Pub Date : 2021-12-01DOI: 10.1109/GLOBECOM46510.2021.9685499
Shun-Meng Huang, Yu-Wen Chen, Jian-Jhih Kuo
Federated learning (FL) enables multiple user de-vices to collaboratively train a global machine learning (ML) model by uploading their local models to the central server for aggregation. However, attackers may upload tampered local models (e.g., label-flipping attack) to corrupt the global model. Existing defense methods focus on outlier detection, but they are computationally intensive and can be circumvented by advanced model tampering. We employ a shuffling-based defense model to isolate the attackers from ordinary users. To explore the intrinsic properties, we simplify the defense model problem and formulate it as a Markov Decision Problem (MDP) to find the optimal policy. Then, we introduce a novel notion, (re)grouping, into the defense model to propose a new cost-efficient defense framework termed SAGE. Experiment results manifest that SAGE can effectively mitigate the impact of attacks in FL by efficiently decreasing the ratio of attacker devices to ordinary user devices. SAGE increases the testing accuracy of the targeted class by at most 40%.
{"title":"Cost-Efficient Shuffling and Regrouping Based Defense for Federated Learning","authors":"Shun-Meng Huang, Yu-Wen Chen, Jian-Jhih Kuo","doi":"10.1109/GLOBECOM46510.2021.9685499","DOIUrl":"https://doi.org/10.1109/GLOBECOM46510.2021.9685499","url":null,"abstract":"Federated learning (FL) enables multiple user de-vices to collaboratively train a global machine learning (ML) model by uploading their local models to the central server for aggregation. However, attackers may upload tampered local models (e.g., label-flipping attack) to corrupt the global model. Existing defense methods focus on outlier detection, but they are computationally intensive and can be circumvented by advanced model tampering. We employ a shuffling-based defense model to isolate the attackers from ordinary users. To explore the intrinsic properties, we simplify the defense model problem and formulate it as a Markov Decision Problem (MDP) to find the optimal policy. Then, we introduce a novel notion, (re)grouping, into the defense model to propose a new cost-efficient defense framework termed SAGE. Experiment results manifest that SAGE can effectively mitigate the impact of attacks in FL by efficiently decreasing the ratio of attacker devices to ordinary user devices. SAGE increases the testing accuracy of the targeted class by at most 40%.","PeriodicalId":200641,"journal":{"name":"2021 IEEE Global Communications Conference (GLOBECOM)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127939652","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 : 2021-12-01DOI: 10.1109/GLOBECOM46510.2021.9685951
S. S. Bacanli, Furkan Cimen, Enas Elgeldawi, D. Turgut
Commercially available unmanned aerial vehicles (UAVs) are usually more affordable and feasible for easy deployment compared to military-level UAVs in civilian applications. However, having a bounded range limits the use of commercially available UAVs in package dropping scenarios. In this paper, we have generated a synthetic dataset for the scenario in which drones or UAVs are used to drop packages to two neighborhoods. The charging and package pick-up station is located between two neighborhoods. By leveraging the synthetic dataset, the location of the charging station is predicted by machine learning techniques given the package request frequency, package dropping times of the UAV, and targeted package delay for the neighborhoods. The results showed that deep neural networks and support vector regressor are more successful in deciding the charging station location.
