Furuno Electric Co. Uses NI AWR Design Environment and NI FlexRIO Platform to Develop Weather Radar in 40 Percent Less Time

 Read in  |   Print

"Compared to a conventional design approach used with similar new product developments, we estimate that we achieved a reduction in development time of more than 40 percent by adopting the NI solution."

- Takuo Kashiwa, Ph. D, Furuno Electric Co., Ltd.

The Challenge:
Developing a weather radar with flexibility in the signal processing unit to accommodate various potential design changes and incorporate a way to verify the system-level performance by co-simulating the digital and analog sections.

The Solution:
Adopting the NI FlexRIO platform for digital section hardware, using graphical system design methodology to accommodate potential design changes in the software, and taking advantage of the co-simulation capability between AWR Visual System Simulator (VSS) and NI LabVIEW software to realize the system-level simulation of digital and analog sections together.

Author(s):
Takuo Kashiwa, Ph. D - Furuno Electric Co., Ltd.
Yasunobu Asada - Furuno Electric Co., Ltd.
Tomonao Kobayashi - Furuno Electric Co., Ltd.

Furuno Electric Co., Ltd. has been steadily growing since introducing the world’s first commercial fish finder in 1948. Using radar and GPS as our core technologies, we have expanded our business with marine electronics as the core products. Frequent short, localized rainstorms, called guerilla rain, and tornadoes have resulted in house flooding and destruction, river flooding, and mudslides in mountainous areas of Japan, which motivated us to shift our focus. We are now focusing on weather radar as a new business area. We believe we can help prevent such natural disasters by using our technology and the knowledge that we have acquired over the years from our marine radar products.

Weather radar systems are designed to predict weather and monitor hurricanes and rain fronts and can be large in size. We set out to develop a compact, low-cost weather radar system design to help predict guerrilla rain and prevent other urban natural disasters (Figure 1). As shown in Figure 2, the basic building blocks are typical for radar. The radar’s main function is to determine the target’s condition by transmitting radio waves and measuring the return time and levels of reflection from the target. For marine radars, the target is mainly marine vessels. The target is rain or snow for weather radars. The function of each block follows:

Operation/Display Unit—Used to operate the weather radar and display the observed information.

Signal Processing Unit—Used to generate transmit waveforms, control signals related to transmit/receive, and sample received signals (A/D conversion).

RF Converter—In the transmit path, the IF signal output from the signal processing unit is upconverted to RF (X band, 9.4 GHz). In the receive path, the RF signal received from the antenna is downconverted to IF.

Power Amplifier—Used to amplify the RF signal received from the RF converter.

Circulator—Used to separate the transmitter path from the receiver path. In the transmitter path, the output signal from the power amplifier is sent to the antenna. In the receiver path, the signal received through the antenna is sent to the RF converter.

Antenna—In the transmitter path, the signal received from the circulator is emitted into the air, and, in the receiver path, the reflected echo from the target is received.

Figure 1. Difference Between Furuno Weather Radar and a Conventional Weather Radar

Figure 2. Basic Weather Radar Configuration

Challenges

We faced two major challenges in the development of the weather radar system.

Need for a development method to flexibly accommodate design changes
This was our first time developing a weather radar. We predicted that there would be various corrections and changes at the prototype and verification stages. However, our main concern was the signal processing unit. Traditionally, we used custom design process that included printed circuit board design, FPGA programming with a hardware description language (HDL), and C programming for software running on the CPU. The custom design process can be when a revision is necessary. Thus, we needed a process to flexibly accommodate revisions.

Need for a unified environment to perform system-level simulation
There was not a practical unified environment to perform the simulation of both digital circuits (signal processing unit) and analog circuits. For this type of RF system, the digital and analog circuits are typically designed independently and then verified in separate environments. The digital and analog circuits are simulated independently. The design of the digital and the analog sections are considered complete when simulations are complete. Thus, the system design is also considered complete. In reality, the system-level verification that integrates both the digital and the analog sections is performed for the first time in the prototype stage. Some simulators for analog circuits can incorporate signal processing code written in C and co-simulate the code and analog circuit together. However, it is difficult and an unrealistic use of resources for an analog circuit engineer, who is not an expert in signal processing, to write such a code.

Features of radar products from various manufacturers depend on the algorithms and intellectual property used in the signal processing unit. To improve system performance, the manufacturer must first verify system performance as the metric by designing an analog circuit matched to the manufacturer’s unique signal processing unit rather than independently trying to improve the analog circuit performance. Co-simulation of the digital signal processing unit and analog section is necessary to realize this process.

