List of Project Publications & Articles
The full list of academic publications done during the project for the support and sharing of the FRACTAL technology development.
Table of Contents
“EDISON: An Edge-native Method and Architecture for Distributed Interpolation“
FRACTAL project relevance: In their article, researchers from the University of Oulu present EDISON – a novel edge-native, distributed interpolation architecture for the smart city networking environment. EDISON’s device layer comprises fixed sensors as well as mobile sensors mounted on vehicles. IoT gateways provide connectivity, store mobile sensor observations, and provide local computational capabilities. The edge layer has edge servers and enhances the fixed sensors with connectivity and further computational capacity. Cloud provides coordination and centralized processing. Is part of the academic study on FRACTAL AI.
“A dark and stormy night: Reallocation storms in edge computing“
FRACTAL project relevance: The article further expands on our earlier anaysis of reallocation storms in edge computing (published in the conference artice “Weathering the reallocation storm: A large-scale analysis of edge server workload” in 2021), a phenomenon which under certain circumstances causes massive unnecessary traffic in large edge networks. This is relevant for WP6 in Fractal. Efficient resource usage in edge computing requires clever allocation of the workload of application components. In this paper, we show that under certain circumstances, the number of superfluous workload reallocations from one edge server to another may grow to a significant proportion of all user tasks—a phenomenon we present as a reallocation storm.
“Research and Education Towards Smart and Sustainable World”
FRACTAL project relevance: This paper proposes a vision for directing research and education in the field of information and communications technology (ICT). Our Smart and Sustainable World vision targets prosperity for the people and the planet through better awareness and control of both human-made and natural environments. The needs of society, individuals, and industries are fulfilled with intelligent systems that sense their environment, make proactive decisions on actions advancing their goals, and perform the actions on the environment.
“MiniFloats on RISC-V Cores: Supporting Scalar and Vector ISA Extensions for Mixed-Precision Floating-Point”
“Weathering the Reallocation Storm: Large-Scale Analysis of Edge Server Workload”
FRACTAL project relevance: The topic of the paper is edge orchestration. In more detail, the article analyses the workloads accumulating on a large deployment of edge servers in a number of scenarios and with different workload reallocation / migration strategies. We identify a phenomenon (named the “reallocation storm”) where a high number of reallocations on the edge network are superfluous, that is, happening unnecessarily. We study the causes behind reallocation storms, and look for ways to avoid them.
“A 10-core SoC with 20 Fine-Grain Power Domains for Energy-Proportional Data-Parallel Processing over a Wide Voltage and Temperature Range”
FRACTAL project relevance: Power gating is a well-known technique to reduce leakage power in large System-on-Chips, in this paper, FRACTAL partner ETH Zurich has explored a fine-grained power gating technique with 20 individually controlled power islands that can be turned on and off on demand saving up to 41% power.
“On-Demand Redundancy Grouping: Selectable Soft-Error Tolerance for a Multicore Cluster“
FRACTAL project relevance: With the shrinking of technology nodes and the use of parallel processor clusters in hostile and critical environments, such as space, run-time faults caused by radiation are a serious cross-cutting concern, also impacting architectural design. This paper introduces an architectural approach to run-time configurable soft-error tolerance at the core level, augmenting a six-core open-source RISC-V cluster with a novel On-Demand Redundancy Grouping (ODRG) scheme.
“MiniFloat-NN and ExSdotp: An ISA Extension and a Modular Open Hardware Unit for Low-Precision Training on RISC-V Cores”
FRACTAL project relevance: Low-precision formats have recently driven major breakthroughs in neural network (NN) training and inference by reducing the memory footprint of the NN models and improving the energy efficiency of the underlying hardware architectures. Narrow integer data types have been vastly investigated for NN inference and have successfully been pushed to the extreme of ternary and binary representations. In contrast, most training- oriented platforms use at least 16-bit floating-point (FP) formats. Lower-precision data types such as 8-bit FP formats and mixed- precision techniques have only recently been explored in hardware implementations. We present MiniFloat-NN, a RISC-V instruction set architecture extension for low-precision NN training, providing support for two 8-bit and two 16-bit FP formats and expanding operations.