{"title":"Placement of Package Delivery Center for UAVs with Machine Learning","authors":"S. S. Bacanli, Furkan Cimen, Enas Elgeldawi, D. Turgut","doi":"10.1109/GLOBECOM46510.2021.9685951","DOIUrl":"https://doi.org/10.1109/GLOBECOM46510.2021.9685951","url":null,"abstract":"Commercially available unmanned aerial vehicles (UAVs) are usually more affordable and feasible for easy deployment compared to military-level UAVs in civilian applications. However, having a bounded range limits the use of commercially available UAVs in package dropping scenarios. In this paper, we have generated a synthetic dataset for the scenario in which drones or UAVs are used to drop packages to two neighborhoods. The charging and package pick-up station is located between two neighborhoods. By leveraging the synthetic dataset, the location of the charging station is predicted by machine learning techniques given the package request frequency, package dropping times of the UAV, and targeted package delay for the neighborhoods. The results showed that deep neural networks and support vector regressor are more successful in deciding the charging station location.","PeriodicalId":200641,"journal":{"name":"2021 IEEE Global Communications Conference (GLOBECOM)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131463672","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 : 2021-12-01DOI: 10.1109/GLOBECOM46510.2021.9685900
You-sheng Zhou, Yang Luo, M. Obaidat, P. Vijayakumar, Xiaojun Wang
With the continuous maturity of Internet of Things (IoT) technology, it has begun to be frequently used in all walks of life to improve people's work efficiency and living standards. The wide use of IoT in the medical field makes it convenient for patients to obtain medical services, and also enables doctors to obtain patients' physical conditions more timely and accurately, so as to formulate more efficient treatment plans. However, when people enjoy the convenience of medical IoT, how to ensure the security of communication and privacy of patients are all problems that cannot be ignored. In order to achieve secure access the network, this paper proposes an anonymous password authenticated key exchange protocol for medical Internet of Things (PAMI), where only a low-entropy password is required to realize the mutual authentication between medical device and telemedicine server, so as to negotiate a high-entropy session key. The security of PAMI is formally proved under the standard model, and the experiment based performance comparison demonstrates that it is more efficient than the existing similar schemes.
{"title":"PAMI-Anonymous Password Authentication Protocol for Medical Internet of Things","authors":"You-sheng Zhou, Yang Luo, M. Obaidat, P. Vijayakumar, Xiaojun Wang","doi":"10.1109/GLOBECOM46510.2021.9685900","DOIUrl":"https://doi.org/10.1109/GLOBECOM46510.2021.9685900","url":null,"abstract":"With the continuous maturity of Internet of Things (IoT) technology, it has begun to be frequently used in all walks of life to improve people's work efficiency and living standards. The wide use of IoT in the medical field makes it convenient for patients to obtain medical services, and also enables doctors to obtain patients' physical conditions more timely and accurately, so as to formulate more efficient treatment plans. However, when people enjoy the convenience of medical IoT, how to ensure the security of communication and privacy of patients are all problems that cannot be ignored. In order to achieve secure access the network, this paper proposes an anonymous password authenticated key exchange protocol for medical Internet of Things (PAMI), where only a low-entropy password is required to realize the mutual authentication between medical device and telemedicine server, so as to negotiate a high-entropy session key. The security of PAMI is formally proved under the standard model, and the experiment based performance comparison demonstrates that it is more efficient than the existing similar schemes.","PeriodicalId":200641,"journal":{"name":"2021 IEEE Global Communications Conference (GLOBECOM)","volume":"146 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132118174","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 : 2021-12-01DOI: 10.1109/GLOBECOM46510.2021.9685276
Hansini Vijayaraghavan, W. Kellerer
The ever-growing wireless networks demand high capacity, have strict latency requirements, and must support diverse communication services. A LiFi-WiFi heterogeneous net-work has proven to be a useful tool to satisfy the growing capacity demand. However, to leverage these co-existing, non-interfering technologies, intelligent resource management schemes have to be developed. To support diverse applications with varying delay and data rate requirements, the resource management scheme should consider the Quality of Service (QoS) while allocating wireless resources. In this work, the downlink wireless resources are allocated to users such that the average network packet delay is minimized. Users that are both capable and not capable of multi-homing are considered and a separate optimization problem is formulated for each case. These problems are then solved using a global Branch and Bound-based solver and a genetic algorithm-based Metaheuristic is also proposed. The algorithms are then evaluated with simulations and the results show that the average network packet delay is significantly lowered and each user's strict QoS requirements are satisfied even in a network with heavy traffic flow.
