Showing 1–4 of 4 results for author: Dshalalow, J H

Discrete Operational Calculus in Delayed Stochastic Games

Authors: J. H. Dshalalow, K. Iwezulu, R. T. White

Abstract: This article deals with classes of antagonistic games with two players. A game is specified in terms of two `hostile' stochastic processes representing mutual attacks upon random times exerting casualties of random magnitudes. The game ends when one of the players is defeated. We target the first passage time $τ_ρ$ of the defeat and the number of casualties to either player upon $τ_ρ$. Here we val… ▽ More This article deals with classes of antagonistic games with two players. A game is specified in terms of two `hostile' stochastic processes representing mutual attacks upon random times exerting casualties of random magnitudes. The game ends when one of the players is defeated. We target the first passage time $τ_ρ$ of the defeat and the number of casualties to either player upon $τ_ρ$. Here we validate our claim of analytic tractability of the general formulas obtained in [1] under various transforms. △ Less

Submitted 22 January, 2019; originally announced January 2019.

Comments: Pre-print

MSC Class: 60G51; 60G55; 60G57; 60K05; 82B41; 91A05; 91A10; 91A60

Journal ref: Neural, Parallel, and Scientific Computations, 24: 55-64 (2016)

Time Sensitive Analysis of Independent and Stationary Increment Processes

Authors: Jewgeni H. Dshalalow, Ryan T. White

Abstract: We study the behavior of independent and stationary increments jump processes as they approach fixed thresholds. The exact crossing time is unavailable because the real-time information about successive jumps is unknown. Instead, the underlying process $A(t)$ is observed only upon a third-party independent point process ${τ_n}$. The observed time series ${A(τ_n)}$ presents crude, delayed data. The… ▽ More We study the behavior of independent and stationary increments jump processes as they approach fixed thresholds. The exact crossing time is unavailable because the real-time information about successive jumps is unknown. Instead, the underlying process $A(t)$ is observed only upon a third-party independent point process ${τ_n}$. The observed time series ${A(τ_n)}$ presents crude, delayed data. The crossing is first observed upon one of the observations, denoted $τ_ν$. We develop and further explore a new technique to revive the real-time paths of $A(t)$ for all $t$ belonging to an interval before the pre-crossing observation, $[0, τ_{ν-1})$, or between the observations just before and just after the crossing, $[τ_{ν-1}, τ_ν)$, as a joint Laplace-Stieltjes transform and probability generating function of $A(τ_{ν-1})$, $A(τ_ν)$, $τ_{ν-1}$, and $τ_ν$. Joint probability distributions are obtained from the transforms in a tractable form and they are applied to modeling of stochastic networks under cyber attacks by accurately predicting their crash. △ Less

Submitted 21 January, 2019; originally announced January 2019.

Comments: Post-print, Journal of Mathematical Analysis and Applications

MSC Class: 60G50; 60G51; 60G52; 60G55; 60G57; 60K05; 60K35; 60K40; 60G25; 90B18; 90B15; 90B25

On Strategic Defense of Stochastic Networks

Authors: Jewgeni H. Dshalalow, Ryan T. White

Abstract: This paper deals with the detection and prediction of losses due to cyber attacks waged on vital networks. The accumulation of losses to a network during a series of attacks is modeled by a 2-dimensional monotone random walk process as observed by an independent delayed renewal process. The first component of the process is associated with the number of nodes (such as routers or operational sites)… ▽ More This paper deals with the detection and prediction of losses due to cyber attacks waged on vital networks. The accumulation of losses to a network during a series of attacks is modeled by a 2-dimensional monotone random walk process as observed by an independent delayed renewal process. The first component of the process is associated with the number of nodes (such as routers or operational sites) incapacitated by successive attacks. Each node has a weight associated with its incapacitation (such as loss of operational capacity or financial cost associated with repair), and the second component models the cumulative weight associated with the nodes lost. Each component has a fixed threshold and crossing of a threshold by either component represents the network entering a critical condition. Results are given as joint functionals of the predicted time of the first observed threshold crossing along with the values of each component upon this time. △ Less

Submitted 21 January, 2019; originally announced January 2019.

Comments: Post-print, 5 figures

MSC Class: 60G50; 60G51; 60G52; 60G55; 60G57; 60K05; 60K35; 60K40; 60G25; 90B18; 90B10; 90B15; 90B25

Journal ref: Stochastic Analysis and Applications, 32:3, 365-396

On Reliability of Stochastic Networks

Authors: Jewgeni H. Dshalalow, Ryan T. White

Abstract: In recent times we hear increasingly often about cyber attacks on various commercial and strategic sites that manage to escape any defense. In this article, we model such attacks on networks via stochastic processes and predict the time of a total or partial failure of a network including the magnitude of losses (such as the number of compromised nodes, lost weights, and a loss of other associated… ▽ More In recent times we hear increasingly often about cyber attacks on various commercial and strategic sites that manage to escape any defense. In this article, we model such attacks on networks via stochastic processes and predict the time of a total or partial failure of a network including the magnitude of losses (such as the number of compromised nodes, lost weights, and a loss of other associated components relative to some fixed thresholds). To make such modeling more realistic we also assume that the information about the attacks is delayed as per random observations. We arrive at analytically and numerically tractable results demonstrated by examples and comparative simulation. △ Less

Submitted 21 January, 2019; originally announced January 2019.

Comments: Pre-print

MSC Class: 60G50; 60G51; 60G55; 60K05; 60K10; 90B15

Journal ref: Neural, Parallel, and Scientific Computations, 21, 141-160