“Improving the Robustness of Redundant Execution with Register File Randomization”
FRACTAL project relevance: This article proposes RFR, a new technique to increase the robustness of redundant execution in the context of safety-critical applications. RFR randomizes the mapping of processors architectural registers to avoid common cause faults in the FRACTAL-based RISCV computing nodes. It supports project activities on HW support for safety.
“The ECSEL FRACTAL Project: A Cognitive Fractal and Secure edge based on a unique Open-Safe-Reliable-Low Power Hardware Platform“
FRACTAL project relevance: This paper has presented the ECSEL FRACTAL project. The aim of the project is to create a reliable computing platform node, realizing a so-called Cognitive Edge under industry standards. The project builds on knowledge of partners gained in current or former EU projects and will demonstrate the newly conceived approaches to co-engineering across use cases spanning Transport and Industrial Control. As the paper is written at the beginning of the project, it focuses rather on the project introduction.
“Time-Triggered Frequency Scaling in Network-on-Chip for Safety-Relevant Embedded Systems”
FRACTAL project relevance: Networks on Chip (NoC) are used in Multiprocessor System on a Chip (MPSoC) architectures and safety-relevant systems as a communication backbone to cope with high communication traffic. Nevertheless, network interfaces and the routers within the NoC add to the power consumed by the chip. This paper presents models and algorithms for low power techniques in NoC based on dynamic frequency scaling for safety-relevant real-time systems.
“Extension of the LISNoC (Network -on-chip) with an AXI”
FRACTAL project relevance: Over the years, Network-on-Chip (NoC) has under-gone a rapid evolution which urges the performance of NoCs to be analyzed thoroughly. Several NoC solutions exist, but the performance of these NoCs are tied to application requirements. Therefore, it has become practical to extend existing NoCs to satisfy particular application requirements. The LISNoC is one such NoCs that is open-source and provides an easily adaptable implementation to extend its features to satisfy different application requirements. Mainly, this work extends the LISNoC to support source-based routing and equips the LISNoC with a new AXI-based network interface.
“Adaptive Scheduling for Time-Triggered Network-on-Chip-Based Multi-Core Architecture Using Genetic Algorithm”
FRACTAL project relevance: In this work, an algorithm for path reconvergence in a multi-schedule graph, enabled by a reconvergence horizon, is presented to manage the state-space explosion problem resulting from an increase in the number of scenarios required for adaptation.
“AI-Based Scheduling for Adaptive Time-Triggered Networks”
FRACTAL project relevance: Time-triggered systems are ideal for safety-critical systems due to the inherent determinism and better fault tolerance. However, the current trend of adaptation in time-triggered systems is typically limited to switching between a small number of precomputed schedules. Artificial neural networks (ANNs) have the potential to overcome this limitation. In this paper, an ANN is implemented to learn schedules to provide adaptation for time-triggered systems while ensuring that collision and precedence constraints are met. In our evaluation, the AI-based scheduler is compared with conventional scheduling algorithms such as list scheduling and genetic algorithm in terms of makespan and computation time. The results show the AI-based scheduler’s potential when increasing the scheduling problem’s complexity.
“Graph Neural Networks Based Meta-scheduling in Adaptive Time-Triggered Systems”
FRACTAL project relevance: Meta-scheduling algorithms are used for adaptation in time-triggered systems as they adapt to different scenarios such as failures or different environmental conditions. Most meta-scheduling algorithms demand a considerable amount of storage space from the host cyber-physical system due to the state-space explosion problem in covering a reasonable number of scenarios. This work deploys the Graph Neural Network (GNN) to learn the multi-schedules from the meta-scheduling algorithm required for adaptation. The results show that the proposed GNN-based meta-scheduling can be suitable for real-time scenario adaptation in cyber-physical systems.
