arXiv subjects

Cunlai Pu

Publications and source records attributed to Cunlai Pu.

At least 19 recordsLinked to original sources

arXiv2026

IGA-LWP: An Iterative Gradient-based Adversarial Attack for Link Weight Prediction

Link weight prediction extends classical link prediction by estimating the strength of interactions rather than merely their existence, and it underpins a wide range of applications such as traffic engineering, social recommendation, and scientific collaboration analysis. However, the robustness of link weight prediction against adversarial perturbations remains largely unexplored.In this paper, we formalize the link weight prediction attack problem as an optimization task that aims to maximize the prediction error on a set of target links by adversarially manipulating the weight values of a limited number of links. Based on this formulation, we propose an iterative gradient-based attack framework for link weight prediction, termed IGA-LWP. By employing a self-attention-enhanced graph autoencoder as a surrogate predictor, IGA-LWP leverages backpropagated gradients to iteratively identify and perturb a small subset of links. Extensive experiments on four real-world weighted networks demonstrate that IGA-LWP significantly degrades prediction accuracy on target links compared with baseline methods. Moreover, the adversarial networks generated by IGA-LWP exhibit strong transferability across several representative link weight prediction models. These findings expose a fundamental vulnerability in weighted network inference and highlight the need for developing robust link weight prediction methods.

cs.SI↗

arXiv2026

CLF-ULP: Cross-Layer Fusion-Based Link Prediction in Dynamic Multiplex UAV Networks

In complex Unmanned Aerial Vehicle (UAV) networks, UAVs can establish dynamic and heterogeneous links with one another for various purposes, such as communication coverage, collective sensing, and task collaboration. These interactions give rise to dynamic multiplex UAV networks, where each layer represents a distinct type of interaction among UAVs. Understanding how such links form and evolve is both of theoretical interest and of practical importance for the control and maintenance of networked UAV systems. In this paper, we first develop a dynamic multiplex network model for UAV networks to characterize their dynamic and heterogeneous link properties. We then propose a cross-layer fusion-based deep learning model, termed CLF-ULP, to predict future inter-UAV links based on historical topology data. CLF-ULP incorporates graph attention networks to extract topological features within each layer and perform a cross-layer attention fusion to capture inter-layer dependencies. Furthermore, a shared-parameter long short-term memory network is employed to model the temporal evolution of each layer. To improve embedding quality and link prediction performance, we develop a joint loss function that considers both intra-layer and inter-layer UAV adjacency. Extensive experiments on simulated UAV datasets under diverse mobility patterns demonstrate that CLF-ULP achieves state-of-the-art performance in predicting links within dynamic multiplex UAV networks.

cs.NI↗

arXiv2025

MUST: Multi-Scale Structural-Temporal Link Prediction Model for UAV Ad Hoc Networks

Link prediction in unmanned aerial vehicle (UAV) ad hoc networks (UANETs) aims to predict the potential formation of future links between UAVs. In adversarial environments where the route information of UAVs is unavailable, predicting future links must rely solely on the observed historical topological information of UANETs. However, the highly dynamic and sparse nature of UANET topologies presents substantial challenges in effectively capturing meaningful structural and temporal patterns for accurate link prediction. Most existing link prediction methods focus on temporal dynamics at a single structural scale while neglecting the effects of sparsity, resulting in insufficient information capture and limited applicability to UANETs. In this paper, we propose a multi-scale structural-temporal link prediction model (MUST) for UANETs. Specifically, we first employ graph attention networks (GATs) to capture structural features at multiple levels, including the individual UAV level, the UAV community level, and the overall network level. Then, we use long short-term memory (LSTM) networks to learn the temporal dynamics of these multi-scale structural features. Additionally, we address the impact of sparsity by introducing a sophisticated loss function during model optimization. We validate the performance of MUST using several UANET datasets generated through simulations. Extensive experimental results demonstrate that MUST achieves state-of-the-art link prediction performance in highly dynamic and sparse UANETs.

