We study the complexity of deciding bisimilarity between non-deterministic processes with explicit primitives for manipulating data values. In particular, we consider a language with valuepassing (input/output of data) and parametric deÿnitions of processes. We distinguish the case in which data can
The complexity of bisimilarity-checking for one-counter processes
✍ Scribed by Antonı́n Kučera
- Publisher
- Elsevier Science
- Year
- 2003
- Tongue
- English
- Weight
- 376 KB
- Volume
- 304
- Category
- Article
- ISSN
- 0304-3975
No coin nor oath required. For personal study only.
✦ Synopsis
We study the problem of bisimilarity-checking between processes of one-counter automata and ÿnite-state processes. We show that deciding weak bisimilarity between processes of one-counter nets (which are 'restricted' one-counter automata where the counter cannot be tested for zero) and ÿnite-state processes is DP-hard. In particular, this means that the problem is both NP and co-NP hard. The same technique is used to demonstrate co-NP-hardness of strong bisimilarity between processes of one-counter nets. Then we design an algorithm which decides weak bisimilarity between processes of one-counter automata and ÿnite-state processes in time which is polynomial for a large subclass of instances, giving a kind of characterization of all hard instances as a byproduct. Moreover, we show how to e ciently estimate the time which is needed to solve a given instance. Finally, we prove that the problem of strong bisimilarity between processes of one-counter automata and ÿnite-state processes is in P.
📜 SIMILAR VOLUMES