Citation Context

...niques for solving systems of linear equations [48]. The PRCTL algorithms in MRMC are described in [1]. CSL model-checking techniques in MRMC stem from [3] and for its reward extension CSRL from [2], =-=[13]-=-, [20]. For time- and reward-bounded until formulae we have implemented two algorithms: one based on discretization [49] and one based on uniformization and path truncation [43]. The algorithms for PR...

Citation Context

... (the Erlangen-Twente Markov Chain Checker) [13] which can both be used for verifying CTMCs against CSL specifications. MRMC also supports rewards-based property specifications through the logic CSRL =-=[6]-=-. Models, in addition to other PRISM outputs such as numerical results or simulation traces, can be imported into more general-purpose tools such as MATLAB and MAPLE. 5. CONCLUSIONS We have illustrate...

Citation Context

...he work of Qiu et al. [16]. Vijayakrishnan et al. [17] proposed to use the more powerful model of Generalized Stochastic Petri Nets (GSPN). The use of Markov Reward Models (MRM) has gained visibility =-=[18, 19]-=- due to their power of modeling and expressiveness in the domain of energy consumption. Cloth et al. [20] presented the efficient use of MRMs in energy-consumption modeling for embedded systems. On th...

Citation Context

...of methods do not scale well to the complex multi-processor systems. Moreover, they are prone to human errors. Computer based simulations or numerical methods are scalable but due to the usage of computer arithmetic and pseudo random numbers and their inherent incompleteness cannot guarantee accurate results. The accuracy of the above mentioned performance analysis is becoming quite important these days due to the increasing usage of multi-processor systems in safety-critical domains like medicine and transportation. Recently, probabilistic model checking has been used to analyze DMRMs (e.g., [4] and [8]). The typical model checking tools are PRISM [16] and MRMC [13], which provide precise system analysis by modeling the stochastic behaviors using probabilistic state machines and exhaustively verifying their probabilistic properties. These tools can be used for performance analysis of multi-processor systems as well. However, some algorithms implemented in these model checking tools are also based on numerical methods. For example, the Power method [15], which is a wellknown iterative method, is applied to compute the steady-state probabilities (or limiting probabilities) of Markov ch...