Biblio

Swarm learning (SL) is an emerging promising decentralized machine learning paradigm and has achieved high performance in clinical applications. SL solves the problem of a central structure in federated learning by combining edge computing and blockchain-based peer-to-peer network. While there are promising results in the assumption of the independent and identically distributed (IID) data across participants, SL suffers from performance degradation as the degree of the non-IID data increases. To address this problem, we propose a generative augmentation framework in swarm learning called SL-GAN, which augments the non-IID data by generating the synthetic data from participants. SL-GAN trains generators and discriminators locally, and periodically aggregation via a randomly elected coordinator in SL network. Under the standard assumptions, we theoretically prove the convergence of SL-GAN using stochastic approximations. Experimental results demonstrate that SL-GAN outperforms state-of-art methods on three real world clinical datasets including Tuberculosis, Leukemia, COVID-19.

Common Vulnerability Scoring System (CVSS) is intended to capture the key characteristics of a vulnerability and correspondingly produce a numerical score to indicate the severity. Important efforts are conducted for building a CVSS stochastic model in order to provide a high-level risk assessment to better support cybersecurity decision-making. However, these efforts consider nothing regarding HPC (High-Performance Computing) networks using a Science Demilitary Zone (DMZ) architecture that has special design principles to facilitate data transition, analysis, and store through in a broadband backbone. In this paper, an HPCvul (CVSS-based vulnerability and risk assessment) approach is proposed for HPC networks in order to provide an understanding of the ongoing awareness of the HPC security situation under a dynamic cybersecurity environment. For such a purpose, HPCvul advocates the standardization of the collected security-related data from the network to achieve data portability. HPCvul adopts an attack graph to model the likelihood of successful exploitation of a vulnerability. It is able to merge multiple attack graphs from different HPC subnets to yield a full picture of a large HPC network. Substantial results are presented in this work to demonstrate HPCvul design and its performance.

In this paper, we consider a discrete-time stochastic Stackelberg game where there is a defender (also called leader) who has to defend a target and an attacker (also called follower). The attacker has a private type that evolves as a controlled Markov process. The objective is to compute the stochastic Stackelberg equilibrium of the game where defender commits to a strategy. The attacker’s strategy is the best response to the defender strategy and defender’s strategy is optimum given the attacker plays the best response. In general, computing such equilibrium involves solving a fixed-point equation for the whole game. In this paper, we present an algorithm that computes such strategies by solving lower dimensional fixed-point equations for each time t. Based on this algorithm, we compute the Stackelberg equilibrium of a security example.

The use of Virtual Machine (VM) migration as support for software rejuvenation was introduced more than a decade ago. Since then, several works have validated this approach from experimental and theoretical perspectives. Recently, some works shed light on the possibility of using the same technique as Moving Target Defense (MTD). However, to date, no work evaluated the availability and security levels while applying VM migration for both rejuvenation and MTD (multipurpose VM migration). In this paper, we conduct a comprehensive evaluation using Stochastic Petri Net (SPN) models to tackle this challenge. The evaluation covers the steady-state system availability, expected MTD protection, and related metrics of a system under time-based multipurpose VM migration. Results show that the availability and security improvement due to VM migration deployment surpasses 50% in the best scenarios. However, there is a trade-off between availability and security metrics, meaning that improving one implies compromising the other.

Smart grid is a new generation of grid that inte-grates traditional grid and grid information system, and infor-mation security of smart grid is an extremely important part of the whole grid. The research of trusted computing technology provides new ideas to protect the information security of the power grid. To address the problem of large deviations in the calculation of credible dynamic thresholds due to the existence of characteristics such as self-similarity and traffic bursts in smart grid information collection, a traffic prediction model based on ARMA and Poisson process is proposed. And the Hurst coefficient is determined more accurately using R/S analysis, which finally improves the efficiency and accuracy of the trusted dynamic threshold calculation.

