If At First You Don't Succeed: Extended Monitorability through Multiple Executions

06/08/2023
by   Antonis Achilleos, et al.
0

This paper investigates the observational capabilities of monitors that can observe a system over multiple runs. We study how the augmented monitoring setup affect the class of properties that can be verified at runtime, focussing on branching-time properties expressed in the modal mu-calculus. Our results show that the setup can be used to systematically extend previously established monitorability limits. We also prove bounds that capture the correspondence between the syntactic structure of a branching-time property and the number of system runs required to conduct the verification.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
02/26/2018

Finite Model Property of Pretransitive Modal Logic

The finite model property of the pretransitive modal logic K_2,3=K⊕ p→ p...
research
11/21/2022

Ain't No Stopping Us Monitoring Now

Not all properties are monitorable. This is a well-known fact, and it me...
research
06/20/2019

Gray-box Monitoring of Hyperproperties (Extended Version)

Many important system properties, particularly in security and privacy, ...
research
11/11/2021

DropGNN: Random Dropouts Increase the Expressiveness of Graph Neural Networks

This paper studies Dropout Graph Neural Networks (DropGNNs), a new appro...
research
08/14/2020

Can determinism and compositionality coexist in RML? (extended version)

Runtime verification (RV) consists in dynamically verifying that the eve...
research
12/15/2022

Runtime Monitoring for Out-of-Distribution Detection in Object Detection Neural Networks

Runtime monitoring provides a more realistic and applicable alternative ...

Please sign up or login with your details

Forgot password? Click here to reset