Shulou(Shulou.com)06/01 Report--

This article analyzes "what does TSN, smart driving and edge computing have to do with it?" The content is detailed and easy to understand. Friends who are interested in "what is the relationship between TSN, smart driving and edge computing" can follow the editor's train of thought to read it deeply. I hope it will be helpful to you after reading. Let's follow the editor to learn more about "what is the relationship between TSN, smart driving and edge computing".

Yesterday, a piece of news was brushed on the screen in the time-deterministic network group: with the opening of Chuangshi technology, SAIC intelligent driving took another crucial step. For time-triggered Ethernet and TTTech, the author has introduced one person, one idea, one company and the network around the world that is about to be changed, and predicts that introducing time-triggered Ethernet will be a major trend in future networks. The cooperation between SAIC and TTTech is a milestone for TTTech to open the domestic market.

On the morning of January 16, Shanghai Chuangshi Automotive Technology Co., Ltd. (referred to as "Chuangshi Technology") held the opening ceremony in the Innovation Port of Shanghai International Automobile City. Shanghai Chuangshi Automotive Technology Co., Ltd. is a joint venture between SAIC Group and Austria TTTech Computertechnik AG (hereinafter referred to as TTTech) with a share ratio of 50.1 to 49.9 on March 13, 2018. Chuangshi Technology is committed to promoting the industrialization of intelligent driving central decision controller (i-ECU) of intelligent network-connected automobile core components, and comprehensively enhance the core technical capabilities in the field of intelligent driving.

Founded in 2009, TTTech, headquartered in Austria, is a unicorn in the field of technology. TTTech serves aerospace, highway, construction machinery, automobile and other industries. In the automotive industry, TTTech serves Audi, BMW, Volkswagen, volvo and many other automotive industry leaders, developing products covering the intelligent driving domain control strategy of Level 3-5.

In addition to time-triggered Ethernet (TTE, strong real-time), time-sensitive network (TSN) is also a choice for on-board control systems. Whether the control system of the car uses TTE or TSN, the result is unknown, but no matter which one is, there is no doubt that the time deterministic joint Ethernet is introduced into the control system of the car. For a long time, the communication bus on the automobile is CAN bus, but with the requirement of autopilot or intelligent driving, the performance of CAN bus, such as communication bandwidth and delay, has gradually exposed its shortcomings. Later, the Flexray bus was invented in 2000 to solve the problem. Today, the core time trigger mechanism and seamless redundancy concept in the TSN standard come from Flexray, but the highest transmission rate of Flexray is only 20Mbps, which is far from meeting the requirements of complete self-driving. In 2006, the IEEE802.1 working Group established the AVB Audio and Video bridging Task Force, and in the following years successfully solved the problem of real-time synchronous data transmission in the audio and video network. This point immediately attracted the attention of people from the fields of automobile and industry. In 2012, the AVB Task Force expanded the application requirements and scope of application of time-deterministic Ethernet in its charter, and changed the name of the Task Force to the current TSN Task Force.

TSN actually refers to a set of "sub-standards" based on specific application requirements under the framework of IEEE802.1 standards, which aims to establish a "universal" time-sensitive mechanism for Ethernet protocols to ensure the time certainty of network data transmission. At the same time, data shaping is used to ensure microsecond latency (typically no more than 250 microseconds per bridge), and FrameReplication and Elimination frames are copied and eliminated to ensure reliable communication regardless of link failures, cable breaks, and other errors. This option ensures that copies of critical traffic can be sent in a non-intersecting path in the network, retaining only any packets that arrive at the destination first, resulting in seamless redundancy and ultra-high reliability, which is especially important for self-driving.

TSN ensures that the requirements of self-driving control can be met through IEEE802.1AS-Rev (time synchronization), IEEE802.1Qbv (low delay queue management and scheduling), IEEE802.1Qbu/802.3br (important data preemption mechanism), IEEE802.1CB (redundancy mechanism to ensure high reliability) and so on. A series of protocols in TSN ensure that all kinds of control systems can be integrated into the IoE scenario of the 5G era, which is the basic requirement of various network upgrades in the 5G era. This is also an important basis for the author to mention the term "the world network that is about to be changed" in the article "one person, one idea, one company and the world network that is about to be changed", and it is also an important reason why the market valuation of TSN technology will reach several trillion yuan in the future. From this point of view, the application scenario of TTE is relatively limited, but some closed situations of strong real-time still need TTE. For a more detailed introduction of vehicle Ethernet TSN, please refer to the article "A brief Analysis of vehicle Ethernet TSN Standard".

