Beam and Beyond: The Role of Reconfigurable Intelligent Surfaces (RIS) in Enhancing mmWave Performance and Phased Array Design

This article is part of a blog series called Beam and Beyond published by Mathworks on everything RF.

Modern wireless systems are pushing the limits of performance, scale, and complexity and engineers are being asked to deliver more and faster. Whether engineers are designing 5G base stations, working on satellite communications, or exploring new spectrum-sharing technologies, one increasingly critical component comes into play: phased arrays.

Phased arrays offer the ability to electronically steer beams without moving parts, enabling faster, more flexible, and more efficient wireless systems. But designing and implementing them isn’t easy. Engineers face a host of challenges from beamforming algorithms and array calibration to hardware constraints and system-level modeling.

The Beam and Beyond quarterly blog series is here to help. The blog series will break down topics shaping the future of wireless and phased array systems through practical examples and offer insights that engineers can apply directly to their work.

The Role of mmWave Frequencies in 5G and Phased Array Design

With 3GPP’s endorsement of mmWave frequencies, engineers can now fit dense antenna arrays into smaller spaces, supporting the high data rates and capacity demands of next-generation networks. But there is a catch! mmWaves are shorter wavelengths and struggle to penetrate most materials, making them reliant on line-of-sight (LOS) links. That makes coverage fragile especially in urban canyons or indoor environments. 

This is where Reconfigurable Intelligent Surfaces (RIS) comes in! In traditional wireless systems, engineers must work around challenges like multipath fading, blockages, and interference. With RIS, engineers can shape the environment to work in their favor. It’s a fundamentally new degree of freedom.

Designing Effective RIS 

This raises important questions for engineers configuring RIS to work effectively: Does the size of the RIS matter? Where should it be placed? What phase shifts are needed to properly adjust the incoming signal? 

To explore these questions, MATLAB can be used to simulate a simple wireless system and test different scenarios.

The scenario: A straightforward 2D wireless setup was created in MATLAB to investigate these factors as shown in Figure 1.

A transmitter (Tx) sends a signal toward a receiver (Rx) but is blocked out by an obstacle

A RIS panel is placed —not in the direct line of sight.

Figure 1: Wireless scenario with a RIS panel. ©2025 The MathWorks, Inc.

Key Modeling Components

1. Set Up Transmitter: Use phased.Platform to define the properties of the transmitter. Setting the transmit power to be 50 mW.

tx = phased.Transmitter('PeakPower',50e-3,'Gain',0);

2. Design the RIS: The RIS is modeled as a uniform rectangular array (URA) using phased.RectangularRIS. You can specify the number of elements, element spacing, and antenna elements.

% Setup surface

Nr = 4;

Nc = 4;

dr = 0.5*lambda;

dc = 0.5*lambda;

% construct surface

ris = phased.RectangularRIS('Size',[Nr Nc],'Spacing',[dr dc],'UnitCell',phased.IsotropicAntennaElement)

Figure 2 shows the response of the designed RIS as visualized with the pattern function:

Figure 2: 3D Pattern Response of RIS panel. ©2025 The MathWorks, Inc.

3. Simulate the Channels: Next, use phased.FreeSpace to model the line-of-sight channels between Tx–RIS and RIS–Rx paths. The reflected signal is a combination of the path loss and the phase-shifted response from each RIS element. The signal propagated is BPSK signal using the function pskmod but you could swap this out for more sophisticated modulations schemes.

4. Combine Signals at the Receiver: Finally, combine the received signals at the receiver and analyze their strength.

Simulation Results

First as a reference, the SNR for a direct path, without any obstacle would be:

With LOS (reference): SNRref = 18.5972

When RIS is added to the environment but only reflects incoming waves without phase control, the improvement is minimal: 

With RIS (untuned): SNRlosris = 18.6079

By tuning the RIS phase shifts, engineers can maximize signal power at the receiver – essentially tuning the phase shifts at each element, so the signal arrives at the receiver at the same phase. For that, engineers need channel information. In this example, a simple channel estimation was performed using a free space path loss model based on distances between the Tx, RIS and Rx. (This can be replaced for a more sophisticated channel estimation model.) When the simulation is rerun with the new phase control coefficients, the SNR is significantly better:

With RIS (tuned): SNRlosrisTuned= 28.4933

Great! Now What About Size? 

Turns out, size matters - quite a bit.

