[Paper Review] On the stability of dual scattering channel schemes

Steffen Hein|ArXiv.org|May 6, 2004

Electromagnetic Simulation and Numerical Methods8 references3 citations

TL;DR

This paper establishes the unconditional stability of dual scattering channel (DSC) schemes by introducing the concept of α-passivity, a generalized energy-like contraction property. It proves that DSC processes generated by α-passive reflection and connection maps remain uniformly bounded under finite excitation, extending TLM's unconditional stability to nonlinear and non-quadratic energy forms.

ABSTRACT

Dual scattering channel (DSC) schemes generalize Johns' TLM algorithm in replacing transmission lines with abstract scattering channels in terms of paired distributions. A well known merit of TLM schemes is unconditional stability, a property that is commonly drawn upon the passivity of linear transmission line networks. So the question arises, if DSC algorithms remain stable in a neat sense. It is shown that a large class of alpha-passive processes are in fact unconditionally stable. The analysis applies to TLM and DSC schemes alike and includes non-linear situations.

Motivation & Objective

• To determine whether dual scattering channel (DSC) schemes, as generalizations of the TLM method, retain unconditional stability.
• To extend the stability analysis of TLM—based on passivity of linear transmission lines—to a broader class of DSC algorithms with abstract scattering channels.
• To formalize stability in terms of a limiting functional α that characterizes contraction properties without requiring quadratic or linear energy forms.
• To provide a general framework applicable to both linear and nonlinear DSC systems, ensuring boundedness under finite excitation.
• To establish that α-passivity of reflection and connection maps is sufficient for uniform boundedness (i.e., stability) of the resulting DSC process.

Proposed method

• Introduces α-passivity as a generalized stability criterion based on contraction with respect to a non-negative limiting functional α.
• Defines causal functions on time-ordered domains I and J, modeling the temporal evolution of DSC processes.
• Uses the space of DSC processes E = {(f,g) ∈ (L²)^H | f and g satisfy switching conditions at τ/2 intervals).
• Derives recurrence relations (12) that couple reflection and connection steps via time-shifted operators T_{-τ/2} and maps F_R, F_C.
• Applies the composition of operators F_C ∘ T_{-τ/2} ∘ F_R ∘ T_{-τ/2} to model the evolution of the process h₁, showing it is α-passive.
• Leverages the fact that products of α-passive operators remain α-passive (Proposition 3.1), enabling inductive stability proof via Theorem 2.1.

Experimental results

Research questions

• RQ1 Can the unconditional stability of TLM schemes be extended to DSC schemes that generalize transmission lines using abstract scattering channels?
• RQ2 Under what conditions on the reflection and connection maps is a DSC process guaranteed to remain uniformly bounded?
• RQ3 Is there a generalized stability criterion that applies beyond quadratic energy forms and includes nonlinear systems?
• RQ4 Can the stability of DSC schemes be proven using a functional α that measures any conserved or dissipative quantity, not necessarily energy?
• RQ5 Does α-passivity of the reflection and connection maps ensure boundedness of the DSC process under finite excitation?

Key findings

• A large class of α-passive processes, including nonlinear and non-quadratic systems, are unconditionally stable.
• The DSC process generated by α-passive reflection and connection maps remains uniformly bounded under any finite excitation.
• The composition F_C ∘ T_{-τ/2} ∘ F_R ∘ T_{-τ/2} is α-passive if F_R and F_C are α-passive, ensuring stability propagation.
• The proof relies on the causality of the maps and the contraction property of α, which generalizes energy conservation or loss.
• Stability is preserved even when the functional α is not quadratic, allowing for diverse physical interpretations (e.g., temperature, velocity).
• The result provides a constructive criterion: if the reflection and connection maps conserve or dissipate a quantity measured by α, the DSC process is stable.