Algorithms for reachability problems on stochastic Markov reward models

Abstract: We study and develop the stochastic Markov reward model (sMRM), which extends the Markov chain where transition time/reward as modelled as random variables. Techniques are presented to enable computing first-passage time distributions (or reward distributions) with high accuracy for only modestly sized systems when solved for on a single computer. In contrast, naive simulations technique scale and are sufficient for a plethora of problems where high accuracy is not an issue, but this work perhaps would be valuable when extremely accurate solutions are demanded if it can be modelled precisely (a potentially big caveat). The work presented is intertwined with theory of temporal logics, although not the major contribution of the work, achieves the connection of this work to the field of automated probabilistic analysis/verification. Although its admittance does obfuscate the algorithmic details that are major, however. We present equations for computing first-passage reward densities, expected value problems, and other reachability problems. The focus was on finding strictly numerical solutions for first-passage time/reward densities. We adapted linear algebra algorithms such as Gaussian elimination, and iterative methods such as the power method, Jacobi and Gauss-Seidel. We provide solutions for both discrete-reward sMRMs, where all rewards discrete (lattice) random variables. And for continuous-reward sMRMs, where all rewards are strictly continuous random variables. Our solutions involve the use of fast Fourier transform (FFT) for faster computation, and we were able to adapt existing quadrature rules for convolution to gain more accurate solutions, rules such as the trapezoid rule, Simpson's rule and Romberg's method.