{"title":"Delay-aware Wireless Resource Allocation and User Association in LiFi-WiFi Heterogeneous Networks","authors":"Hansini Vijayaraghavan, W. Kellerer","doi":"10.1109/GLOBECOM46510.2021.9685276","DOIUrl":"https://doi.org/10.1109/GLOBECOM46510.2021.9685276","url":null,"abstract":"The ever-growing wireless networks demand high capacity, have strict latency requirements, and must support diverse communication services. A LiFi-WiFi heterogeneous net-work has proven to be a useful tool to satisfy the growing capacity demand. However, to leverage these co-existing, non-interfering technologies, intelligent resource management schemes have to be developed. To support diverse applications with varying delay and data rate requirements, the resource management scheme should consider the Quality of Service (QoS) while allocating wireless resources. In this work, the downlink wireless resources are allocated to users such that the average network packet delay is minimized. Users that are both capable and not capable of multi-homing are considered and a separate optimization problem is formulated for each case. These problems are then solved using a global Branch and Bound-based solver and a genetic algorithm-based Metaheuristic is also proposed. The algorithms are then evaluated with simulations and the results show that the average network packet delay is significantly lowered and each user's strict QoS requirements are satisfied even in a network with heavy traffic flow.","PeriodicalId":200641,"journal":{"name":"2021 IEEE Global Communications Conference (GLOBECOM)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132461757","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 : 2021-12-01DOI: 10.1109/GLOBECOM46510.2021.9685893
Hung T. Nguyen, Roberto Morabito, Kwang Taik Kim, M. Chiang
Edge computing has revolutionized the world of mobile and wireless networks world thanks to its flexible, secure, and performing characteristics. Lately, we have witnessed the increasing use of it to make more performing the deployment of machine learning (ML) techniques such as federated learning (FL). FL was debuted to improve communication efficiency compared to conventional distributed machine learning (ML). The original FL assumes a central aggregation server to aggregate locally optimized parameters and might bring reliability and latency issues. In this paper, we conduct an in-depth study of strategies to replace this central server by a flying master that is dynamically selected based on the current participants and/or available resources at every FL round of optimization. Specifically, we compare different metrics to select this flying master and assess consensus algorithms to perform the selection. Our results demonstrate a significant reduction of runtime using our flying master FL framework compared to the original FL from measurements results conducted in our EdgeAI testbed and over real 5G networks using an operational edge testbed.
{"title":"On-the-fly Resource-Aware Model Aggregation for Federated Learning in Heterogeneous Edge","authors":"Hung T. Nguyen, Roberto Morabito, Kwang Taik Kim, M. Chiang","doi":"10.1109/GLOBECOM46510.2021.9685893","DOIUrl":"https://doi.org/10.1109/GLOBECOM46510.2021.9685893","url":null,"abstract":"Edge computing has revolutionized the world of mobile and wireless networks world thanks to its flexible, secure, and performing characteristics. Lately, we have witnessed the increasing use of it to make more performing the deployment of machine learning (ML) techniques such as federated learning (FL). FL was debuted to improve communication efficiency compared to conventional distributed machine learning (ML). The original FL assumes a central aggregation server to aggregate locally optimized parameters and might bring reliability and latency issues. In this paper, we conduct an in-depth study of strategies to replace this central server by a flying master that is dynamically selected based on the current participants and/or available resources at every FL round of optimization. Specifically, we compare different metrics to select this flying master and assess consensus algorithms to perform the selection. Our results demonstrate a significant reduction of runtime using our flying master FL framework compared to the original FL from measurements results conducted in our EdgeAI testbed and over real 5G networks using an operational edge testbed.","PeriodicalId":200641,"journal":{"name":"2021 IEEE Global Communications Conference (GLOBECOM)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130088192","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 : 2021-12-01DOI: 10.1109/GLOBECOM46510.2021.9685336
Yanjiao Chen, Runmin Ou, Y. Deng, Xiaoyan Yin
With recent advances in the study of biometrics, gait analysis has drawn much attention for its potential use in forensics, surveillance, and legal systems. In this paper, we present WIAGE, a contactless and non-intrusive gait-based age estimation system, which leverages wireless sensing to perform gait analysis to infer the age of individuals. Traditional age estimation systems either require users to carry wearable devices that are inconvenient or rely on image processing that is computationally intensive and sensitive to lighting conditions and occlusion. In contrast, WIAGE utilizes the incumbent WiFi infrastructure to infer the age of users with minimal interference to their activities. We adopt a series of signal processing techniques to recover clear gait patterns from the noisy WiFi signals and extract the most relevant features from steps that can be used for robust age estimation. The experimental results show that WIAGE can achieve an age estimation accuracy of 95.2% for 23 users, which demonstrates the feasibility and effectiveness of our proposed system.
