Abstract:Machine learning ensembles combine multiple base models to produce a more accurate output. They can be applied to a range of machine learning problems, including anomaly detection. In this paper, we investigate how to maximize the composability and scalability of an FPGA-based streaming ensemble anomaly detector (fSEAD). To achieve this, we propose a flexible computing architecture consisting of multiple partially reconfigurable regions, pblocks, which each implement anomaly detectors. Our proof-of-concept design supports three state-of-the-art anomaly detection algorithms: Loda, RS-Hash and xStream. Each algorithm is scalable, meaning multiple instances can be placed within a pblock to improve performance. Moreover, fSEAD is implemented using High-level synthesis (HLS), meaning further custom anomaly detectors can be supported. Pblocks are interconnected via an AXI-switch, enabling them to be composed in an arbitrary fashion before combining and merging results at run-time to create an ensemble that maximizes the use of FPGA resources and accuracy. Through utilizing reconfigurable Dynamic Function eXchange (DFX), the detector can be modified at run-time to adapt to changing environmental conditions. We compare fSEAD to an equivalent central processing unit (CPU) implementation using four standard datasets, with speed-ups ranging from $3\times$ to $8\times$.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to improve the composability and scalability of the FPGA - based streaming integrated anomaly detection system. Specifically, the authors propose fSEAD (Field - Programmable Gate Array - based Streaming Ensemble Anomaly Detection Library), which is a composable and low - latency FPGA library aiming to achieve more efficient, flexible and powerful anomaly detection through hardware acceleration. ### Main Problems: 1. **Limitations of Existing CPU Implementations**: - Although anomaly detection libraries on the CPU (such as PyOD, SUOD, etc.) provide good API interfaces, their sequential execution characteristics do not match the parallel processing requirements, resulting in performance bottlenecks. - Existing FPGA designs can provide higher performance, but are far less flexible and customizable than software implementations. 2. **Requirements for Streaming Data Processing**: - Streaming anomaly detection needs to achieve real - time processing under limited memory, processing power and time constraints, and be able to adapt to changes in data distribution (i.e., concept drift). - Traditional batch - processing methods are not suitable for real - time systems because their computational resource requirements and latency increase as the batch size increases. 3. **Challenges of Hardware Acceleration**: - FPGA hardware acceleration can significantly improve performance, but existing FPGA implementations usually lack flexibility and it is difficult to dynamically adjust or add new detection algorithms. - How to achieve efficient parallel processing on the FPGA while maintaining the flexibility and scalability of the system is an important research topic. ### Solutions: To address the above challenges, the authors propose fSEAD, which has the following features: 1. **Modularity and Composability**: - fSEAD adopts the design of partial reconfigurable regions (pblocks), and each pblock can implement different anomaly detection algorithms. - By connecting these pblocks through the AXI switch network, data can be dynamically switched and routed at runtime, thus achieving flexible combination and expansion. 2. **High Performance and Low Latency**: - It uses high - level synthesis (HLS) to generate optimized FPGA instances, supporting three advanced streaming integrated anomaly detection algorithms: Loda, RS - Hash and xStream. - Through hardware acceleration and parallel processing, fSEAD has a speed increase of 3 to 8 times on multiple standard datasets. 3. **Online Reconfiguration**: - Using dynamic function exchange (DFX) technology, fSEAD can be reconfigured during idle time to adapt to different application scenarios and environmental changes. 4. **Ease of Use and Integration**: - It provides a Python interface (such as PYNQ partial overlays), allowing users to easily compose and compare different anomaly detection algorithms. In conclusion, this paper solves the deficiencies of existing anomaly detection systems in terms of performance, flexibility and real - time processing capabilities by proposing fSEAD, providing an innovative solution for efficient anomaly detection of streaming data.

fSEAD: a Composable FPGA-based Streaming Ensemble Anomaly Detection Library

AnRAD: A Neuromorphic Anomaly Detection Framework for Massive Concurrent Data Streams

Selective ensemble method for anomaly detection based on parallel learning

FPGA Implementation of Feature Detection Algorithm Based on High Level Synthesis

AD 2 S: Adaptive anomaly detection on sporadic data streams

SECLEDS: Sequence Clustering in Evolving Data Streams via Multiple Medoids and Medoid Voting

Exploiting Stable Data Dependency in Stream Processing Acceleration on FPGAs

An Ensemble Learning Approach for In-situ Monitoring of FPGA Dynamic Power

Nanosecond anomaly detection with decision trees and real-time application to exotic Higgs decays

A Ubiquitous Machine Learning Accelerator With Automatic Parallelization on FPGA

An efficient FPGA implementation of Mahalanobis distance-based outlier detection for streaming data

Distributed Inference over Decision Tree Ensembles on Clusters of FPGAs

SATAY: A Streaming Architecture Toolflow for Accelerating YOLO Models on FPGA Devices

Fast Memory-efficient Anomaly Detection in Streaming Heterogeneous Graphs

Efficient HW/SW Co-design of FPGA Accelerator to Detect Anomaly Attacks in Smart Grids

Midas: Microcluster-Based Detector of Anomalies in Edge Streams

SALAD: A split active learning based unsupervised network data stream anomaly detection method using autoencoders

FPGA Deployment of LFADS for Real-time Neuroscience Experiments

Unsupervised real-time anomaly detection for streaming data

FLAIRS: FPGA-Accelerated Inference-Resistant & Secure Federated Learning