A simulation was set up to study how the data rate changes as the RIS size increases. The number of elements was varied from 50 to 500 (in steps of 50), and link performance was simulated for different RIS sizes:

Ncparam = 5:5:50;

SNRlosris_param,SNRris_param,SNRlos_param] = helperRISSimulation(xt,Ncparam,chanAPToUE,chanAPToRIS,chanRISToUE,...         ...pos_ap,pos_ris,pos_ue,v,fc,c,SNRref); % Function estimates SNR for RIS of size Ncparam. Combine signals from LOS channel and reflection from RIS

plot(Ncparam.'*10,log2(1+db2pow([SNRlosris_param;SNRris_param;SNRlos_param])));

legend('LOS+RIS','RIS','LOS');

xlabel('Number of Elements');

ylabel('Data Rate (bps/Hz)')

Figure 3: Link Performance vs No. of RIS elements. ©2025 The MathWorks, Inc. 

Figure 3 shows eye-opening results. At smaller sizes, the RIS doesn’t pull its weight, the performance barely nudges past the baseline LOS link. But once a certain threshold in element count is crossed, the link quality shoots up. Essentially, a RIS needs to reach critical mass before it starts to meaningfully boost performance.

The most fascinating bit? A large enough RIS can carry the system even without a direct line of sight. That’s powerful. It’s like having a flexible, passive repeater that can intelligently bend the signal around obstacles and the bigger it gets, the better it performs. This has big implications for indoor systems, dense urban areas, or satellite communication scenarios with intermittent LOS.

Impact of RIS Placement on Performance

But size isn’t the only lever here, placement plays a huge role too.

Another test was conducted by keeping the RIS size fixed while sweeping its position along the x-axis between the base station and the mobile user. The code for this process was straightforward.

dbrparam = 5:5:50;

SNRris_param= helperRISPlacementSimulation(xt,dbrparam,risreflector,...

   ...chanAPToUE,chanAPToRIS,chanRISToUE,pos_ap,pos_ue,v,fc,c); % Function estimates SNR for RIS at location dbrparam. 

plot(dbrparam,SNRris_param);

xlabel('RIS Position along X (m)');

ylabel('SNR (dB)')

Figure 4: SNR vs. Placement of RIS panel. ©2025 The MathWorks, Inc. 

 The resulting SNR curve told a clear story (Figure 4): The RIS is most effective when it’s placed either close to the base station or close to the user. Positioning it in the middle? Not ideal. This intuitively makes sense—placing it near the transmitter or receiver gives the system more control over how the signal is launched or received, leading to higher gains.

A small RIS in the wrong spot won’t do much. But a well-sized, strategically placed RIS? That could be the difference between a dropped call and a seamless high-speed connection.

What Was Learned Along the Way

Working with RIS simulations in MATLAB provides a clearer understanding of the nuances behind this technology. Some findings align with expectations, while others offer new insights:

• Size makes a big difference. While it was anticipated that more elements would improve performance, the data rate improvements observed after the RIS surpassed a certain size threshold highlight that it’s not just about having an RIS but having one large enough to make a difference.
• Placement is less intuitive than expected. Rather than positioning the RIS in the middle for balanced results, placing it closer to the transmitter or receiver yields better gains, an important consideration for future propagation models.
• RIS doesn’t need LOS to be helpful. A well-configured RIS can maintain the connection despite blocked line-of-sight paths, which is especially valuable when working with higher frequencies like mmWave or satellite communications, where blockages are common.

Looking Ahead: A Shift in How Signals Are Managed

RIS is unlikely to replace large antennas or beamforming in the near future, but it represents a subtle shift that could change how wireless environments are approached. Instead of combating reflections and blockages, these challenges could be leveraged as advantages.

While there is still much to learn, this simulation serves as a valuable initial step. Further exploration, including real-world deployment studies, will provide deeper insights. The idea of transforming walls, ceilings, or even satellites into programmable surfaces that guide signals precisely where needed offers an exciting vision for the future.

These and many other topics will be explored in upcoming editions of the Beam and Beyond blog series. Each quarterly installment will delve into the challenges and opportunities in phased arrays and advanced wireless systems, illustrating complex concepts through practical examples like the one presented here, including the next quarterly blog topic on Integrated Sensing and Communications (ISAC). Continue following this blog series to learn more, and if there are specific topics or questions you’d like addressed, please email me directly at [email protected] or connect on LinkedIn.

Related Resources:

For more about modeling and simulating RIS surfaces, visit the links below:

• Radar Sensing with Reconfigurable Intelligent Surfaces (RIS) (Example): This example shows how an RIS can help maintain a wireless communication link under challenging conditions like urban environments and indoor spaces where walls and furniture block signals. 
• Create a RIS surface in MATLAB (Documentation): This documentation provides detailed information for users on how to create a RIS surface in MATLAB. 
• Electromagnetic Analysis of Reconfigurable Intelligent Surface (Example): This example shows how to model the response of a RIS using full-wave electromagnetic simulation.
• Phased Array Geometries and Analysis (Documentation): This documentation provides an overview of the types of antenna or microphone array configurations supported by the Phased Array System Toolbox and Antenna Toolbox in MATLAB.