FPGA-based automotive ECU design fully complies with AUTOSAR and ISO 26262 standards

The automotive industry utilizes reconfigurable hardware technology to flexibly integrate in-vehicle functions.

Today's automakers are adding more and more advanced features to automotive electronic control units (ECUs) to improve the driving experience, enhance safety, and of course expect to exceed the sales of competing products. In this case, the Automotive Open System Architecture (AUTOSAR) program and the functional safety standard ISO26262 are rapidly becoming the technical and architectural foundation for automotive ECU design.

In order to meet the increasing functional demands of new models, the density of automotive electronics is increasing, and FPGA manufacturers are continually introducing larger devices. These devices integrate all applications and are less power consuming and competitively priced than previous generation devices. This trend means that reconfigurable computing technology will be further promoted and applied in the automotive industry.

We have pioneered a groundbreaking approach to designing an automotive ECU that meets both AUTOSAR and ISO 26262 standards using a programmable FPGA device instead of an MCU-based platform as the basis for the ECU. Our design approach explores key features of reconfigurable hardware such as parallelism, customizability, flexibility, redundancy, and versatility. After the conceptual design is complete, we want to implement the design in the prototype. To this end, the Xilinx ZynqTM-7000 scalable processing platform is ideal. This FPGA platform combines the ARM® dual-core CortexTM-A9 MPCore hard processor with the 28 nm Xilinx 7 Series programmable logic device with dynamic partially reconfigurable functionality to meet the required requirements. It is also equipped with an on-chip communication controller commonly used in in-vehicle networks such as CAN and Ethernet.

Emerging applications

Current automotive computing capabilities are distributed by means of ECUs interconnected by a communication network. In the next few years, such computing power is expected to increase further due to the rise of new applications in motor vehicles. These new applications include safety and driver assistance, inter-vehicle communication, comfort and control, in-car entertainment and a host of hybrid electric technologies. There is no doubt that the number of vehicle electronics is expected to increase. According to analysts' forecasts, the size of the automotive applications semiconductor market will grow at a compound annual growth rate (CAGR) of 8% over the next five years. One of the fastest growing segments involves microcontrollers (MCUs) and programmable logic devices such as field-programmable gate arrays (FPGAs).

While the number and advancement of in-vehicle functionality is increasing, designing and managing these systems has become increasingly complex, and automakers believe it is necessary to take an effective approach to solving this problem. The result is that today's AUTOSAR and ISO 26262 standards are affecting the architecture, design, and deployment of real-world automotive ECU hardware and software systems (see sidebar).

The AUTOSAR standard, developed by several automotive manufacturers in 2003, aims to define a standard system software architecture for ECUs distributed in vehicles. The purpose of the ISO 26262 standard is to focus on functional safety, essentially to avoid or detect and handle faults, thereby mitigating the effects of failures and preventing violations of any existing system safety objectives. Functional safety has become one of the key issues in automotive development with the introduction of new safety-critical features such as driver assistance or dynamic control. The ISO 26262 standard was approved in 2011 and supports the development of hardware and software security.

Therefore, the entire ECU design and development process is managed by standards that require systemic processes. Our job is to design a cost-effective embedded computing platform that uses reconfigurable hardware technology to optimize the system architecture.

system structure

The AUTOSAR and ISO 26262 standards are primarily from a software development perspective and are oriented to microcontroller-based computing platforms. However, the combination of hardware/software co-design and reconfigurable computing technology can bring many advantages to this area. Although standard MCUs are often the best choice for automotive ECU hardware platforms, with the ever-decreasing cost of new FPGAs and the integration of hard-core processors within some FPGA products, FPGA devices are becoming widely used in this market. The ideal solution. In addition, the trend to continuously integrate new embedded features in automobiles has also raised the need for parallel computing architectures. This is especially true in today's in-vehicle infotainment industry where high-speed digital signal processing is opening the door to FPGA technology. Programmable logic vendors like Xilinx and EDA tool vendors like MathWorks have shown a clear interest in this area.

