Marcin Jurdzinski

Small Progress Measures for Solving Parity Games

In this paper we develop a new algorithm for deciding the winner in parity games, and hence also for the modal µ-calculus model checking. The design and analysis of the algorithm is based on a notion of game progress measures: they are witnesses for winning strategies in parity games. We characterize game progress measures as pre-fixed points of certain monotone operators on a complete lattice. As a result we get the existence of the least game progress measures and a straightforward way to compute them. The worst-case running time of our algorithm matches the best worst-case running time bounds known so far for the problem, achieved by the algorithms due to Browne et al., and Seidl. Our algorithm has better space complexity: it works in small polynomial space; the other two algorithms have exponential worst-case space complexity.

Deciding the winner in parity games is in UP ∩ co-UP

Abstract We observe that the problem of deciding the winner in mean payoff games is in the complexity class UP ∩ co-UP. We also show a simple reduction from parify games to mean payoff games. From this it follows that deciding the winner in parity games and the modal μ-calculus model checking are in UP ∩ co-UP.

A Discrete Strategy Improvement Algorithm for Solving Parity Games

A discrete strategy improvement algorithm is given for constructing winning strategies in parity games, thereby providing also a new solution of the model-checking problem for the modal μ-calculus. Known strategy improvement algorithms, as proposed for stochastic games by Hoffman and Karp in 1966, and for discounted payoff games and parity games by Puri in 1995, work with real numbers and require solving linear programming instances involving high precision arithmetic. In the present algorithm for parity games these difficulties are avoided by the use of discrete vertex valuations in which information about the relevance of vertices and certain distances is coded. An efficient implementation is given for a strategy improvement step. Another advantage of the present approach is that it provides a better conceptual understanding and easier analysis of strategy improvement algorithms for parity games. However, so far it is not known whether the present algorithm works in polynomial time. The long standing problem whether parity games can be solved in polynomial time remains open.

A deterministic subexponential algorithm for solving parity games

The existence of polynomial time algorithms for the solution of parity games is a major open problem. The fastest known algorithms for the problem are randomized algorithms that run in subexponential time. These algorithms are all ultimately based on the randomized subexponential simplex algorithms of Kalai and of Matousek, Sharir and Welzl. Randomness seems to play an essential role in these algorithms. We use a completely different, and elementary, approach to obtain a deterministic subexponential algorithm for the solution of parity games. Our deterministic algorithm is almost as fast as the randomized algorithms mentioned above.

Simple Stochastic Parity Games

Many verification, planning, and control problems can be modeled as games played on state-transition graphs by one or two players whose conflicting goals are to form a path in the graph. The focus here is on simple stochastic parity games, that is, two-player games with turn-based probabilistic transitions and ω-regular objectives formalized as parity (Rabin chain) winning conditions. An efficient translation from simple stochastic parity games to nonstochastic parity games is given. As many algorithms are known for solving the latter, the translation yields efficient algorithms for computing the states of a simple stochastic parity game from which a player can win with probability 1.

How much memory is needed to win infinite games

We consider a class of infinite two-player games on finitely coloured graphs. Our main question is: given a winning condition, what is the inherent blow-up (additional memory) of the size of the I/O automata realizing winning strategies in games with this condition. This problem is relevant to synthesis of reactive programs and to the theory of automata on infinite objects. We provide matching upper and lower bounds for the size of memory needed by winning strategies in games with a fixed winning condition. We also show that in the general case the LAR (latest appearance record) data structure of Gurevich and Harrington is optimal. Then we propose a more succinct way of representing winning strategies by means of parallel compositions of transition systems. We study the question: which classes of winning conditions admit only polynomial-size blowup of strategies in this representation.

Mean-payoff parity games

Games played on graphs may have qualitative objectives, such as the satisfaction of an /spl omega/-regular property, or quantitative objectives, such as the optimization of a real-valued reward. When games are used to model reactive systems with both fairness assumptions and quantitative (e.g., resource) constraints, then the corresponding objective combines both a qualitative and a quantitative component. In a general case of interest, the qualitative component is a parity condition and the quantitative component is a mean-payoff reward. We study and solve such mean-payoff parity games. We also prove some interesting facts about mean-payoff parity games which distinguish them both from mean-payoff and from parity games. In particular, we show that optimal strategies exist in mean-payoff parity games, but they may require infinite memory.

A Deterministic Subexponential Algorithm for Solving Parity Games

The existence of polynomial time algorithms for the solution of parity games is a major open problem. The fastest known algorithms for the problem are randomized algorithms that run in subexponential time. These algorithms are all ultimately based on the randomized subexponential simplex algorithms of Kalai and of Matoušek, Sharir and Welzl. Randomness seems to play an essential role in these algorithms. We use a completely different, and elementary, approach to obtain a deterministic subexponential algorithm for the solution of parity games. Our deterministic algorithm is almost as fast as the randomized algorithms mentioned above.

On Nash Equilibria in Stochastic Games

We study infinite stochastic games played by n-players on a finite graph with goals given by sets of infinite traces. The games are stochastic (each player simultaneously and independently chooses an action at each round, and the next state is determined by a probability distribution depending on the current state and the chosen actions), infinite (the game continues for an infinite number of rounds), nonzero sum (the players’ goals are not necessarily conflicting), and undiscounted. We show that if each player has a reachability objective, that is, if the goal for each player i is to visit some subset Ri of the states, then there exists an e-Nash equilibrium in memoryless strategies, for every e >0. However, exact Nash equilibria need not exist. We study the complexity of finding such Nash equilibria, and show that the payoff of some e-Nash equilibrium in memoryless strategies can be e-approximated in NP.

Model checking probabilistic timed automata with one or two clocks

Probabilistic timed automata are an extension of timed automata with discrete probability distributions.We consider model-checking algorithms for the subclasses of probabilistic timed automata which have one or two clocks. Firstly, we show that PCTL probabilistic model-checking problems (such as determining whether a set of target states can be reached with probability at least 0.99 regardless of how nondeterminism is resolved) are PTIME-complete for one clock probabilistic timed automata, and are EXPTIME-complete for probabilistic timed automata with two clocks. Secondly, we show that the model-checking problem for the probabilistic timed temporal logic PTCTL is EXPTIME-complete for one clock probabilistic timed automata. However, the corresponding model-checking problem for the subclass of PTCTL which does not permit both (1) punctual timing bounds, which require the occurrence of an event at an exact time point, and (2) comparisons with probability bounds other than 0 or 1, is PTIME-complete.