cs.LG↗

arXiv2024

Line Graph Neural Networks for Link Weight Prediction

In real-world networks, predicting the weight (strength) of links is as crucial as predicting the existence of the links themselves. Previous studies have primarily used shallow graph features for link weight prediction, limiting the prediction performance. In this paper, we propose a new link weight prediction method, namely Line Graph Neural Networks for Link Weight Prediction (LGLWP), which learns deeper graph features through deep learning. In our method, we first extract the enclosing subgraph around a target link and then employ a weighted graph labeling algorithm to label the subgraph nodes. Next, we transform the subgraph into the line graph and apply graph convolutional neural networks to learn the node embeddings in the line graph, which can represent the links in the original subgraph. Finally, the node embeddings are fed into a fully-connected neural network to predict the weight of the target link, treated as a regression problem. Our method directly learns link features, surpassing previous methods that splice node features for link weight prediction. Experimental results on six network datasets of various sizes and types demonstrate that our method outperforms state-of-the-art methods.

cs.SI↗

arXiv2023

Epidemic spreading in wireless sensor networks with node sleep scheduling

Wireless Sensor Networks (WSNs) have become widely used in various fields like environmental monitoring, smart agriculture, and health care. However, their extensive usage also introduces significant vulnerabilities to cyber viruses. Addressing this security issue in WSNs is very challenging due to their inherent limitations in energy and bandwidth to implement real-time security measures. To tackle the virus issue, it is crucial to first understand how it spreads in WSNs. In this brief, we propose a novel epidemic spreading model for WSNs, integrating the susceptible-infected-susceptible (SIS) epidemic spreading model and node probabilistic sleep scheduling--a critical mechanism for optimizing energy efficiency. Using the microscopic Markov chain (MMC) method, we derive the spreading equations and epidemic threshold of our model. We conduct numerical simulations to validate the theoretical results and investigate the impact of key factors on epidemic spreading in WSNs. Notably, we discover that the epidemic threshold is directly proportional to the ratio of node sleeping and node activation probabilities.

cs.NI↗

arXiv2021

NC-MOPSO: Network centrality guided multi-objective particle swarm optimization for transport optimization on networks

Transport processes are universal in real-world complex networks, such as communication and transportation networks. As the increase of the traffic in these complex networks, problems like traffic congestion and transport delay are becoming more and more serious, which call for a systematic optimization of these networks. In this paper, we formulate a multi-objective optimization problem (MOP) to deal with the enhancement of network capacity and efficiency simultaneously, by appropriately adjusting the weights of edges in networks. To solve this problem, we provide a multi-objective evolutionary algorithm (MOEA) based on particle swarm optimization (PSO), namely network centrality guided multi-objective PSO (NC-MOPSO). Specifically, in the framework of PSO, we propose a hybrid population initialization mechanism and a local search strategy by employing the network centrality theory to enhance the quality of initial solutions and strengthen the exploration of the search space, respectively. Simulation experiments performed on network models and real networks show that our algorithm has better performance than four state-of-the-art alternatives on several most-used metrics.

cs.NI↗

arXiv2020

Cost Restrained Hybrid Attacks in Power Grids

The frequent occurrences of cascading failures in power grids have been receiving continuous attention in recent years. An urgent task for us is to understand the cascading failure vulnerability of power grids against various kinds of attacks. We consider a cost restrained hybrid attack problem in power grids, in which both nodes and links are targeted with a limited total attack cost. We propose an attack centrality metric for a component (node or link) based on the consequence and cost of the removal of the component. Depending on the width of cascading failures considered, the attack centrality can be a local or global attack centrality. With the attack centrality, we further provide a greedy hybrid attack, and an optimal hybrid attack with the Particle Swarm Optimization (PSO) framework. Simulation results on IEEE bus test data show that the optimal hybrid attack is more efficient than the greedy hybrid attack. Furthermore, we find counterintuitively that the local centrality based algorithms are better than the global centrality based ones when the cost constraint is considered in the attack problem.

eess.SY↗

arXiv2019

Node similarity distribution of complex networks and its application in link prediction

