User-Centric Cooperative Transmission in Intelligent Reflecting Surface Assisted Networks

The intelligent reflecting surface (IRS) exhibits no- table benefits in flexibility, low power consumption, and cost- effectiveness, offering a promising prospect for extensive deploy- ment. However, the performance of the IRS-assisted networks can be hindered by double path-loss fading, resulting in limited serv- ing region and IRS selection constraints. Therefore, it becomes imperative to propose a more rational network architecture that surpasses the current limitations of IRS. This paper proposes a user-centric IRS group cooperative serving scheme, which integrates the benefits of user-side IRS selection and multi- IRS cooperative transmission, and establishes a comprehensive system-level performance analysis framework under the proposed scheme via stochastic geometry. Specifically, IRS cooperative group is decided by an adjustable radius
C
centered on each user and the IRSs within this circle collaborate to provide services to the user. The average signal strength under cooperative transmission approximated via moment matching, where we propose a general framework to calculate the high order moments of combined channel of IRS-reflected paths as a random sum. In addition, our proposed architecture exacerbates the irregularity of network topology, and more accurate interference Laplace functions by quantifying the irregular interference partitioning is obtained. Numerical results demonstrate that under our proposed user-centric IRS group cooperative serving scheme, the coverage probability can be improved by 3.66x when signal-to- interference-plus-noise ratio (SINR) equals to 0dB, compared to traditional network with no IRSs
.