts mainly focus on finding elephant flows, and limited works on identifying active elephant flows suffer from low accuracy and heavy overheads. To address this issue, this paper proposes ActiveGuardian to identify active elephant flows with high accuracy, low memory, and high throughput. The key idea is to intelligently separate potential elephant flows from mice flows, and guard and report the information of active elephant flows in every time window. To obtain high accuracy, we devise a filtering module that adaptively filters unnecessary flows with low arrival rates, by applying an Adaptive Counter Update strategy. To achieve high memory utilization, we design a Leapfrog Hashing algorithm for the guarding module to effectively solve hash collisions. Lastly, we perform theoretical derivation on the false positive and the error bound of ActiveGuardian, and experimental evaluations on its performance with real network traffic traces. The experimental results show that ActiveGuardian achieves higher accuracy (99.65%) with identical memory sizes, and higher throughput (26.53Mps) than the state-of-the-art solutions.

Triadic signatures of global human mobility networks.

Global refugee and migrant flows form complex networks with serious consequences for both sending and receiving countries as well as those in between. While several basic network properties of these networks have been documented, their finer structural character remains under-studied. One such structure is the triad significance profile (TSP). In this study, the TSPs of global refugee and migrant flow networks are assessed. Results indicate that the migrant flow network's size and TSP remain stable over the years; its TSP shares patterns with social networks such as trade networks. In contrast, the refugee network has been more dynamic and structurally unstable; its TSP shares patterns with networks in the information-processing superfamily, which includes many biological networks. Our findings demonstrate commonality between migrant and social networks as well as between refugee and biological networks, pointing to possible interdisciplinary collaboration-e.g., application of biological network theories to refugee network dynamics-, potentially furthering theoretical development with respect to both network theory and theories on human mobility.

Uncovering structural characteristics of virtual water flow network in India

India plays a pivotal role in global water resource management. Existing studies have not well characterized structural characteristics of virtual water flow network in India (e.g., critical sectors, transactions, and supply chain paths) in the context of global supply chains. This may impede the efficiency of policy measures on the mitigation of India's water shortage. This study identified critical transmission sectors, intersectoral transactions, supply chain paths, and communities of India's virtual water flow network, by combining the global environmentally extended multi-regional input-output model, structural path analysis, betweenness concept, and community detection method. Results show that in addition to domestic water-intensive sectors, the transmission sectors, intersectoral transactions, and supply chain paths associated with domestic service sectors (e.g., the “Hotel and restaurant services” and “Other land transportation services” sectors) and foreign sectors (mainly chemical industries, light industries, and services in large manufacturing entities) also play important roles in transmitting India's water uses. The importance of critical transmission sectors and transactions related to foreign regions (e.g., the rest of Asia-Pacific Region and China) have increased significantly during 1995–2020. Moreover, major communities mostly involve sectors from different regions. These findings can provide hotspots for India's water resource management in terms of improving the productivity and strengthening the interregional and intersectoral cooperation.

Interplay of network topologies in aviation delay propagation: A complex network and machine learning analysis

In this study, the fundamental characteristics of flight delay propagation and the key factors influencing such propagation are investigated. Three distinct types of networks were constructed: an aviation network, a traffic flow network, and a delay propagation network. Employing complex network theory, an analysis of the fundamental topological attributes of each network was conducted, exploring the interrelations among these attributes. Findings reveal a notable resemblance between the network topology attributes of the delay propagation network and the traffic flow network. The delay propagation network overall adheres to scale-free network attributes, with a few dominant major airports playing a pivotal role in the delay propagation process. Furthermore, it was uncovered that smaller airports are more susceptible to the influence of delay propagation than their larger counterparts. Delays, in a general sense, exhibit a pronounced tendency to aggregate and spread extensively within the realm of smaller airports. Regarding delay propagation, airport traffic emerges as the primary factor precipitating this phenomenon, with a feature importance score reaching 0.889. An escalation in airport traffic significantly amplifies the extent of delay propagation, yet concurrently, this escalation is not incessant but stabilizes beyond a certain threshold.

