Pub Date : 2022-07-01DOI: 10.1109/ICDCS54860.2022.00129
Yung-Tsai Weng, H. Nguyen
The energy consumption of Internet Data Centers (IDCs) is rapidly increasing and may account for 20.9% of global energy consumption in 2030. This huge demand will change the power system operations significantly in the future because IDCs are different from the traditional loads in power systems [1] . Specifically, IDCs are energy-intensive loads that can dominate and alter the nearby power flow directions, thus posing challenges to the regulation of power systems. Also, working loads migration across IDCs at different locations and time slots can disturb the real-time power balance in power systems. Besides, IDCs’ intensive electricity demand rising following the expansion of IDCs might not be met due to supply limits of the power infrastructure. Moreover, IDCs scattered in a power grid can introduce stress and overload "weak" power transmission lines as well as cause other operational violations in power systems, such as voltages and frequency. All these effects will become more pronounced with more and larger IDCs.
{"title":"Interdependence Analysis and Co-optimization of Scattered Data Centers and Power Systems","authors":"Yung-Tsai Weng, H. Nguyen","doi":"10.1109/ICDCS54860.2022.00129","DOIUrl":"https://doi.org/10.1109/ICDCS54860.2022.00129","url":null,"abstract":"The energy consumption of Internet Data Centers (IDCs) is rapidly increasing and may account for 20.9% of global energy consumption in 2030. This huge demand will change the power system operations significantly in the future because IDCs are different from the traditional loads in power systems [1] . Specifically, IDCs are energy-intensive loads that can dominate and alter the nearby power flow directions, thus posing challenges to the regulation of power systems. Also, working loads migration across IDCs at different locations and time slots can disturb the real-time power balance in power systems. Besides, IDCs’ intensive electricity demand rising following the expansion of IDCs might not be met due to supply limits of the power infrastructure. Moreover, IDCs scattered in a power grid can introduce stress and overload \"weak\" power transmission lines as well as cause other operational violations in power systems, such as voltages and frequency. All these effects will become more pronounced with more and larger IDCs.","PeriodicalId":225883,"journal":{"name":"2022 IEEE 42nd International Conference on Distributed Computing Systems (ICDCS)","volume":"374 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123320310","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-01DOI: 10.1109/ICDCS54860.2022.00117
Jian Kang, Alian Yu, Wei Jiang, D. Lin
With the advances in autonomous vehicles and intelligent intersection management systems, traffic lights may be replaced by optimal travel plans calculated for each passing vehicle in the future. While these technological advancements are envisioned to greatly improve travel efficiency, they are still facing various challenging security hurdles since even a single deviation of a vehicle from its assigned travel plan could cause a serious accident if the surrounding vehicles do not take necessary actions in a timely manner. In this paper, we propose a novel security mechanism namely NWADE which can be integrated into existing autonomous intersection management systems to help detect malicious vehicle behavior and generate evacuation plans. In the NWADE mechanism, we introduce the neighborhood watch concept whereby each vehicle around the intersection will serve as a watcher to report or verify the abnormal behavior of any nearby vehicle and the intersection manager. We propose a blockchain-based verification framework to guarantee the integrity and trustworthiness of the individual travel plans optimized for the entire intersection. We have conducted extensive experimental studies on various traffic scenarios, and the experimental results demonstrate the practicality, effectiveness, and efficiency of our mechanism.
