Reconfigurable Computing for Smart Vehicles

The number of electronic control units (ECUs) in modern vehicles has continuously increased in last few decades. In addition, advanced functionalities antilock braking systems, adaptive cruise control and climate control put higher computational demand on ECUs, which further increases the design complexity of automotive control systems. Multicore processors having multiple processing units are integrated on a single chip, have emerged to be the main computing controllers not only for high-end servers but also for embedded control systems. Using multicore processors, more centralized architecture designs can be adopted for automotive control systems. The recent initiative on AUTOmotive Open System ARchitecture (AUTOSAR) has established several standards for automotive software and hardware designs, which include guidelines for designing centralized architecture with multicore ECUs for automotive control systems. With AUTOSAR, it is expected that computational control tasks of different functions can share one ECU or run on any ECU connected with in-vehicle network like CAN and FlexRay. Also, In this paper, we describe some key analytical components that together characterize the hierarchical bus structures composed of FlexRay, CANbus and LINbus. Our analysis shows that the practical constraints imposed by standards such as AUTOSAR can lead to higher levels of schedulable resource utilization. This reduces both the overall component count and cost, while facilitating easy enhancements.

Consolidating tasks to a smaller number of electronic control units (ECUs) is an important strategy for optimizing costs and resources in the automotive industry. In our research, we aim to enable ECU consolidation by migrating tasks at runtime between different ECUs, which adds redundancy and fail-safety capabilities to the system. In this paper, we present a setup with a generalistic and modular architecture that allows for integrating and testing different ECU architectures and machine learning (ML) models. As part of a holistic testbed, we introduce a collection of reproducible tasks, as well as a toolchain that controls the dynamic migration of tasks depending on ECU status and load. The migration is aided by the machine learning predictions on the schedulability analysis of possible future task distributions. To demonstrate the capabilities of the setup, we show its integration with FreeRTOS-based ECUs and two ML models—a long short-term memory (LSTM) network and a spiking neural network—along with a collection of tasks to distribute among the ECUs. Our approach shows a promising potential for machine-learning-based schedulability analysis and enables a comparison between different ML models.

It is seen that the number of Electronic Control Units (ECUs) in the automotive has grown significantly in recent years. In today's luxury cars, up to 2500 signals are exchanged by up to 70-80 ECUs. The number communication channels implement required realize such systems is about n2, here n is the number of ECUs in the automobile. This complexity requires the use of communication networks between them and to manage the functionality of these ECUs the Real-Time Operating Systems (RTOS) is used. This real world scenario has motivated us to give the hands on experience to students by introducing integrated course projects, which includes the use of Real-Time Operating Systems (RTOS) in automotive embedded systems. This activity also involves creating an industry like environment for execution of the project. Deployment of more advanced control strategies and integration of sub modules were the major technical outcomes of the activity. For executing this activity, the problem statement given was divided into sub-problems and each sub problem is converted to sub module/ECU depicting the industry scenario. The concepts of automotive embedded systems were used during sub module development and during integration the concepts of RTOS were used. The activity provided the method and tool to facilitate the integration of different electronic subsystems coming from various suppliers. The activity created the feel of real world industry environment amongst the students and enabled to achieve various technical and professional outcomes of ABET. The processes of achieving these outcomes are discussed.

Modern vehicles represent a complex distributed cyber-physical system that simultaneously handles critical functions like drive-by-wire systems, non-critical functions like window/door control, and compute intensive multimedia functions. Distributed electronic control units (ECUs), which integrate processing elements and supporting peripherals (network interfaces, memory), implement a variety of functions in software, and information is exchanged between ECUs and sensors/actuators over in-vehicle networks. As the complexity of applications rises with increasing automation, extensive hardware support is required in the form of multicore processors or special purpose hardware accelerators to offer required levels of performance. Additional features also drive an increase in the number of ECUs since new functions are rarely consolidated on existing ECUs. Furthermore, network interfaces implemented as ASICs or dedicated logic must be adapted to handle increased communication loads, consuming power and requiring more infrastructure support in terms of cabling and weight. The increasing number of safety-critical functions further impacts complexity if existing one-to-one redundancy schemes are applied. Rising automation also poses news challenges like security, with researchers showing that internal networks are easily manipulated with catastrophic effects and total loss of control. Our research aims to address these challenges using architectural enhancements that are transparent at the computational and network levels, leveraging the capabilities of reconfigurable hardware. We present advanced ECU architectures with extended network capabilities, apply these in the context of safety critical systems, explore ways of extending these schemes to offer advanced security features, and show how such advanced systems can be validated in hardware. Our work represents an advancement in the state of the art with regard to applying FPGAs in vehicular systems. 7.3 Summary . . . . .