Long‐range quantum energy teleportation and distribution on a hyperbolic quantum network

AbstractTeleporting energy to remote locations is new challenge for quantum information science and technology. Developing a method for transferring local energy in laboratory systems to remote locations will enable non‐trivial energy flows in quantum networks. From the perspective of quantum information engineering, we propose a method for distributing local energy to a large number of remote nodes using hyperbolic geometry. Hyperbolic networks are suitable for energy allocation in large quantum networks since the number of nodes grows exponentially. To realise long‐range quantum energy teleportation (QET), we propose a hybrid method of quantum state telepotation and QET. By transmitting local quantum information through quantum teleportation and performing conditional operations on that information, QET can theoretically be realized independent of geographical distance. The method we present will provide new insights into new applications of future large‐scale quantum networks and potential applications of quantum physics to information engineering.

The role of occupancy on traffic flow in a multiple-loop network

A dynamic model for traffic flow is proposed to analyze the impact of occupancy in a multiple-loop network with a single intersection. The graph representation of multiple-loop lines is obtained utilizing the cell-transmission model, and consequently, the density equations are derived. The macroscopic fundamental diagrams are investigated for different cases of occupancy and the fraction of vehicles. For the double-loop lines, equilibrium densities, traffic currents, and travel time are obtained theoretically and validated via simulation. It is observed that the equilibrium densities increase linearly against the mean density for an equal fraction of vehicles while the traffic currents remain symmetric, unlike the case of an unequal fraction of vehicles. The travel time varies inversely with respect to the occupancy of the loop line. Additionally, when the fraction of vehicles and occupancy at both nodes are different, a critical value of mean density is obtained below (above), in which the travel time at one node is less (more) than the other. To analyze the traffic flow on multiple-loop lines, stability analysis is conducted using the tools from cooperative theory and repelling boundaries. As a result, the vehicular densities converge to a unique equilibrium point irrespective of the initial densities in the multiple-loop network. The theoretical results agree well with the simulation results.

Scheduling calculations for a multiprocessor system in real time

The problem of scheduling computations in a multiprocessor system is considered for the case when, at some time instants, requests for the execution of job packages with known characteristics are received. Interrupts and switching from one processor to another are allowed. In the first formulation, the composition of all complexes and the characteristics of tasks are known in advance. In the second setting, this information becomes known only at the time of each request. It is required to determine whether there is an admissible schedule for the total set of jobs and build it in case of a positive answer. A setting is studied in which, in addition to processors, there is a non-renewable resource. A polynomial algorithm for solving the problem is developed, based on the construction of a network flow model and the search for the maximum flow.

Calculation of probabilistic harmonic power flow based on improved three-point estimation method and maximum entropy as distributed generators access to distribution network

Uncertainty and volatility are the features of distributed generation (DG), which can impact greatly on the operational status of distribution network. Thus, an improved three-point estimation method and maximum entropy theory (ITPEM&ME) for calculating the probabilistic harmonic power flow in distribution network is proposed in this study. Firstly, the probabilistic model is constructed according to the uncertainty of input variables and the improved three-point estimate method is utilized to convert a probabilistic problem into a deterministic problem. Secondly, in order to obtain the distribution of the random harmonics in distribution network, the maximum entropy theory is adopted to reconstruct the statistical characteristics after the deterministic calculation is realized for each harmonic voltages. The case of an IEEE-33 system under various conditions is conducted. The results by the different algorithms indicate that the proposed method ITPEM&ME has a smaller error than the traditional method, as well as the fitting effect of the probability density function (PDF) is much closer to the practical engineering. It is feasible and accurate in dealing with probabilistic harmonic power flow problems.