In order to take full advantage of the reconfigurable hardware in automotive applications, we will focus on one of the most important ECUs in the automotive computing network for deploying end-user functions - the "body controller module". The potential of technology. The ECU, also known as the "body domain controller", is responsible for synthesizing and controlling the main electronic body functions in the vehicle, such as windshield wipers/spray systems, lights, window regulators, engine ignition/extinguishing, exterior rear view Mirror and central locking. Our goal is to design an AUTOSAR-compliant ECU system with safety-critical features on the FPGA platform.

Actual situation

If automakers want to cost-effectively manage increasingly complex vehicle functions, the standardization of the ECU system architecture promoted by AUTOSAR is the only way. It enables highly integrated and distributed software components distributed across the ECU. The main purpose of AUTOSAR is to define a unified ECU architecture that separates hardware from software. In this way, AUTOSAR can improve software reuse by defining hardware-independent interfaces. In other words, software components written in accordance with the AUTOSAR standard can be run on any vendor's microcontroller as long as they are properly integrated into the AUTOSAR-compliant operating environment.

This feature gives car manufacturers greater flexibility. Thanks to the plug-and-play features inherent in the AUTOSAR standard, automakers can transparently change versions of the same software modules developed by different vendors across the entire vehicle platform without negatively impacting the rest of the car. . The final hardware and software implementations are highly independent of each other. This separation is achieved by interconnecting the abstraction layers through the API of standard software. Figure 1 is an exploded view of the functional layers defined by AUTOSAR.

Figure 1 AUTOSAR hierarchical model from MCU to application layer
Figure 1 AUTOSAR hierarchical model from MCU to application layer

The bottom is shown in black as the hardware or physical layer and consists of the MCU itself (the CPU and some of the standard peripherals connected to it). Above the microcontroller is the base software (BSW), which is divided into three layers: a pink microcontroller abstraction layer (MCAL), a green ECU abstraction layer (ECUAL) and a complex driver, and a purple service layer (SRV). These three layers are organized into multiple columns or protocol stacks (memory, communication, input/output, etc.).

Close to the hardware components is the microcontroller abstraction layer. As the name suggests, this layer is an abstraction of the MCU. The purpose of this layer is to provide a hardware-independent API that handles the hardware peripherals in the microcontroller. The upper layer of the microcontroller abstraction layer is the ECU abstraction layer, which is responsible for abstracting other smart devices on the ECU development board, generally in direct contact with the MCU (for example, system voltage regulators, intelligent switching controllers, configurable communication transceivers, etc.) . The next third layer is the service layer. This layer is basically hardware-independent, and its role is to handle the different types of background services required. For example, network services, system watchdog NVRAM processing or management. Through these three layers, AUTOSAR defines a set of basic software functions. This software feature supports all the functions of the advanced abstraction layers of automotive ECUs on a specific hardware platform.

The fourth layer is the operating environment (RTE), which provides communication services for application software. It consists of a set of signals (transmitter/receiver ports) and services (client and server ports) that are accessible from the BSW layer and the application layer (APP) above. The RTE abstracts the application from the underlying software and clearly outlines the software stack architecture that separates the generic exchangeable software code (APP) from the specific hardware-related code (BSW). In other words, RTE can separate software applications from hardware platforms. Therefore all software modules running on the RTE are platform independent.

On top of RTE, the software architecture moves from tiering to component-based through the application layer. The features are primarily packaged in software components (SWC). Therefore, standardization of the AUTOSAR software component interface is a central part of supporting the scalability and portability of ECUs across different vehicle platforms. In addition to complex drivers, the AUTOSAR standard specifies the APIs and features of these components. The SWC communicates with other modules (inter-ECU or internal) only through the operating environment.

As ECUs continue to integrate more and more features, FPGA devices are a smart alternative to single-core or multi-core MCUs. By grasping the different levels of AUTOSAR in general, you can anticipate the advantages that designers can bring to deploy this architecture in programmable logic. The following section provides a more in-depth look at how our design implements a solution based on custom static hardware (flash or SRAM-based FPGA technology) and then extends this approach to a runtime reconfigurable hardware implementation (based on SRAM). Part of the reconfigurable FPGA).

Led Lighting

Led Lighting,Aluminum Material Reflector,Anodized Parabolic Reflector,Parabolic Grow Light Reflector

Yangzhou Huadong Can Illuminations Mould Manufactory Co., Ltd. , https://www.light-reflectors.com

This entry was posted in on