arXiv subjects

Takanori Hara

Publications and source records attributed to Takanori Hara.

8 recordsLinked to original sources

arXiv2026

Optimization of Model Splitting, Placement, and Chaining for Multi-hop Split Learning and Inference

Service Function Chaining (SFC) establishes efficient communication paths by ensuring that traffic traverses a predefined sequence of network functions in a specified order to meet particular service requirements. Inspired by this concept, we have proposed an SFC-based architecture for multi-hop split learning (MSL) and split inference (MSI), facilitating distributed AI applications to effectively route smashed data across multi-hop networks. However, the multi-hop environment presents new challenges, including (1) determining optimal cut points, (2) deploying split sub-models on appropriate computing nodes, and (3) routing smashed data through the underlying communication networks while adhering to service requirements. To address these challenges, we formulate an Integer Linear Programming (ILP) model to jointly optimize model splitting, placement, and chaining (data routing) in the SFC-based MSL/MSI architecture, aiming to minimize end-to-end inference or training latency. Additionally, we propose a Block Coordinate Descent (BCD)-based heuristic algorithm to efficiently solve the problem. Comprehensive evaluations demonstrate the effectiveness and characteristics of the proposed formulation and algorithm.

cs.NI↗

arXiv2025

Service Function Chaining Architecture for Multi-hop Split Inference and Learning

Service Function Chaining (SFC) is a networking technique that ensures traffic traverses a predefined sequence of service functions, realizing arbitrary network services through dynamic and efficient communication paths. Inspired by this concept, we propose an SFC-based architecture for Multi-hop Split Inference (MSI), where split sub-models are interpreted as service functions and their composition forms a service chain representing the global model. By leveraging SFC, the proposed architecture dynamically establishes communication paths for split sub-models, ensuring efficient and adaptive execution. Furthermore, we extend this architecture to Multi-hop Split Learning (MSL) by applying SFC to the bidirectional communication required for training tasks. To realize the proposed architecture, we design Neural Service Functions (NSFs) to execute split sub-models as transparent TCP proxies and integrate them with Segment Routing over IPv6 (SRv6) and the extended Berkeley Packet Filter (eBPF)-based SFC proxy. This integration ensures efficient ML processing over dynamic routing while maintaining compatibility with existing applications. Evaluation results demonstrate that (1) the proposed architecture is feasible for both MSI and MSL; (2) it is particularly suitable for real-time inference in MSI scenarios with small mini-batch sizes; (3) it supports dynamic path reconfiguration, enabling adaptive responses to changing network conditions while minimizing the impact of control mechanisms on inference and learning processes.

cs.NI↗

arXiv2025

Data Preservation in High Energy Physics

Data preservation significantly increases the scientific output of high-energy physics experiments during and after data acquisition. For new and ongoing experiments, the careful consideration of long-term data preservation in the experimental design contributes to improving computational efficiency and strengthening the scientific activity in HEP through Open Science methodologies. This contribution is based on 15 years of experience of the DPHEP collaboration in the field of data preservation and focuses on aspects relevant for the strategic programming of particle physics in Europe: the preparation of future programs using data sets preserved from previous similar experiments (e.g. HERA for EIC), and the use of LHC data long after the end of the data taking. The lessons learned from past collider experiments and recent developments open the way to a number of recommendations for the full exploitation of the investments made in large HEP experiments.

hep-ex↗

arXiv2024

Eigenvalue Based Active User Enumeration for Grant-Free Access Under Carrier Frequency Offsets

This paper investigates a grant-free non-orthogonal multiple access (GF-NOMA) system in the presence of carrier frequency offsets. We propose two schemes for enumerating active users in such a GF-NOMA system, which is equivalent to estimating the sparsity level. Both schemes utilize a short common pilot and the eigenvalues of the sample covariance matrix of the received signal. The two schemes differ in their treatment of noise variance: one exploits known variance information, while the other is designed to function without this knowledge. Simulation results demonstrate the effectiveness of the proposed schemes in terms of the normalized root-mean-squared error.