“Ternarized TCN for μJ/Inference Gesture Recognition from DVS Event Frames”
FRACTAL project relevance: Dynamic Vision Sensors (DVS) offer the opportunity to scale the energy consumption in image acquisition proportionally to the activity in the captured scene by only transmitting data when the captured image changes. Their potential for energy-proportional sensing makes them highly attractive for severely energy-constrained sensing nodes at the edge.
“Systematic Prevention of On-Core Timing Channels by Full Temporal Partitioning“
FRACTAL project relevance: Microarchitectural timing channels enable unwanted information flow across security boundaries, violating fundamental security assumptions. They leverage timing variations of several state-holding microarchitectural components and have been demonstrated across instruction set architectures and hardware implementations. We find that a complete, systematic, ISA-supported erasure of all non-architectural core components is the most effective implementation while featuring a low implementation effort, a minimal performance overhead of approximately 2%, and negligible hardware costs.
“Leveraging the PULP Platform to Build Reliable Systems“
FRACTAL project relevance: PULP platform is used and validated for different use case and system applications within the project.
“The “Great Beauty” of Diversity: Smart Totems to Promote Gender Uniqueness”
FRACTAL project relevance: This paper presents an Internet-of-Things solution represented by smart totems for advertisement and wayfinding services within advanced ICT-based shopping malls, that are conceived as a sentient space. To take into account the specificity of final users and to increase awareness of the gender uniqueness in the design of Internet-of-Things, a User-Centered Design methodology has been used.
“Situation Awareness for Autonomous Vehicles Using Blockchain-Based Service Cooperation”
FRACTAL project relevance: To solve the issues of trust and latency in data sharing between stakeholders within the Intelligent Traffic Systems, we propose a decentralized framework that enables smart contracts between traffic data producers and consumers based on blockchain.
“Understanding and Mitigating Memory Interference in FPGA-based HeSoCs”
FRACTAL project relevance: Like most high-end embedded systems, FPGA-based systems-on-chip (SoC) are increasingly adopting heterogeneous designs, where CPU cores, the configurable logic and other ICs all share interconnect and main memory (DRAM) controller. Our experimental results show that: i) memory interference can slow down CPU tasks by up to 16×in the tested FPGA-based SoCs; ii) CMRI allows to exploit more than 40% of the memory bandwidth avail-able to FPGA accelerators (normally completely unused in PREM-like schemes), keeping the slowdown due to interference below 10%.
“Unboxing the Sand: on Deploying Safety Measures in the Programmable Logic of COTS MPSoCs”
FRACTAL project relevance: The lack of sufficient hardware support for functional safety precludes the full adoption of many Commercial Off-the-Shelf (COTS) MPSoCs in safety-related systems, such as those in the aerospace industry. Some recent MPSoCs come along with programmable logic (PL), primarily intended to offload some specific complex functions that can be much more efficiently implemented in hardware than in software, hence being such PL a kind-of-sandbox fully mastered by ASIC cores outside the PL. The early work presented in this paper already provides specific monitoring, diversity, and controlling strategies to allow PL take over safety-related functionalities.
“SafeSoftDR: A Library to Enable Software-based Diverse Redundancy for Safety-Critical Tasks”
FRACTAL project relevance: This paper presents SafeSoftDR, a library providing a standard interface to deploy software-based lockstepped execution across non-natively lockstepped cores relieving end-users from having to manage the burden to create redundant processes, copying input/output data, and performing result comparison. Our library has been tested on x86-based Linux and is currently being integrated on top of an open-source RISC-V platform targeting safety-related applications, hence offering a convenient environment for safety-critical applications.
“Adaptive Histogram Equalization in Diabetic Retinopathy Detection”
FRACTAL project relevance: Diabetic retinopathy is a dangerous pathology that can ultimately lead to permanent blindness. Recent studies have proved the feasibility of automatic diagnosis systems, supporting specialists, from eye fundus images, based on deep networks. A careful image equalization turned out to be important in obtaining good performances. However, state-of-the-art algorithms for image equalization require expensive parameter tuning which may limit the adoption of such support systems in practice. In this paper, we propose a learning-based approach to adaptively select the right equalization parameters for each image. This approach allows significant reduction in the inference time without necessarily sacrificing the accuracy. A preliminary empirical evaluation confirms the advantages of the proposed method.
