Renesas unveils radar solution to bolster ADAS, autonomous driving
Renesas Electronics Corporation, a premier supplier of advanced semiconductor solutions, has announced its first RH850-based, 32bit, automotive radar microcontroller (MCU) series, the RH850/V1R. It will deliver the high performance and features required for enabling future advanced driver assistance systems (ADAS) and autonomous driving. The RH850/V1R-M, the first product from the RH850/V1R Series, includes a digital signal processor (DSP, Note 1) and high speed serial interfaces and is specifically designed for middle-to longrange radars.
Software and tools including evaluation boards will also become available to enable the system developers to start their development immediately with the RH850/V1R-M solution. With the expansion of ADAS and autonomous driving, sensors are fast becoming a key technology. Additionally, high precision sensing becomes critical in realising future autonomous driving to meet the increasing requirements of range resolution, separation of objects and precision in measurement of velocity. This requires increased numbers of antennas and boosting of the signal processing performance.
To address these needs, Renesas launched the new RH850/V1R-M MCU specifically designed for radar applications in ADAS. The new MCU includes optimized, programmable digital signal processing, a dual core at 320 megahertz (MHz) with industryleading high speed flash of 2 megabyte (MB), a 2 MB internal RAM, while meeting the industry’s highest temperature requirements.
Key features
High-performance DSP with flexible programmability for improved radar signal processing performance and increased sensing accuracy. The RH850/V1R-M MCU is designed to have optimised hardware acceleration in the DSP. The optimised DSP allows the system developers to process the raw data into target objects efficiently, separately from the safety relevant classification and tracking done in the CPUs. Renesas’ high performance DSP performs radar specific algorithms such as fast fourier transforms (FFTs, Note2), beamforming, windowing, channel calibration, peak search, at high speed and with low power consumption. The DSP offers high flexible programmability and Renesas specifically developed a DSP math library for automotive radar sensors to support system developers in their algorithm development.
Renesas low-cost radar sensors
The RH850/V1R-M is developed using Renesas’ world leading 40-nanometer (nm) embedded flash (eFlash) process technology, which has a proven track record in terms of re-write cycles, the industry’s fastest random access operation speeds and high reliability. It also has the merit of low power consumption as the transistors are smaller lowering parasitic capacitances. The process specification of RH850/ V1R-M also fulfills the highest temperature requirements in the industry (T-junction 150°C). The embedded flash brings advantage to the system designer by offering higher integration using less PCB space and better real time behavior.
Dual core at 320MHz high performance, integrated 2MB large capacity RAM, 2MB highest speed flash and highest temperature requirements. The new RH850/V1R-M features 2 G3MH CPU cores operating at 320 megahertz (MHz). The G3MH is a superscalar reduced instruction set computer (RISC) architecture with two 7-stage integer pipelines, which allows execution of 2 different instructions at the same time. Each G3MH core achieves the performance of 3.2 DMIPS/MHz (Note 3).
The RH850/V1R-M also includes 2MB industry-highest speed flash based on Renesas world-leading automotive 40nm embedded flash technology. By incorporating a large 2MB capacity RAM, the RH850/V1R-M handles all specific calculations on radar cube data such as range and velocity FFTs, digital beam forming, constant false alarm rate (CFAR, Note 4) and peak detection.
Pricing, availability
Samples of the RH850/V1R-M and the DSP math library will be available from second half of 2017. Samples will be tentatively priced at US$30 per unit. Mass production is scheduled to begin in November 2018. A flexible programmable complete toolchain with C/C++ compiler, debugger, simulation models and detailed performance profiling tools will also be available.