Low-Complexity Planar Beyond-Diagonal RIS Architecture Design Using Graph Theory

Matteo Nerini, Zheyu Wu · 2026

Read Low-Complexity Planar Beyond-Diagonal RIS Architecture Design Using Graph Theory by Matteo Nerini, Zheyu Wu (2026) 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-2601-03831.txt.

Introduction

Matteo Nerini and three co-authors use graph theory to design smarter reconfigurable intelligent surfaces, panels that steer wireless signals by tuning their elements. Beyond-diagonal surfaces interconnect elements through tunable components for extra flexibility, but too many interconnections force complex multi-layer circuit boards that are hard to fabricate. The letter characterizes which interconnection patterns fit on a double-layer board, calling them planar-connected architectures.

Among those buildable patterns, the authors identify the ones with the most degrees of freedom, which should perform best under practical constraints. Graph theory gives the characterization: the wiring pattern is a graph, and planarity decides what a two-layer board can host. For future wireless networks counting on such surfaces, designs that respect fabrication limits from the start are the ones that will actually ship.

Catalog record

CREATORMatteo Nerini, Zheyu Wu
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE26 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-03831

Browse the full library · About the Tower of Babel

Full text

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

Low-Complexity Planar Beyond-Diagonal RIS Architecture Design Using Graph Theory Thanks: M. Nerini, Z. Wu, and B. Clerckx are with the Department of Electrical and Electronic Engineering, Imperial College London, London SW7 2AZ, U.K. (e-mail: {m.nerini20, zheyu.wu, b.clerckx}@imperial.ac.uk). Thanks: S. Shen is with the State Key Laboratory of Internet of Things for Smart City and Department of Electrical and Computer Engineering, University of Macau, Macau, China. (e-mail: [email protected]) Matteo Nerini, Zheyu Wu Abstract Reconfigurable intelligent surfaces (RISs) enable programmable control of the wireless propagation environment and are key enablers for future networks. Beyond-diagonal RIS (BD-RIS) architectures enhance conventional RIS by interconnecting elements through tunable impedance components, offering greater flexibility with higher circuit complexity. However, excessive interconnections between BD-RIS elements require multi-layer printed circuit board (PCB) designs, increasing fabrication difficulty. In this letter, we use graph theory to characterize the BD-RIS architectures that can be realized on double-layer PCBs , denoted as planar-connected RISs. Among the possible planar-connected RISs, we identify the ones with the most degrees of freedom, expected to achieve the best performance under practical constraints. I Introduction Reconfigurable intelligent surface (RIS) is a technology designed to control wireless communication environments [ 1 , 2 , 3 ] . An RIS is a surface consisting of multiple elements with reconfigurable reflecting properties that can be deployed in the propagation environment and used to shape signal propagation. By steering the incident signal towards the intended direction and suppressing interference, RIS offers a promising green solution for next-generation communication networks such as 6G. An RIS has been commonly implemented by connecting each RIS element to a tunable impedance, and individually controlling the reflection coefficient of each element. The conventional RIS architecture, referred to as single-connected RIS [ 4 ] , has been generalized by interconnecting the RIS elements to each other through tunable impedance components in [ 4 ] , leading to the general family of beyond-diagonal RIS (BD-RIS). When all RIS elements are interconnected to each other, the resulting architecture is denoted as fully-connected RIS [ 4 ] . To trade flexibility and circuit complexity, the group-connected RIS was also proposed, where the RIS elements are divided into groups of equal size and are interconnected with each other if and only if within the same group [ 4 ] . The presence of interconnections enables BD-RIS to have additional capabilities than conventional RIS, such as the possibility to achieve full-space coverage [ 5 , 6 ] . A fundamental question in BD-RIS is whether the additional circuit complexity justifies the performance improvement. Therefore, it is crucial to design BD-RIS architectures that have low circuit complexity and achieve high performance gains at the same time. This can be done by modeling BD-RIS architectures as graphs, and using graph theory to gain insights into their circuit complexity and performance [ 7 ] . Following this graph theoretical modeling, two BD-RIS architectures have been proposed