Fraunhofer IIS is going to present their proof-of-concept demonstrator for an ASA transceiver implementation including live stream from the lab. We will illustrate the uses of an ASA SerDes link in classic P2P connections and show, how the ASA standard and Fraunhofer implementation is pushing the envelope of performance and efficiency with regard to available data rates, high-power-over-cable, EMC performance and testing support. Furthermore, we are going to talk about the possible second generation use cases which employ daisy-chaining applications, duplication for high availability, distributed timed sensor fusion, integration with other IVN technologies as well as extension of encapsulated formats.

We are amid a tremendous transformation in the automotive industry. The consumer demand for greater vehicle functionality is spurring a lot of complex electronics in cars with increased number of electronic control unit (ECUs). Latest automotive in-vehicle networks contain more devices and components than ever before to produce, transmit and receive data at higher rates. With all this growth and variety of applications to support from the channel perspective, one question that the industry hears a lot is what to use, Single-ended solutions or Differential solutions? This presentation will review the pros and cons for each one, followed by the SI & EMI/EMC performance parameters from an assembly for each category (single-ended (coax cable terminated with coax connectors on both ends of the cable) and differential (Shielded Twisted Pair or Shielded Parallel Pair with fully shielded differential connectors on both ends of the cable). Everyone has heard of one being used more predominantly in some specific applications like cameras for example, but this presentation will help understand why that is. Knowing these performance differences will help the industry come up with technical reasonings to use one over the other in certain applications. After understanding the performance limitations for each solution, this will allow us to use them in the right applications and create overall solutions that are most optimal in terms of relative cost, performance, and ease of implementation. It will also allow us to create standards and specifications limits for channel in a manner that is reasonable for the various applications of the Automotive world.

The evolution of autonomous driving vehicles will result in fundamental changes in requirements for automotive architectures. Recently discussed approaches show a trend towards centralization of computation power on one side and the need for redundancy and decentralized aggregation of high-speed sensor and camera data on the other side. This development requires robust data transmission across the car. The discussion will include implications on the component limits to fulfill future application needs related to bandwidth, robustness and integration on Board Level.From the cabling hardware and harness point of view, a robust system design is needed to meet the requirements for future SerDes applications. Especially for high data rates of 10 Gbps and above, the automotive industry enters new ground and there is the question about the maximum data rate that can reliably be transmitted via electrical wires in a car. The presentation will answer this question by describing the physical effects and limitations of the link segment. Analysis and simulation results will be discussed about the available data capacity of wire-based transmission channels depending on parameters such as link topology, performance of the cabling components and constraints regarding electromagnetic compatibility. Sufficient margin is required in the definition of electrical parameters for connectors and cables to allow environmental variations, process tolerances and changes of harness lengths including different numbers and locations of inline connectors in a link. In order to perform a realistic channel analysis it is essential to extract the high-frequency characteristics of mated connector pairs from measurements. The talk will illustrate the use of advanced measurement methods to overcome the effects of test fixtures. The discussion will conclude, how the components match to current and future application needs, being able to enhance in car communication as the nerve cord of future car architectures.

New technologies like advanced driver-assistance, increase automation level in modern cars. Implementation of these technologies requires reliable multi-gigabit/s communications between scores of in-vehicle sensors and control units. Since data rates of the mainstream car communication technologies are several orders of magnitude lower than necessary, novel communication protocols arise. Automotive Ethernet symmetric protocols provide a backbone for vehicle communications. Other applications like data acquisition from cameras and LIDARs or sending data to displays, motivate development of asymmetric protocols, such as Automotive SerDes. To allow simple interoperability between chipsets, new protocols require more complex and rigorous compliance testing than the earlier automotive standards. Such testing includes measuring quality of the transmitted signal and characterization of the communication channel. Compliance testing should also take into account the fact that in automotive environment signals undergo severe disturbances due to power losses, noise and crosstalk. Even if a signal was perfect at the transmitter, it would get distorted when reaching receiver, and recovering data from this degraded signal can be a challenge. Thus, to ensure successful data communication, transmitter testing is not sufficient and receiver physical layer (Rx PHY) testing is necessary. Rx PHY Test specifications are defined today for most of the multi-gigabit communication protocols. The principal goal of Rx PHY tests is the verification of the bit error rate under physical layer disturbances such as crosstalk or symbol rate deviations. Further tests can improve the reliability. In the first part of this talk, BitifEye would like to share what we have learnt from our 15 years of experience in developing Rx PHY Testing solutions, describe the most important tests required for a proper characterisation of the receiver and the benefits they bring for both test houses and design engineers. Rx PHY tests are usually more complicated than Transmitter PHY tests and involve more equipment. While for testing transmitter it is sufficient to have an off-the-shelf oscilloscope that analyse the signal produced by the transmitter (eye characteristic), Rx tests require to recreate a compliant transmitter, capable of generating a controlled distorted signal, and to analyse the impact of the distortions on the transmission. Generator can be either a part of a Bit-Error-Rate Tester (BERT), an Arbitrary Waveform Generator or a media convertor, translating the signal into the given standard. Error detector is either a part of BERT or a built-in block of the Device-Under-Test. After describing the tests for characterising a receiver, BitifEye will present the advantages and drawbacks of the 3 possible approaches to transmitter emulation, the challenges in designing a proper Rx PHY testing and offer the audience the elements to decide on the most suitable solutions for SerDes Alliance.

