Difference between revisions of "Resource:Seminar"

From MobiNetS
Jump to: navigation, search
m
 
(100 intermediate revisions by 4 users not shown)
Line 1: Line 1:
{{SemNote
{{SemNote
|time='''2023-02-20 9:30'''
|time='''2024-11-1 10:30-12:00'''
|addr=4th Research Building A527-B
|addr=4th Research Building A533
|note=Useful links: [[Resource:Reading_List|Readling list]]; [[Resource:Seminar_schedules|Schedules]]; [[Resource:Previous_Seminars|Previous seminars]].
|note=Useful links: [[Resource:Reading_List|📚 Readling list]]; [[Resource:Seminar_schedules|📆 Schedules]]; [[Resource:Previous_Seminars|🧐 Previous seminars]].
}}
}}


===Latest===
===Latest===
{{Latest_seminar
{{Latest_seminar
|abstract = Visible light communications (VLC) is a good candidate technology for the 6th generation (6G) wireless communications. Red, green, and blue (RGB) light-emitting diodes (LEDs) based VLC has become an important research branch due to its low price and high reliability. However, the saturation of photodiode (PD) caused by the ambient background light may seriously degrade the bit error rate (BER) performance of an RGB-VLC system's three spatially uncoupled information streams (i.e., red, green, and blue LEDs can transmit different data packets simultaneously) in practical applications. To mitigate the ambient light interference in point-to-point RGB-VLC systems, we propose, PNC-VLC, a network-coded scheme that uses two LEDs with the same color at the transmitter to transmit two different data streams and we make use of the naturally overlapped signals at the receiver to formulate physical-layer network coding (PNC). The adaptivity of PNC-VLC could effectively improve the BER degradation problem caused by the saturation of PD under the influence of ambient light. We conducted simulations based on the parameters of commercial off-the-shelf (COTS) products to prove the superiority of the PNC-VLC under the influence of four typical illuminants. Simulation results show that the PNC-VLC system can maintain a better and more stable system BER performance under different ambient background light conditions. Remarkably, with 2/3 throughput efficiency, PNC-VLC can bring 133.3% gain to the BER performance when compared with RGB-VLC under the Illuminant A interference model, making it a good option for VLC applications with unpredictable ambient background interferences.
|abstract = Sparsely-activated Mixture-of-Expert (MoE) layers have found practical applications in enlarging the model size of large-scale foundation models, with only a sub-linear increase in computation demands. Despite the wide adoption of hybrid parallel paradigms like model parallelism, expert parallelism, and expert-sharding parallelism (i.e., MP+EP+ESP) to support MoE model training on GPU clusters, the training efficiency is hindered by communication costs introduced by these parallel paradigms. To address this limitation, we propose Parm, a system that accelerates MP+EP+ESP training by designing two dedicated schedules for placing communication tasks. The proposed schedules eliminate redundant computations and communications and enable overlaps between intra-node and inter-node communications, ultimately reducing the overall training time. As the two schedules are not mutually exclusive, we provide comprehensive theoretical analyses and derive an automatic and accurate solution to determine which schedule should be applied in different scenarios. Experimental results on an 8-GPU server and a 32-GPU cluster demonstrate that Parm outperforms the state-of-the-art MoE training system, DeepSpeed-MoE, achieving 1.13× to 5.77× speedup on 1296 manually configured MoE layers and approximately 3× improvement on two real-world MoE models based on BERT and GPT-2.
|confname=IEEE Photonics Journal 2023
|confname =INFOCOM‘24
|link=https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=10028767
|link = https://ieeexplore.ieee.org/abstract/document/10621327
|title=Physical-Layer Network Coding Enhanced Visible Light Communications Using RGB LEDs
|title= Parm: Efficient Training of Large Sparsely-Activated Models with Dedicated Schedules
|speaker=Jiahui}}
|speaker=Mengqi
|date=2024-11-1
}}
{{Latest_seminar
{{Latest_seminar
|abstract = Mobile edge computing (MEC), as a key ingredient of the 5G ecosystem, is envisioned to support demanding applications with stringent latency requirements. The basic idea is to deploy servers close to end-users, e.g., on the network edge-side instead of the remote cloud. While conceptually reasonable, we find that the operational 5G is not coordinated with MEC and thus suffers from intolerable long response latency. In this work, we propose Tutti, which couples 5G RAN and MEC at the user space to assure the performance of latency-critical video analytics. To enable such capacity, Tutti precisely customizes the application service demand by fusing instantaneous wireless dynamics from the 5G RAN and application-layer content changes from edge servers. Tutti then enforces a deadline-sensitive resource provision for meeting the application service demand by real-time interaction between 5G RAN and edge servers in a lightweight and standard-compatible way. We prototype and evaluate Tutti on a software-defined platform, which shows that Tutti reduces the response latency by an average of 61.69% compared with the existing 5G MEC system, as well as negligible interaction costs.
|abstract = HD map is a key enabling technology towards fully autonomous driving. We propose VI-Map, the first system that leverages roadside infrastructure to enhance real-time HD mapping for autonomous driving. The core concept of VI-Map is to exploit the unique cumulative observations made by roadside infrastructure to build and maintain an accurate and current HD map. This HD map is then fused with on-vehicle HD maps in real time, resulting in a more comprehensive and up-to-date HD map. By extracting concise bird-eye-view features from infrastructure observations and utilizing vectorized map representations, VI-Map incurs low compute and communication overhead. We conducted end-to-end evaluations of VI-Map on a real-world testbed and a simulator. Experiment results show that VI-Map can construct decentimeter-level (up to 0.3 m) HD maps and achieve real-time (up to a delay of 42 ms) map fusion between driving vehicles and roadside infrastructure. This represents a significant improvement of 2.8× and 3× in map accuracy and coverage compared to the state-of-the-art online HD mapping approaches. A video demo of VI-Map on our real-world testbed is available at https://youtu.be/p2RO65R5Ezg.
|confname=Mobicom 2022
|confname=Mobicom'23
|link=https://dl.acm.org/doi/pdf/10.1145/3498361.3539765
|link = https://dl.acm.org/doi/abs/10.1145/3570361.3613280
|title=Tutti: coupling 5G RAN and mobile edge computing for latency-critical video analytics
|title= VI-Map: Infrastructure-Assisted Real-Time HD Mapping for Autonomous Driving
|speaker=Silience}}
|speaker=Wangyang
 
