mm-wave FMCW radar antennas are a system-level design decision, not simply a matter of choosing a radiating element. Antenna efficiency, transmitter-to-receiver isolation, packaging, interconnect losses and array geometry all affect how a compact radar performs. For industrial sensing and other short- to medium-range applications, silicon-based radar offers a route to highly integrated hardware, but antenna integration introduces trade-offs that engineers and system integrators must evaluate together.
Why antenna integration matters in mm-wave FMCW radar
Frequency-modulated continuous-wave (FMCW) radar transmits and receives at the same time. The receiver therefore has to detect reflected signals while some of the transmitted energy can couple into its receive path. Insufficient isolation between transmit (TX) and receive (RX) antennas can contribute to receiver saturation and degrade the radar’s usable signal quality.
Silicon BiCMOS technology enables compact, integrated mm-wave front ends, and the small antenna dimensions at these frequencies make on-chip integration possible. Placing an antenna close to the radar circuitry can reduce the need for an off-chip mm-wave transition. But the silicon substrate’s electrical properties can reduce antenna radiation efficiency, while package materials and structures can support surface waves or substrate-guided modes. These losses may divert energy from the intended radiation direction and affect the radiation pattern.
Planar antenna implementations can also incur conductor and dielectric losses. Consequently, antenna performance depends not only on the radiating structure, but on its substrate, package, interconnect and surrounding layout.
On-chip antennas: compact integration with package-level challenges
On-chip antennas can support compact, lower-cost radar integration and avoid a separate chip-to-antenna connection. Their main challenge is efficiency: energy can be dissipated in the silicon or propagate through the substrate and package instead of radiating as a useful space wave.
One approach examined for a 60 GHz single-chip FMCW radar was a cavity-backed on-chip monopole. The design work also optimized the packaging environment, including interconnects and the thicknesses of the substrate and protective epoxy. These elements influence guided-wave propagation and antenna behavior, so package design is part of the RF design process rather than a purely mechanical consideration.
Experimental work with three integrated antennas reported 2 dBi gain and 25 dB of isolation between the TX and RX antennas. These results illustrate the value of considering antenna efficiency, radiation pattern and channel isolation together when designing an on-chip radar assembly.
PCB antenna arrays and TX-RX isolation
For applications requiring higher gain or array-based beam characteristics, a printed circuit board (PCB) provides more area for antenna elements and isolation structures. The trade-off is the transition between the radar IC and the off-chip antenna. A bond-wire connection can provide a low-cost interconnect, but its impedance mismatch can cause power loss unless the transition is addressed in the design.
Electromagnetic bandgap (EBG) structures offer a passive method for reducing coupling between antennas. In the described PCB approach, EBG patterns were used to suppress surface-wave propagation and improve isolation between transmit and receive arrays. Suppressing these waves can also help reduce radiation-pattern ripples caused when waves propagating through a finite substrate scatter at its edges and interact with the antenna’s direct radiation.
A series transmission-line section was used to compensate for the impedance mismatch associated with the bond-wire transition. In the reported prototype, an eight-element series-fed PCB array achieved 12 dBi gain and 40 dB isolation between TX and RX arrays. The transition achieved a -10 dB impedance match across 57 to 64 GHz, the radar’s operating bandwidth. These findings show how antenna layout, isolation structures and interconnect matching can be treated as a coordinated integration problem.
MIMO radar and sparse array design
When electronic scanning or improved angular resolution is needed, multiple-input multiple-output (MIMO) radar offers an alternative to a conventional phased array. By combining signals from multiple transmit and receive antennas, a MIMO system can form a virtual array. This can provide a larger effective aperture than a phased array with the same number of physical antenna elements, because MIMO arrangements can use wider element spacing than the traditional half-wavelength spacing associated with phased arrays.
Physical constraints on a single radar chip can limit the available antenna positions. A sparse, non-regular array can help address this limitation by selecting antenna locations to increase spatial resolution while managing array characteristics such as sidelobes. The design described used multiple single-chip radars with three on-chip antennas per chip. Synchronization was controlled using a 50 MHz reference generated outside the chips by crystal oscillators; the design work identified this approach as more suitable for coherent beam steering and phase-noise reduction than sharing a 60 GHz reference directly.
For engineers, the practical point is that MIMO performance depends on more than antenna count. Array placement, synchronization, signal processing and the physical limits of the integrated hardware all shape the resulting radar system.
Choosing an antenna approach
The appropriate architecture depends on the system’s integration and performance priorities:
- On-chip antennas favor compact integration, but require careful management of substrate and package losses.
- PCB arrays provide room for higher-gain structures and passive isolation features, while introducing an RF transition that must be matched.
- EBG structures can help reduce coupling and suppress surface waves, but need to be designed for the relevant antenna layout and frequency range.
- MIMO arrays can support a larger virtual aperture, with array geometry and chip synchronization central to system operation.
Conclusion
Successful mm-wave FMCW radar antenna integration requires coordinated decisions across the antenna, package, interconnect and radar architecture. On-chip designs prioritize compactness but must address silicon and package losses. PCB arrays offer opportunities for gain and passive isolation, provided the chip-to-board transition is managed. For scanning and spatial resolution, MIMO and sparse-array methods add flexibility while making synchronization and geometry important design considerations. Treating these elements as one system helps technical teams align antenna choices with the requirements of the radar application.