Impact of Spatial Correlation on Link Selection and Delay Performance in Multi-RIS Networks

Reconfigurable Intelligent Surface (RIS) technology has emerged as a key enabler for enhancing the performance of wireless communication networks. We investigate the deployment of multiple RISs within a cellular coverage area to enable diverse link options. We analytically characterize the impact of spatial correlation among RIS elements and derive the signal-to-noise ratio (SNR) distribution for the direct link to the base station (BS) and the cascaded RIS-assisted link, under Nakagami-m fading. Building on this analysis, a comprehensive framework is developed to evaluate the average packet delay, addressing the gap in delay analysis under spatial correlation for multi-RIS systems. Moreover, we derive a closed-form condition to identify when the direct link outperforms RIS-assisted transmission, thus guiding link selection and RIS-user association based on delay performance. Numerical results highlight the substantial effect of spatial correlation on delay analysis and demonstrate that neglecting this factor can severely underestimate network performance by leading to suboptimal link selections. A high degree of correlation concentrates channel energy into a dominant eigenmode for single-user transmission, thereby enhancing the effective SNR. Furthermore, the results show that the proposed delay-based selection approach accurately evaluates feasible links with the aim of minimizing the average packet delay, achieving superior performance compared to benchmark methods.