TY - UNPB

T1 - A Faster Algorithm for Solving One-Clock Priced Timed Games

AU - Hansen, Thomas Dueholm

AU - Ibsen-Jensen, Rasmus

AU - Miltersen, Peter Bro

PY - 2012

Y1 - 2012

N2 - One-clock priced timed games is a class of two-player, zero-sum, continuous-time games that was defined and thoroughly studied in previous works. We show that one-clock priced timed games can be solved in time m 12^n n^(O(1)), where n is the number of states and m is the number of actions. The best previously known time bound for solving one-clock priced timed games was 2^(O(n^2+m)), due to Rutkowski. For our improvement, we introduce and study a new algorithm for solving one-clock priced timed games, based on the sweep-line technique from computational geometry and the strategy iteration paradigm from the algorithmic theory of Markov decision processes. As a corollary, we also improve the analysis of previous algorithms due to Bouyer, Cassez, Fleury, and Larsen; and Alur, Bernadsky, and MadhusudanWe present a construction of log-depth formulae for various threshold functions based on atomic threshold gates of constant size. From this, we build a new family of linear secret sharing schemes that are multiplicative, scale well as the number of players increases and allows to raise a shared value to the characteristic of the underlying field without interaction. Some of these schemes are in addition strongly multiplicative. Our formulas can also be used to construct multiparty protocols from protocols for a constant number of parties. In particular we implement black-box multiparty computation over non-Abelian groups in a way that is much simpler than previously known and we also show how to get a protocol in this setting that is efficient and actively secure against a constant fraction of corrupted parties, a long standing open problem. Finally, we show a negative result on usage of our scheme for pseudorandom secret sharing as defined by Cramer, Damgård and Ishai

AB - One-clock priced timed games is a class of two-player, zero-sum, continuous-time games that was defined and thoroughly studied in previous works. We show that one-clock priced timed games can be solved in time m 12^n n^(O(1)), where n is the number of states and m is the number of actions. The best previously known time bound for solving one-clock priced timed games was 2^(O(n^2+m)), due to Rutkowski. For our improvement, we introduce and study a new algorithm for solving one-clock priced timed games, based on the sweep-line technique from computational geometry and the strategy iteration paradigm from the algorithmic theory of Markov decision processes. As a corollary, we also improve the analysis of previous algorithms due to Bouyer, Cassez, Fleury, and Larsen; and Alur, Bernadsky, and MadhusudanWe present a construction of log-depth formulae for various threshold functions based on atomic threshold gates of constant size. From this, we build a new family of linear secret sharing schemes that are multiplicative, scale well as the number of players increases and allows to raise a shared value to the characteristic of the underlying field without interaction. Some of these schemes are in addition strongly multiplicative. Our formulas can also be used to construct multiparty protocols from protocols for a constant number of parties. In particular we implement black-box multiparty computation over non-Abelian groups in a way that is much simpler than previously known and we also show how to get a protocol in this setting that is efficient and actively secure against a constant fraction of corrupted parties, a long standing open problem. Finally, we show a negative result on usage of our scheme for pseudorandom secret sharing as defined by Cramer, Damgård and Ishai

M3 - Working paper

BT - A Faster Algorithm for Solving One-Clock Priced Timed Games

ER -