{"title":"WIAGE: A Gait-based Age Estimation System Using Wireless Signals","authors":"Yanjiao Chen, Runmin Ou, Y. Deng, Xiaoyan Yin","doi":"10.1109/GLOBECOM46510.2021.9685336","DOIUrl":"https://doi.org/10.1109/GLOBECOM46510.2021.9685336","url":null,"abstract":"With recent advances in the study of biometrics, gait analysis has drawn much attention for its potential use in forensics, surveillance, and legal systems. In this paper, we present WIAGE, a contactless and non-intrusive gait-based age estimation system, which leverages wireless sensing to perform gait analysis to infer the age of individuals. Traditional age estimation systems either require users to carry wearable devices that are inconvenient or rely on image processing that is computationally intensive and sensitive to lighting conditions and occlusion. In contrast, WIAGE utilizes the incumbent WiFi infrastructure to infer the age of users with minimal interference to their activities. We adopt a series of signal processing techniques to recover clear gait patterns from the noisy WiFi signals and extract the most relevant features from steps that can be used for robust age estimation. The experimental results show that WIAGE can achieve an age estimation accuracy of 95.2% for 23 users, which demonstrates the feasibility and effectiveness of our proposed system.","PeriodicalId":200641,"journal":{"name":"2021 IEEE Global Communications Conference (GLOBECOM)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130263539","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 : 2021-12-01DOI: 10.1109/GLOBECOM46510.2021.9685259
Ruyang Wang, Chunyan Zang, Peng He, Yaping Cui, D. Wu
Mobile edge computing (MEC) enables resource-constrained mobile devices (MDs) to offload their tasks onto nearby edge servers. However, there exists a profit allocation problem between users and edge nodes (ENs) due to the limi-tations of ENs computing capacity and spectrum resources. In this paper, we propose an auction pricing-based MEC offloading strategy to maximize the profit of ENs. Firstly, we design an overall auction process using the binary offloading model by considering MDs battery capacity, basic profit, and tasks tolerable delay. Secondly, the bidding willingness of MDs in each round of auction are given on the premise of effectively ensuring users rationality. Finally, an auction pricing-based task offloading strat-egy is proposed, in which the winner of a single-round auction can offload its computation task to the ES. Simulation results verify the performance of the proposed strategy. Compared with the VA algorithm, the profit obtained by ENs has increased by 23.8%.