In the last few decades, advances in vehicular technology have necessitated large increases in the number of Electronic Control Units (ECUs) within vehicles. These ECUs are required to communicate with other ECUs in order to achieve full functionality. The communication networks that connect the ECUs have been shown to pose significant security risks, as they can be the target for attacks. Establishing security in vehicles against attacks is a complex topic that has generated much interest as a focus of research. In this article, we will highlight current security approaches in terms of hardware modules and physically unclonable functions (PUFs).

More and more electronic control functions have been added to vehicles. The number of electronic control unit (ECU) has also been gradually increased. Calibration work is an important part of ECU development. At present, the calibration software on the market follows the ASAP protocol standard. It reads and extracts the text-based A2L file. This A2L file is plaintext, which probably has a serious data security risk in the process of multi-user calibration. This paper presents and designs a calibration system with A2L files encryption and decryption based on 3DES algorithm. ECU calibration experiment results showed that based on the encryption and decryption method effectively ensures the A2L file reliability and security.

Due to the rising number of electronic control units (ECU) in a vehicle and the growing complexity of the related software installed, a fast and efficient system for updating software is needed. Wireless software updates similar to firmware over the air updates for smartphones can be a suitable solution to solve this issue. In this paper we propose a wireless update system based on an IEEE 802.11s mesh network and describe related high-level requirements for such a system. Additionally the prototype of a wireless vehicle interface (WVI) is described. This interface is needed to maintain the wireless link as well as to forward the received data to the in-vehicle communication system and finally to the ECU. Existing diagnostic standards are applied to transfer and install the new software on the ECU. Furthermore, IEEE 802.11s-based network nodes and the WVI prototype are used to evaluate the applicability of IEEE 802.11s for a wireless update system used in the vehicle development phase. We performed indoor measurements as well as measurements inside two different vehicles to evaluate the influence of the shielding properties of a vehicle. The results of these measurements show that the used setup consisting of the WVI prototype and other IEEE 802.11s based nodes can be used to realize a wireless update system and is able to fulfil the defined system requirements.

Modern technologies lead to more sophisticated hardware, while software is becoming more complex. These trends are widely present in consumer electronics and do not bypass automotive electronics either. There is an evident recent growth in in-vehicle infotainment, telematics, advanced driver assistance systems (ADASs) and cluster development. The number of electronic control units (ECUs) in vehicle constantly grows. Since typical vehicle ECU is providing one function per vehicle, it becomes harder for manufacturers to manage these ECUs due to diverse nature of the system, hence a rising demand for ECU consolidation exists. With the availability of sophisticated hardware, powerful system-on-chips (SoCs) can be used for multiple functions inside a vehicle. The transition toward less ECUs is an ongoing process, in which software needs to be aligned first and then transferred to the same SoC. This paper presents the software platform for heterogeneous immersive in-vehicle environments, providing a step in software consolidation, by allowing same abstractions for diverse applications executing on various hardware platforms. It proposes a framework for the scalable development of ADAS from consumer level to different automotive safety levels, provides unified access toward algorithm building blocks, multi-sensor real-time environment and easy integration of algorithms, thus enabling shorter development time.

In the current approach of automotive electronic system design, the multicore processors have prevailed to achieve high computing performance at low thermal dissipation. Multicore processors offer functional parallelism that helps in meeting the safety critical requirements of vehicles. The number of Electronic Control Units (ECUs) in high-end cars could be reduced by conglomerating more functions into a multicore ECU. AUTOSAR stack has been designed to support the applications developed for multicore ECUs. The real challenges lie in adapting new design methods while developing sophisticated applications with multicore constraints. It is imperative to utilize the most of multicore computational capability towards enhancing the overall performance of ECUs. In this context, the scheduling of the real-time multitasking software components by the operating system is one of the challenging issues to be addressed. Here, the state-of-the-art scheduling algorithm is reviewed and its merits and limitations are identified. A hybrid scheduler has been proposed, tested and compared with the state-of-the-art algorithm that offers better performance in terms of CPU utilization, average response time and deadline missing rate both in normal and high load conditions.