We selected a solution from National Instruments to solve these challenges. For each of the challenges, we took the approach described below.

Adoption of Graphical System Design

For the first challenge, we implemented the signal processing unit with the NI FlexRIO platform. The NI FlexRIO platform consists of three components: the NI FlexRIO FPGA module that includes a reconfigurable FPGA, the adapter module that provides high-performance I/O, and the PXI system. All the software is programmable with NI LabVIEW system design software. Using the NI LabVIEW FPGA Module, we can also graphically program the FPGA with LabVIEW. We can accommodate design revisions by writing a graphical code in LabVIEW with fixed hardware configurations using this graphical system design method.

Another advantage of using NI FlexRIO hardware was the extreme ease of developing a data acquisition mechanism. Since this was our first weather radar application, we did not know what kind of data to expect for rain and snow conditions. Thus, we wanted to have a system that could save a large amount of data to the hard disk drive (HDD) during the prototyping stage. However, with traditional design methodologies, it would take extensive efforts to design it from the beginning. The data can be saved easily on the HDD without extra design work with the PXI-based NI FlexRIO hardware.

Using Co-Simulation Between AWR VSS and LabVIEW

For the second challenge, we used the co-simulation capability of the AWR Visual System Simulator (VSS) suite and LabVIEW. VSS was the wireless communication systems design software we were using for the design of the analog section. By using this co-simulation capability, the LabVIEW code could be called and executed from VSS. The analog circuits in VSS and the LabVIEW code were co-simulated by importing the LabVIEW code written by our digital engineer for the signal processing unit. This meant we could not only verify the performance of the analog section alone, but also verify the system performance that includes the signal processing unit as well. There were no technical obstacles for the analog designers during this process. By directly using the LabVIEW code for the NI FlexRIO, we could verify the system performance in advance.

Realizing a More Than 40 Percent Reduction in Development Time

We could design, prototype, evaluate, and correct the features and characteristics of our weather radar using the methods we mentioned so far (Figure 3 and Figure 4). The time it took to achieve this was remarkable. Compared to a conventional design approach used with similar new product developments, we estimate that we achieved a reduction in development time of more than 40 percent by adopting the NI solution.

Also, with a conventional approach for the digital section, board-level problems typically resulted in rework from the beginning. By adopting the NI FlexRIO platform, we did not have any hardware changes. In addition, developing everything in the LabVIEW system design environment was very effective from a human resource perspective. We did not need domain experts such as a board designer, an HDL programmer for the FPGA, or a C programmer for the host software. Moreover, we developed a system-level simulation combining analog and digital sections that was previously not practically feasible.

Figure 3. External View of the Developed Weather Radars

Figure 4. Image Acquired By the Weather Radar.*

Future Work

We are in the process of commercializing the weather radar. Our weather radar can perform an observation as a single unit, but we plan to develop a more advanced system that can perform simultaneous observations with multiple radar. In addition, we want to develop a system to predict precipitation from the observed data and develop a solution that will help reduce the damage from disasters.

We have high expectations for the NI solution, not only for the weather radar, but for our future design work as well. The common perception is that LabVIEW is used for instrument control. Lately, however, LabVIEW is being used more as a tool for signal processing in product design. With a text-based language, it is a very complicated task to implement signal processing functions, but with graphical development in LabVIEW, that complexity can be greatly reduced. The co-simulation capability of LabVIEW with the AWR VSS further increases the significance of using LabVIEW. In fact, the usefulness of LabVIEW is becoming widely recognized within our company, with increased number of users centered in the research department. On the analog design side, we plan to reuse the joint VSS and LabVIEW methodology for applications completely different from the weather radar. We hope to replicate this success internally.

*This image was obtained by performing image processing on 3D observation data acquired from the weather radar. The stagnating rain can be observed over the skies of Osaka Bay and Southern Osaka.

Bookmark and Share


Explore the NI Developer Community

Discover and collaborate on the latest example code and tutorials with a worldwide community of engineers and scientists.

‌Check‌ out‌ the‌ NI‌ Community


Who is National Instruments?

National Instruments provides a graphical system design platform for test, control, and embedded design applications that is transforming the way engineers and scientists design, prototype, and deploy systems.

‌Learn‌ more‌ about‌ NI