Security evaluation can be performed using a variety of analysis methods, such as attack trees, attack graphs, threat propagation models, stochastic Petri nets, and so on. These methods analyze the effect of attacks on the system, and estimate security attributes from different perspectives. However, they require information from experts in the application domain for properly capturing the key elements of an attack scenario: i) the attack paths a system could be subject to, and ii) the different characteristics of the possible adversaries. For this reason, some recent works focused on the generation of low-level security models from a high-level description of the system, hiding the technical details from the modeler.In this paper we build on an existing ontology framework for security analysis, available in the ADVISE Meta tool, and we extend it in two directions: i) to cover the attack patterns available in the CAPEC database, a comprehensive dictionary of known patterns of attack, and ii) to capture all the adversaries’ profiles as defined in the Threat Agent Library (TAL), a reference library for defining the characteristics of external and internal threat agents ranging from industrial spies to untrained employees. The proposed extension supports a richer combination of adversaries’ profiles and attack paths, and provides guidance on how to further enrich the ontology based on taxonomies of attacks and adversaries.

When it comes to cryptographic random number generation, poor understanding of the security requirements and “mythical aura” of black-box statistical testing frequently leads it to be used as a substitute for cryptanalysis. To make things worse, a seemingly standard document, NIST SP 800–22, describes 15 statistical tests and suggests that they can be used to evaluate random and pseudorandom number generators in cryptographic applications. The Chi-nese standard GM/T 0005–2012 describes similar tests. These documents have not aged well. The weakest pseudorandom number generators will easily pass these tests, promoting false confidence in insecure systems. We strongly suggest that SP 800–22 be withdrawn by NIST; we consider it to be not just irrelevant but actively harmful. We illustrate this by discussing the “reference generators” contained in the SP 800–22 document itself. None of these generators are suitable for modern cryptography, yet they pass the tests. For future development, we suggest focusing on stochastic modeling of entropy sources instead of model-free statistical tests. Random bit generators should also be reviewed for potential asymmetric backdoors via trapdoor one-way functions, and for security against quantum computing attacks.

Probabilistic model checking is a useful technique for specifying and verifying properties of stochastic systems including randomized protocols and reinforcement learning models. However, these methods rely on the assumed structure and probabilities of certain system transitions. These assumptions may be incorrect, and may even be violated by an adversary who gains control of some system components. In this paper, we develop a formal framework for adversarial robustness in systems modeled as discrete time Markov chains (DTMCs). We base our framework on existing methods for verifying probabilistic temporal logic properties and extend it to include deterministic, memoryless policies acting in Markov decision processes (MDPs). Our framework includes a flexible approach for specifying structure-preserving and non structure-preserving adversarial models. We outline a class of threat models under which adversaries can perturb system transitions, constrained by an ε ball around the original transition probabilities. We define three main DTMC adversarial robustness problems: adversarial robustness verification, maximal δ synthesis, and worst case attack synthesis. We present two optimization-based solutions to these three problems, leveraging traditional and parametric probabilistic model checking techniques. We then evaluate our solutions on two stochastic protocols and a collection of Grid World case studies, which model an agent acting in an environment described as an MDP. We find that the parametric solution results in fast computation for small parameter spaces. In the case of less restrictive (stronger) adversaries, the number of parameters increases, and directly computing property satisfaction probabilities is more scalable. We demonstrate the usefulness of our definitions and solutions by comparing system outcomes over various properties, threat models, and case studies.

Low-frequency oscillation (LFO) is a security and stability issue that the power system focuses on, measurement data play an important role in online monitoring and analysis of low-frequency oscillation parameters. Aiming at the problem that the measurement data containing noise affects the accuracy of modal parameter identification, a VMD-SSI modal identification algorithm is proposed, which uses the variational modal decomposition algorithm (VMD) for noise reduction combined with the stochastic subspace algorithm for identification. The VMD algorithm decomposes and reconstructs the initial signal with certain noise, and filters out the noise signal. Then, the optimized signal is input into stochastic subspace identification algorithm(SSI), the modal parameters is obtained. Simulation of a three-machine ninenode system verifies that the VMD-SSI mode identification algorithm has good anti-noise performance.

The proliferation of autonomous and connected vehicles on our roads is increasingly felt. However, the problems related to the optimization of the energy consumed, to the safety, and to the security of these do not cease to arise on the tables of debates bringing together the various stakeholders. By focusing on the security aspect of such systems, we can realize that there is a family of problems that must be investigated as soon as possible. In particular, those that may manifest as the system expands. Therefore, this work aims to model and simulate the behavior of a system of autonomous and connected vehicles in the face of a malware invasion. In order to achieve the set objective, we propose a model to our system which is inspired by those used in epidimology, such as SI, SIR, SIER, etc. This being adapted to our case study, stochastic processes are defined in order to characterize its dynamics. After having fixed the values of the various parameters, as well as those of the initial conditions, we run 100 simulations of our system. After which we visualize the results got, we analyze them, and we give some interpretations. We end by outlining the lessons and recommendations drawn from the results.