{"title":"NWADE: A Neighborhood Watch Mechanism for Attack Detection and Evacuation in Autonomous Intersection Management","authors":"Jian Kang, Alian Yu, Wei Jiang, D. Lin","doi":"10.1109/ICDCS54860.2022.00117","DOIUrl":"https://doi.org/10.1109/ICDCS54860.2022.00117","url":null,"abstract":"With the advances in autonomous vehicles and intelligent intersection management systems, traffic lights may be replaced by optimal travel plans calculated for each passing vehicle in the future. While these technological advancements are envisioned to greatly improve travel efficiency, they are still facing various challenging security hurdles since even a single deviation of a vehicle from its assigned travel plan could cause a serious accident if the surrounding vehicles do not take necessary actions in a timely manner. In this paper, we propose a novel security mechanism namely NWADE which can be integrated into existing autonomous intersection management systems to help detect malicious vehicle behavior and generate evacuation plans. In the NWADE mechanism, we introduce the neighborhood watch concept whereby each vehicle around the intersection will serve as a watcher to report or verify the abnormal behavior of any nearby vehicle and the intersection manager. We propose a blockchain-based verification framework to guarantee the integrity and trustworthiness of the individual travel plans optimized for the entire intersection. We have conducted extensive experimental studies on various traffic scenarios, and the experimental results demonstrate the practicality, effectiveness, and efficiency of our mechanism.","PeriodicalId":225883,"journal":{"name":"2022 IEEE 42nd International Conference on Distributed Computing Systems (ICDCS)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121457316","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Public goods projects, including open source technology, client development, and blockchain knowledge education, play an important role in the flourishing blockchain ecosystem. Accordingly, decision making for public goods funding is a key issue in the studies of the blockchain ecosystem. This work develops a human oracle protocol approach, involved with public goods projects, experts, and funders, as a solution to the public goods investment problem on blockchain. In our human oracle, funders contribute their investments, which are stored in a funding pool. Experts provide investment advice on public goods projects based on their experience. Decisions made by the human oracle on the amount of support from the funding pool are based on experts’ reputation. The reputation of each expert is updated by the performance of the project’s implementation in comparison to her advice. That is, better investment performance brings a higher reputation. Besides being applied to static model, our human oracle can also be extended to accommodate dynamic settings, in which the experts might leave or join the decision-making process. We introduce a regret bound to measure the effectiveness of our human oracle. Theoretically, we prove an upper regret bound for both static and dynamic models, and prove its tightness with an asymptotically equal lower bound. Empirically, we show that our oracle’s investment decision is close to the optimal investment in hindsight.
{"title":"Funding Public Goods with Expert Advice in Blockchain System","authors":"Jichen Li, Yukun Cheng, Wenhan Huang, Mengqian Zhang, Jiarui Fan, Xiaotie Deng, Jan Xie","doi":"10.1109/ICDCS54860.2022.00026","DOIUrl":"https://doi.org/10.1109/ICDCS54860.2022.00026","url":null,"abstract":"Public goods projects, including open source technology, client development, and blockchain knowledge education, play an important role in the flourishing blockchain ecosystem. Accordingly, decision making for public goods funding is a key issue in the studies of the blockchain ecosystem. This work develops a human oracle protocol approach, involved with public goods projects, experts, and funders, as a solution to the public goods investment problem on blockchain. In our human oracle, funders contribute their investments, which are stored in a funding pool. Experts provide investment advice on public goods projects based on their experience. Decisions made by the human oracle on the amount of support from the funding pool are based on experts’ reputation. The reputation of each expert is updated by the performance of the project’s implementation in comparison to her advice. That is, better investment performance brings a higher reputation. Besides being applied to static model, our human oracle can also be extended to accommodate dynamic settings, in which the experts might leave or join the decision-making process. We introduce a regret bound to measure the effectiveness of our human oracle. Theoretically, we prove an upper regret bound for both static and dynamic models, and prove its tightness with an asymptotically equal lower bound. Empirically, we show that our oracle’s investment decision is close to the optimal investment in hindsight.","PeriodicalId":225883,"journal":{"name":"2022 IEEE 42nd International Conference on Distributed Computing Systems (ICDCS)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116855753","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-01DOI: 10.1109/ICDCS54860.2022.00041
Shaoke Xi, Fuliang Li, Xingwei Wang
Enforcing scheduling policies at end-hosts with software schedulers suffers from high CPU consumption, low throughput, and inaccuracy. Offloading scheduling functions to the network interface card (NIC) provides a promising direction to address these problems. However, existing efforts in scheduling offloading suffer from inflexible on-NIC packet schedulers, which cannot execute complex hierarchies of network policies. In this paper, we present FlowValve, the first parallel packet scheduler for Network Processor (NP)-based SmartNICs that offloads critical functions of Linux traffic control, including packet classifying and scheduling. The key insight behind FlowValve is to abstract inherent queues attached to the NIC interface (wire side) as a single FIFO queue and perform specialized tail drop to mix the FIFO queue with expected flow proportions. FlowValve takes advantage of on-chip multi-core parallelism and hardware accelerations to produce high throughput. Meanwhile, it substantially reduces CPU and memory burdens on end-hosts. We prototype FlowValve on a Netronome Agilio SmartNIC and demonstrate its effectiveness against non-offloaded kernel schedulers and DPDK QoS Scheduler. We find that FlowValve outperforms both in accurately enforcing network policies while driving line rate performance (i.e., 40Gbps), which contributes to saving at least two CPU cores.
