Difference between revisions of "Resource:Seminar"

From MobiNetS
Jump to: navigation, search
m
(wenliang updates seminars)
Line 8: Line 8:


{{Latest_seminar
{{Latest_seminar
|abstract = Underwater wireless sensor networks (UWSNs) have emerged as an enabling technology for aquatic monitoring. However, data delivery in UWSNs is challenging, due to the harsh aquatic environment and characteristics of the underwater acoustic channel. In recent years, underwater nodes with multi-modal communication capabilities have been proposed to create communication diversity and improve data delivery in UWSNs. Nevertheless, less attention has been devoted to the design of networking protocols leveraging multi-modal communication capabilities of underwater nodes. In this paper, we propose a novel stochastic model for the study of opportunistic routing (OR) in multi-modal UWSNs. We also design two candidate set selection heuristics, named OMUS-E and OMUS-D, for the joint selection of the most suitable acoustic modem for data transmission and next-hop forwarder candidate nodes at each hop, aimed to reduce the energy consumption and improve the network data delivery ratio in multi-modal UWSNs, respectively. Numerical results showed that both proposed heuristics reduced the energy consumption by 65%, 70%, and 75% as compared to the DBR, HydroCast, and GEDAR classical related work protocols, while maintaining a similar data delivery ratio. Furthermore, the proposed solutions outperformed the CAPTAIN routing protocol in terms of data delivery ratio, while maintaining comparable energy consumption.
|abstract = Resource provisioning in multi-tenant stream processing systems faces the dual challenges of keeping resource utilization high (without over-provisioning), and ensuring performance isolation. In our common production use cases, where streaming workloads have to meet latency targets and avoid breaching service-level agreements, existing solutions are incapable of handling the wide variability of user needs. Our framework called Cameo uses fine-grained stream processing (inspired by actor computation models), and is able to provide high resource utilization while meeting latency targets. Cameo dynamically calculates and propagates priorities of events based on user latency targets and query semantics. Experiments on Microsoft Azure show that compared to state-of-the-art, the Cameo framework: i) reduces query latency by 2.7X in single tenant settings, ii) reduces query latency by 4.6X in multi-tenant scenarios, and iii) weathers transient spikes of workload.
|confname= TWC 2021
|confname= NSDI 2021
|link=https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=939476
|link=https://www.usenix.org/system/files/nsdi21spring-xu.pdf
|title=OMUS: Efficient Opportunistic Routing in Multi-Modal Underwater Sensor Networks
|title=Move Fast and Meet Deadlines: Fine-grained Real-time Stream Processing with Cameo
|speaker=Xianyang
|speaker=Jianfei
}}
}}
{{Latest_seminar
{{Latest_seminar
|abstract = LoRa, as a representative Low-Power Wide-Area Network (LPWAN) technology, can provide long-range communication for battery-powered IoT devices with a 10-year lifetime. LoRa links in practice, however, experience high dynamics in various environments. When the SNR falls below the threshold (e.g., in the building), a LoRa device disconnects from the network. We propose Falcon, which addresses the link dynamics by enabling data transmission for very low SNR or even disconnected LoRa links. At the heart of Falcon, we reveal that low SNR LoRa links that cannot deliver packets can still introduce interference to other LoRa transmissions. Therefore, Falcon transmits data bits on the low SNR link by selectively interfering with other LoRa transmissions. We address practical challenges in Falcon design. We propose a low-power channel activity detection method to detect other LoRa transmissions for selective interference. To interfere with the so-called interference-resilient LoRa, we accurately estimate the time and frequency offsets on LoRa packets and propose an adaptive frequency adjusting strategy to maximize the interference. We implement Falcon, all using commercial off-the-shelf LoRa devices, and extensively evaluate its performance. The results show that Falcon can provide reliable communication links for disconnected LoRa devices and achieves the SNR boundary upto 7.5 dB lower than that of standard LoRa.
|abstract = Federated Learning (FL) is a decentralized machine-learning paradigm, in which a global server iteratively averages the model parameters of local users without accessing their data. User heterogeneity has imposed significant challenges to FL, which can incur drifted global models that are slow to converge. Knowledge Distillation has recently emerged to tackle this issue, by refining the server model using aggregated knowledge from heterogeneous users, other than directly averaging their model parameters. This approach, however, depends on a proxy dataset, making it impractical unless such a prerequisite is satisfied. Moreover, the ensemble knowledge is not fully utilized to guide local model learning, which may in turn affect the quality of the aggregated model. Inspired by the prior art, we propose a data-free knowledge distillation} approach to address heterogeneous FL, where the server learns a lightweight generator to ensemble user information in a data-free manner, which is then broadcasted to users, regulating local training using the learned knowledge as an inductive bias. Empirical studies powered by theoretical implications show that, our approach facilitates FL with better generalization performance using fewer communication rounds, compared with the state-of-the-art.
|confname= MobiCom 2021
|confname= ICML 2021
|link= https://dl.acm.org/doi/pdf/10.1145/3447993.3483250
|link= https://arxiv.org/pdf/2105.10056.pdf
|title=Combating link dynamics for reliable lora connection in urban settings
|title=Data-Free Knowledge Distillation for Heterogeneous Federated Learning
|speaker=Wangxiong
|speaker=Jianqi
}}
{{Latest_seminar
|abstract = The revolution of online shopping in recent years demands corresponding evolution in delivery services in urban areas. To cater to this trend, delivery by the crowd has become an alternative to the traditional delivery services thanks to the advances in ubiquitous computing. Notably, some studies use public transportation for crowdsourcing delivery, given its low-cost delivery network with millions of passengers as potential couriers. However, multiple practical impact factors are not considered in existing public-transport-based crowdsourcing delivery studies due to a lack of data and limited ubiquitous computing infrastructures in the past. In this work, we design a crowdsourcing delivery system based on public transport, considering the practical factors of time constraints, multi-hop delivery, and profits. To incorporate the impact factors, we build a reinforcement learning model to learn the optimal order dispatching strategies from massive passenger data and package data. The order dispatching problem is formulated as a sequential decision making problem for the packages routing, i.e., select the next station for the package. A delivery time estimation module is designed to accelerate the training process and provide statistical delivery time guarantee. Three months of real-world public transportation data and one month of package delivery data from an on-demand delivery platform in Shenzhen are used in the evaluation. Compared with existing crowdsourcing delivery algorithms and widely used baselines, we achieve a 40% increase in profit rates and a 29% increase in delivery rates. Comparison with other reinforcement learning algorithms shows that we can improve the profit rate and the delivery rate by 9% and 8% by using time estimation in action filtering. We share the data used in the project to the community for other researchers to validate our results and conduct further research.1 [1].
|confname= IMWUT 2021
|link= https://dl.acm.org/doi/pdf/10.1145/3478117
|title=A City-Wide Crowdsourcing Delivery System with Reinforcement Learning
|speaker=Wenjie
}}
}}


