| RIZKY PRATAMA HUDHAJANTO | D, 中間発表 | ネットワークシステム学 | 岡田 実, | 林 優一, | 東野 武史, | DUONG QUANG THANG, | CHEN Na |
|
title: Symbol Level Precodng in Satellite Communication System
abstract: Integrating terrestrial and satellite communication systems is promising approach for achieving global coverage in 5G and beyond. However, there are still many challenges to be faced when implementing this satellite system. In the lates satellite technology, namely High-Throughput Satellites (HTS) using Multibeam schemes, interference arises and leads to performance degradation. In this research presentation, we will discuss how Symbol Level Precoding (SLP) can serve as a solution to overcome this interference problem. Additionally, we will discuss our original SLP which enhances performance and reduces system complexity, special designed for Satellite Communication System. language of the presentation: ENGLISH | |||||||
| BIN BHUTTO ADIL | M, 2回目発表 | サイバーレジリエンス構成学 | 門林 雄基, | 岡田 実, | 妙中 雄三, | ||
|
title: Performance Challenges in Containerized Networking: Latency, Jitter, and Bandwidth Beyond 5G
abstract: The advent of Network Function Virtualization (NFV) has transitioned network processing from specialized hardware to adaptable commodity servers, a shift that continues to evolve. Emerging industrial applications, the Internet of Things (IoT), and technologies such as augmented, virtual, and mixed reality (AR/VR/MR) demand networks capable of handling event-based operations and middleware with exceptionally low and predictable latency, posing significant performance optimization challenges for packet processing in layered infrastructures. This study explores the implementation challenges of such network infrastructures using general-purpose hardware, driven by the flexibility to realize telco-cloud and the potential to reduce electronic waste. Focusing on containerized network functions (CNFs) within the NFV framework, we investigate performance limitations, particularly high jitter and throughput variation in packet forwarding. We conducted ten experiments using a network function implemented with the industry-standard user-space I/O architecture DPDK in bare-metal and containerized environments, within a 40 GbE setup, measuring throughput, latency, and jitter across various packet sizes, traffic rates, and system configurations. Our findings indicate that adjusting CPU settings can significantly enhance throughput for CNFs, albeit with a potential increase in jitter. CNFs were found to be feasible for latency-sensitive tasks under conditions of low traffic and specific packet sizes, and with careful system-level configuration, they show promise for use in beyond 5G cloud-native networking. This research provides insights into the effect of CPU configurations on the latency and jitter of CNFs, examining the correlation between CNF performance and CPU performance and idle states, and identifying bottlenecks within the CNF architecture that lead to poorer performance compared to bare-metal NFs. The results offer valuable guidance for designing high-performance CNFs with manageable latency and jitter. language of the presentation: English | |||||||
| SYAFRUDIN RAIS AKHDAN | M, 1回目発表 | ヒューマンロボティクス | 和田 隆広, | 松原 崇充, | 織田 泰彰, | 劉 海龍 | |
|
title: *** PRIORITIZING HUMAN PATHWAYS IN THE DEVELOPMENT OF QUADRUPEDAL GUIDE ROBOTS *** abstract: *** Due to the limited availability of guide dogs and the high cost of training, quadrupedal robots offer a more affordable and faster training alternative. This research develops a guide robot system that prioritizes human movement using a human-robot interaction model and the Point LIO SLAM algorithm for robot localization, combined with OpenPose for human movement detection. Experiments indicate that human movement tracking can be performed accurately despite challenges such as lateral movement and loss of detection during aggressive robot maneuvers. The study concludes that integrating path planning for human using human movement detection and a rigid harness can significantly improve user accuracy and safety, by focusing research on how robots guide humans to stay on a specified path. Future developments aim to shift path planning focus from the robot to the human, optimizing overall guidance effectiveness. *** language of the presentation: *** English *** | |||||||
| 松本 暖樹 | M, 1回目発表 | ヒューマンロボティクス | 和田 隆広 | 松原 崇充, | 織田 泰彰, | 劉 海龍 | |
|
title:Effect of Displaying Internal Mechanical State of Robot to Human on Human-Robot Handovers
abstract:Previous research revealed that presenting the stiffness of a robot hand to a human forearm as a tightening force with a newly developed interface achieved smooth handover of an object between a human and a robot. The purpose of the present study is to investigate whether the method is effective for a variety of control algorithms of robot hands, which use different levels of information related to the handover and takeover. A experiment with participants demonstrated that presenting the stiffness of the robot hand not only reduces time at receiving but also improves handing accuracy, thereby alleviating the burden on humans in the workplace. language of the presentation:Japanese 発表題目:人間へのロボット関節剛性呈示がロボット間の物体受け渡しに及ぼす効果 —異なるロボット制御アルゴリズム間の比較— 発表概要:先行研究ではロボットハンドの剛性を人の前腕に締め付ける力として提示することで，人間-ロボット間の物体受け渡しの効率が向上することが示されている．本研究では人間とロボットの物体の受け渡しにおいて，異なる状態遷移アルゴリズムのロボットに対して，人間へのロボット関節剛性提示が効果的であるかどうかを調査することが目的である．被験者実験によりロボットハンドの剛性を提示することで，受け取り時の反応時間の短縮や，作業者の主観的な動作タイミングの分かりやすさ及びスムーズさの向上，物体を落とす不安感の減少が示唆された. | |||||||