Helping to address and study the challenges faced by automakers, suppliers and vehicle researchers, the level of sophisticated debug and data logger has never been greater. To properly understand what has happened minutes prior to a disengagement, we must be able to record video data with its detection classifications to ensure we can reconstruct and better understand the events leading up to the disengagement. Being able to record and synchronize multiple data sources like Video, CAN-FD and Ethernet will help address these design challenges. Our most precious asset is time, and test tracks are eager to sell you as much of it as you need. But are OEM’s, designers and suppliers utilizing this time properly. With the amount of data being generated in a vehicle today, having a solution that address the recording and the storage challenges will help speed the pace of development. This presentation will help show how to calculate the amount of bandwidth and storage needs when dealing with transmitting a cameras RAW images in various pixel formats. Together being able to capture raw vision data along with other communication buses like CAN-FD and Ethernet helps engineers properly evaluate developments and disengagements helping to speed along development in a reliable and robust approach.

One of the biggest challenges in the implementation of autonomous driving (AD) is validation and testing. Data-driven development and testing are getting more and more popular due to the high costs involved in test drives. In-vehicle data logging features dozens of Automotive Ethernet and Gbit/s SerDes sensor data (camera, radar, lidar) channels, and requires high-end logging solutions and large storage. This data has to be managed, labeled, and replayed in the lab and in the cloud. Another option to test AD hardware and software is hardware-in-the-loop (HIL) testing with virtual/synthetic sensor data. This test method usually occurs in later development phases. Automotive SerDes interfaces are widely used for cameras and will become increasingly popular for lidar and radar sensors. This talk comprehensively addresses the challenges and possible solutions using integrated tooling for AD validation and testing. A wide variety of relevant sensor interfaces are covered, both proprietary (e.g., Maxim GMSL, TI FPD-Link) and standardized (e.g., MIPI CSI-2 and D-PHY).

Design and integration complexity has continued to increase along with the growing number of cameras, sensors and displays needed for autonomous driving, ADAS and infotainment systems. This complexity is further compounded by the stringent requirements for functional safety and security in the automotive environment. With the completion of its MIPI A-PHY “long reach” automotive physical layer specification, MIPI Alliance has created the foundation for what will be an end-to-end system designed to allow native connections to its widely used MIPI CSI-2 and DSI-2 specifications for camera and display, as well as other third-party protocols, while also incorporating functional safety and security across all layers. This presentation will provide an overview of how MIPI is addressing these system-level issues, highlight the components of the system and describe how this standardized approach will benefit the automotive ecosystem.

The Automotive SerDes Market is really a brown field market. Asymmetric (Multi-) Gbps-Links, specialized on the transmission of uncompressed video from camera to vision processor, or from central media server to displays exist today as proprietary point solutions from a variety of suppliers. In the past, this has been accepted by OEMs, as the complete system, consisting of video source and sink, has often been designed and tested end-to-end by the same Tier-1 supplier. However, now we see several mega trends are changing the landscape drastically. OEMs have to reconsider their E/E-Architectures and supply chain to achieve electrification, service orientation and autonomous driving. As a result, the industry is standardizing on hardware to enable configurability in software. Like for other in-vehicle-networks (IVN) as well, this change has a huge effect on the requirements for automotive SerDes links. The most urgent requirement, to enable interoperability and less dependency on complete system suppliers, is the introduction of a SerDes standard itself. This goal is currently pushed ahead by the Automotive SerDes Alliance (ASA). SerDes Links are used in a variety of systems with completely different requirements. Speed grade, interface, latency and data type requirements vary depending on application and use case. Safety and security requirements depend on autonomous driving level and OEM choice, and solutions need to scale. This is essential to enable the evolution of E/E architectures from Domain based to Zonal. This paper presents how the standardized ASA SerDes will address those key requirements and enable efficient solutions for any future E/E architecture.

