Abstract:Barrier certificates, serving as differential invariants that witness system safety, play a crucial role in the verification of cyber-physical systems (CPS). Prevailing computational methods for synthesizing barrier certificates are based on semidefinite programming (SDP) by exploiting Putinar Positivstellensatz. Consequently, these approaches are limited by the Archimedean condition, which requires all variables to be bounded, i.e., systems are defined over bounded domains. For systems over unbounded domains, unfortunately, existing methods become incomplete and may fail to identify potential barrier certificates. In this paper, we address this limitation for the unbounded cases. We first give a complete characterization of polynomial barrier certificates by using homogenization, a recent technique in the optimization community to reduce an unbounded optimization problem to a bounded one. Furthermore, motivated by this formulation, we introduce the definition of homogenized systems and propose a complete characterization of a family of non-polynomial barrier certificates with more expressive power. Experimental results demonstrate that our two approaches are more effective while maintaining a comparable level of efficiency.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the incompleteness problem of synthesizing barrier certificates on unbounded domains. Specifically, the existing methods based on semidefinite programming (SDP) have limitations when dealing with systems on unbounded domains because these methods rely on Putinar's Positivstellensatz, which requires that all variables must be bounded. For systems defined on unbounded domains, existing methods may not be able to identify potential barrier certificates, resulting in conservative or incomplete results. ### Background and Problem Description 1. **Importance of Barrier Certificates** - As differential invariants, barrier certificates are used to prove the safety of systems, and they play a crucial role especially in verifying the safety of cyber - physical systems (CPS). - Existing computational methods are mainly based on semidefinite programming (SDP) and use Putinar's Positivstellensatz to synthesize barrier certificates. 2. **Limitations of Existing Methods** - Putinar's Positivstellensatz requires that all variables must be bounded, that is, the system is defined on a bounded domain. - For systems defined on unbounded domains, existing methods become incomplete and may not be able to identify potential barrier certificates. ### Main Contributions of the Paper 1. **Clearly Distinguish Soundness and Completeness** - The authors clearly distinguish the soundness and completeness of the semidefinite programming representation of barrier certificates, which is often overlooked in existing work. 2. **Propose a Complete Representation on Unbounded Domains** - Using the homogenization technique, the authors propose the first complete semidefinite programming representation for synthesizing polynomial barrier certificates on unbounded domains. 3. **Extend to Non - polynomial Barrier Certificates** - The definition of a homogenized system is introduced, and a class of non - polynomial barrier certificates with stronger expressive power is considered. - A complete semidefinite programming representation of this class of non - polynomial barrier certificates is proposed. ### Experimental Results - The authors implemented algorithms based on the existing incomplete representation and the two newly proposed complete representations. - These algorithms were tested on a set of unbounded - domain benchmarks adapted from the literature. - The experimental results show that the two complete representations have higher expressive power while maintaining similar efficiency. ### Conclusions - This paper solves the incompleteness problem of barrier certificate synthesis on unbounded domains by introducing the homogenization technique and the new definition of non - polynomial barrier certificates. - The experimental results show that the new method can identify barrier certificates more effectively while maintaining computational efficiency. ### Key Concepts - **Barrier Certificate**: A real - valued function whose zero - sub - level set serves as a differential invariant, separating the initial state set and the unsafe region. - **Homogenization**: An optimization technique that transforms an unbounded optimization problem into a bounded one. - **Semidefinite Programming**: An optimization technique for solving polynomial optimization problems. Through these contributions, this paper provides a new and more effective solution for system safety verification on unbounded domains.

On Completeness of SDP-Based Barrier Certificate Synthesis over Unbounded Domains

Generalizing SDP-Based Barrier Certificate Synthesis to Unbounded Domains by Dropping Archimedean Condition.

Barrier certificates revisited

Synthesizing Invariant Barrier Certificates via Difference-of-Convex Programming

A New Barrier Certificate for Safety Verification of Hybrid Systems

Data-driven verification and synthesis of stochastic systems via barrier certificates

Compositional Construction of Control Barrier Certificates for Large-Scale Stochastic Switched Systems

From Small-Gain Theory to Compositional Construction of Barrier Certificates for Large-Scale Stochastic Systems

Synthesizing Barrier Certificates of Neural Network Controlled Continuous Systems Via Approximations.

Probabilistic Safety Verification of Stochastic Hybrid Systems Using Barrier Certificates.

Exponential-Condition-Based Barrier Certificate Generation for Safety Verification of Hybrid Systems

A New Safety Criterion for Dynamics Systems by Barrier Function Based on Forward Invariant Set

Darboux-type barrier certificates for safety verification of nonlinear hybrid systems.

Encoding Inductive Invariants As Barrier Certificates: Synthesis Via Difference-of-convex Programming.

A Linear Programming Relaxation Based Approach for Generating Barrier Certificates of Hybrid Systems.

Co-Buchi Barrier Certificates for Discrete-time Dynamical Systems

Compositional synthesis of control barrier certificates for networks of stochastic systems against ω -regular specifications

Closure Certificates

A Semi-Algebraic Framework for Verification and Synthesis of Control Barrier Functions

A Novel Approach for Solving the BMI Problem in Barrier Certificates Generation

Verification of Quantum Systems using Barrier Certificates