At European Microwave Week 2025, Electronic Specifier’s Editor, Mick Elliott speaks to Giorgia Zucchelli, RF and Mixed Signal Product Manager about the solutions supplied by MathWorks that can accelerate development.
Founded in 1984, the company is known for its Matlab and Simulink software that is used in the engineering and scientific fields.
“[MathWorks] has a long history of accelerating the pace of engineering and science. We are here today … primarily to talk about how MATLAB and Simulink can help the acceleration of the development of wireless communication and radar systems.”
Headquartered in Natick, Massachusetts, with “offices in 34 different locations around the globe” and around “6,500 people … mostly in development,” MathWorks is a privately-owned company, “funded by engineers for engineers.”
Discussing some of the latest tools and workflows, Zucchelli said that one of the key reasons for MathWorks’ attendance at European Microwave Week was to enable engineers to cover the entire design workflow.
“Our mantra is really trying to cover the entire antenna-to-bits span,” Zucchelli said. Using MATLAB and Simulink, engineers can model RF chains, run harmonic balance analysis, and generate models automatically.
“This is very important because it helps … when I don’t know exactly how things work, I want to have different ways to validate the results. I like the duality of analysis and simulation … they land on the same results. This gives me confidence that I know what I’m doing.”
Alongside the RF Toolbox, MathWorks also provides an Antenna Toolbox with a “full wave electromagnetic solver … very suitable for analysing radiating structures in air.” Engineers can quickly design, scale, and analyse antennas, then expand to arrays and full system-level models.
“Step by step you build up on the complexity and this process … is really the process of model-based design,” Zucchelli explained.
System-level modelling extends to features such as beamforming. “You can add phase shifters or time delays to steer the beam in a specific direction, and you can verify all of this again with electromagnetic simulation.”
MathWorks’ tools also allow integration of PCB components and full standard-compliant 5G waveforms.
“Once you stream a real signal that might have 100, 400, 800MHz of bandwidth … this really tells you is the system working yes or no.”
Turning to the challenges RF engineers face, Zucchelli noted the impact of higher frequencies and wider bandwidths: “Nowadays a lot of the systems are moving to higher and higher frequencies and larger bandwidth. This makes it harder for RF engineers because certain effects that in the past you could ignore, [but] you cannot ignore nowadays.
The traditional design methodology based on prototyping, rapid prototyping, trial and error, tweaking in the lab, becomes almost prohibitive. Using models and simulation … is a way to anticipate how the system works.”
Talking about AI and the role it’s playing in antenna characterisation and system optimisation, Zucchelli said: “We are talking a lot about AI because it’s a little bit of a hype.” However, MathWorks takes “a pragmatic approach to AI” – not as a universal solution, but as a targeted tool. A key example is antenna pattern reconstruction.
“System engineers … only get two orthogonal cuts [of the antenna pattern] … How do they go from limited amount of information to a full 3D radiation pattern? … This is also a perfect problem that can be solved with AI … In Antenna Toolbox you’ll find a neural network … pre-trained with 100,000 patterns of antennas … Users can just use this pre-trained network to infer the 3D pattern from two orthogonal cuts … It just works and provides better results than analytical methods.”
MathWorks has validated this work with Rohde & Schwarz using real measurements. Looking ahead, Zucchelli said that the company is also exploring AI in areas such as optimisation and electromagnetic analysis.
“MathWorks is a company founded by engineers to deliver solutions to engineers,” she concluded.
Watch the video below to learn more.
