The Role of AI With Dynamic Modulation and Sparsity in High Performance Imaging Radar
Please note: This session is part of AI Day. Learn more and register for the full event here.
Today, the state-of-the-art solution to enable high performance imaging radar is to increase the antenna count and processing capabilities of the system. Fundamentally, this approach is not an optimal solution for automotive and robotics applications as the imaging radar available on the market use the same principles as basic automotive radar. In contrast, utilizing AI with dynamic modulation and sparsity in multiple dimensions can create a significantly higher-performance radar system. This approach enables performance to be scaled, not by the antenna count, but rather by the compute capabilities of the radar system, which yields a more optimal and higher-performance solution.
4 Key Takeaways:
1. There are tradeoffs associated with state-of-the-art imaging radar systems.
2. AI with dynamic modulation can be used to enable high Doppler resolution and range resolution.
3. Sparsity in both hardware and software can be leveraged to achieve better angular resolution.
4. These approaches enhance performance and have wide applicability in both edge AI and central domain control AI processing architectures for automotive and robotics applications.
Paul Dentel is a Senior Technical Product Manager at Ambarella focusing on radar systems. He has been working with frequency-modulated continuous-wave (FMCW) radar systems since 2016. That experience includes productizing Texas Instruments’ first...
Executive Vice-President, Brand Director, Embedded Franchise, OpenSystems Media
Richard Nass’ key responsibilities include setting the direction for all aspects of OpenSystems Media’s Embedded and IoT portfolios, including digital, print, and live events. Previously, Nass was the Brand Director or UBM’s award-winning Design...