It is not just the number of displays in vehicles that is increasing globally, but also the size and resolution of the panels. The larger size and the resolution of the panel will open the door for new applications, and new applications always demand further enhancement of the hardware in turn; larger resolution, larger size. Therefore, it is becoming a very challenging, but interesting market for those suppliers who provide display components such as display controllers, timing controllers, source drivers, panels, and backlights. On the other hand, the Automotive industry is currently shifting towards the domain architecture where all the different applications are executed via a central ECU. More and more functions are required to be integrated in the cockpits of the future, such as safety, high resolution support and peripheral system control. This means the central ECU must also become powerful enough to drive such multiple displays remotely. Thus, the central ECU is getting more powerful, and the display component is becoming more demanding. This means the connection between them must also evolve: SerDes. The choice of SerDes is therefore very important not only for the technical level, but also for the architecture and the concept level. When we look at the trend of cockpit displays nowadays, some car OEMs are showing unique “Pillar to Pillar” concepts already. And it won’t be a big surprise anymore to see, for example, 4-5 panels around the cockpit. Here we can see two fundamental requirements; high bandwidth to drive up to e.g. 8K x 1K, and a daisy chain capability, because it is very unlikely that the domain ECU is supporting 5 video outputs. With the increase in available compute power, the domain architecture is expected to yield towards a more centralized software driven architecture, for Ethernet type structures. However, display streams with their high data bandwidth and low latency requirements will retain a dedicated data path. Nonetheless there will be significant interaction between the two areas, often sharing the same physical medium. Thus, SerDes will need to support increasing video rates as well as more generic and slower network standards such as Ethernet. For example, it will be important to consider how the display applications can be combined with camera applications in such an architecture. Considered from such a market trend, several key requirements for SerDes can be identified; payload bandwidth, preferred standard for the sideband signal, the interface to the peripheral devices, noise robustness, etc. We are pleased to make a presentation focusing on the display applications in order to identify several mandatory requirements for the future automotive SerDes technology, as well as the requirements and the challenges on this application.

The requirements for the camera system are getting more and more complicated. As required detection range/angle increases, the required resolution of imager goes from 1-2Mpix to 8Mpix, and will be over 10Mpix in the future. Also, since HDR performance is the key characteristics of automotive cameras, bit-depth per each pixel is much higher than other applications. On the other hand, camera manufacturers are required to minimize the product size and BOM cost. we believe integrating serializer into the imager could provide a solution for these challenges. In this presentation, The key benefits and challenges of imager integrated serializer, will be discussed.

Automobiles are subject to electrical overstress (EOS) events in the form of switching transients, load dump, electrostatic discharge (ESD), and lightning. Modern automobiles utilize multiple sensors, cameras and displays in advance diver-assist systems (ADAS), which rely on IC’s with operating speeds up to 10Gbps manufactured on increasingly smaller device geometries. These IC’s are extremely sensitive to the effects of transient over-voltages. Adding on-chip protection becomes a costly if not impossible task. Therefore, external protection devices designed to absorb high-energy surges are often required to supplement vulnerable components. Traditional protection devices such as p-n junction ESD diodes, zener diodes, and polymers are not adequate to fully protect and maintain signal integrity in today’s automotive, high-speed SerDes interfaces. These solutions can yield a false sense of security or worse, may interfere with the normal operation of the circuit. As design geometries shrink and Bus speeds continue to increase, engineering TVS diodes is more challenging than ever. Most modern TVS devices are intricate and multi-junction, designed for specific applications. This paper will discuss advances in protection device designs that make them suitable for protection of sensitive components in in high-speed ADAS interfaces as well as the device characteristics to consider when incorporating automotive qualified TVS into high-speed designs.

Automotive Applications like ADAS and V2X are adding complex Embedded Electronics in the car. These complex Embedded Electronics and sensors are interconnected with High speed Serial standards like Automotive Ethernet and network standard like ASA. To meet safety and reliability requirement, Automotive SERDES standards need to be validated not only at component but also at system level. Along with Compliance test, user need to perform ECU system performance test like Electromagnetic Noise, Signal Quality, Timing measurement for Gateway etc. Above test requires multiple Test and measurement tools along with Test software. In this presentation, we will introduce Vehicle Network Test Automation platform which integrates and automates component and system level test tools with single User interface. We will cover compliance test tools and system level test tools required for High speed Automotive SERDES validation. We will introduce Automated platform which can help users uncover some of the system level issues at early stages.