in [ 7 ] , i.e., the forest- and tree-connected RISs, that are proven to be the least complex architectures achieving the same performance of group- and fully-connected RISs in single-user systems. With a focus on single-user systems, the fundamental limits of the trade-off between circuit complexity and performance achievable with BD-RIS have been derived in closed-form in [ 8 ] , and also extended to dual-polarized systems in [ 9 ] . Further literature has also extended the results of [ 7 ] to multi-user systems, proposing stem-connected RISs [ 10 , 11 ] and band-connected RISs [ 12 ] as BD-RISs achieving the same performance as fully-connected RISs with much reduced circuit complexity. Previous research on the design of BD-RIS architectures aimed at minimizing the circuit complexity of the BD-RIS measured by the number of tunable impedance components [ 7 ] - [ 12 ] . However, when the number of interconnections is too large, they inevitably cross each other, requiring a multi-layer printed circuit board (PCB) design with vias to realize those crossings. This increases the implementation cost, time, and probability of fabrication failure, posing a practical challenge. A significantly more convenient approach is to use a double-layer PCB , where one layer has the BD-RIS interconnections and the other serves as the ground plane. In this work, we identify the requirements that a BD-RIS architecture must satisfy to be realizable on a double-layer PCB , in which one layer is entirely reserved for the ground plane. Furthermore, we characterize the BD-RIS architectures that include the largest number of tunable components while remaining compatible with double-layer PCB implementation. Specifically, the contributions of this letter are as follows. First , we use graph theory to characterize the BD-RIS architectures that can be implemented in a double-layer PCB , which we refer to as planar-connected RISs. Second , we examine existing BD-RIS architectures to determine whether they are planar-connected. Third , we characterize the planar-connected RISs with the most degrees of freedom, denoted as maximal-planar-connected RISs, which are expected to achieve the best performance with the constraint of double-layer PCB . II BD-RIS Graph Theoretical Model Consider an N -element BD-RIS, where the RIS elements are connected to an N -port reconfigurable microwave network, commonly represented through its scattering matrix \boldsymbol{\Theta}\in\mathbb{C}^{N\times N} [ 4 ] . According to multiport network theory [ 13 , Chapter 4] , \boldsymbol{\Theta} can be expressed as a function of the admittance matrix of the BD-RIS \mathbf{Y}\in\mathbb{C}^{N\times N} as \boldsymbol{\Theta}=\left(\mathbf{I}+Z_{0}\mathbf{Y}\right)^{-1}\left(\mathbf{I}-Z_{0}\mathbf{Y}\right), (1) where Z_{0} is the reference impedance, set to Z_{0}=50\;\Omega . In this work, we consider the admittance matrix representation since it can be directly linked to the tunable impedance (or admittance) components implementing the BD-RIS, assumed to be reciprocal. Specifically, denoting as Y_{n} the tunable admittance connecting the n th RIS element to ground and as Y_{n,m}=Y_{m,n} the tunable admittance interconnecting the n th and the m th RIS elements, we have \left[\mathbf{Y}\right]_{n,m}=\begin{cases}-Y_{n,m}&n\neq m\\ Y_{n}+\sum_{k\neq n}Y_{n,k}&n=m\end{cases}, (2) for n,m=1,\ldots,N . From ( 2 ), we observe that if the n th and m th RIS elements are not interconnected, i.e., Y_{n,m}=0 , the (n,m) th entry of the admittance matrix \mathbf{Y} is zero, i.e., [\mathbf{Y}]_{n,m}=0 . To model the general circuit topology of a BD-RIS, we resort to graph theory by briefly recalling the model developed in [ 7 ] . Following this model, the circuit topology of any BD-RIS is represented through a graph \mathcal{G}=(\mathcal{V},\mathcal{E}) , where \mathcal{V} and \mathcal{E} are the vertex set and the edge set of \mathcal{G} , respectively. The vertex set is given by the indexes of the RIS elements, i.e. \mathcal{V}=\{v_{1},v_{2},\ldots,v_{N}\} , and the edge set is given by \mathcal{E}=\left\{\left(v_{n},v_{m}\right)|\>v_{n},v_{m}\in\mathcal{V},\>Y_{n,m}\neq 0,\>n\neq m\right\}, (3) which means that vertices v_{n} and v_{m} are connected by an edge if and only if there is a tunable admittance interconnecting the n th and m th RIS elements. III Planar-Connected RIS In this section, we exploit the graph theoretical model of BD-RIS to identify the BD-RIS architectures that are implementable in a double-layer PCB , where one layer is occupied by the interconnections and the other is the ground. Such architectures are practically useful since having a PCB with more layers increases the implementation complexity. We begin by observing that a BD-RIS can be implemented into a double-layer PCB when its

Read the complete text