Jobs/M2 2017 MC Spot Generic Acc
|Algorithms for generic acceptance|
M2 2017 MC Spot Generic Acc
5-6 months in 2018
|General presentation of the field||
The Spot library (http://spot.lrde.epita.fr/) contains many algorithms for translating LTL formulas into omega-automata, and to simplify these formulas and automata. It also provides support to build an LTL model checker, through algorithms testing whether the language of such an automaton is empty or not.
Omega-automata are used to represent languages over infinite words. These automata can be defined using numerous acceptance conditions such as Büchi, co-Büchi, Rabin, Streett, Parity, etc. Spot actually supports these acceptances condition in a generic way and allow them to be combined as desired.
Experience of C++, notions of automata (not necessarily omega-automata).
The goal is to implement several algorithms that transform automata using one acceptance condition into automata using another acceptance condition. Some of these conversions are essential to some decision procedures.
For instance, we currently have an implement of Krishnan's algorithm for converting a deterministic Rabin automata into an equivalent deterministic Büchi automaton when such an automaton exist. We believe the procedure could be extended to handle deterministic "Generic Rabin" as input. Implementing such a new transformation would give us an efficient way to decide whether some LTL formula is realizable by a deterministic Büchi automaton, because we can already transform LTL into generic Rabin.
|Benefit for the candidate||
Krishnan et al.'s "Deterministic ω-Automata vis-a-vis Deterministic Büchi Automata" (ISAAC'94)
C. Löding. "Methods for the transformation of omega-automata: Complexity and connection to second order logic" (Master Thesis 1998)
U. Boker. "Why these automata types" (LPAR 2018)
|Place||LRDE: How to get to us|
1000 € gross / month
|Future work opportunities||
The internship can easily be prolonged as a PhD thesis.
< adl at lrde.epita.fr >