Domestic network companies, including Huawei, have carried out TSN and other aspects of research, constantly looking for whether to find the basic conditions mentioned in the early days of the establishment of the TSN standard to achieve the Internet of everything: a general interface standard, hoping to achieve seamless access to the network of equipment made by different manufacturers through the standard. On the other hand, how to reduce the network load and cloud pressure by enhancing the ability of edge nodes, the concept of edge computing arises at the historic moment.

On January 3, 2019, at the second European Edge Computing Forum (ECF), Huawei reached a cooperation intention with Analog Devices,ARM,Bombardier,B&R Automation,Fraunhofer Institute for Open Communication Systems (FOKUS), German Edge Cloud (GEC), German Research Center for Artificial Intelligence (DFKI), HARTING IT,IBM,Intel,KUKA, National Instruments,Renesas Electronics,Schneider Electric,Software AG,Spirent,TTTech and other partners to jointly establish the European Edge Computing Industry Alliance (Edge Computing Consortium Europe). Referred to as ECCE). ECCE aims to provide a full range of edge computing industry cooperation platform for manufacturers and organizations in the fields of intelligent manufacturing, operators, enterprises and IoT, through industry coordination, open innovation and demonstration promotion, deepen the digital transformation of the industry, and jointly promote the vigorous development of the edge computing industry. As a bridge between the physical and digital world, edge computing is a distributed open platform that integrates the core capabilities of network, computing, storage and application on the edge side of the network close to the object or data source, and provides edge intelligence services nearby. to meet the key needs of industry digital in agile connection, real-time business, data optimization, application intelligence, security and privacy protection. According to analysts, edge-side data processing will grow rapidly, and by 2025, 75% of enterprise-generated data will be created and processed outside the data center or cloud, far more than less than 20% today. Stefan Schonegger, Vice President of Product Strategy and Innovation at Berkeley: "We believe that edge computing is critical to the factories of the future." Processing data near the data source would ideally complement the public cloud-based Internet of things solution. Bergale is committed to rapidly expanding our Edge portfolio and providing the market with a comprehensive range of OPC UA and TSN-based Edge products. ECCE is helping to align our activities with the broad ecosystem of partners, which helps accelerate adoption and drive innovation. "

Intelligent network combined car and autopilot is a typical application scenario of edge computing.

With the development of machine vision, deep learning and sensor technology, the function of automobile is no longer limited to the traditional means of travel and transportation, but gradually becomes an intelligent, interconnected computing system. We call this new type of car intelligent network-connected vehicle (Connected and Autonomous Vehicles,CAVs). The emergence of intelligent network combined car has given birth to a series of new application scenarios, such as autopilot, vehicle networking and intelligent transportation. Intel's 2016 report pointed out that the data generated by a self-driving vehicle in a day is 4TB, which cannot be uploaded to the cloud for processing and needs to be stored and calculated in edge nodes (cars).

Autopilot computing scene is undoubtedly one of the hottest research directions at present. There are classic autopilot algorithm evaluation data set KITTI and classical visual algorithms for different autopilot stages around this scene. At present, there are many computing platforms for CAVs scenarios in the industry, such as NVIDIA DRIVE PX2 and XilinxZynq UltraScale+ ZCU106. At the same time, a lot of cutting-edge work in academia also began to explore the system design of the edge computing platform in the CAVs scenario. Liu divides autopilot into three processing phases: Sensing, Perception and Decision-making, and compares the execution effects of the three stages on different heterogeneous hardware, thus summarizing the matching rules between autopilot tasks and execution hardware. Lin and others compared the latency and power consumption of the three core applications in the perception phase, location (Localization), identification (Object Detection) and tracking (Object Tracking) in different combinations of GPUs,FPGAs and ASICs, to guide researchers to design an end-to-end autopilot computing platform. In addition to the hardware system architecture design, there is also a kind of research that has launched a complete software stack to help researchers implement autopilot systems, such as Baidu's Apollo and Japan's Waseda University's Autoware.

After the technical reserve in recent years, edge computing has been widely valued and recognized by governments, academia and industry at home and abroad. Now it is time to blossom and bring economic benefits. Thanks to the rapid development of seven key technologies: network, isolation technology, architecture, operating system, algorithm execution framework, data processing platform and security and privacy, edge computing technology has become mature and plays a role in many application scenarios. If this trend continues, edge computing will produce greater spillover effect, become the adhesive of various industries and the catalyst for the development of intelligent industry, and promote the upgrading and transformation of the whole industrial system. But at the same time, we must also see that it is difficult to integrate these related technologies, such as TSN, edge computing, autopilot intelligent algorithms and so on.

So much for the relationship between TSN, smart driving and edge computing. I hope the above content can improve everyone. If you want to learn more knowledge, please pay more attention to the editor's updates. Thank you for following the website!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.