1: \begin{abstract}

2: Reconfigurable Intelligent Surfaces (RISs) have emerged as a promising

3: technology for enhancing satellite communication systems by manipulating

4: the phase of electromagnetic waves. This study addresses optimising

5: phase shift values ( $\phi_{R}$) in RIS networks under ideal and non-ideal conditions. For ideal scenarios, we introduce a novel approach that simplifies the traditional optimisation methods for

6: determining the optimal value. Leveraging trigonometric identities

7: and the law of cosines, we create a more tractable formulation for

8: the received power that allows for efficient optimisation of $\phi_{R}$.

9: However, practical applications often grapple with non-ideal conditions.

10: These conditions can introduce phase errors, significantly affecting

11: the received signal and overall system performance. To accommodate

12: these complexities, our optimisation framework extends to include

13: phase errors, which are modelled as a uniform distribution. To solve

14: this optimisation problem, we propose a stochastic framework that

15: harnesses the Monte Carlo method to consider all plausible phase error

16: values. Furthermore, we employ the Broyden\textendash Fletcher\textendash Goldfarb\textendash Shanno

17: (BFGS) algorithm, an iterative method known for its efficacy. This

18: algorithm systematically updates $\phi_{R}$ values, incorporating

19: the gradient of the objective function and Hessian matrix approximations.

20: The algorithm also monitors convergence to balance computational complexity

21: and accuracy. The results of the theoretical analysis are illustrated

22: with several examples. As herein demonstrated, the proposed solution

23: offers profound insights into the impacts of phase errors on RIS system

24: performance. It also unveils innovative optimisation strategies for

25: real-world satellite communication scenarios under diverse conditions.

26: \end{abstract}

27: