The complexity of synthesizing elementary net systems relative to natural parameters

作者:

Highlights:

摘要

Elementary net systems (ENS) are the most fundamental class of Petri nets. Given a labeled transition system (TS) A, feasibility is the NP-complete decision problem whether A can be synthesized into an ENS. We analyze the impact of state degree, the number of allowed successors and predecessors of states in A, and event manifoldness, the amount of occurrences that an event can have in A, on the computational complexity of feasibility and the related event state separation property (ESSP) and state separation property (SSP). Feasibility, ESSP and SSP are NP-complete for TSs with state degree one and event manifoldness not less than three and for TSs with state degree and event manifoldness at least two. As we also show that SSP becomes tractable for TSs with state degree one and event manifoldness two, the only cases left open are ESSP and feasibility for the same input restriction.

论文关键词:Elementary net systems,Petri net synthesis,NP-completeness,Parameterized complexity

论文评审过程:Received 2 March 2019, Revised 26 November 2019, Accepted 12 December 2019, Available online 14 January 2020, Version of Record 14 February 2020.

论文官网地址:https://doi.org/10.1016/j.jcss.2019.12.007