{"title":"FlowValve: Packet Scheduling Offloaded on NP-based SmartNICs","authors":"Shaoke Xi, Fuliang Li, Xingwei Wang","doi":"10.1109/ICDCS54860.2022.00041","DOIUrl":"https://doi.org/10.1109/ICDCS54860.2022.00041","url":null,"abstract":"Enforcing scheduling policies at end-hosts with software schedulers suffers from high CPU consumption, low throughput, and inaccuracy. Offloading scheduling functions to the network interface card (NIC) provides a promising direction to address these problems. However, existing efforts in scheduling offloading suffer from inflexible on-NIC packet schedulers, which cannot execute complex hierarchies of network policies. In this paper, we present FlowValve, the first parallel packet scheduler for Network Processor (NP)-based SmartNICs that offloads critical functions of Linux traffic control, including packet classifying and scheduling. The key insight behind FlowValve is to abstract inherent queues attached to the NIC interface (wire side) as a single FIFO queue and perform specialized tail drop to mix the FIFO queue with expected flow proportions. FlowValve takes advantage of on-chip multi-core parallelism and hardware accelerations to produce high throughput. Meanwhile, it substantially reduces CPU and memory burdens on end-hosts. We prototype FlowValve on a Netronome Agilio SmartNIC and demonstrate its effectiveness against non-offloaded kernel schedulers and DPDK QoS Scheduler. We find that FlowValve outperforms both in accurately enforcing network policies while driving line rate performance (i.e., 40Gbps), which contributes to saving at least two CPU cores.","PeriodicalId":225883,"journal":{"name":"2022 IEEE 42nd International Conference on Distributed Computing Systems (ICDCS)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114549362","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-01DOI: 10.1109/ICDCS54860.2022.00103
Jiachen Wang, Tianyu Zhang, Dawei Shen, Xiao Hu, Song Han
Industrial wireless networks (IWNs) are being increasingly deployed in the field to serve as the network fabrics for various industrial Internet-of-Things (IIoT) applications. Given that IWNs typically operate in noisy and harsh environments, frequently occurring network dynamics post huge challenges for IWN resource management especially when the network scales up. Existing centralized and distributed network management solutions either suffer from large communication overhead and time delay, or introduce schedule collisions which unnecessarily degrade the system performance. To address these problems, this work proposes a novel HierArchical Resource Partitioning framework (HARP), to provide dynamic resource management in IWNs. By hierarchically partitioning and allocating resources for the links in the network, HARP enables distributed collision-free resource allocation. HARP enables rapid adjustment of the partitions in the presence of network dynamics with modest communication overhead. The effectiveness of HARP is validated and evaluated through both simulation studies and testbed experiments on a 50-node multi-channel multi-hop 6TiSCH network.