=== History ===
=== History ===
{{Resource:Previous_Seminars}}
{{Resource:Previous_Seminars}}

Revision as of 08:08, 9 December 2021

Time: 2021-11-26 8:40
Address: Main Building B1-612
Useful links: Readling list; Schedules; Previous seminars.

Latest

  1. [NSDI 2021] Move Fast and Meet Deadlines: Fine-grained Real-time Stream Processing with Cameo, Jianfei
    Abstract: Resource provisioning in multi-tenant stream processing systems faces the dual challenges of keeping resource utilization high (without over-provisioning), and ensuring performance isolation. In our common production use cases, where streaming workloads have to meet latency targets and avoid breaching service-level agreements, existing solutions are incapable of handling the wide variability of user needs. Our framework called Cameo uses fine-grained stream processing (inspired by actor computation models), and is able to provide high resource utilization while meeting latency targets. Cameo dynamically calculates and propagates priorities of events based on user latency targets and query semantics. Experiments on Microsoft Azure show that compared to state-of-the-art, the Cameo framework: i) reduces query latency by 2.7X in single tenant settings, ii) reduces query latency by 4.6X in multi-tenant scenarios, and iii) weathers transient spikes of workload.
  2. [ICML 2021] Data-Free Knowledge Distillation for Heterogeneous Federated Learning, Jianqi
    Abstract: Federated Learning (FL) is a decentralized machine-learning paradigm, in which a global server iteratively averages the model parameters of local users without accessing their data. User heterogeneity has imposed significant challenges to FL, which can incur drifted global models that are slow to converge. Knowledge Distillation has recently emerged to tackle this issue, by refining the server model using aggregated knowledge from heterogeneous users, other than directly averaging their model parameters. This approach, however, depends on a proxy dataset, making it impractical unless such a prerequisite is satisfied. Moreover, the ensemble knowledge is not fully utilized to guide local model learning, which may in turn affect the quality of the aggregated model. Inspired by the prior art, we propose a data-free knowledge distillation} approach to address heterogeneous FL, where the server learns a lightweight generator to ensemble user information in a data-free manner, which is then broadcasted to users, regulating local training using the learned knowledge as an inductive bias. Empirical studies powered by theoretical implications show that, our approach facilitates FL with better generalization performance using fewer communication rounds, compared with the state-of-the-art.

History

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=2218"