The power and sophistication of modern radar systems is growing by leaps and bounds. These systems already offer capabilities that couldn’t have been dreamed of just a few short years ago. Furthermore, ongoing developments mean that these systems are poised to revolutionize sensing applications. However, there is a data bandwidth bottleneck that must be overcome to take full advantage of these state-of-the-art technologies.
Beamforming is a signal processing technique that employs a phased array of antennas to provide directional signals during transmission or reception. This means you can control the direction of the signal (or multiple signals) without having to physically move the antenna.
Beamforming can be accomplished using analog, digital, or hybrid techniques. Digital beamforming uses computationally intensive digital signal processing (DSP) algorithms to control the signals being fed to, and received from, the antenna array.
Digital beamforming provides many advantages over its analog counterpart, including simplifying the RFIC design and enabling better scalability by moving the beamforming operations to the digital domain. This is critical to providing flexibility in forming the beam patterns and supporting multiple parallel data streams. In the case of radar applications, this means that signals from multiple directions can be monitored and measured simultaneously. Digital beamforming is also much better at handling interference. If a radar system is being subject to jamming from a hostile source, for example, digital beamforming can null this out, adapting its own beam patterns to accomplish the mission.
Digital beamforming provides capabilities that are simply not achievable using analog or hybrid beamforming approaches. The bottleneck is the vast amounts of digital data that must be conveyed back and forth between the antenna array and the compute (processor) array.
Digital beamforming enables antenna array system scaling to higher frequencies called millimeter wave (mmWave) by simplifying the RFIC front-end circuits. Using mmWave conveys multiple advantages. In addition to supporting higher bandwidth, its shorter wavelengths mean more antennas can be packed closer together. This means radar arrays can monitor larger numbers of objects, and much smaller objects, due to higher fidelity and resolution.
However, this also results in a “double whammy”— as the bandwidth increases the number of samples increases, and as the wavelength decreases the number of antennas can be more densely packed. All of this results in a quadratic increase in bandwidth density needs. In turn, this demands a new interconnect solution between the antenna array and the compute array (also between processing elements forming the compute array).
The obvious solution is to use an optical transport mechanism because optical interconnect offers high-bandwidth, low-latency, and low-power interconnections that are resistant to electromagnetic interference (EMI). However, it is not sufficient to simply take existing devices (CPUs, DSPs, FPGAs, ASICs, RFICs, etc.) and augment them with external optical interconnects. To achieve the highest possible transmission speeds and bandwidths, it is required for the optical interconnect to be incorporated inside the device packages.
In-package optical I/O enables chip-to-chip communication across a wide range of distances, from millimeters up to kilometers. This opens the door to a wealth of possibilities for new phased-array radar architectures, such as size, weight, and power (SWaP)-friendly disaggregated implementations. Some real-world examples of these possibilities are as follows:
Aerospace: In the case of radar in an aircraft, for example, sensor arrays can be mounted in the nose and the tips of the wings, and in-package optical I/O can be used to convey the data directly from RFIC chips in the array to a compute array located in the middle of the plane. Combining the data from multiple, widely distributed sensors provides a greater field of view (FoV) and higher resolution.
Ground and sea radar deployments: In-package optical I/O enables disaggregated sensing and processing for end uses like battlespace interconnectivity. Aggregating data from multiple distributed sensing arrays provides a more complete and holistic picture, thereby maximizing situational awareness.
Search and rescue: In-package optical I/O facilitates increased numbers of simultaneous beams, resulting in substantial improvements in spatial coverage. In turn, this increases the probability of detection and reduces the time taken to find people who are in distress or in imminent danger.
Ayar Labs has developed TeraPHY™ in-package optical I/O chiplets that designers of other devices (CPUs, GPUs, FPGAs, ASICs, RFICs, etc.) can incorporate in their system-in-package (SiP) modules. These chiplets are complemented by Ayar Labs’ SuperNova™ advanced light source (Figure 1).
Figure 1. TeraPHY optical I/O chiplets and SuperNova light sources unleash the potential of new digital beamforming architectures.
The combination of TeraPHY chiplets and the SuperNova light source is set to disrupt the traditional performance, cost, and efficiency curves of digital beamforming applications by delivering up to 1000x bandwidth density improvements at one tenth the power compared to traditional copper-based electrical interconnect technologies.
To learn more about advanced phased array radar architectures, please join us at the 2022 IEEE International Symposium on Phased Array Systems and Technology. Lockheed Martin Corporation, one of our strategic investors, will be presenting a paper we co-wrote, Converged RF Phased Arrays Enabled by Silicon Photonics, on Thursday, Oct 13, at 2:50 PM EDT.
Unleashing Opportunities with Optical I/O for the Aerospace Industry | Solution brief
Optical Interconnects for Future Advanced Antenna Systems: Architectures, Requirements and Technologies | Technical paper
Six Reasons the Aerospace Sector Needs Optical I/O Solutions | Blog
Optical I/O Chiplets Eliminate Bottlenecks to Unleash Innovation | Technical paper
Official U.S. Navy Page, CC BY 2.0, Creative Commons, via Wikimedia Commons.