{"title":"HARP: Hierarchical Resource Partitioning in Dynamic Industrial Wireless Networks","authors":"Jiachen Wang, Tianyu Zhang, Dawei Shen, Xiao Hu, Song Han","doi":"10.1109/ICDCS54860.2022.00103","DOIUrl":"https://doi.org/10.1109/ICDCS54860.2022.00103","url":null,"abstract":"Industrial wireless networks (IWNs) are being increasingly deployed in the field to serve as the network fabrics for various industrial Internet-of-Things (IIoT) applications. Given that IWNs typically operate in noisy and harsh environments, frequently occurring network dynamics post huge challenges for IWN resource management especially when the network scales up. Existing centralized and distributed network management solutions either suffer from large communication overhead and time delay, or introduce schedule collisions which unnecessarily degrade the system performance. To address these problems, this work proposes a novel HierArchical Resource Partitioning framework (HARP), to provide dynamic resource management in IWNs. By hierarchically partitioning and allocating resources for the links in the network, HARP enables distributed collision-free resource allocation. HARP enables rapid adjustment of the partitions in the presence of network dynamics with modest communication overhead. The effectiveness of HARP is validated and evaluated through both simulation studies and testbed experiments on a 50-node multi-channel multi-hop 6TiSCH network.","PeriodicalId":225883,"journal":{"name":"2022 IEEE 42nd International Conference on Distributed Computing Systems (ICDCS)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134012584","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-01DOI: 10.1109/ICDCS54860.2022.00119
Chenxi Qiu, Sourabh Yadav, A. Squicciarini, Qing Yang, Song Fu, Juanjuan Zhao, Chengzhong Xu
To enable self-driving without a human driver, an autonomous vehicle needs to perceive its surrounding obstacles using onboard sensors, of which the perception accuracy might be limited by their own sensing range. An effective way to improve vehicles’ perception accuracy is to let nearby vehicles exchange their sensor data so that vehicles can detect obstacles beyond their own sensing ranges, called cooperative perception. The shared sensor data, however, might disclose the sensitive information of vehicles’ passengers, raising privacy and safety concerns (e.g. stalking or sensitive location leakage).In this paper, we propose a new data-sharing policy for the cooperative perception of autonomous vehicles, of which the objective is to minimize vehicles’ information disclosure without compromising their perception accuracy. Considering vehicles usually have different desires for data-sharing under different traffic environments, our policy provides vehicles autonomy to determine what types of sensor data to share based on their own needs. Moreover, given the dynamics of vehicles’ data-sharing decisions, the policy can be adjusted to incentivize vehicles’ decisions to converge to the desired decision field, such that a healthy cooperation environment can be maintained in a long term. To achieve such objectives, we analyze the dynamics of vehicles’ data-sharing decisions by resorting to the game theory model, and optimize the data-sharing ratio in the policy based on the analytic results. Finally, we carry out an extensive trace-driven simulation to test the performance of the proposed data-sharing policy. The experimental results demonstrate that our policy can help incentivize vehicles’ data-sharing decisions to the desired decision fields efficiently and effectively.
{"title":"Distributed Data-Sharing Consensus in Cooperative Perception of Autonomous Vehicles","authors":"Chenxi Qiu, Sourabh Yadav, A. Squicciarini, Qing Yang, Song Fu, Juanjuan Zhao, Chengzhong Xu","doi":"10.1109/ICDCS54860.2022.00119","DOIUrl":"https://doi.org/10.1109/ICDCS54860.2022.00119","url":null,"abstract":"To enable self-driving without a human driver, an autonomous vehicle needs to perceive its surrounding obstacles using onboard sensors, of which the perception accuracy might be limited by their own sensing range. An effective way to improve vehicles’ perception accuracy is to let nearby vehicles exchange their sensor data so that vehicles can detect obstacles beyond their own sensing ranges, called cooperative perception. The shared sensor data, however, might disclose the sensitive information of vehicles’ passengers, raising privacy and safety concerns (e.g. stalking or sensitive location leakage).In this paper, we propose a new data-sharing policy for the cooperative perception of autonomous vehicles, of which the objective is to minimize vehicles’ information disclosure without compromising their perception accuracy. Considering vehicles usually have different desires for data-sharing under different traffic environments, our policy provides vehicles autonomy to determine what types of sensor data to share based on their own needs. Moreover, given the dynamics of vehicles’ data-sharing decisions, the policy can be adjusted to incentivize vehicles’ decisions to converge to the desired decision field, such that a healthy cooperation environment can be maintained in a long term. To achieve such objectives, we analyze the dynamics of vehicles’ data-sharing decisions by resorting to the game theory model, and optimize the data-sharing ratio in the policy based on the analytic results. Finally, we carry out an extensive trace-driven simulation to test the performance of the proposed data-sharing policy. The experimental results demonstrate that our policy can help incentivize vehicles’ data-sharing decisions to the desired decision fields efficiently and effectively.","PeriodicalId":225883,"journal":{"name":"2022 IEEE 42nd International Conference on Distributed Computing Systems (ICDCS)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132056413","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-01DOI: 10.1109/ICDCS54860.2022.00132
Jagnyashini Debadarshini, S. Saha
An efficient Home Area Network (HAN) acts as a base of an Advanced Metering Infrastructure (AMI). A HAN not only facilitates AMI with efficient real-time monitoring of the electricity consumption but also manages the load profile of the whole system. However, the existing works on implementing HAN are mostly centralized and suffer from well-known problems. In this work, we propose an IoT-based efficient decentralized strategy using synchronous transmission to practically realize HAN. An inter-device coordination strategy is proposed to minimize the peak load as well as reduce the sudden changes in the overall system without compromising the user’s requirements. Through experiments over IoT-testbeds, we demonstrate that the proposed strategy can reduce the peak load upto 50% and reduce the load variations upto 58% for even a high and random rate of requests for execution of power-hungry house appliances.