Recent years have seen an exponential rise in complex software-driven functionality in vehicles, leading to a rising number of electronic control units (ECUs), network capabilities, and interfaces. These expanded capabilities also bring-in new planes of vulnerabilities making intrusion detection and management a critical capability; however, this can often result in more ECUs and network elements due to the high computational overheads. In this paper, we present a consolidated ECU architecture incorporating an Intrusion Detection System (IDS) for Automotive Controller Area Network (CAN) along with traditional ECU functionality on an off-the-shelf hybrid FPGA device, with near-zero overhead for the ECU functionality. We propose two quantised multi-layer perceptrons (QMLP's) as isolated IDSs for detecting a range of attack vectors including Denial-of-Service, Fuzzing and Spoofing, which are accelerated using off-the-shelf deep-learning processing unit (DPU) IP block from Xilinx, operating fully transparently to the software on the ECU. The proposed models achieve the state-of-the-art classification accuracy for all the attacks, while we observed a 15x reduction in power consumption when compared against the GPU-based implementation of the same models quantised using Nvidia libraries. We also achieved a 2.3x speed up in per-message processing latency (at 0.24 ms from the arrival of a CAN message) to meet the strict end-to-end latency on critical CAN nodes and a 2.6x reduction in power consumption for inference when compared to the state-of-the-art IDS models on embedded IDS and loosely coupled IDS accelerators (GPUs) discussed in the literature.

The modern car is increasingly connected. That connection is magnified by the presence of a large number of electronic control units (ECUs). The communication between the ECUs of a modern car is assured by the Controller Area Network (CAN) bus system. Despite its importance, the CAN bus system is bereft of security mechanisms making it vulnerable to numerous security attacks. When an attacker succeeds in compromising the ECUs, they can take control and stop the engine, disable the brakes, turn the lights on/off, etc. An intrusion detection system (IDS) can be deployed as an appropriate security measure to detect the malicious network traffic in the CAN bus system. In this paper, we propose a Convolutional Neural Network (CNN)-based network attacks IDS for protecting the CAN bus system. For efficiency reasons, we generated our own datasets from three car models. Our experiment results demonstrate that our classifier is efficient for detecting the CAN bus system attacks, and it performs with a high accuracy of 99.99% and a detection rate of 0.99.

In the automotive domain the permanent increase in functionality led to a vast number of electronic control units (ECUs) in today’s cars, but packaging and network bandwidth demands became problematic in the last years. Thus it is vital to integrate more functions per ECU and shift in-car-networking complexity into software. To master this challenge, it is essential to find local and global optimization possibilities, which includes practical software component partitioning strategies while not overlooking the multiplicity of influence factors as well as smart software modules that help to reduce the energy demand wherever possible.

In recent years, the need for better comfort and safety in automobiles leads to the drastic increase in the number of electronic control units in automobiles. These extra features in a vehicle increase the complexity of the software in ECUs. Model-Based Software Development (MBSD) is widely used in automotive software development in order to reduce the cost and time meanwhile ensuring the safety of the product. This chapter explains the concept of MBSD in the automotive domain. This chapter also discusses the tools and techniques available in the market for Model-Based Software Development with provision of auto code generation. MATLAB/Simulink/Stateflow-based modeling approach is described with a brief introduction to Unified Modeling Language.

Identifying in-vehicle electronic control units based on voltage characteristics has been the subject of extensive research in cybersecurity. However, the results reported so far generally depend on restricted datasets and supervised learning. In this work, we show that clustering, i.e., unsupervised learning, of voltage characteristics, is in fact more challenging when done on a larger pool of electronic control units as several out-of-the-box clustering methods and metrics will fail to determine the correct number of clusters when exerted over a large dataset. To overcome this issue, we propose a new methodology that takes advantage of domain-specific constraints, which guide the search toward the correct number of electronic control units in a car, or even in a larger pool of units from several cars. We introduce two new metrics: correctness, which measures the success ratio with respect to the constraints, and divergence, which measures the consistency of the clustering, and show that they provide a strong indication for the optimal number of clusters. In this specific context, both metrics prove to be more reliable than the widely used Silhouette score, Davies-Bouldin and Calinski-Harabas indexes. We successfully test our methodology on the largest dataset available today for in-vehicle voltage characteristics and discover new insights regarding the number of devices.

Data Reduction (DR) algorithms can play very important role in case of in-vehicle networks. The information exchanged between Electronic Control Units (ECUs) on Controller Area Network (CAN) tends to become large. This is due to increase in number of ECUs on the bus. CAN bus with its limited bandwidth may get overloaded due to increase in data traffic. In this paper authors propose new DR algorithm namely, Boundary of Fifteen Compression (BFC) algorithm. It is shown that, BFC provides improvement in compression ratio as compared to earlier proposed Quotient Remainder Compression (QRC) algorithm and Enhanced Data Reduction (EDR) algorithm. Also BFC is less computation intensive as compared to QRC and EDR.