Jaehan Kim

Jaehan Kim

Auto-generated profile, based on publicly available information

Jaehan Kim

School of Electrical Engineering, KAIST, Daejeon, South Korea

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
5 December 2022
Verified
Conference Paper
Open access
MSMinjae Seo
Minjae Seo
JKJaehan Kim
Jaehan Kim
EMEduard Marin
Eduard Marin
MYMyoungsung You
Myoungsung You
TPTaejune Park
Taejune Park
JKJinwoo Kim
Jinwoo Kim
Software-defined wide area network (SD-WAN) has emerged as a new paradigm for steering a large-scale network flexibly by adopting distributed software-defined network (SDN) controllers. The key to building a logically centralized but physically distributed control-plane is running diverse cluster management protocols to achieve consistency through an exchange of control traffic. Meanwhile, we observe that the control traffic exposes unique time-series patterns and directional relationships due to the operational structure even though the traffic is encrypted, and this pattern can disclose confidential information such as control-plane topology and protocol dependencies, which can be exploited for severe attacks. With this insight, we propose a new SD-WAN fingerprinting system, called Heimdallr. It analyzes periodical and operational patterns of SD-WAN cluster management protocols and the context of flow directions from the collected control traffic utilizing a deep learning-based approach, so that it can classify the cluster management protocols automatically from miscellaneous control traffic datasets. Our evaluation, which is performed in a realistic SD-WAN environment consisting of geographically distant three campus networks and one enterprise network shows that Heimdallr can classify SD-WAN control traffic with ≥ 93%, identify individual protocols with ≥ 80% macro F-1 scores, and finally can infer control-plane topology with ≥ 70% similarity.

23 September 2024 in IEEE Access
Verified
Journal Article
Open access
SLSuyeol Lee
Suyeol Lee
JKJaehan Kim
Jaehan Kim
MSMinjae Seo
Minjae Seo
SNSeung Ho Na
Seung Ho Na
SSSeungwon Shin
Seungwon Shin
JKJinwoo Kim
Jinwoo Kim
Cryptocurrencies have increasingly been used as a medium for illicit financial activities by criminals. Annually, billions of dollars’ worth of Bitcoin penetrate cryptocurrency exchanges. Despite the critical need for advanced Bitcoin financial forensics to investigate these criminal activities, no novel methods have been developed to detect illicit Bitcoin operations. Existing approaches to identifying illegal Bitcoin activity are fundamentally limited due to their inadequate consideration of graph data. To address these limitations, we present a novel approach, Hyperedge Classification, to detect illegal transactions with greater precision. This approach introduces a novel cluster-based Hyperedge-Node Switching technique, which enables effective hyperedge classification and visualization of hyperedge relationships. Additionally, we propose a framework named CENSor, which offers more powerful and robust detection capabilities compared to traditional techniques for both illegal entity detection and illegal transaction detection. Our cluster-based Hyperedge-Node Switching technique demonstrates a tenfold improvement in the F1-score compared to previous graph-based methods. Moreover, CENSor visualizes the Bitcoin cluster graph and the Hyperedge-Node switched graph, highlighting the importance of utilizing appropriate graph information in Bitcoin analysis.

27 September 2024 in IEEE Access
Verified
Journal Article
Open access
MSMinjae Seo
Minjae Seo
JKJaehan Kim
Jaehan Kim
MYMyoungsung You
Myoungsung You
SSSeungwon Shin
Seungwon Shin
JKJinwoo Kim
Jinwoo Kim
Blockchain technology has ushered in a transformative paradigm of decentralized and transparent systems, offering innovative solutions across diverse sectors. While these systems strive for unparalleled transparency and trustlessness in a fully distributed framework, permissionless blockchains, such as Bitcoin and Ethereum, encounter vulnerabilities due to their intrinsically public nature. Addressing these vulnerabilities, the emergence of permissioned blockchains presents a fortified alternative, incorporating rigorous access controls and authentication protocols to ensure participation exclusivity and transaction confidentiality. Nevertheless, a keen observation reveals that, despite encryption, the operational traffic within these blockchains manifests distinct time-series patterns and operational relations during sensitive data exchanges. Such patterns hold the potential to inadvertently expose critical details about the network, encompassing its topology and the operational dependencies among nodes. In light of this revelation, we introduce a pioneering blockchain fingerprinting mechanism, denoted as gShock. This system meticulously analyzes periodic patterns and the context of operational relations from the collected blockchain network traffic. It employs a Graph Neural Network (GNN)-based model, adept at capturing the intricate characteristics innate to specialized blockchain operations. Through empirical experiments conducted in a realistic permissioned blockchain environment, comprising various nodes, we ascertain that gShock demonstrates a remarkable proficiency in classifying blockchain operational traffic with an F-1 score of ≥ 96% and identifying individual dependencies with a macro F-1 score of ≥ 93%.

20 May 2024 in IEEE Transactions on Cloud Computing
Verified
Journal Article
Closed access
MYMyoungsung You
Myoungsung You
MSMinjae Seo
Minjae Seo
JKJaehan Kim
Jaehan Kim
SSSeungwon Shin
Seungwon Shin
JNJaehyun Nam
Jaehyun Nam
Containers have become the predominant virtualization technique for deploying microservices in cloud environments. However, container networking, critical for microservice functionality, often introduces significant overhead and resource consumption, potentially degrading the performance of microservices. This challenge arises from the complexity of the software-based network data plane, responsible for network virtualization and access control within container traffic. To tackle this challenge, we propose Hyperion , a novel hardware-based container networking system that prioritizes high performance and security. Leveraging smartNICs, commonly found in cloud environments, Hyperion implements a fully-functional container network data plane, encompassing network virtualization and access control. It also has the capability to dynamically optimize its data plane for agile responses to frequent changes in container environments, ensuring up-to-date data plane operation. This hardware-based design empowers Hyperion to significantly improve the overall container networking performance without relying on the host system resources. Notably, Hyperion seamlessly integrates with existing containerized applications without necessitating modifications. Our evaluation shows that compared to state-of-the-art solutions, Hyperion achieves significant improvements in HTTP container communication latency and throughput by up to 2.25x and 4.3x, respectively. Furthermore, it reduces CPU utilization associated with container networking by up to 4x.