A new Domain Name System (DNS) cache poisoning attack aiming at clients has emerged recently. It induced cloud users to visit fake web sites and thus reveal information such as account passwords. However, the design of current DNS defense architecture does not formally consider the protection of clients. Although the DNS traffic encryption technology can alleviate this new attack, its deployment is as slow as the new DNS architecture. Thus we propose a lightweight adaptive intelligent defense strategy, which only needs to be deployed on the client without any configuration support of DNS. Firstly, we model the attack and defense process as a static stochastic game with incomplete information under bounded rationality conditions. Secondly, to solve the problem caused by uncertain attack strategies and large quantities of game states, we adopt a deep reinforcement learning (DRL) with guaranteed monotonic improvement. Finally, through the prototype system experiment in Alibaba Cloud, the effectiveness of our method is proved against multiple attack modes with a success rate of 97.5% approximately.

In this paper, we consider a sequential stochastic Stackelberg game with two players, a leader, and a follower. The follower observes the state of the system privately while the leader does not. Players play Stackelberg equilibrium where the follower plays best response to the leader's strategy. In such a scenario, the leader has the advantage of committing to a policy that maximizes its returns given the knowledge that the follower is going to play the best response to its policy. Such a pair of strategies of both the players is defined as Stackelberg equilibrium of the game. Recently, [1] provided a sequential decomposition algorithm to compute the Stackelberg equilibrium for such games which allow for the computation of Markovian equilibrium policies in linear time as opposed to double exponential, as before. In this paper, we extend that idea to the case when the state update dynamics are not known to the players, to propose an reinforcement learning (RL) algorithm based on Expected Sarsa that learns the Stackelberg equilibrium policy by simulating a model of the underlying Markov decision process (MDP). We use particle filters to estimate the belief update for a common agent that computes the optimal policy based on the information which is common to both the players. We present a security game example to illustrate the policy learned by our algorithm.

In this paper, we proposed a path planning algorithm based on Q-learning model to simulate an environment model, which is suitable for the complex environment. A virtual simulation platform has been built to complete the experiments. The experimental results show that the algorithm proposed in this paper can be effectively applied to the solution of vehicle routing problems in the complex environment.

Moving Target Defense (MTD) is a defensive mechanism based on dynamic system reconfiguration to prevent or thwart cyberattacks. In the last years, considerable progress has been made regarding MTD approaches for virtualized environments, and Virtual Machine (VM) migration is the core of most of these approaches. However, VM migration produces system downtime, meaning that each MTD reconfiguration affects system availability. Therefore, a method for a combined evaluation of availability and security is of utmost importance for VM migration-based MTD design. In this paper, we propose a Stochastic Reward Net (SRN) for the probability of attack success and availability evaluation of an MTD based on VM migration scheduling. We study the MTD system under different conditions regarding 1) VM migration scheduling, 2) VM migration failure probability, and 3) attack success rate. Our results highlight the tradeoff between availability and security when applying MTD based on VM migration. The approach and results may provide inputs for designing and evaluating MTD policies based on VM migration.

Underwater Wireless Communication Network (UWCN) is highly emerging in recent times due to the broad variety of underwater applications ranging from disaster prediction, environmental resource monitoring, military security surveillance and assisted navigation. Since the kind of accuracy these applications demands from the dynamic underwater environment is really high, so there is a need for effective way of study underwater communication networks. Usually underwater networks can be studied with the help of actual underwater testbed or with the model of the underwater network. Studying the underwater system with the actual underwater testbed is costly. The effective way of analysis can be done by creating a mathematical model of underwater systems. Queuing theory is one of the most popular mathematical theories used for conventional circuit switched networks whereas it can’t be applied for modeling modern packet switched networks which has high variability compared to that of circuit switched networks. So this paper presents Stochastic Network Calculus (SNC) as the mathematical theory for modeling underwater communication networks. Underlying principles and basic models provided by SNC for analyzing the performance graduates of UWCN is discussed in detail for the benefit of researchers looking for the effective mathematical theory for modeling the system in the domain of underwater communication.

