MIMO Systems Aided by Microwave Linear Analog Computers: Capacity-Achieving Architectures with Reduced Circuit Complexity
Matteo Nerini, Bruno Clerckx · 2025
Read MIMO Systems Aided by Microwave Linear Analog Computers: Capacity-Achieving Architectures with Reduced Circuit Complexity by Matteo Nerini, Bruno Clerckx (2025) 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-2506-15052.txt.
Introduction
Matteo Nerini and Bruno Clerckx tackle the cost of very large antenna arrays in a 2025 paper on wireless systems. Future networks may use huge arrays, a step beyond today's massive MIMO systems. One way to cut cost is to shape signals with analog circuits at microwave frequencies, instead of doing all the work in digital form.
Fully linked analog networks can reach full capacity, but the parts they need grow with the square of the array size. The authors use graph theory to design a sparser family of networks, which they call stem-connected. Their design still reaches full capacity, yet the part count grows only in step with the array size. Math and computer tests both support the claim. The result points to large arrays that are cheaper to build.
Catalog record
| CREATOR | Matteo Nerini, Bruno Clerckx |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 78 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-15052
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
MIMO Systems Aided by Microwave Linear Analog Computers: Capacity-Achieving Architectures with Reduced Circuit Complexity Matteo Nerini, Bruno Clerckx Abstract To meet the demands of future wireless networks, antenna arrays must scale from massive multiple-input multiple-output (MIMO) to gigantic MIMO , involving even larger numbers of antennas. To address the hardware and computational cost of gigantic MIMO , several strategies are available that shift processing from the digital to the analog domain. Among them, microwave linear analog computers (MiLACs) offer a compelling solution by enabling fully analog beamforming through reconfigurable microwave networks. Prior work has focused on fully-connected MiLACs , whose ports are all interconnected to each other via tunable impedance components. Although such MiLACs are capacity-achieving, their circuit complexity, given by the number of required impedance components, scales quadratically with the number of antennas, limiting their practicality. To solve this issue, in this paper, we propose a graph theoretical model of MiLAC facilitating the systematic design of lower-complexity MiLAC architectures. Leveraging this model, we propose stem-connected MiLACs as a family of MiLAC architectures maintaining capacity-achieving performance while drastically reducing the circuit complexity. Besides, we optimize stem-connected MiLACs with a closed-form capacity-achieving solution. Our theoretical analysis, confirmed by numerical simulations, shows that stem-connected MiLACs are capacity-achieving, but with circuit complexity that scales linearly with the number of antennas, enabling high-performance, scalable, gigantic MIMO . I Introduction As wireless systems evolve toward sixth-generation (6G) , the demand for ultra-high data rates, low latency, and pervasive connectivity continues to increase, pushing beyond the capabilities of fifth-generation (5G) technologies. While massive multiple-input multiple-output (MIMO) systems, typically including tens of antennas, have proven successful in 5G [ 1 , 2 ] , 6G requires a further substantial increase in the number of antennas. This emerging paradigm, denoted as gigantic MIMO , envisions MIMO architectures with antenna arrays scaling to the order of thousands of antennas [ 3 , 4 ] . Such extreme scaling offers significant enhancements in spatial multiplexing and beamforming resolution, which are critical to meet 6G targets in the anticipated upper mid-band spectrum (7-24 GHz). However, traditional digital MIMO architectures are not suited to this scale due to the prohibitive cost, power consumption, and complexity of assigning one radio frequency (RF) chain per antenna, each requiring high-resolution analog-to-digital converters (ADCs) / digital-to-analog converters (DACs) and mixers. This challenge has steered research efforts toward analog-domain solutions, which can provide more scalable, cost-effective, and energy-efficient alternatives to fully digital MIMO architectures. Hybrid digital-analog MIMO architectures originally emerged to reduce the demand of RF chains while preserving high performance through the use of analog phase shifters [ 5 , 6 ] . Another technology classically used to change the radiation pattern of an antenna array operating