Monitoring Temporal Properties of Continuous Signals
read more
Citations
A Brief Account of Runtime Verification
Robust satisfaction of temporal logic over real-valued signals
Robustness of temporal logic specifications for continuous-time signals
Runtime Verification for LTL and TLTL
Model predictive control with signal temporal logic specifications
References
Model checking
A Temporal Logic of Nested Calls and Returns
An Automata-Theoretic Approach to Automatic Program Verification
Temporal Verification of Reactive Systems: Safety
Related Papers (5)
Frequently Asked Questions (12)
Q2. What are the future works in "Monitoring temporal properties of continuous signals" ?
To treat such properties the authors need to extend the architecture of the monitor beyond Boolean filters to include arithmetical blocks, integrators, etc. – Frequency domain properties: these properties are not temporal in their sense but speak about the spectrum of the signal via some transform such as Fourier or wavelets. In the future versions the authors will develop a tighter coupling between the monitor and the simulator where the monitor is a Matlab block that can influence the choice of sampling points in order to detect changes in the Boolean abstractions. Some of these transforms can be done only offline and some can be dome partially online using a shifting time window. The authors will also work on integration with other simulators used in control and circuit design.
Q3. What is the preferred method of verification for systems outside the scope of automatic verification tools?
For systems which are outside the scope of automatic verification tools, either due to the incorporation of unbounded variables (numbers, queues) or simply due to size, simulation/testing is still the preferred validation method.
Q4. What is the meaning of a time-bounded until operator?
The satisfaction relation (s, t) |= ϕ, indicating that signal s satisfies ϕ starting from position t, is defined inductively as follows:(s, t) |= p ↔ πp(s)[t] = T (s, t) |= ¬ϕ ↔ (s, t) |= ϕ (s, t) |= ϕ1 ∨ ϕ2 ↔ (s, t) |= ϕ1 or (s, t) |= ϕ2 (s, t) |= ϕ1U[a,b]ϕ2 ↔ ∃t′ ∈ [t+ a, t+ b] (s, t′) |= ϕ2 and ∀t′′ ∈ [t, t′], (s, t′′) |= ϕ1 Note that their definition of the semantics of the time-bounded until operator differs slightly from its conventional definition since it requires a time instant t′ ∈ [t+a, t+ b] where both (s, t′) |= ϕ2 and (s, t′) |= ϕ1.
Q5. What is the main contribution of this work?
The main contribution of this work is the definition of a temporal logic for specifying properties of dense-time real-valued signals and the automatic generation of property monitors for this language.
Q6. How do you get a deterministic acceptor for a formula?
For discrete systems there are various ways to obtain a deterministic acceptor for a formula, e.g. by applying subset construction to the non-deterministic automaton obtained using a tableau-based translation method.
Q7. What is the only work the authors are aware of concerning monitoring of dense time properties?
The only work the authors are aware of concerning monitoring of dense time properties is that of [BBKT04] who propose an automatic generation of real-time (analog or digital) observers from timed automaton specifications.
Q8. What is the definition of temporal logic?
Temporal logic [MP95] is a rigorous formalism for specifying desired behaviors of discrete systems such as programs or digital circuits.
Q9. What is the definition of a MITL?
Let U = {μ1, . . . , μn} be a collection of predicates, effective functions of the form μi : Rm → B. An STL(U) formula is an MITL[a,b] formula over the atomic propositions μ1(x), . . . μn(x).
Q10. What is the future of the monitor?
In the future versions the authors will develop a tighter coupling between the monitor and the simulator where the monitor is a Matlab block that can influence the choice of sampling points in order to detect changes in the Boolean abstractions.
Q11. What is the common version of real-time temporal logic?
Some restricted versions of real-time temporal logic already appear in some tools, for example, the specification of real-time properties in MaCS is based on a logic that supports timestamped instantaneous events and conditions which have a duration between two events.
Q12. What is the reason to prefer an online procedure for monitoring bounded variability signals?
To monitor them the authors need to pass the signal first through the transform in question and then check whether the result satisfies the formula.