Which is Better: Simulated or Measured Performance for Simple Antennas?

Accurately measuring antenna patterns and gain is often a challenge. It requires not only expensive test equipment, but also skilled and knowledgeable workers. A third problem arises from the test equipment itself: it is impossible to avoid the interactions introduced by the test equipment and test fixtures and/or equipment holders.

Antenna pattern and gain measurements are quite unlike most microwave component testing. This is because it needs to be performed in an open space environment withwell controlled minimum reflections. There are three categories of widely used antenna pattern and gain measurement systems: Far Field Range, Near Field Range, and Compact Range. Generally these systems are unwieldly and expensive, regardless of the frequency band. If a Vector Network Analyzer (VNA) or Synthesizer and Spectrum Analyzer are added, the cost becomes unaffordable. This is especially true for millimeter wave bands. Due to these limitations, most businesses cannot afford their own antenna range and so they outsourcce the service instead. So there's a looming question: Is antenna gain measurement for simple antennas even a necessity in these modern times? On the other hand, measuring pattern is a very delicate and detailed task. It requires the same skilled and knowledgeable workers, generally full-fledged microwave engineers, that are also familiar with VNAs, proficient in the setup and alignment of the Antenna Under Test (AUT) and Transmitter Antenna. Any misalignment or reflection caused by the text fixtures or any part of the testing environment can introduce measurement errors and uncertainties in the data.

However, modern EM solvers are mature. Today there are various commercially available simulators: HFSS, CST, and uWave Wizard. These simulators can take into account all aspects, including the material and finishing properties, in order to give very accurate simulation results. There are also mechanical design tools, such as Solidworks, and others that are capable of taking the simulated structure and realizing the physical antenna structures without losing the dimension accuracies. With the help of these modern manufacturing tools, such as CNN machine centers, wire EDM, EDM Sinkers, and 3D printers, we can achieve the required accuracy.

Due to the above reasons, the industry has accepted using simulated antenna patterns and gains for simple aperture antennas are very accurate. These include rectangular horns, circular horns, lens corrected antennas, scalar antennas, Gaussian antennas, Cassegrain antennas, reflector antennas, and omnidirectional antennas. As long as the processes lined out above are used, the antenna hardware will create the patterns and gains as predicted by the simulator. In contrast to the simulator data, measured data will often be distorted due to setup limitations and the operator's abilities. The measured patterns and gains are often less accurate than the simulated data. Based on this conclusion, SAGE Millimeter highly recommends using simulated antenna gains and patterns instead of measured data for the simple aperture antennas mentioned above.

To further this conclusion, SAGE Millimeter performed an investigation by measuring the patterns and gains for an E-Band Gain Horn Antenna. The results are shown below.

From these findings, one can see that the measured data and simulated data are very similar, but the simulated data is more accurate. This is because the patterns should be more symmetrical due to the antenna also being symmetrical.

I do not think that this is the right question. Simulation is better when we need it and the measurement is the final proof of the design. These two design stages should not be confused.