Novel Stability Criteria for Discrete and Hybrid Systems via Ramanujan Inner Products

Shyam Kamal, Sunidhi Pandey, Thach Ngoc Dinh · 2025

Read Novel Stability Criteria for Discrete and Hybrid Systems via Ramanujan Inner Products by Shyam Kamal, Sunidhi Pandey, Thach Ngoc Dinh (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-13690.txt.

Introduction

Shyam Kamal and 3 co-authors introduce a Ramanujan inner product as a new yardstick for stability of discrete and hybrid systems. Classic checks rely on Euclidean distance, but number-based sums carry structure that standard tools ignore. Their framework builds a norm and conditions directly from Ramanujan summation and its arithmetic links.

New epsilon-delta style criteria state when a system stays stable under the Ramanujan measure, with stronger robustness guarantees claimed than the standard route. All main results are proven in full. Numerical simulations on sample systems back the theory, suggesting number traits of a system's dynamics can inform how engineers judge its stability.

Catalog record

CREATORShyam Kamal, Sunidhi Pandey, Thach Ngoc Dinh
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE34 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-13690

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Novel Stability Criteria for Discrete and Hybrid Systems via Ramanujan Inner Products Shyam Kamal, Sunidhi Pandey, Thach Ngoc Dinh Abstract This paper introduces a Ramanujan inner product and its corresponding norm, establishing a novel framework for the stability analysis of hybrid and discrete-time systems as an alternative to traditional Euclidean metrics. We establish new ϵ \epsilon - δ \delta stability conditions that utilize the unique properties of Ramanujan summations and their relationship with number-theoretic concepts. The proposed approach provides enhanced robustness guarantees and reveals fundamental connections between system stability and arithmetic properties of the system dynamics. Theoretical results are rigorously proven, and simulation results on numerical examples are presented to validate the efficacy of the proposed approach. I Introduction Stability analysis is a cornerstone of control theory, with Lyapunov’s direct method, input-to-state stability (ISS), and small-gain theorems serving as fundamental tools for characterizing system behavior [ 5 , 2 ] . Traditional stability criteria predominantly employ Euclidean inner products and norms, which measure energy and convergence in a purely geometric sense. While powerful, these metrics may overlook structural properties of signals and dynamics that arise in modern control systems, especially those involving hybrid switching, sampled-data control, or discrete scheduling mechanisms. In such systems, the underlying state evolution may possess arithmetic or combinatorial characteristics, such as periodic switching sequences or residue-class-dependent dynamics, which are not optimally captured by Euclidean norms. Motivated by this observation, we propose a novel framework for stability analysis based on Ramanujan inner products and associated Ramanujan norms. Ramanujan sums c_{q}(n) , introduced by Srinivasa Ramanujan in 1918 [ 4 ] , are arithmetic sums of exponential functions that capture divisibility and residue-class information. These sums have been widely studied in number theory and have found applications in signal processing [ 7 , 1 ] , where they are used to reveal hidden periodicities and quasi-periodic structures in integer-indexed data. The key idea of this paper is to leverage Ramanujan sums to construct a new inner product space for system trajectories, thereby defining a norm that is sensitive to arithmetic structure. This perspective is particularly relevant for hybrid and discrete-time systems, where the behavior of solutions is strongly influenced by discrete jumps and sampling instants. By employing a Ramanujan-inspired inner product, we obtain stability criteria that are tuned to periodic, almost-periodic, and arithmetic progression-based dynamics. Our results include the Ramanujan stability criterion for discrete-time and hybrid systems, giving small-gain-like conditions expressed in the Ramanujan metric. The contributions of this paper are threefold: (a) we introduce the Ramanujan inner product and associated norm into the control-theoretic context, (b) we develop Lyapunov-based and ISS-like stability conditions using this norm, and (iii) we demonstrate that systems that are Ramanujan stable enjoy enhanced robustness against structured disturbances that exhibit arithmetic patterns, such as periodic packet drops or cyclostationary noise. This paper is organized as follows. Section II presents preliminaries. Section III establishes notation for Ramanujan inner products and norms. Section IV presents Ramanujan stability criterion for discrete-time nonlinear systems with structured arithmetic disturbances. Section V extends these results to hybrid systems and also establishes Ramanujan-Lyapunov stability conditions for hybrid systems. Section VI is dedicated to the validation of the proposed formulation using simulation results. Finally, the conclusion of this work is stated. II Mathematical Preliminaries II-A