|date=2024-11-1
 
}}
 
=== History ===


{{Resource:Previous_Seminars}}
{{Resource:Previous_Seminars}}

Latest revision as of 05:48, 1 November 2024

Time: 2024-11-1 10:30-12:00
Address: 4th Research Building A533
Useful links: 📚 Readling list; 📆 Schedules; 🧐 Previous seminars.

Latest

  1. [INFOCOM‘24] Parm: Efficient Training of Large Sparsely-Activated Models with Dedicated Schedules, Mengqi
    Abstract: Sparsely-activated Mixture-of-Expert (MoE) layers have found practical applications in enlarging the model size of large-scale foundation models, with only a sub-linear increase in computation demands. Despite the wide adoption of hybrid parallel paradigms like model parallelism, expert parallelism, and expert-sharding parallelism (i.e., MP+EP+ESP) to support MoE model training on GPU clusters, the training efficiency is hindered by communication costs introduced by these parallel paradigms. To address this limitation, we propose Parm, a system that accelerates MP+EP+ESP training by designing two dedicated schedules for placing communication tasks. The proposed schedules eliminate redundant computations and communications and enable overlaps between intra-node and inter-node communications, ultimately reducing the overall training time. As the two schedules are not mutually exclusive, we provide comprehensive theoretical analyses and derive an automatic and accurate solution to determine which schedule should be applied in different scenarios. Experimental results on an 8-GPU server and a 32-GPU cluster demonstrate that Parm outperforms the state-of-the-art MoE training system, DeepSpeed-MoE, achieving 1.13× to 5.77× speedup on 1296 manually configured MoE layers and approximately 3× improvement on two real-world MoE models based on BERT and GPT-2.
  2. [Mobicom'23] VI-Map: Infrastructure-Assisted Real-Time HD Mapping for Autonomous Driving, Wangyang
    Abstract: HD map is a key enabling technology towards fully autonomous driving. We propose VI-Map, the first system that leverages roadside infrastructure to enhance real-time HD mapping for autonomous driving. The core concept of VI-Map is to exploit the unique cumulative observations made by roadside infrastructure to build and maintain an accurate and current HD map. This HD map is then fused with on-vehicle HD maps in real time, resulting in a more comprehensive and up-to-date HD map. By extracting concise bird-eye-view features from infrastructure observations and utilizing vectorized map representations, VI-Map incurs low compute and communication overhead. We conducted end-to-end evaluations of VI-Map on a real-world testbed and a simulator. Experiment results show that VI-Map can construct decentimeter-level (up to 0.3 m) HD maps and achieve real-time (up to a delay of 42 ms) map fusion between driving vehicles and roadside infrastructure. This represents a significant improvement of 2.8× and 3× in map accuracy and coverage compared to the state-of-the-art online HD mapping approaches. A video demo of VI-Map on our real-world testbed is available at https://youtu.be/p2RO65R5Ezg.

History

2024

2023

2022

2021

2020

  • [Topic] [ The path planning algorithm for multiple mobile edge servers in EdgeGO], Rong Cong, 2020-11-18

2019

2018

2017

Instructions

请使用Latest_seminar和Hist_seminar模板更新本页信息.

    • 修改时间和地点信息
    • 将当前latest seminar部分的code复制到这个页面中
    • 将{{Latest_seminar... 修改为 {{Hist_seminar...，并增加对应的日期信息|date=
    • 填入latest seminar各字段信息
    • link请务必不要留空，如果没有link则填本页地址 https://mobinets.org/index.php?title=Resource:Seminar
  • 格式说明
    • Latest_seminar:

{{Latest_seminar
|confname=
|link=
|title=
|speaker=
}}

    • Hist_seminar

{{Hist_seminar
|confname=
|link=
|title=
|speaker=
|date=
}}

Retrieved from "http://mobinets.cn/site/index.php?title=Resource:Seminar&oldid=3215"