A scalable and analytically tractable probabilistic model for the cascading failure dynamics in power grids is constructed while retaining key physical attributes operating characteristics of grid. The approach based upon extracting a reduced abstraction large-scale using small number aggregate state variables modeling system continuous-time Markov chain. represent critical power-grid attributes, which have been shown, from prior simulation-based historical-data-based analysis, to strongly...

Cloud computing is being widely accepted and utilized in the business world. From perspective of businesses utilizing cloud, it critical to meet their customers' requirements by achieving service-level-objectives. Hence, ability accurately characterize optimize cloud-service performance great importance. In this paper a stochastic multi-tenant framework proposed model service customer requests cloud infrastructure composed heterogeneous virtual machines. Two metrics are mathematically...

Human operators play a key role in the reliable operation of critical infrastructures. However, human may take actions that are far from optimum. This can be due to various factors affecting operators' performance time-sensitive and situations such as reacting contingencies with significant monetary social impacts. In this paper, an analytic framework is proposed based on Markov chains for modeling dynamics cascading failures power grids. The model captures effects behavior quantified by...

Current power grids suffer periodic disturbances that may trigger cascades of component failures, which, in turn, can result blackouts different scales. Understanding cascading failures and the ability to predict risk are therefore great value designing control outage management systems for future smart grids. In this paper, a novel probabilistic approach is proposed model dynamics The characterizes probability reaching blackout an arbitrary size any time interval starting from initial grid...

Graph-based multi-view clustering encodes data into sample affinities to find consensus representation, effectively overcoming heterogeneity across different views. However, traditional affinity measures tend collapse as the feature dimension expands, posing challenges in estimating a unified alignment that reveals both crossview and inner relationships. To tackle this challenge, we propose achieve uniform via representation coregularization. First, are encoded by popular dyadic recent...

Communication networks in the smart grids play an important role reliability of power delivery. When grid is stressed as a result failures due to natural disasters and attacks (e.g., High Altitude Electro-Magnetic Pulse (HEMP) Weapons Mass Destruction (WMD's)), successive interdependent may propagate within across these leading large-scale cascading-failures blackouts. Failures can affect communication network, which turn, grid, so on. In this paper, we study influence network functionality...

Modern power grids rely heavily on their control systems operating over communication networks to mitigate the effects of stresses in grid. However, cascading phenomenon and blackouts remain possible if initial disturbances grid are accompanied by other system vulnerabilities such as failures that transmit implement critical signals. Understanding important design future smart grids. In this paper coupled with load shedding mechanism power-grid behavior revealed means simulation. particular,...

Various large-scale failures and attacks (e.g., natural disasters, high-altitude electromagnetic pulses (HEMPs), weapons of mass destruction (WMDs), etc.) can instigate initial critical network components router, generator) in a cyber-physical infrastructure. In addition, the intra-dependency among various inside communication inter-dependency with supporting power supply nodes contribute to propagation failures. Under probabilistic networks' physical infrastructures, this paper, we develop...

A novel consensus-based protocol is developed for estimating the load information at nodes in a distributed computing system operating over partially connected communication network. The challenge such estimation process arises from existence of tangible delays exchange information, and fact that loads are dynamic since continue to execute their while ongoing. utilizes concept trust weight each node has about any other node, based on number hops between, periodically form an updated estimate...

Cascading failures in the power grid refer to a chain of events triggered by an initial event such as single or combination generator, transmission line, communication control action. The can be attributed natural disaster intentional human-made attack. In this paper, cascading are analyzed under various conditions. First, Gaussian, circular and linear stressor(s) used model probability line failure due stressor(s). Second, Monte-Carlo simulations analyze impact based on patterns. reported...

A novel probabilistic multi-tenant model is developed to characterize the service performance of cloud systems. The considers essential cloud-system characteristics including virtualization, multi-tenancy and heterogeneity physical servers. Given model, three virtual machine mapping algorithms are proposed. Of particular interest max-load-first algorithm, which firstly maps largest VM, in terms workload size a user's request it serves, fastest server system. Monte-Carlo simulation results...

Cascading failures in a power grid result successive of transmission lines that may eventually affect large portions grid. In this paper, parametric model is proposed represents the propagation line At first, by optimal DC power-flow simulations IEEE 118 bus system, different types data on are collected, which can capture time evolution cascade. The probability given capacity fails found to be well approximated power-law (the higher lower probability). Then based these data, discrete-time...

This paper presents a general framework for optimal task reallocation in heterogeneous distributed-computing systems and offers rigorous analytical model the stochastic execution time of workload. The takes into account heterogeneity nature tasks' service transfer times, servers' failure as well an arbitrary task-reallocation policy. service, times are assumed to have general, age-dependent (non-exponential) distributions, resulting tandem distributed queuing system with non-Markovian...

In the last couple of years ARM-based servers have been gradually adopted by cloud service providers and utilized in data centers. Such tendency may provide great business opportunities for various companies industry. Hence, ability to timely track development trend server ecosystem (ASE) from technical perspective is importance. this paper level ASE quantitatively assessed based on open-source analysis. particular, statistical extracted 42 Linux kernels analyze process ASE. Furthermore, an...

In the last couple of years ARM-based servers have been gradually deployed in global data centers. Such tendency may provide great business opportunities for companies IT industry. Hence, ability to track development process server ecosystems is value. this paper a feasible method proposed evaluate level based on analyzing open-source information obtained from Linux kernels. particular, source code and git commit logs kernels are utilized characterize three major ecosystems, including x86,...