Time‐based DDoS attack detection through hybrid LSTM‐CNN model architectures: An investigation of many‐to‐one and many‐to‐many approaches

SummaryInternet data thefts, intrusions and DDoS attacks are some of the big concerns for the network security today. Detection of these anomalies, is gaining tremendous impetus with the development of machine learning and artificial intelligence. Even now researchers are shifting the base from machine learning to the deep neural architectures with auto‐feature selection capabilities. We in this paper propose multiple deep neural network architectures which can select, co‐learn and teach the gradients of the neural network by itself with no human intervention. This is what we call as meta‐learning. The models are configured in both many to one and many to many design architectures. We combine long short‐term memory (LSTM), bi‐directional long short‐term memory (BiLSTM), convolutional neural network (CNN) layers along with attention mechanism to achieve the higher accuracy values among all the available deep learning model architectures. LSTMs overcomes the vanishing and exploding gradient problem of RNN and attention mechanism mimics the human cognitive attention that screens the network flow to obtain the key features for network traffic classification. In addition, we also add multiple convolutional layers to get the key features for network traffic classification. We get the time series analysis of the traffic done for the possibility of a DDoS attack without using any feature selection techniques and without balancing the dataset. The performance analysis is done based on confusion matrix scores, that is, accuracy, false alarm rate (FAR), sensitivity, specificity, false‐positive rate (FPR), F1 score, area under curve (AUC) analysis and loss functions on well‐known public benchmark KDD Cup'99 data set. The results of our experiments reveal that our models outperform existing techniques, showing their superiority in performance.

Simulation approach with MC-reduction for multi-state flow network reliability estimation

Simulation approach with MC-reduction for multi-state flow network reliability estimation

Enhancing the Thermal Performance of Shape Memory Polymers: Designing a Minichannel Structure.

This research proposes a numerical approach to improve the thermal performance of shape memory polymers (SMPs) while their mechanical properties remain intact. Sixteen different 3D minichannel structures were numerically designed to investigate the impact of embedded water flow in microchannel networks on the thermal response and shape recovery of SMPs. This work employs two approaches, each with different physics: approach A focuses on solid mechanics analysis and, accordingly, thermal analysis in solids without considering the fluid. approach B tackles solid and fluid mechanics analysis and thermal analysis in both solid and fluid subdomains, which inherently calls for fluid-structure coupling in a uniform procedure. Finally, the results of these two approaches are compared to predict the SMP's thermal and mechanical behavior. The structural designs are then analyzed in terms of their shape recovery speed, recovery ratio, and recovery parameters. The results indicate that isotropic structures thermally outperform their anisotropic counterparts, exhibiting improved thermal characteristics and faster shape recovery. Additionally, it was observed that polymeric structures with a low volume fraction of embedded branches thermally perform efficiently. The findings of this study predict that the geometrical angle between the main branch and sub-branches of SMP favorably impacts the enhancement of thermal characteristics of the structure, accelerating its shape recovery. Approach B accelerates the shape recovery rate in SMPs due to fluid flow and uniform heat transfer within the structures.

Optimizing fire emergency evacuation routes in underground coal mines: A lightweight network flow approach

In underground mine fires, the presence of smoke, toxic gasses, and high heat can significantly hinder the evacuees from identifying the optimal path to safety. This study presents a framework that couples a mine fire simulator software with a Ford-Fulkerson algorithm to model fire evacuations as a minimum-cost flow problem. The fire-induced risks are quantified based on the MSHA safety standards and a user-friendly FFA computes the evacuation routes. By accumulating the quantified effect of the risk exposure and updating the network depending on the mine conditions, different safe evacuation routes are identified. The algorithm is demonstrated through fire simulation data acquired from a model of the VentSim™ DESIGN software. Airflow quantity, air quality, heat, carbon monoxide concentration, and visibility obstruction are acquired from the simulations and processed through the proposed algorithm. Based on the distribution of the quantified hazards and the MSHA safety standards, the algorithm outputs optimal evacuation paths. The computed evacuation routes minimize the exposure to the fire-induced hazards while at the same time prioritize the shortest routes. The presented framework can be used for evaluating ventilation designs and emergency plans as well as in real-time self-evacuation in mine emergencies.