cs.IT↗

arXiv2024

Dealing with Imbalanced Classes in Bot-IoT Dataset

With the rapidly spreading usage of Internet of Things (IoT) devices, a network intrusion detection system (NIDS) plays an important role in detecting and protecting various types of attacks in the IoT network. To evaluate the robustness of the NIDS in the IoT network, the existing work proposed a realistic botnet dataset in the IoT network (Bot-IoT dataset) and applied it to machine learning-based anomaly detection. This dataset contains imbalanced normal and attack packets because the number of normal packets is much smaller than that of attack ones. The nature of imbalanced data may make it difficult to identify the minority class correctly. In this thesis, to address the class imbalance problem in the Bot-IoT dataset, we propose a binary classification method with synthetic minority over-sampling techniques (SMOTE). The proposed classifier aims to detect attack packets and overcome the class imbalance problem using the SMOTE algorithm. Through numerical results, we demonstrate the proposed classifier's fundamental characteristics and the impact of imbalanced data on its performance.

cs.CR↗

arXiv2016

Status Report of the DPHEP Collaboration: A Global Effort for Sustainable Data Preservation in High Energy Physics

Data from High Energy Physics (HEP) experiments are collected with significant financial and human effort and are mostly unique. An inter-experimental study group on HEP data preservation and long-term analysis was convened as a panel of the International Committee for Future Accelerators (ICFA). The group was formed by large collider-based experiments and investigated the technical and organizational aspects of HEP data preservation. An intermediate report was released in November 2009 addressing the general issues of data preservation in HEP and an extended blueprint paper was published in 2012. In July 2014 the DPHEP collaboration was formed as a result of the signature of the Collaboration Agreement by seven large funding agencies (others have since joined or are in the process of acquisition) and in June 2015 the first DPHEP Collaboration Workshop and Collaboration Board meeting took place. This status report of the DPHEP collaboration details the progress during the period from 2013 to 2015 inclusive.

hep-ex↗

arXiv2013

Belle II Experiment Network and Computing

The Belle experiment, part of a broad-based search for new physics, is a collaboration of approximately 400 physicists from 55 institutions across four continents. The Belle detector is located at the KEKB accelerator in Tsukuba, Japan. The Belle detector was operated at the asymmetric electron-positron collider KEKB from 1999-2010. The detector accumulated more than 1/ab of integrated luminosity corresponding to more than 2 PB of data near 10 GeV center-of-mass energy. Recently, KEK has initiated a $400 million accelerator upgrade to be called SuperKEKB, designed to produce instantaneous and integrated luminosity two orders of magnitude greater than KEKB. The new international collaboration at SuperKEKB is called Belle II. The first data from Belle II/SuperKEKB is expected in 2015. In October 2012, senior members of the Belle II collaboration gathered at PNNL to discuss the computing and networking requirements of the Belle II experiment with ESnet staff and other computing and networking experts. The day-and-a-half-long workshop characterized the instruments and facilities used in the experiment, the process of science for Belle II, and the computing and networking equipment and configuration requirements to realize the full scientific potential of the collaboration's work. The requirements identified at the Belle II Experiment Requirements workshop are summarized in this report.

physics.ins-det↗

arXiv2012

Status Report of the DPHEP Study Group: Towards a Global Effort for Sustainable Data Preservation in High Energy Physics

Data from high-energy physics (HEP) experiments are collected with significant financial and human effort and are mostly unique. An inter-experimental study group on HEP data preservation and long-term analysis was convened as a panel of the International Committee for Future Accelerators (ICFA). The group was formed by large collider-based experiments and investigated the technical and organisational aspects of HEP data preservation. An intermediate report was released in November 2009 addressing the general issues of data preservation in HEP. This paper includes and extends the intermediate report. It provides an analysis of the research case for data preservation and a detailed description of the various projects at experiment, laboratory and international levels. In addition, the paper provides a concrete proposal for an international organisation in charge of the data management and policies in high-energy physics.

hep-ex↗