{"title":"Collaborative Load Management in Smart Home Area Network","authors":"Jagnyashini Debadarshini, S. Saha","doi":"10.1109/ICDCS54860.2022.00132","DOIUrl":"https://doi.org/10.1109/ICDCS54860.2022.00132","url":null,"abstract":"An efficient Home Area Network (HAN) acts as a base of an Advanced Metering Infrastructure (AMI). A HAN not only facilitates AMI with efficient real-time monitoring of the electricity consumption but also manages the load profile of the whole system. However, the existing works on implementing HAN are mostly centralized and suffer from well-known problems. In this work, we propose an IoT-based efficient decentralized strategy using synchronous transmission to practically realize HAN. An inter-device coordination strategy is proposed to minimize the peak load as well as reduce the sudden changes in the overall system without compromising the user’s requirements. Through experiments over IoT-testbeds, we demonstrate that the proposed strategy can reduce the peak load upto 50% and reduce the load variations upto 58% for even a high and random rate of requests for execution of power-hungry house appliances.","PeriodicalId":225883,"journal":{"name":"2022 IEEE 42nd International Conference on Distributed Computing Systems (ICDCS)","volume":"385 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124781237","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
One-hop multicasting (OHM) of high-volume sensor data is essential for cooperative autonomous driving applications. While millimeter-Wave (mmWave) bands can be utilized for high-bandwidth OHM data transmission, it is very challenging for individual vehicles to find and communicate with a proper neighbor in a fully distributed and highly dynamic scenario. In this paper, we propose a fully distributed OHM scheme in vehicular networks, called mmV2V, which consists of three highly integrated protocols. Specifically, synchronized vehicles first conduct a probabilistic neighbor discovery procedure, in which randomly divided transmitters (or receivers) clockwise scan (or listen to) the surroundings in pace with heterogeneous Tx (or Rx) beams. In this way, the vast majority of neighbors can be identified in a few repeated rounds. Furthermore, vehicles negotiate with each of their neighbors about the optimal communication schedule in evenly distributed slots. Finally, each agreed pair of neighboring vehicles start high data rate transmissions with refined beams. We conduct extensive simulations and the results demonstrate that mmV2V can achieve a high completion ratio in rigid OHM tasks under various traffic conditions.
