Autonomous vehicles (AVs) require consistently low-latency computations; however, operating system (OS) CPU scheduler can lead to high tail latency that threatens timely decision-making and safety. General-purpose OS schedulers prioritize fairness and throughput over individual task deadlines, posing challenges for complex AV workloads with mixed-criticality tasks. Despite extensive research on CPU scheduling, reinforcement learning (RL) has not been explored at the OS level due to feasibility concerns. This paper presents SCHED, the first RL-based CPU scheduling framework optimized for OS integration. SCHED learns a scheduling policy via reinforcement learning and deploys it as a lightweight, decision-tree-based scheduler. We demonstrate that SCHED remains sufficiently fast for OS-level integration compared to existing OS schedulers. Furthermore, we validate SCHED on a realistic autonomous vehicle pipeline, demonstrating its practical potential in maintaining low latency and ensuring safer, more responsive AV operations.

Self-Organizing Maps (SOM) promise a lightweight approach for multivariate time series anomaly detection in lightweight autonomous embedded systems. However, the enormous volume of time series data from autonomous systems testing requires huge SOMs with impractical search overhead. We present KDTree-SOM that effectively optimizes the winner node search for huge SOMs by reconstructing the SOM as a k-dimensional tree (kd-tree). KDTree-SOM achieves on average a 4 × inference time reduction for huge SOMs while achieving up to 95% anomaly detection accuracy with only KB-level memory overhead, demonstrating its potential for anomaly detection in lightweight autonomous embedded platforms.

The continuing advancement of storage technology has introduced ultra-low latency (ULL) SSDs that feature 20 μs or less access latency. Therefore, the context switching overhead of interrupts has become more pronounced on these SSDs, prompting consideration of polling as an alternative to mitigate this overhead. At the same time, the high price of ULL SSDs is a major issue preventing the wide adoption of polling.

We present ZoneTrace , a runtime monitoring tool for the Flash-friendly File System (F2FS) on Zoned Namespace (ZNS) Solid-state Drives (SSDs). ZNS SSD organizes its storage into zones of sequential write access. Due to ZNS SSD’s sequential write nature, F2FS is a log-structured file system that has recently been adopted to support ZNS SSDs. To present the space management with the zone concept between F2FS and the underlying ZNS SSD, we developed ZoneTrace , a tool that enables users to visualize and analyze the space management of F2FS on ZNS SSDs. ZoneTrace utilizes the extended Berkeley Packet Filter (eBPF) to trace the updated segment bitmap in F2FS and visualize each zone space usage accordingly. Furthermore, ZoneTrace is able to analyze on file fragmentation in F2FS and provides users with informative fragmentation histogram to serve as an indicator of file fragmentation. Using ZoneTrace ’s visualization, we are able to identify the current F2FS space management scheme’s inability to fully optimize space for streaming data recording in autonomous systems, which leads to serious file fragmentation on ZNS SSDs. Our evaluations show that ZoneTrace is lightweight and assists users in getting useful insights for effortless monitoring on F2FS with ZNS SSD with both synthetic and realistic workloads. We believe ZoneTrace can help users analyze F2FS with ease and open up space management research topics with F2FS on ZNS SSDs.

This paper summarizes the talks of a special session on the IPF 2.0 project, a collaborative German-US research project that leverages self-awareness principles for the self-management of distributed systems of autonomous multiprocessor systems-on-chip (MPSoCs).

The sequential write nature of ZNS SSDs makes them very well-suited for log-structured file systems. The Flash-Friendly File System (F2FS), is one such log-structured file system and has recently gained support for use with ZNS SSDs. The large F2FS over-provisioning space for ZNS SSDs greatly reduces the garbage collection (GC) overhead in the log-structured file systems. Motivated by this observation, we explore the trade-off between disk utilization and over-provisioning space, which affects the garbage collection process, as well as the user application performance. To address the performance degradation in write-intensive workloads caused by GC overhead, we propose a modified free segment-finding policy and a Parallel Garbage Collection (P-GC) scheme for F2FS that efficiently reduces GC overhead. Our evaluation results demonstrate that our P-GC scheme can achieve up to 42% performance enhancement with various workloads.

Linux prevents errors due to physical memory limits by swapping out active application memory from main memory to secondary storage. Swapping degrades application performance due to swap-in/out latency overhead. To mitigate the swapping overhead in periodic applications, we present ProSwap: a period-aware proactive and adaptive swapping policy for em-bedded systems. ProSwap exploits application periodic behavior to proactively swap-out rarely-used physical memory pages, creating more space for active processes. A flexible memory reclamation time-window enables adaptation to memory limitations that vary between applications. We demonstrate ProSwap's efficacy for an autonomous vehicle application scenario executing multi-application pipelines, and show that our policy achieves up to 1.26×performance gain via proactive swapping.

