Millimeter-wave transceivers can support two distinct functions relevant to industrial and technical systems: sensing an environment with radar and moving data over a short-range wireless link. Operating from 30 to 300 GHz, millimeter-wave technology offers access to broad frequency bandwidths, but realizing that potential depends on more than the radio-frequency integrated circuit. Antennas, signal generation, receiver design and packaging must be considered together.
A pair of transceiver designs described in Andrew Townley’s Berkeley dissertation illustrates the engineering tradeoffs. One is a phased-array radar transceiver intended for applications such as gesture recognition; the other is a wideband communication transceiver designed around a 120 GHz carrier. Together, they show how millimeter-wave transceivers address sensing and communication while presenting different system-level design requirements.
Why millimeter-wave transceivers use high carrier frequencies
Wireless capacity depends on both signal-to-noise ratio and bandwidth. Raising signal-to-noise ratio can support more complex modulation, but it calls for higher transmit power or lower receiver noise, and its contribution to capacity diminishes as the ratio increases. Using more bandwidth is another route to higher data rates.
Bandwidth must be considered relative to carrier frequency. A given absolute bandwidth represents a smaller fractional bandwidth at a higher carrier frequency, making broad channels more practical to implement. This is a central motivation for millimeter-wave communication. For radar, bandwidth also matters because greater signal bandwidth can improve the ability to distinguish targets that are close together.
FMCW radar for millimeter-wave sensing
The radar design uses frequency-modulated continuous-wave (FMCW) operation. Rather than transmitting isolated pulses, an FMCW system continuously transmits a signal whose frequency changes over time. The received echo is a delayed, attenuated version of that transmitted signal. Comparing the two produces an intermediate-frequency signal related to the round-trip delay, from which target distance can be determined.
Linear frequency sweeps are useful because they allow range information to appear as an IF tone, without requiring the receiver to process a rapidly changing, wideband waveform at every instant. The overall sweep bandwidth still matters for distance resolution, but the frequency changes gradually. This can ease instantaneous bandwidth demands on the transceiver and baseband chain.
The radar implementation described in the dissertation combines four transmitters and four receivers with shared frequency-generation circuitry. Its phased-array architecture uses multiple antenna paths, allowing the transmitted beam to be steered by controlling signal phase. The fabricated radar IC was packaged with eight antennas on a 1.2-by-1.2-centimeter module. Measurements included array characterization, beam steering and radar operation, demonstrating the importance of treating the chip and antenna module as parts of one system.
Wideband communication places different demands on the radio
Radar estimates the environment from echoes; a communication transceiver must encode and recover data. The second design targets a short-range wireless link using a 120 GHz dual-channel transceiver. Its architecture explores how broad bandwidth can support higher data throughput, while maintaining the signal quality needed to distinguish modulation symbols.
The transmitter uses digital-to-analog conversion and quadrature signal paths to create modulated signals, followed by a power amplifier. On reception, a low-noise amplifier, mixer and baseband amplification chain convert and amplify the incoming signal. The dissertation presents simulations of this chain and examines modulation formats including QPSK, 16QAM and 64QAM. These are design results, not evidence that every modulation mode or target data rate will be achieved in a deployed system.
Higher-order modulation carries more information per symbol but requires the receiver to distinguish more closely spaced signal states. That places demands on signal quality and the transceiver’s linearity. The design discussion therefore considers both circuit behavior and complete-link requirements, rather than treating bandwidth alone as a guarantee of higher throughput.
Integration challenges extend beyond the IC
At millimeter-wave frequencies, interconnects and packaging can strongly affect system behavior. The dissertation identifies local oscillator (LO) generation and distribution as a challenge: generating the carrier directly at very high frequency can be difficult, while distributing a shared LO across multiple channels can require carefully designed routing and buffers.
Packaging and antenna integration also affect performance. Both transceiver projects use flip-chip attachment to connect the IC to a PCB or antenna module, making the chip-to-board transition part of the RF design. For the communication system, the antenna PCB design includes attention to substrate properties, antenna behavior and surface waves. These details can influence how effectively a signal reaches the antenna and how consistently the assembled system behaves.
Technology selection involves its own tradeoffs. CMOS is suited to digitally intensive systems, while SiGe BiCMOS can be attractive for analog-heavy functions such as power amplification. The appropriate choice depends on the balance of RF performance, digital processing and integration required by the application.
Practical implications for automation teams
For engineers assessing millimeter-wave sensing or wireless links, the work points to several system-level questions:
- Define the function first. Radar and communication use different waveforms, receiver processing and performance measures.
- Match bandwidth to the task. Radar bandwidth affects distance resolution; communication bandwidth is one factor in achievable data capacity.
- Plan the signal chain as a whole. LO generation and distribution, transmit power, receiver noise and baseband bandwidth interact.
- Include packaging and antennas in early design decisions. Chip-to-board transitions, antenna structures and substrate effects are not separate from RF performance.
- Distinguish measured results from simulations and targets. The radar design includes reported measurements, while many communication-chain results are presented as simulations.
These considerations are relevant when evaluating whether millimeter-wave technology fits a sensing or connectivity requirement. The application’s needed range, resolution or data throughput must be weighed against the complexity of the RF and antenna implementation.
Conclusion
Millimeter-wave transceivers offer a route to broadband sensing and short-range wireless communication, but their value comes from coordinated system design. FMCW radar uses frequency sweeps to derive range from echoes, while wideband communication relies on a transmitter and receiver capable of preserving modulated data. In both cases, circuit architecture, LO distribution, antennas and packaging shape the result. For industrial automation decision-makers, the practical lesson is to evaluate the complete signal path against the application—not just the carrier frequency or headline bandwidth.