Classical Stability Notions II-A 1 Stability of Discrete-Time Nonlinear Systems Consider the following nonlinear discrete-time system x_{k+1}=f(x_{k},u_{k}),\qquad k\in\mathbb{N} (1) where x_{k}\in\mathcal{X}\subseteq\mathbb{R}^{n} is the state vector, u_{k}\in\mathcal{U}\subseteq\mathbb{R}^{m} is the input, and f:\mathcal{X}\times\mathcal{U}\to\mathcal{X} is a nonlinear state transition map assumed to be locally Lipschitz. The equilibrium point be x^{\star}=0 , satisfying f(0,0)=0 . Definition 1 . [ 3 ] Consider the dynamics ( 1 ). The zero solution is said to be stable if for every \epsilon>0 , there exists a \delta>0 such that \|x_{0}\|_{2}<\delta\quad\Rightarrow\quad\|x_{k}\|_{2}<\epsilon,\qquad\forall k\in\mathbb{N}. If, in addition to being stable, the state trajectory converges to the equilibrium as k\to\infty , i.e., \lim_{k\to\infty}x_{k}=0, then the equilibrium point is said to be asymptotically stable. II-A 2 Stability of Nonlinear Hybrid Systems Consider the following nonlinear hybrid system \mathcal{S}:\dot{x}=f(x,u),~x\in C,\qquad x^{+}=g(x,u),~x\in D (2) where x\in\mathcal{X}\subseteq\mathbb{R}^{n} denotes the system state and u\in\mathcal{U}\subseteq\mathbb{R}^{m} represents the input. The set C\subseteq\mathcal{X} is called the flow set and characterizes the states where the system evolves according to the continuous-time dynamics governed by the locally Lipschitz function f:\mathcal{X}\times\mathcal{U}\to\mathbb{R}^{n} . The set D\subseteq\mathcal{X} is the jump set which specifies the states where a discrete transition occurs, with the jump map g:\mathcal{X}\times\mathcal{U}\to\mathcal{X} determining the state immediately after the jump. Solutions of the hybrid system are defined on a hybrid time domain \mathcal{T}\subset\mathbb{R}_{\geq 0}\times\mathbb{N} , where each solution trajectory x(t,j) is parameterized by the pair (t,j) . Here t represents the usual continuous time during flows, while j counts the number of discrete jumps that have occurred up to time t . Here, The classical notion of stability is defined with respect to an equilibrium point x^{\star}=0 , and is expressed using the Euclidean norm. Definition 2 . [ 2 ] Consider the hybrid system ( 2 ). The equilibrium point x^{\star}=0 is said to be stable if for every \epsilon>0 , there exists a \delta>0 such that \|x(0,0)\|_{2}<\delta\quad\Rightarrow\quad\|x(t,j)\|_{2}<\epsilon,\quad\forall(t,j)\in\mathcal{T}. This definition naturally extends the well-known concept of Lyapunov stability from continuous-time systems to hybrid systems by explicitly considering the hybrid time domain. If, in addition, the solution satisfies \lim_{(t,j)\to\infty}x(t,j)=0 , the equilibrium is said to be asymptotically stable. III Ramanujan Inner Product In this section, we define the Ramanujan inner product and associated norm based on Ramanujan’s sum. Definition 3 (Ramanujan Sum) . For a positive integer q and an integer n , the Ramanujan sum is defined as c_{q}(n)=\sum_{\begin{subarray}{c}k=1\\ (k,q)=1\end{subarray}}^{q}e^{2\pi i\frac{k}{q}n} (3) where (k,q) denotes the greatest common divisor (gcd) of k and q . In words, the sum is taken over all integers k between 1 and q that are coprime to q . The function c_{q}(n) is periodic in n with period q , and it is always a real-valued number despite being expressed as a sum of complex exponentials. Definition 4 . [ 8 ] [Ramanujan Inner Product] Let a=\{a_{n}\} and b=\{b_{n}\} be two sequences in \ell^{2}(\mathbb{Z}) , i.e., \sum_{n=-\infty}^{\infty}|a_{n}|^{2}<\infty and \sum_{n=-\infty}^{\infty}|b_{n}|^{2}<\infty . The Ramanujan inner product between a and b is defined as \displaystyle\langle a,b\rangle_{\mathrm{R}}=\lim_{Q\to\infty}\frac{1}{Q}\sum_{q=1}^{Q}\frac{1}{q}\sum_{r=0}^{q-1}c_{q}(r)\sum_{\begin{subarray}{c}n\in\mathbb{Z}\\ n\equiv r\,(\mathrm{mod}\,q)\end{subarray}}a_{n}\,\overline{b_{n}}. (4) where c_{q}(r) is the classical Ramanujan sum. This definition groups the terms of each sequence according to their residue class modulo q, then weights them with the arithmetic kernel c_{q}(r) . In this way, the inner product performs a number-theoretic correlation, analogous to Fourier inner products but aligned with arithmetic periodicities rather than trigonometric frequencies. The associated Ramanujan norm of a sequence a is defined in the usual way by \|a\|_{\mathrm{R}}=\sqrt{\langle a,a\rangle_{\mathrm{R}}}, (5) which provides a measure of the “energy” of the sequence in the Ramanujan sense. III-A Properties of the Ramanujan Norm Theorem 1 . Let \mathcal{S} be the class of complex sequences a=(a_{n})_{n\in\mathbb{Z}} that admit a Ramanujan expansion a(n)=\sum_{d\geq 1}\alpha_{d}\,c_{d}(n) , with coefficient sequence (\alpha_{d})_{d\geq 1} satisfying the weighted square–summability condition \sum_{d\geq 1}\varphi(d)\,|\alpha_{d}|^{2}<\infty, where c_{d}(n) denotes the classical Ramanujan sum and \varphi denotes Euler’s totient function. For a,b\in\mathcal{S} with coefficients (\alpha_{d}) and (\beta_{d}) , define the Ramanujan inner product by Parseval-like identity \langle a,b\rangle_{\mathrm{R}}:=\sum_{d\geq 1}\varphi(d)\,\alpha_{d}\,\overline{\beta_{d}}. The associated function \|a\|_{R}:=\sqrt{\langle a,a\rangle_{R}} defines a norm

Read the complete text