“BAFFI: a bit-accurate fault injector for improved dependability assessment of FPGA prototypes“
FRACTAL project relevance: FPGA-based fault injection (FFI) is an indispensable technique for verification and dependability assessment of FPGA designs and prototypes. Existing FFI tools make use of Xilinx essential bits technology to locate the relevant fault targets in FPGA configuration memory (CM). Most FFI tools treat essential bits as black-box, while few of them are able to filter essential bits on the area basis in order to selectively target design components contained within the predefined Pblocks. This approach, however, remains insufficiently precise since the granularity of Pblocks in practice does not reach the smallest design components. This paper proposes an open-source FFI tool that enables much more fine-grained FFI experiments for Xilinx 7-series and Ultrascale+ FPGAs. Through case studies we show how the proposed tool can be applied to different kinds of DUTs: from small-footprint microcontrollers, up to multicore RISC-V SoC.
“Autonomy and Intelligence in the Computing Continuum: Challenges, Enablers, and Future Directions for Orchestration”
FRACTAL project relevance: The main focus of the article is on the AI-based orchestration in the device-edge-cloud computing continuum.
“Triggering Conditions Analysis and Methodology for validation of ADAS/AD functions: A use case study”
FRACTAL project relevance: Covers safety analysis topics related to UC7 (SPIDER autonomous robot).
“Spatial dependency in Edge-native Artificial Intelligence“
FRACTAL project relevance: Covers safety analysis topics related to UC7 (SPIDER autonomous robot).This thesis studies edge AI – an important part of work in the project, needed to deliver the knowledge for building the FRACTAL system. A nascent field of research combining edge computing and artificial intelligence. A particular focus in the thesis is on spatial dependencies, which quantify the similarity of observations in the spatial dimension. Spatial dependencies are prominent in edge AI due to the local nature of edge service users, the computational resources, as well as many of the observed data-generating processes.
“RISC-V as ultra-low power microcontroller“
“Advanced Reinforcement Learning based Thermal Management Strategy for Battery Electric Vehicles”
FRACTAL project relevance: The article covers the application of reinforcement learning to control strategies for electric vehicle thermal management and is related to Use Case 2 (Automotive air path control) of the project.
“Sentient Spaces: Intelligent Totem Use Case in the ECSEL FRACTAL Project“
“Advanced Vehicle Thermal Management using Ray and RLlib“
“Empirical Characterization of Wireless Connectivity Performance for Cognitive Edge IoT Nodes“
FRACTAL project relevance: Cognitive edge nodes are becoming increasingly important for various Internet of Things (IoT) applications, requiring reliable, efficient and ubiquitous communication. This paper evaluates the performance of direct cellular (5G) and IEEE 802.11-based Wireless Local Area Network (WLAN) technology for cognitive edge nodes supporting network architectures potential for FRACTAL edge platform. The FRACTAL edge platform is flexible, scalable and supports different wireless technologies, making it a suitable platform for implementing cognitive edge nodes. The study assesses the network performance in terms of throughput, latency, and power consumption for three different network architectures. The findings reveal that IEEE 802.11 technology is more energy-efficient and favourable for latency for peer-to-peer communication scenarios, while 5G technology demonstrates high throughput for communication between a test node and an upper-tier edge node.
“Empirical Characterization of Wireless Connectivity Performance for Cognitive Edge IoT Nodes“
FRACTAL project relevance: We present our approach for the development, validation, and deployment of a data-driven decision-making function for the automated control of a vehicle. The decision-making function, based on an artificial neural network is trained to steer the mobile robot SPIDER towards a predefined, static path to a target point while avoiding collisions with obstacles along the path. The training is conducted by means of proximal policy optimization (PPO), a state-of-the-art algorithm from the field of reinforcement learning. The controller is deployed on a FPGA-based development platform, the FRACTAL platform, and integrated into the SPIDER software stack for validation.