Stojan Trajanovski

Robustness envelopes of networks

We study the robustness of networks under node removal, considering random node failure, as well as targeted node attacks based on network centrality measures. Whilst both of these have been studied in the literature, existing approaches tend to study random failure in terms of average-case behavior, giving no idea of how badly network performance can degrade purely by chance. Instead of considering average network performance under random failure, we compute approximate network performance probability density functions as functions of the fraction of nodes removed. We find that targeted attacks based on centrality measures give a good indication of the worst-case behavior of a network. We show that many centrality measures produce similar targeted attacks and that a combination of degree centrality and eigenvector centrality may be enough to evaluate worst-case behavior of networks. Finally, we study the robustness envelope and targeted attack responses of networks that are rewired to have high- and low-degree assortativities, discovering that moderate assortativity increases confer more robustness against targeted attacks whilst moderate decreases confer more robustness against random uniform attacks.

Finding critical regions and region-disjoint paths in a network

Due to their importance to society, communication networks should be built and operated to withstand failures. However, cost considerations make network providers less inclined to take robustness measures against failures that are unlikely to manifest, like several failures coinciding simultaneously in different geographic regions of their network. Considering networks embedded in a two-dimensional plane, we study the problem of finding a critical region-a part of the network that can be enclosed by a given elementary figure of predetermined size-whose destruction would lead to the highest network disruption. We determine that only a polynomial, in the input, number of nontrivial positions for such a figure needs to be considered and propose a corresponding polynomial-time algorithm. In addition, we consider region-aware network augmentation to decrease the impact of a regional failure. We subsequently address the region-disjoint paths problem, which asks for two paths with minimum total weight between a source (s) and a destination (d) that cannot both be cut by a single regional failure of diameter D (unless that failure includes s or d). We prove that deciding whether region-disjoint paths exist is NP-hard and propose a heuristic region-disjoint paths algorithm.

Decentralized Protection Strategies Against SIS Epidemics in Networks

Defining an optimal protection strategy against viruses, spam propagation, or any other kind of contamination process is an important feature for designing new networks and architectures. In this paper, we consider decentralized optimal protection strategies when a virus is propagating over a network through an SIS epidemic process. We assume that each node in the network can fully protect itself from infection at a constant cost, or the node can use recovery software, once it is infected. We model our system using a game-theoretic framework and find pure, mixed equilibria, and the Price of Anarchy in several network topologies. Further, we propose a decentralized algorithm and an iterative procedure to compute a pure equilibrium in the general case of a multiple communities network. Finally, we evaluate the algorithms and give numerical illustrations of all our results.

Complete game-theoretic characterization of SIS epidemics protection strategies

Defining an optimal protection strategy against viruses, spam propagation or any other kind of contamination process is an important feature for designing new networks and architectures. In this work, we consider decentralized optimal protection strategies when a virus is propagating over a network through a Susceptible Infected Susceptible (SIS) epidemic process. We assume that each node in the network can fully protect itself from infection at a constant cost, or the node can use recovery software, once it is infected. We model our system using a game theoretic framework. Based on this model, we find pure and mixed equilibria, and evaluate the performance of the equilibria by finding the Price of Anarchy (PoA) in several network topologies. Finally, we give numerical illustrations of our results.

Finding critical regions in a network

It is important that our vital networks (e.g., infrastructures) are robust to more than single-link failures. Failures might for instance affect a part of the network that resides in a certain geographical region. In this paper, considering networks embedded in a two-dimensional plane, we study the problem of finding a critical region - that is, a part of the network that can be enclosed by a given elementary figure (a circle, ellipse, rectangle, square, or equilateral triangle) with a predetermined size - whose removal would lead to the highest network disruption. We determine that there is a polynomial number of non-trivial positions for such a figure that need to be considered and, subsequently, we propose a polynomial-time algorithm for the problem. Simulations on realistic networks illustrate that different figures with equal area result in different critical regions in a network.

Availability-based path selection and network vulnerability assessment

In data-communication networks, network reliability is of great concern to both network operators and customers. On the one hand, the customers care about receiving reliable services and, on the other hand, for the network operators it is vital to determine the most vulnerable parts of their network. In this article, we first study the problem of establishing a connection over at most k partially link-disjoint paths and for which the total availability is no less than i¾?0<i¾?i¾?1. We analyze the complexity of this problem in generic networks, shared-risk link group networks and multilayer networks. We subsequently propose a polynomial-time heuristic algorithm and an exact integer nonlinear program for availability-based path selection. The proposed algorithms are evaluated in terms of acceptance ratio and running time. Subsequently, in the three aforementioned types of networks, we study the problem of finding a set of network cuts for which the failure probability of its links is largest.

Detection of spatially-close fiber segments in optical networks

Spatially-close network fibers have a significant chance of failing simultaneously in the event of man-made or natural disasters within their geographic area. Network operators are interested in the proper detection and grouping of any existing spatially-close fiber segments, to avoid service disruptions due to simultaneous fiber failures. Moreover, spatially-close fibers can further be differentiated by computing the intervals over which they are spatially close. In this paper, we propose (1) polynomial-time algorithms for detecting all the spatially-close fiber segments of different fibers, (2) a polynomial-time algorithm for finding the spatially-close intervals of a fiber to a set of other fibers, and (3) a fast exact algorithm for grouping spatially-close fibers using the minimum number of distinct risk groups. All of our algorithms have a fast running time when simulated on three real-world network topologies.

ILIGRA: An Efficient Inverse Line Graph Algorithm

This paper presents a new and efficient algorithm, IligraLIGRA, for inverse line graph construction. Given a line graph H, ILIGRA constructs its root graph G with the time complexity being linear in the number of nodes in H. If ILIGRA does not know whether the given graph H is a line graph, it firstly assumes that H is a line graph and starts its root graph construction. During the root graph construction, ILIGRA checks whether the given graph H is a line graph and ILIGRA stops once it finds H is not a line graph. The time complexity of ILIGRA with line graph checking is linear in the number of links in the given graph H. For sparse line graphs of any size and for dense line graphs of small size, numerical results of the running time show that ILIGRA outperforms all currently available algorithms.

Generating graphs that approach a prescribed modularity

Modularity is a quantitative measure for characterizing the existence of a community structure in a network. A networks modularity depends on the chosen partitioning of the network into communities, which makes finding the specific partition that leads to the maximum modularity a hard problem. In this paper, we prove that deciding whether a graph with a given number of links, number of communities, and modularity exists is NP-complete and subsequently propose a heuristic algorithm for generating graphs with a given modularity. Our graph generator allows constructing graphs with a given number of links and different topological properties. The generator can be used in the broad field of modeling and analyzing clustered social or organizational networks.

Optimization problems in correlated networks

AbstractBackgroundSolving the shortest path and min-cut problems are key in achieving high-performance and robust communication networks. Those problems have often been studied in deterministic and uncorrelated networks both in their original formulations as well as in several constrained variants. However, in real-world networks, link weights (e.g., delay, bandwidth, failure probability) are often correlated due to spatial or temporal reasons, and these correlated link weights together behave in a different manner and are not always additive, as commonly assumed.MethodsIn this paper, we first propose two correlated link weight models, namely (1) the deterministic correlated model and (2) the (log-concave) stochastic correlated model. Subsequently, we study the shortest path problem and the min-cut problem under these two correlated models.Results and ConclusionsWe prove that these two problems are NP-hard under the deterministic correlated model, and even cannot be approximated to arbitrary degree in polynomial time. However, these two problems are solvable in polynomial time under the (constrained) nodal deterministic correlated model, and can be solved by convex optimization under the (log-concave) stochastic correlated model.