Dielectric Resonator Antennas – When Your 5G Base Stations Need that Extra Performance
Base station antenna design is usually evaluated with one metric – the price. Therefore, many of the typical antenna concepts used in LTE and earlier networks are based on dipoles or patch antennas, which are easy to manufacture and can be very affordable with low-cost substrates. However, when stepping into the fifth generation of mobile communications, the requirements for network performance are more crucial than ever.
As for performance, everyone is usually talking about higher speeds. But let’s emphasize something else for a change – reliability.
One way to improve reliability is to have faster and more accurate beam steering, which makes efficient MIMO systems very attractive. However, achieving the desired level of efficiency might need better solutions and antennas than what is currently available.
One antenna technology which has been hiding from the wireless industry for a long time is the dielectric resonator antenna, DRA. Although DRAs have been used in satellite communications, radar systems and WiFi, they have not truly penetrated to the telecom market where mass-manufacturing induces tight requirements on the technology. Lately, several telecom companies, such as Kathrein, have proposed to use DRAs in new 5G tech.
From a performance point of view, however, there are true benefits – DRAs obtain significantly wider bandwidth and higher efficiency compared to microstrip antennas. Many recent studies have focused on DRAs, highlighting also other benefits such as low cross polarization, high isolation between mmWave MIMO antenna elements and wide beam steering capabilities.
To get some more insights on DRAs, I talked to Jiyoun Munn, who works as a Technical Product Manager at COMSOL and has a long experience in antenna design. I was wondering what has limited the use of DRAs in the past.
- Probably quality control. The mechanical tolerance for cutting the right size resonator within acceptable frequency response variation has been challenging. Another factor relates to material properties – the dielectric constant of a circuit board material may vary more than 3% while the bandwidth of a microstrip antenna can be as low as 1.4 %. With exactly the same design, antennas fabricated from the same material but at different times will have different performance.
In other words, the narrow bandwidth and uneven material have made DRAs difficult to use in the past. Another factor limiting DRA success has been the rather bulky design at low frequencies which makes manufacturing difficult and expensive.
But… That was then, this is now.
With new, increasing frequencies coming into play in 5G as well as new materials capable of very low material tolerances, the wireless world is welcoming DRAs with open arms. Using injection moldable thermoplastics with dielectric constants above 10 and very low losses, DRAs can combine excellent performance with mass manufacturability at new 5G frequencies such as 3.5 GHz or 28 GHz. DRAs also give design freedom regarding antenna shape and feed.
This brings us to the importance of simulating antenna performance with the right software during initial concept validation.
- COMSOL Multiphysics® software can address an arbitrary shape of DRAs without sacrificing accuracy. When the material is lossy, the heat dissipation can also be captured in the same simulation environment without any hassle. The capability to handle anisotropic and nonlinear material properties is another charm, Mr. Munn explains.
Another benefit is that the antenna design is accessible to colleagues and customers through the software “app” during the whole antenna development cycle.
Whether you are working below 6 GHz or at mmWave frequencies, the benefits of using DRAs to improve your network performance are considerable: Better bandwidth, higher efficiency and improved antenna isolation leading to increased coverage, more accurate beam steering and, in the end, more reliable networks. With new injection moldable high-permittivity materials the limitations in DRA manufacturability are long gone.
The final limitation might be the one in our minds: we are slow to get rid of old habits – and old technology.