Cal structural refinement relation amongst reactive models is Robin Milner’s simulation relation. We generalize this relation to various quantitative distances, such as simulation failure distance, which measures the frequency with which a program SYN115 manufacturer violates a requirement; simulation tolerance distance, which measures the degree to which the program is extra constrained than the requirement; and simulation robustness distance, which measures the degree to which the system can be relaxed without the need of violating the requirement. Every of those simulation distances define quantitative measures of fitness with respect to each functional and nonfunctional system needs. Subject 3: Composing and refining distance measures Assume-guarantee proof decomposition, needs preserving abstraction and abstraction refinement, at the same time as symbolic data structures are some of the most significant principles for organizing the qualitative analysis of reactive models with massive and unbounded state spaces. We generalize these paradigms towards the quantitative setting where distances need to be added when composing, and approximated monotonically when abstracting or refining. Topic 4: Measuring method robustness and designing robust systems The (informal) notion of robustness is an crucial design and style criterion for artificial systems, also as a characteristic attribute of quite a few natural systems. On the list of primary added benefits of a quantitative theory of systems is that robustness can be formalized as mathematical continuity, in the sense that a method is robust iff little adjustments inside the input cause only little modifications within the output [4]. A variation of this theme is the fact that, within a robust method, couple of violations of input assumptions should trigger only couple of violations of output needs. We create and study such theories of system robustness. Topic 5: Quantitative measures in multicore and cloud computing We define and evaluate quantitative measures that attempt to capture modern elements of concurrent computation. For multicore applications, we want to quantify the performance of different thread synchronization principles, which include the trade-off involving fine-grained atomicity, which provides greater flexibility by permitting many interleavings of distinctive threads, and coarse-grained atomicity, which causes reduce overheads for context switching. For information center and cloud applications, we desire to quantify the resource price and utilization of numerous processor assignment and job scheduling policies. Subject 6: Quantitative models in systems biology Whilst there have been a number of extensively noticed attempts to unleashthe enormous potential of reactive modeling in biology, by far the most usually employed mechanistic models in cell biology are nonetheless primitive by the requirements of personal computer science. We think that a quantitative framework for reactive modeling won’t only enhance acceptance by biologists, but also supply techniques for measuring the fitness of reactive models with respect to experimental data, and for synthesizing models of maximal fitness. We collaborate with systems and cell biologists to achieve this job.2 The state with the art We develop on perform from quite a few diverse subdisciplines of laptop or computer science. First, there is a massive physique of mature investigation in qualitative reactive modeling and verification. Second, the qualitative setting has been extended in a number of quantitative directions, most notably inside the directions of (i) real-time and hybrid systems, (ii) probabilistic systems, and.