pForest: In-Network Inference with Random Forests

Busse-Grawitz, Coralie; Meier, Roland; Dietmüller, Alexander; Bühler, Tobias; Vanbever, Laurent

doi:10.48550/arxiv.1909.05680

Cited by 10 publications

(18 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This code word is the result of the feature tables lookup. Therefore the number of pipeline stages consumed by a decision tree is independent of the depth of the tree, overcoming previous limitations [12,29].…”

Section: Decision Treementioning

confidence: 99%

“…Implementing binary neural networks was explored in N2Net [44] and BaNaNa Split [40], with limited performance benefits. pForest [12], SwitchTree [29] and NERDS [55] explored mapping random forests, with only pForest attempting ASIC implementation. Their methodology is different, encoding each decision tree in separate tables, with a table for every tree level, achieving lower scalability (e.g., depth of 4 in pForest).…”

Section: Related Workmentioning

confidence: 99%

“…The last class of solutions has focused on implementing Random Forests on switches. These solutions had failed to scale on a switch-ASIC [12] or were independent of most constraints by running in a software environment or on a NIC [29,55]. Harnessing the power of network processing for ML within commodity switches remains a challenge.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

IIsy: Practical In-Network Classification

Zheng¹,

Xiong²,

Bui³

et al. 2022

Preprint

View full text Add to dashboard Cite

The rat race between user-generated data and data-processing systems is currently won by data. The increased use of machine learning leads to further increase in processing requirements, while data volume keeps growing. To win the race, machine learning needs to be applied to the data as it goes through the network. In-network classification of data can reduce the load on servers, reduce response time and increase scalability.In this paper, we introduce IIsy, implementing machine learning classification models in a hybrid fashion using offthe-shelf network devices. IIsy targets three main challenges of in-network classification: (i) mapping classification models to network devices (ii) extracting the required features and (iii) addressing resource and functionality constraints. IIsy supports a range of traditional and ensemble machine learning models, scaling independently of the number of stages in a switch pipeline. Moreover, we demonstrate the use of IIsy for hybrid classification, where a small model is implemented on a switch and a large model at the backend, achieving near optimal classification results , while significantly reducing latency and load on the servers.

show abstract

Section: Decision Treementioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

IIsy: Practical In-Network Classification

Zheng¹,

Xiong²,

Bui³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…In-device machine learning (ML) offloading: Work that focuses on the programmable data-plane for switches based on Protocol Independent Switch Architecture (PISA) [22,36,38,45] tries to cope with the limited resources that are available for the implementation of ML models. For example [45] adapt several ML models to the match-action table model in P4, while [15] implementing a random forest model to classify traffic. In both cases, the benefits of running the analytics in the data-plane are limited since no immediate packet forwarding decision derives from the analytics.…”

Section: Related Workmentioning

confidence: 99%

FENXI: Deep-learning Traffic Analytics at the Edge

Gallo,

Finamore,

Simon

et al. 2021

Preprint

View full text Add to dashboard Cite

Live traffic analysis at the first aggregation point in the ISP network enables the implementation of complex traffic engineering policies but is limited by the scarce processing capabilities, especially for Deep Learning (DL) based analytics. The introduction of specialized hardware accelerators i.e., Tensor Processing Unit (TPU), offers the opportunity to enhance processing capabilities of network devices at the edge. Yet, to date, no packet processing pipeline is capable of offering DL-based analysis capabilities in the data-plane, without interfering with network operations.In this paper, we present FENXI, a system to run complex analytics by leveraging TPU. The design of FENXI decouples forwarding operations and traffic analytics which operates at different granularities i.e., packet and flow levels. We conceive two independent modules that asynchronously communicate to exchange network data and analytics results, and design data structures to extract flow level statistics without impacting per-packet processing. We prototyped and evaluated FENXI on general-purpose servers considering both both adversarial and realistic network conditions. Our analysis shows that FENXI can sustains 100 Gbps line rate traffic processing requiring only limited resources, while also dynamically adapting to variable network conditions. CCS CONCEPTS• Hardware → Analysis and design of emerging devices and systems; • Networks → In-network processing.

show abstract

“…Researchers have so far demonstrated the offloading of algorithms such as Support Vector Machines (SVM) 1 [59,64,67], Naïve Bayes (NB) [64,67], K-means (KM) [19,59,64,67], Decision Trees (DT) [64,67,68], Random Forest (RF) [7,36,63,65,67,68], and (Binary) Neural Networks (NN) [48,50,54,56,57,59,70]. These works have provided proofs-of-concept and laid the foundation of in-network ML.…”

Section: Introductionmentioning

confidence: 99%

Automating In-Network Machine Learning

Zheng¹,

Zang²,

Hong³

et al. 2022

Preprint

View full text Add to dashboard Cite

Using programmable network devices to aid in-network machine learning has been the focus of significant research. However, most of the research was of a limited scope, providing a proof of concept or describing a closed-source algorithm. To date, no general solution has been provided for mapping machine learning algorithms to programmable network devices. In this paper, we present Planter, an opensource, modular framework for mapping trained machine learning models to programmable devices. Planter supports a wide range of machine learning models, multiple targets and can be easily extended. The evaluation of Planter compares different mapping approaches, and demonstrates the feasibility, performance, and resource efficiency for applications such as anomaly detection, financial transactions, and quality of experience. The results show that Planterbased in-network machine learning algorithms can run at line rate, have a negligible effect on latency, coexist with standard switching functionality, and have no or minor accuracy trade-offs.

show abstract

pForest: In-Network Inference with Random Forests

Cited by 10 publications

References 0 publications

IIsy: Practical In-Network Classification

IIsy: Practical In-Network Classification

FENXI: Deep-learning Traffic Analytics at the Edge

Automating In-Network Machine Learning

Contact Info

Product

Resources

About