in the analog domain is reconfigurable antennas [ 7 ] . The idea of reconfigurable antennas traces back to the 1930s, when they were realized through mechanically moving parts. Since the 1990s, reconfigurable antennas have been realized by interconnecting parts of the antenna through tunable components, such as varactors, mechanical RF switches, or semiconductor RF switches. To synergize the benefits of hybrid beamforming and the recent advances in reconfigurable antennas, the tri-hybrid MIMO architecture has been recently proposed [ 8 ] , where the signal is processed in three stages: digitally, in the analog domain by phase shifts, and in the electromagnetic (EM) domain by reconfigurable antennas. Beamforming in the analog domain can also be achieved through the use of intelligent reflecting surfaces (IRSs) , also known as reconfigurable intelligent surfaces (RISs) . A RIS is a surface made of a multitude of elements with reconfigurable scattering capabilities, which can be integrated at the transceiver device to reconfigure its effective radiation pattern [ 9 , 10 ] . To enhance the flexibility of RIS , the concept of beyond diagonal RIS (BD-RIS) has emerged, where the RIS elements are interconnected to each other through tunable interconnections [ 11 , 12 , 13 ] . Furthermore, the use of multiple stacked RISs has been explored to increase adaptability, giving rise to stacked intelligent metasurface (SIM) technology [ 14 , 15 ] . While [ 9 ] - [ 15 ] have considered reconfigurable surfaces that passively manipulate the incident EM signal, dynamic metasurface antennas (DMAs) have emerged to actively manipulate how the EM signal is radiated and received [ 16 ] . Beyond diagonal DMAs (BD-DMAs) have been recently proposed as an advanced version of DMAs , where tunable interconnections between the meta-atoms enable reconfigurable coupling and hence increase the performance [ 17 ] . While analog-domain solutions are particularly appealing for gigantic MIMO beamforming, they commonly suffer from two limitations. First, they typically still rely on digital pre-processing at the transmitter and digital post-processing at the receiver, which could require large hardware and computational complexity in gigantic MIMO . Second, beamforming in the analog domain generally suffers from limited reconfigurability and cannot achieve the same performance as fully digital beamforming. To address both limitations, microwave linear analog computer (MiLAC) has recently emerged as a promising approach for gigantic MIMO systems [ 18 , 19 ] . A MiLAC is a multiport microwave network made of tunable impedance components, having input and output ports. At the input ports, a signal is applied, which is processed by the MiLAC in the analog domain as it propagates within the microwave network. Thus, at the output ports, the processed signal can be read. Exploiting these analog processing capabilities, we can use a MiLAC at the transmitter side to precode the transmitted symbols by feeding them to the MiLAC input ports through RF chains, and connecting the MiLAC output ports to the transmitting antennas. Similarly, at the receiver side, we can use a MiLAC to combine the received signal by connecting the receiving antennas to the MiLAC input ports and sampling the signal at the MiLAC output ports via RF chains. By precoding and combining the signal fully in the analog domain, MiLAC offers five distinct advantages for gigantic MIMO in terms of hardware and computational complexity [ 19 ] . First , it offers the same flexibility as digital beamforming, and hence maximum performance. Second , it only requires as many RF chains as the transmitted symbols, i.e., data streams. Third , since the signals carried and sampled at the RF chains are the actual symbols, low-resolution ADCs and DACs are sufficient. Fourth , it does not require any computation at each symbol time since the symbols are precoded and combined fully in the analog domain. Fifth , it can perform zero-forcing beamforming (ZFBF) and minimum mean square error (MMSE) combining with significantly reduced computational complexity, as shown in [ 19 ] . In detail, ZFBF and MMSE combining can be executed with a computational complexity growing quadratically with the number of antennas, rather than cubically as in digital MIMO [ 19 ] . To analyze the fundamental performance limits of MiLAC , in [ 18 ] and [ 19 ] it has been assumed that its tunable impedance components can be arbitrarily reconfigured. A more practical design has been considered in [ 20 ] , where a MiLAC made of lossless and reciprocal tunable impedance components has been considered. Interestingly, even under