Resource-Aware Verification Using Randomized Exploration of Large State Spaces

Abstract

Abstract. Exhaustive verification often suffers from the state-explosion problem, where the reachable state space is too large to fit in main memory. For this reason, and because of disk swapping, once the main memory is full very little progress is made, and the process is not scalable. To alleviate this, partial verification methods have been proposed, some based on randomized exploration, mostly in the form of random walks. In this paper, we enhance partial, randomized state-space exploration methods with the concept of resource-awareness: the exploration algorithm is made aware of the limits on resources, in particular memory and time. We present a memory-aware algorithm that by design never stores more states than those that fit in main memory. We also propose criteria to compare this algorithm with similar other algorithms. We study properties of such algorithms both theoretically on simple classes of state spaces and experimentally on some preliminary case studies.

Sequential and Parallel Algorithms for the State Space Exploration

“…Exploration is a computing process which determines a sequence of actions making it possible to achieve a desired goal. A good exploration means the achievement and the storage of a large number of states of a system without exceeding the available memory resources [1]. The state space can be described by an initial state and a set of transitions.…”

Modeling, Verification, and Testing Using Timed and Hybrid Automata

“…See [104] for a detailed discussion. The idea of memory-aware verification is also exploited in [1], where a class of randomized exploration algorithms are introduced that use a parameter N representing the number of states that the algorithm is allowed to maintain at any given time during its execution. Given a model, N can be computed as Having N as an upper bound, many different randomized exploration algorithms can be tried out, depending on how two main policies are defined: how, given the current state of the algorithm, to pick which node to explore next (the select function), and how, given a selected node, to pick a successor of this node and update the state (the update function).…”

“…Because of this dependence, obtaining analytical formulas for the above criteria is a very difficult tasks. Even for simple graphs such as regular trees, it can be non-trivial [1]. On the other hand, experimental results can often be obtained much more easily, e.g., see [84,1].…”

“…Even for simple graphs such as regular trees, it can be non-trivial [1]. On the other hand, experimental results can often be obtained much more easily, e.g., see [84,1]. This is an exciting field of research and we expect it to become more popular in the near future, because of its high relevance in industrial practice.…”

Uniform Monte-Carlo Model Checking

“…Random exploration of a model is a classical approach in simulation [3] and testing (see for instance [26,6]) and more recently in model-checking [21,8,20,1]. A usual way to explore a model at random is to use isotropic random walks: given the states of the model and their successors, an isotropic random walk is a succession of states where at each step, the next state is drawn uniformly at random among the successors, or, as in [11] the next transition is drawn uniformly at random among the outgoing transitions.…”