Over the years, quantifying the similarity of nodes has been a hot topic in complex networks, yet little has been known about the distributions of node-similarity. In this paper, we consider a typical measure of node-similarity called the common neighbor based similarity (CNS). By means of the generating function, we propose a general framework for calculating the CNS distributions of node sets in various complex networks. In particular, we show that for the Erdös-Rényi (ER) random network, the CNS distribution of node sets of any particular size obeys the Poisson law. We also connect the node-similarity distribution to the link prediction problem. We found that the performance of link prediction depends solely on the CNS distributions of the connected and unconnected node pairs in the network. Furthermore, we derive theoretical solutions of two key evaluation metrics in link prediction: i) precision and ii) area under the receiver operating characteristic curve (AUC). We show that for any link prediction method, if the similarity distributions of the connected and unconnected node pairs are identical, the AUC will be $0.5$. The theoretical solutions are elegant alternatives of the traditional experimental evaluation methods with nevertheless much lower computational cost.

cs.SI↗

arXiv2019

Vulnerability Assessment of Power Grids Based on Both Topological and Electrical Properties

In modern power grids, a local failure or attack can trigger catastrophic cascading failures, which make it challenging to assess the attack vulnerability of power grids. In this Brief, we define the $K$-link attack problem and study the attack vulnerability of power grids under cascading failures. Particularly, we propose a link centrality measure based on both topological and electrical properties of power grids. According to this centrality, we propose a greedy attack algorithm and an optimal attack algorithm. Simulation results on standard IEEE bus test data show that the optimal attack is better than the greedy attack and the traditional PSO-based attack in fracturing power grids. Moreover, the greedy attack has smaller computational complexity than the optimal attack and the PSO-based attack with an adequate attack efficiency. Our work helps to understand the vulnerability of power grids and provides some clues for securing power grids.

eess.SY↗

arXiv2018

Robustness of link prediction under network attacks

While link prediction in networks has been a hot topic over the years, its robustness has not been well discussed in literature. In this paper, we study the robustness of some mainstream link prediction methods under various kinds of network attack strategies, including the random attack (RDA), centrality based attacks (CA), similarity based attacks (SA), and simulated annealing based attack (SAA). Through the variation of precision, a typical evaluation index of link prediction, we find that for the SA and SAA, a small fraction of link removals can significantly reduce the performance of link prediction. In general, the SAA has the highest attack efficiency, followed by the SA and then CA. Interestingly, the performance of some particular CA strategies, such as the betweenness based attacks (BA), are even worse than the RDA. Furthermore, we discover that a link prediction method with high performance probably has lower attack robustness, and the vice versa.

physics.soc-ph↗

arXiv2017

Bufferless transmission in complex networks

Complex bufferless networks such as on-chip networks and optical burst switching networks haven't been paid enough attention in network science. In complex bufferless networks, the store and forward mechanism is not applicable, since the network nodes are not allowed to buffer data packets. In this paper, we study the data transmission process in complex bufferless networks from the perspective of network science. Specifically, we use the Price model to generate the underlying network topological structures. We propose a delivery queue based deflection mechanism, which accompanies the efficient routing protocol, to transmit data packets in bufferless networks. We investigate the average deflection times, packets loss rate, average arrival time, and how the network topological structure and some other factors affect these transmission performances. Our work provides some clues for the architecture and routing design of bufferless networks.

cs.NI↗

arXiv2017

Information transmission on hybrid networks

Many real-world communication networks often have hybrid nature with both fixed nodes and moving modes, such as the mobile phone networks mainly composed of fixed base stations and mobile phones. In this paper, we discuss the information transmission process on the hybrid networks with both fixed and mobile nodes. The fixed nodes (base stations) are connected as a spatial lattice on the plane forming the information-carrying backbone, while the mobile nodes (users), which are the sources and destinations of information packets, connect to their current nearest fixed nodes respectively to deliver and receive information packets. We observe the phase transition of traffic load in the hybrid network when the packet generation rate goes from below and then above a critical value, which measures the network capacity of packets delivery. We obtain the optimal speed of moving nodes leading to the maximum network capacity. We further improve the network capacity by rewiring the fixed nodes and by considering the current load of fixed nodes during packets transmission. Our purpose is to optimize the network capacity of hybrid networks from the perspective of network science, and provide some insights for the construction of future communication infrastructures.