When executing a deep neural network (DNN), its model parameters are loaded into GPU memory before execution, incurring a significant GPU memory burden. There are studies that reduce GPU memory usage by exploiting CPU memory as a swap device. However, this approach is not applicable in most embedded systems with integrated GPUs where CPU and GPU share a common memory. In this regard, we present Demand Layering, which employs a fast solid-state drive (SSD) as a co-running partner of a GPU and exploits the layer-by-layer execution of DNNs. In our approach, a DNN is loaded and executed in a layer-by-layer manner, minimizing the memory usage to the order of a single layer. Also, we developed a pipeline architecture that hides most additional delays caused by the interleaved parameter loadings alongside layer executions. Our implementation shows a 96.5% memory reduction with just 14.8% delay overhead on average for representative DNNs. Furthermore, by exploiting the memory-delay tradeoff, near-zero delay overhead (under 1 ms) can be achieved with a slightly increased memory usage (still an 88.4% reduction), showing the great potential of Demand Layering.

Internet of Things (IoT) paradigm raises challenges for devising efficient strategies that offload applications to the fog or the cloud layer while ensuring the optimal response time for a service. Traditional computation offloading policies assume the response time is only dominated by the execution time. However, the response time is a function of many factors including contextual parameters and application characteristics that can change over time. For the computation offloading problem, the majority of existing literature presents efficient solutions considering a limited number of parameters (e.g., computation capacity and network bandwidth) neglecting the effect of the application characteristics and dataflow configuration. In this paper, we explore the impact of the computation offloading on total application response time in three-layer IoT systems considering more realistic parameters, e.g., application characteristics, system complexity, communication cost, and dataflow configuration. This paper also highlights the impact of a new application characteristic parameter defined as Output–Input Data Generation (OIDG) ratio and dataflow configuration on the system behavior. In addition, we present a proof-of-concept end-to-end dynamic computation offloading technique, implemented in a real hardware setup, that observes the aforementioned parameters to perform real-time decision-making.

Deep-learning-based intelligent services have become prevalent in cyber-physical applications, including smart cities and health-care. Deploying deep-learning-based intelligence near the end-user enhances privacy protection, responsiveness, and reliability. Resource-constrained end-devices must be carefully managed to meet the latency and energy requirements of computationally intensive deep learning services. Collaborative end-edge-cloud computing for deep learning provides a range of performance and efficiency that can address application requirements through computation offloading. The decision to offload computation is a communication-computation co-optimization problem that varies with both system parameters (e.g., network condition) and workload characteristics (e.g., inputs). However, deep learning model optimization provides another source of tradeoff between latency and model accuracy. An end-to-end decision-making solution that considers such computation-communication problem is required to synergistically find the optimal offloading policy and model for deep learning services. To this end, we propose a reinforcement-learning-based computation offloading solution that learns optimal offloading policy considering deep learning model selection techniques to minimize response time while providing sufficient accuracy. We demonstrate the effectiveness of our solution for edge devices in an end-edge-cloud system and evaluate with a real-setup implementation using multiple AWS and ARM core configurations. Our solution provides 35% speedup in the average response time compared to the state-of-the-art with less than 0.9% accuracy reduction, demonstrating the promise of our online learning framework for orchestrating DL inference in end-edge-cloud systems.

Self-driving systems execute an ensemble of different self-driving workloads on embedded systems in an end-to-end manner, subject to functional and performance requirements. To enable exploration, optimization, and end-to-end evaluation on different embedded platforms, system designers critically need a benchmark suite that enables flexible and seamless configuration of self-driving scenarios, which realistically reflects real-world self-driving workloads’ unique characteristics. Existing CPU and GPU embedded benchmark suites typically (1) consider isolated applications, (2) are not sensor-driven, and (3) are unable to support emerging self-driving applications that simultaneously utilize CPUs and GPUs with stringent timing requirements. On the other hand, full-system self-driving simulators (e.g., AUTOWARE, APOLLO) focus on functional simulation, but lack the ability to evaluate the self-driving software stack on various embedded platforms. To address design needs, we present Chauffeur, the first open-source end-to-end benchmark suite for self-driving vehicles with configurable representative workloads. Chauffeur is easy to configure and run, enabling researchers to evaluate different platform configurations and explore alternative instantiations of the self-driving software pipeline. Chauffeur runs on diverse emerging platforms and exploits heterogeneous onboard resources. Our initial characterization of Chauffeur on different embedded platforms – NVIDIA Jetson TX2 and Drive PX2 – enables comparative evaluation of these GPU platforms in executing an end-to-end self-driving computational pipeline to assess the end-to-end response times on these emerging embedded platforms while also creating opportunities to create application gangs for better response times. Chauffeur enables researchers to benchmark representative self-driving workloads and flexibly compose them for different self-driving scenarios to explore end-to-end tradeoffs between design constraints, power budget, real-time performance requirements, and accuracy of applications.