Lateral movement attack performs malicious activities by infecting part of a network system first and then moving laterally to the left system in order to compromise more computers. It is widely used in various sophisticated attacks and plays a critical role. This paper aims to quantitatively analyze the transient security and dependability of a critical network system under lateral movement attacks, whose intruding capability increases with the increasing number of attacked computers. We propose a survivability model for capturing the system and adversary behaviors from the time instant of the first intrusion launched from any attacked computer to the other vulnerable computers until defense solution is developed and deployed. Stochastic Reward Nets (SRN) is applied to automatically build and solve the model. The formulas are also derived for calculating the metrics of interest. Simulation is carried out to validate the approximate accuracy of our model and formulas. The quantitative analysis can help network administrators make a trade-off between damage loss and defense cost.

Mobile Edge Computing may become a prevalent platform to support applications where mobile devices have limited compute, storage, energy and/or data privacy concerns. In this paper, we study the efficient provisioning and management of compute resources in the Edge-to-Cloud continuum for different types of real-time applications with timeliness requirements depending on application-level update rates and communication/compute delays. We begin by introducing a highly stylized network model allowing us to study the salient features of this problem including its sensitivity to compute vs. communication costs, application requirements, and traffic load variability. We then propose an online decentralized service placement algorithm, based on estimating network delays and adapting application update rates, which achieves high service availability. Our results exhibit how placement can be optimized and how a load-balancing strategy can achieve near-optimal service availability in large networks.

Edge computing (EC) is a new developing computing technology where data are collected, and analysed nearer to the edge or sources of the data. Cloud to the edge, intelligent applications and analytics are part of the IoT applications and technology. Edge computing technology aims to bring cloud computing features near to edge devices. For time-sensitive applications in cloud computing, architecture massive volume of data is generated at the edge and stored and analysed in the cloud. Cloud infrastructure is a composition of data centres and large-scale networks, which provides reliable services to users. Traditional cloud computing is inefficient due to delay in response, network delay and congestion as simultaneous transactions to the cloud, which is a centralised system. This paper presents a literature review on cloud-based edge computing technologies for delay-sensitive applications and suggests a conceptual model of edge computing architecture. Further, the paper also presents the implementation of QoS support edge computing paradigm in Python for further research to improve the latency and throughput for time-sensitive applications.

Stochastic computing is based on probability concepts which are different from conventional mathematical operations. Advantages of stochastic computing in the fields of neural networks and digital image processing have been reported in literature recently. Arithmetic operations especially multiplications can be performed either by logical AND gates in unipolar format or by EXNOR gates in bipolar format in stochastic computation. Stochastic computing is inherently fault-tolerant and requires fewer logic gates to implement arithmetic operations. Long computing time and low accuracy are the main drawbacks of this system. In this presentation, to reduce hardware requirement and delay, modified stochastic multiplication using AND gate array and multiplexer are used for the design of Finite Impulse Response Filter cores. Performance parameters such as area, power and delay for FIR filter using modified stochastic computing methods are compared with conventional floating point computation.

Kernel embeddings of distributions have recently gained significant attention in the machine learning community as a data-driven technique for representing probability distributions. Broadly, these techniques enable efficient computation of expectations by representing integral operators as elements in a reproducing kernel Hilbert space. We apply these techniques to the area of stochastic optimal control theory and present a method to compute approximately optimal policies for stochastic systems with arbitrary disturbances. Our approach reduces the optimization problem to a linear program, which can easily be solved via the Lagrangian dual, without resorting to gradient-based optimization algorithms. We focus on discrete- time dynamic programming, and demonstrate our proposed approach on a linear regulation problem, and on a nonlinear target tracking problem. This approach is broadly applicable to a wide variety of optimal control problems, and provides a means of working with stochastic systems in a data-driven setting.