cs.NI↗

arXiv2016

Multiple predator based capture process on complex networks

The predator/prey (capture) problem is a prototype of many network-related applications. We study the capture process on complex networks by considering multiple predators from multiple sources. In our model, some lions start from multiple sources simultaneously to capture the lamb by biased random walks, which are controlled with a free parameter $α$. We derive the distribution of the lamb's lifetime and the expected lifetime $\left\langle T\right\rangle $. Through simulation, we find that the expected lifetime drops substantially with the increasing number of lions. We also study how the underlying topological structure affects the capture process, and obtain that locating on small-degree nodes is better than large-degree nodes to prolong the lifetime of the lamb. Moreover, dense or homogeneous network structures are against the survival of the lamb.

physics.soc-ph↗

arXiv2016

Entropy-based link prediction in weighted networks

Information entropy has been proved to be an effective tool to quantify the structural importance of complex networks. In the previous work (Xu et al, 2016 \cite{xu2016}), we measure the contribution of a path in link prediction with information entropy. In this paper, we further quantify the contribution of a path with both path entropy and path weight, and propose a weighted prediction index based on the contributions of paths, namely Weighted Path Entropy (WPE), to improve the prediction accuracy in weighted networks. Empirical experiments on six weighted real-world networks show that WPE achieves higher prediction accuracy than three typical weighted indices.

cs.SI↗

arXiv2016

Traffic dynamics on dynamical networks: The connection between network lifetime and traffic congestion

For many power-limited networks, such as wireless sensor networks and mobile ad hoc networks, maximizing the network lifetime is the first concern in the related designing and maintaining activities. We study the network lifetime from the perspective of network science. In our dynamic network, nodes are assigned a fixed amount of energy initially and consume the energy in the delivery of packets. We divided the network traffic flow into four states: no, slow, fast, and absolute congestion states. We derive the network lifetime by considering the state of the traffic flow. We find that the network lifetime is generally opposite to traffic congestion in that the more congested traffic, the less network lifetime. We also find the impacts of factors such as packet generation rate, communication radius, node moving speed, etc., on network lifetime and traffic congestion.

cs.NI↗

arXiv2015

Link prediction based on path entropy

Information theory has been taken as a prospective tool for quantifying the complexity of complex networks. In this paper, we first study the information entropy or uncertainty of a path using the information theory. Then we apply the path entropy to the link prediction problem in real-world networks. Specifically, we propose a new similarity index, namely Path Entropy (PE) index, which considers the information entropies of shortest paths between node pairs with penalization to long paths. Empirical experiments demonstrate that PE index outperforms the mainstream link predictors.

physics.soc-ph↗

arXiv2015

Information transport in multiplex networks

In this paper, we study information transport in multiplex networks comprised of two coupled subnetworks. The upper subnetwork, called the logical layer, employs the shortest paths protocol to determine the logical paths for packets transmission, while the lower subnetwork acts as the physical layer, in which packets are delivered by the biased random walk mechanism characterized with a parameter $α$. Through simulation, we obtain the optimal $α$ corresponding to the maximum network lifetime and the maximum number of the arrival packets. Assortative coupling is better than the random coupling and the disassortative coupling, since it achieves much better transmission performances. Generally, the more homogeneous the lower subnetwork, the better the transmission performances are, which is opposite for the upper subnetwork. Finally, we propose an attack centrality for nodes based on the topological information of both subnetworks, and further investigate the transmission performances under targeted attacks. Our work helps to understand the spreading and robustness issues of multiplex networks and provides some clues about the designing of more efficient and robust routing architectures in communication systems.

physics.soc-ph↗

arXiv2015

Bounded link prediction for very large networks

Evaluation of link prediction methods is a hard task in very large complex networks because of the inhibitive computational cost. By setting a lower bound of the number of common neighbors (CN), we propose a new framework to efficiently and precisely evaluate the performances of CN-based similarity indices in link prediction for very large heterogeneous networks. Specifically, we propose a fast algorithm based on the parallel computing scheme to obtain all the node pairs with CN values larger than the lower bound. Furthermore, we propose a new measurement, called self-predictability, to quantify the performance of the CN-based similarity indices in link prediction, which on the other side can indicate the link predictability of a network.

physics.soc-ph↗