Potential Improvement of GK2A Clear-Sky Atmospheric Motion Vectors Using the Convolutional Neural Network Model

In this study, we propose a new approach to improve the accuracy of the horizontal atmospheric motion vector (AMV) in cloud-free skies and its forecasting. We adapted the optical flow of the convolutional neural network (CNN) framework model using two 10-min interval infrared images at water vapor channels (centered at 6.3, 7.0, and 7.3 μm\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\mu m$$\\end{document}) from the Korean geostationary satellite GEO-KOMPSAT-2A (GK2A). Since all pixels had seamless AMVs calculated by CNN (CNN AMVs), we could also predict AMVs using the linear regression method. The tracking performance of the CNN-based algorithm was validated using AMVs retrieved from GK2A (GK2A AMVs) by estimating the difference between those values and the ECMWF (European Centre for Medium-Range Weather Forecasts) Reanalysis v5 (ERA5) wind data over Korea in 2022. CNN AMVs showed similar or better root-mean-square vector differences (RMSVDs) than GK2A AMVs (12.33–12.86 vs. 15.89–19.96 m/s). The RMSVDs of the forecasted AMVs were 2.74, 2.95, 3.41, and 4.79 m/s at lead times of 10, 20, 30, and 60 min, respectively. Consequently, our method showed higher accuracy for tracking motion in the production of AMVs and succeeded in forecasting AMVs. We expect that such potential improvements in computational accuracy for operational GK2A AMVs will contribute to increased accuracy when forecasting meteorological phenomena related to wind.

Research on the joint adjustment model of regional water resource network based on the network flow theory

Abstract This study uses the network flow theory to optimize regional water resource allocation. In order to solve the problem of inefficient utilization of water resources with decentralized decision-making by different administrative units, a regional water resource networking and joint dispatching model with multi-objective nonlinear characteristics based on the network flow theory (hereinafter referred to as the network flow model) is constructed in the study. The network flow model was simulated and applied in the Xin-Sheng area of the Cao'e River, a tributary of the Qiantang River, and the results of the study showed that the network flow model scheduling increased significantly in efficiency compared with the current conventional scheduling, with an increase of 35.24 and 9.91% in the water resource utilization rate in the two typical years of 2019 and 2022, respectively, and showed that 2022, which has less rainfall, has a better effect than 2019. The study concludes that the network flow model can effectively improve the efficiency of water resource utilization, solve the problem of water resource imbalance between cities in the region, and play a positive role in the construction of the national water network.

Trade flows of Central Asian states within the framework of the Shanghai Cooperation Organization

Aim. To form a systematic understanding of the structure of trade of Central Asian states within the framework of the Shanghai Cooperation Organization (SCO).Objectives. To analyze the commodity structure of trade flows and their geographical distribu- tion; to identify key characteristics of trade flow networks of Central Asian states within the SCO.Methods. Statistical analysis of the geographical, commodity, sectoral structure of exports and imports of goods, as well as the degree of their technological capacity.Results. It can be stated that there are practically no high-tech groups of goods among the significant commodity groups in the trade of Central Asian SCO member states with their part- ners in this organization. Moreover, to a large extent the growth of trade turnover of the countries is conditioned by the intensification of supplies of various raw materials to China (oil, ore, textile and agrarian raw materials). The pattern of “raw materials for raw materials” trade exchange often persists, especially between the Central Asian states proper; moreover, these countries are to a large extent competitors, as the commodity structure of their export flows largely coincides (in particular, in terms of raw materials and agrarian products). Conclusions. The current situation hampers the development of industrial cooperation between the SCO countries. The fact that the “One Belt - One Road” initiative is based on the established trade specializations of the countries may, on the one hand, “revive” trade relations in the region, on the other hand, consolidate the role of Central Asian states as suppliers of raw ma- terials for China.