{"title":"mmV2V: Combating One-hop Multicasting in Millimeter-wave Vehicular Networks","authors":"Jiangang Shen, Hongzi Zhu, Yunxiang Cai, Bangzhao Zhai, Xudong Wang, Shan Chang, Haibin Cai, M. Guo","doi":"10.1109/ICDCS54860.2022.00076","DOIUrl":"https://doi.org/10.1109/ICDCS54860.2022.00076","url":null,"abstract":"One-hop multicasting (OHM) of high-volume sensor data is essential for cooperative autonomous driving applications. While millimeter-Wave (mmWave) bands can be utilized for high-bandwidth OHM data transmission, it is very challenging for individual vehicles to find and communicate with a proper neighbor in a fully distributed and highly dynamic scenario. In this paper, we propose a fully distributed OHM scheme in vehicular networks, called mmV2V, which consists of three highly integrated protocols. Specifically, synchronized vehicles first conduct a probabilistic neighbor discovery procedure, in which randomly divided transmitters (or receivers) clockwise scan (or listen to) the surroundings in pace with heterogeneous Tx (or Rx) beams. In this way, the vast majority of neighbors can be identified in a few repeated rounds. Furthermore, vehicles negotiate with each of their neighbors about the optimal communication schedule in evenly distributed slots. Finally, each agreed pair of neighboring vehicles start high data rate transmissions with refined beams. We conduct extensive simulations and the results demonstrate that mmV2V can achieve a high completion ratio in rigid OHM tasks under various traffic conditions.","PeriodicalId":225883,"journal":{"name":"2022 IEEE 42nd International Conference on Distributed Computing Systems (ICDCS)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125354727","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-01DOI: 10.1109/ICDCS54860.2022.00089
Xueyu Wu, Cho-Li Wang
Federated Averaging (FedAvg) and its variants are prevalent optimization algorithms adopted in Federated Learning (FL) as they show good model convergence. However, such optimization methods are mostly running in a synchronous flavor which is plagued by the straggler problem, especially in the real-world FL scenario. Federated learning involves a massive number of resource-weak edge devices connected to the intermittent networks, exhibiting a vastly heterogeneous training environment. The asynchronous setting is a plausible solution to fulfill the resources utilization. Yet, due to data and device heterogeneity, the training bias and model staleness dramatically downgrade the model performance. This paper presents KAFL, a fast-K Asynchronous Federated Learning framework, to improve the system and statistical efficiency. KAFL allows the global server to iteratively collect and aggregate (1) the parameters uploaded by the fastest K edge clients (K-FedAsync); or (2) the first M updated parameters sent from any clients (Mstep-FedAsync). Compared to the fully asynchronous setting, KAFL helps the server obtain a better direction toward the global optima as it collects the information from at least K clients or M parameters. To further improve the convergence speed of KAFL, we propose a new weighted aggregation method which dynamically adjusts the aggregation weights according to the weight deviation matrix and client contribution frequency. Experimental results show that KAFL achieves a significant time-to-target-accuracy speedup on both IID and Non-IID datasets. To achieve the same model accuracy, KAFL reduces more than 50% training time for five CNN and RNN models, demonstrating the high training efficiency of our proposed framework.
联邦平均法（FedAvg）及其变体是联邦学习（FL）中普遍采用的优化算法，因为它们显示出良好的模型收敛性。然而，这类优化方法大多以同步方式运行，存在 "散兵游勇"（standggler）问题，尤其是在现实世界的 FL 场景中。联盟学习涉及大量连接到间歇性网络的资源薄弱的边缘设备，呈现出巨大的异构训练环境。异步设置是一种合理的资源利用解决方案。然而，由于数据和设备的异构性，训练偏差和模型僵化会大大降低模型性能。本文提出了快速异步联合学习框架 KAFL，以提高系统和统计效率。KAFL 允许全局服务器迭代收集和汇总（1）最快的 K 个边缘客户端上传的参数（K-FedAsync）；或（2）任何客户端发送的前 M 个更新参数（Mstep-FedAsync）。与完全异步设置相比，KAFL 可以帮助服务器获得更好的全局最优方向，因为它至少收集了 K 个客户端或 M 个参数的信息。为了进一步提高 KAFL 的收敛速度，我们提出了一种新的加权聚合方法，该方法可根据权重偏差矩阵和客户端贡献频率动态调整聚合权重。实验结果表明，KAFL 在 IID 数据集和非 IID 数据集上都实现了显著的目标准确率加速。为了达到相同的模型精度，KAFL 为五个 CNN 和 RNN 模型减少了 50% 以上的训练时间，这表明我们提出的框架具有很高的训练效率。
