On the solutions of second order difference equations with variable coefficients

Shirali Kadyrov,Alibek Orynbassar

2024-02-02

Abstract:In this article we study solutions to second order linear difference equations with variable coefficients. Under mild conditions we provide closed form solutions using finite continued fraction representations. The proof of the results are elementary and based on factoring a quadratic shift operator. As an application, we obtain two new generalized continued fraction formulas for the mathematical constant $\pi^2$.

Number Theory

What problem does this paper attempt to address?

The problem that this paper attempts to solve is about the solution method of second - order linear difference equations in the case of variable coefficients. Specifically, the authors focus on difference equations in the following form:

\[ y_n - b_n y_{n - 1} - a_n y_{n - 2}=f_n, \quad n\geq1 \]

Here, the coefficients \( a_n \) and \( b_n \) are variables in the complex number field. When the coefficients are constants, the expression of the solution can be found by solving the characteristic equation. However, when the coefficients are variables, the form of the solution becomes complex, and the existing research methods provide limited insights.

### Main contributions
1. **Closed - form solution**: The authors propose a method using finite continued fractions and give a closed - form solution to the above - mentioned difference equation. A closed - form solution refers to an explicit mathematical expression that can directly provide the solution at any discrete time point without the need for iterative or recursive calculations.

2. **Auxiliary sequences**: By introducing two auxiliary sequences \( (c_n) \) and \( (d_n) \) that satisfy the following conditions:
\[
c_n + d_{n + 1}=b_n, \quad c_n d_n=-a_n
\]
the authors derive the expression of the solution:
\[
y_n=\sum_{i = 1}^{n + 1}\left(\sum_{j = 1}^{i - 1}\frac{f_j}{\prod_{k = j + 1}^{i - 1}c_k}+(y_0 - d_1 y_{- 1})\prod_{k = 1}^{i - 1}c_k\right)\prod_{j = i + 1}^{n + 1}d_j+y_{- 1}\prod_{j = 1}^{n + 1}d_j, \quad \forall n\geq1
\]

3. **Applications**: As an application, the authors use this method to verify two new conjectures about the mathematical constant \(\pi^2\) obtained from the website www.ramanujanmachine.com and derive two new generalized continued fraction formulas:
\[
\frac{8}{\pi^2}=b_0+\frac{a_1}{b_1+\frac{a_2}{b_2+\cdots}}
\]
where \( b_n = 3n(n + 1)+1 \) and \( a_n=-(2n - 1)n^3 \).

\[
\frac{18}{\pi^2}=b_0+\frac{a_1}{b_1+\frac{a_2}{b_2+\cdots}}
\]
where \( b_n=n(5n + 6)+2 \) and \( a_n=-4n^4 + 2n^3 \).

### Conclusion
This paper successfully solves the problem of the closed - form solution of second - order linear difference equations in the case of variable coefficients by introducing auxiliary sequences and the continued fraction representation method. This method not only provides a new tool for the study of difference equations but also reveals the interesting connection between difference equations and transcendental numbers. Future research can further explore the application of this method in higher - order difference equations and its applicable conditions.