The accelerated penetration rate of renewable energy sources (RES) brings environmental benefits at the expense of increasing operation cost and undermining the satisfaction of the N-1 security criterion. To address the latter issue, this paper extends the state of the art, i.e. deterministic AC security-constrained optimal power flow (SCOPF), to capture two new dimensions: RES stochasticity and inter-temporal constraints of emerging sources of flexibility such as flexible loads (FL) and energy storage systems (ESS). Accordingly, the paper proposes and solves for the first time a new problem formulation in the form of stochastic multi-period AC SCOPF (S-MP-SCOPF). The S-MP-SCOPF is formulated as a non-linear programming (NLP). It computes optimal setpoints in day-ahead operation of flexibility resources and other conventional control means for congestion management and voltage control. Another salient feature of this paper is the comprehensive and accurate modelling: AC power flow model for both pre-contingency and post-contingency states, joint active/reactive power flows, inter-temporal resources such as FL and ESS in a 24-hours time horizon, and RES uncertainties. The applicability of the proposed model is tested on 5-bus (6 contingencies) and 60 bus Nordic32 (33 contingencies) systems.

We propose a novel theoretical approach for solving a Stochastic Security Game using augmented Markov Decison Processes and an experimental evaluation. Most of the previous works mentioned in the literature focus on Linear Programming techniques seeking Strong Stackelberg Equilibria through the defender and attacker’s strategy spaces. Although effective, these techniques are computationally expensive and tend to not scale well to very large problems. By fixing the set of the possible defense strategies, our approach is able to use the well-known augmented MDP formalism to compute an optimal policy for an attacker facing a defender patrolling. Experimental results on fully observable cases validate our approach and show good performances in comparison with optimistic and pessimistic approaches. However, these results also highlight the need of scalability improvements and of handling the partial observability cases.

Vehicular ad-hoc networks (VANETs) are one of the most stochastic networks in terms of node movement patterns. Due to the high speed of vehicles, nodes form temporary clusters and shift between clusters rapidly, which limits the usable computational complexity for quality of service (QoS) and security enhancements. Hence, VANETs are one of the most insecure networks and are prone to various attacks like Masquerading, Distributed Denial of Service (DDoS) etc. Various algorithms have been proposed to safeguard VANETs against these attacks, which vary concerning security and QoS performance. These algorithms include linear rule-checking models, software-defined network (SDN) rules, blockchain-based models, etc. Due to such a wide variety of model availability, it becomes difficult for VANET designers to select the most optimum security framework for the network deployment. To reduce the complexity of this selection, the paper reviews statistically investigate a wide variety of modern VANET-based security models. These models are compared in terms of security, computational complexity, application and cost of deployment, etc. which will assist network designers to select the most optimum models for their application. Moreover, the paper also recommends various improvements that can be applied to the reviewed models, to further optimize their performance.

Cyber-Physical Systems(CPS) have been experiencing a fast-growing process in recent decades, and related security issues also have become more important than ever before. To design an efficient defensive policy for operators and controllers is the utmost task to be considered. In this paper, a stochastic game-theoretic model is developed to study a CPS security problem by considering the interdependence between cyber and physical spaces of a CPS. The game model is solved with Minimax Q-learning for finding the mixed strategies equilibria. The numerical simulation revealed that the defensive factors and attack cost can affect the policies adopted by the system. From the perspective of the operator of a CPS, increasing successful defense probability in the phrase of disruption will help to improve the probability of defense strategy when there is a correlation between the cyber layer and the physical layer in a CPS. On the contrary side, the system defense probability will decrease as the total cost of the physical layer increases.

We present a scalable optimization algorithm and its parallel implementation for two-stage stochastic programming problems of large-scale, particularly the security constrained optimal power flow models routinely used in electrical power grid operations. Such problems can be prohibitively expensive to solve on industrial scale with the traditional methods or in serial. The algorithm decomposes the problem into first-stage and second-stage optimization subproblems which are then scheduled asynchronously for efficient evaluation in parallel. Asynchronous evaluations are crucial in achieving good balancing and parallel efficiency because the second-stage optimization subproblems have highly varying execution times. The algorithm employs simple local second-order approximations of the second-stage optimal value functions together with exact first- and second-order derivatives for the first-stage subproblems to accelerate convergence. To reduce the number of the evaluations of computationally expensive second-stage subproblems required by line search, we devised a flexible mechanism for controlling the step size that can be tuned to improve performance for individual class of problems. The algorithm is implemented in C++ using MPI non-blocking calls to overlap computations with communication and boost parallel efficiency. Numerical experiments of the algorithm are conducted on Summit and Lassen supercomputers at Oak Ridge and Lawrence Livermore National Laboratories and scaling results show good parallel efficiency.