{"title":"KAFL: Achieving High Training Efficiency for Fast-K Asynchronous Federated Learning","authors":"Xueyu Wu, Cho-Li Wang","doi":"10.1109/ICDCS54860.2022.00089","DOIUrl":"https://doi.org/10.1109/ICDCS54860.2022.00089","url":null,"abstract":"Federated Averaging (FedAvg) and its variants are prevalent optimization algorithms adopted in Federated Learning (FL) as they show good model convergence. However, such optimization methods are mostly running in a synchronous flavor which is plagued by the straggler problem, especially in the real-world FL scenario. Federated learning involves a massive number of resource-weak edge devices connected to the intermittent networks, exhibiting a vastly heterogeneous training environment. The asynchronous setting is a plausible solution to fulfill the resources utilization. Yet, due to data and device heterogeneity, the training bias and model staleness dramatically downgrade the model performance. This paper presents KAFL, a fast-K Asynchronous Federated Learning framework, to improve the system and statistical efficiency. KAFL allows the global server to iteratively collect and aggregate (1) the parameters uploaded by the fastest K edge clients (K-FedAsync); or (2) the first M updated parameters sent from any clients (Mstep-FedAsync). Compared to the fully asynchronous setting, KAFL helps the server obtain a better direction toward the global optima as it collects the information from at least K clients or M parameters. To further improve the convergence speed of KAFL, we propose a new weighted aggregation method which dynamically adjusts the aggregation weights according to the weight deviation matrix and client contribution frequency. Experimental results show that KAFL achieves a significant time-to-target-accuracy speedup on both IID and Non-IID datasets. To achieve the same model accuracy, KAFL reduces more than 50% training time for five CNN and RNN models, demonstrating the high training efficiency of our proposed framework.","PeriodicalId":225883,"journal":{"name":"2022 IEEE 42nd International Conference on Distributed Computing Systems (ICDCS)","volume":"86 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126136406","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2022-07-01DOI: 10.1109/ICDCS54860.2022.00017
Ernst Althaus, P. Berenbrink, A. Brinkmann, Rebecca Steiner
Solid State Drives (SSDs) have replaced magnetic disks in many application areas, as they provide very high performance for arbitrary access patterns. Nevertheless, data written to a physical page has to be erased before a page can be rewritten. The corresponding garbage collection (GC) process can only be performed on a block granularity, where a block includes many pages, impacting both the performance and lifetime of an SSD. The cost of a GC process is typically measured in terms of its write amplification, i.e., the number of blocks internally written by the SSD divided by the number of write requests of the host.Several GC heuristics have been proposed to optimize the write amplification of SSDs. These heuristics have been mostly empirically evaluated, while no thorough theoretical results are available on the optimality of GC algorithms even for seemingly simple cases like uniform and independent access distributions.In this work, we theoretically investigate the GREEDY GC strategy for uniformly independently distributed write accesses. We therefore model the garbage collection process on SSDs as a stochastic process and prove that the expected write amplification incurred by the GREEDY GC strategy is at most that of any other online GC strategy.
{"title":"On the Optimality of the Greedy Garbage Collection Strategy for SSDs","authors":"Ernst Althaus, P. Berenbrink, A. Brinkmann, Rebecca Steiner","doi":"10.1109/ICDCS54860.2022.00017","DOIUrl":"https://doi.org/10.1109/ICDCS54860.2022.00017","url":null,"abstract":"Solid State Drives (SSDs) have replaced magnetic disks in many application areas, as they provide very high performance for arbitrary access patterns. Nevertheless, data written to a physical page has to be erased before a page can be rewritten. The corresponding garbage collection (GC) process can only be performed on a block granularity, where a block includes many pages, impacting both the performance and lifetime of an SSD. The cost of a GC process is typically measured in terms of its write amplification, i.e., the number of blocks internally written by the SSD divided by the number of write requests of the host.Several GC heuristics have been proposed to optimize the write amplification of SSDs. These heuristics have been mostly empirically evaluated, while no thorough theoretical results are available on the optimality of GC algorithms even for seemingly simple cases like uniform and independent access distributions.In this work, we theoretically investigate the GREEDY GC strategy for uniformly independently distributed write accesses. We therefore model the garbage collection process on SSDs as a stochastic process and prove that the expected write amplification incurred by the GREEDY GC strategy is at most that of any other online GC strategy.","PeriodicalId":225883,"journal":{"name":"2022 IEEE 42nd International Conference on Distributed Computing Systems (ICDCS)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128327741","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：