{"title":"Auction Pricing-Based Task Offloading Strategy for Cooperative Edge Computing","authors":"Ruyang Wang, Chunyan Zang, Peng He, Yaping Cui, D. Wu","doi":"10.1109/GLOBECOM46510.2021.9685259","DOIUrl":"https://doi.org/10.1109/GLOBECOM46510.2021.9685259","url":null,"abstract":"Mobile edge computing (MEC) enables resource-constrained mobile devices (MDs) to offload their tasks onto nearby edge servers. However, there exists a profit allocation problem between users and edge nodes (ENs) due to the limi-tations of ENs computing capacity and spectrum resources. In this paper, we propose an auction pricing-based MEC offloading strategy to maximize the profit of ENs. Firstly, we design an overall auction process using the binary offloading model by considering MDs battery capacity, basic profit, and tasks tolerable delay. Secondly, the bidding willingness of MDs in each round of auction are given on the premise of effectively ensuring users rationality. Finally, an auction pricing-based task offloading strat-egy is proposed, in which the winner of a single-round auction can offload its computation task to the ES. Simulation results verify the performance of the proposed strategy. Compared with the VA algorithm, the profit obtained by ENs has increased by 23.8%.","PeriodicalId":200641,"journal":{"name":"2021 IEEE Global Communications Conference (GLOBECOM)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133862287","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 : 2021-12-01DOI: 10.1109/GLOBECOM46510.2021.9685987
Shuyang Fang, Zhengchuan Chen, Zhong Tian, Yunjian Jia, Min Wang
For timeliness-sensitive applications in Internet of things (IoT) systems, it is critical to efficiently allocate transmission resources such that the freshness of information updates can be improved. This paper focuses on timely status updating in an orthogonal frequency division multiplexing-based IoT systems, in which all devices update status to one data center by sharing available bandwidth. To improve the timeliness of updates, a resource allocation optimization problem is formulated, based on finite blocklength (FBL) transmission and the age of information (AoI) metric. Two suboptimal policies, namely, fixed time slot policy and fixed blocklength policy, along with an iterative optimization algorithm, and an approximate optimal policy are presented for addressing the optimal resource allocation. By comparing the performance of different policies, it is shown that the iterative algorithm and the approximate optimal policy outperforms the other two suboptimal policies, and closely approaches the global optimal resource allocation.
{"title":"Resource Allocation for Age of Information Minimization in An OFDM Status Update System","authors":"Shuyang Fang, Zhengchuan Chen, Zhong Tian, Yunjian Jia, Min Wang","doi":"10.1109/GLOBECOM46510.2021.9685987","DOIUrl":"https://doi.org/10.1109/GLOBECOM46510.2021.9685987","url":null,"abstract":"For timeliness-sensitive applications in Internet of things (IoT) systems, it is critical to efficiently allocate transmission resources such that the freshness of information updates can be improved. This paper focuses on timely status updating in an orthogonal frequency division multiplexing-based IoT systems, in which all devices update status to one data center by sharing available bandwidth. To improve the timeliness of updates, a resource allocation optimization problem is formulated, based on finite blocklength (FBL) transmission and the age of information (AoI) metric. Two suboptimal policies, namely, fixed time slot policy and fixed blocklength policy, along with an iterative optimization algorithm, and an approximate optimal policy are presented for addressing the optimal resource allocation. By comparing the performance of different policies, it is shown that the iterative algorithm and the approximate optimal policy outperforms the other two suboptimal policies, and closely approaches the global optimal resource allocation.","PeriodicalId":200641,"journal":{"name":"2021 IEEE Global Communications Conference (GLOBECOM)","volume":"37 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133940139","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 : 2021-12-01DOI: 10.1109/GLOBECOM46510.2021.9685356
Ali El Amine, O. Brun, Slim Abdellatif, Pascal Berthou
We consider the SFC embedding (SFCE) problem in the Slice as a Service (SlaaS) model. In this model, a slice provider leases resources from multiple cloud and network providers in order to instantiate the Service Function Chain (SFC) requested by a slice tenant. As the slice provider has no visibility on the infrastructures of the resource providers, in which resources may be purchased and released quite rapidly, it has to query them to determine what are the possible allocations and their costs. We show that when there are many resource providers and many VNFs composing the SFC, the number of queries to be made for discovering a minimum cost SFC embedding grows quickly, leading to excessively long deployment times. In order to reduce the latter quantity, we propose to query resource providers strategically, rather than collecting the information on all possible allocations at once. We provide bounds on the number of queries to be made in this approach, and propose to exploit a Shortest Path Discovery algorithm in order to reduce this number of queries and thus the SFC deployment time. Our numerical results suggest that this algorithm is fairly efficient, and that the deployment times can be significantly shortened, in particular when initial estimates of allocation costs can be provided by the slice provider.
我们在切片即服务(SlaaS)模型中考虑了SFC嵌入(SFCE)问题。在这个模型中，一个片提供程序从多个云和网络提供程序租用资源，以便实例化片承租者请求的服务功能链(Service Function Chain, SFC)。由于片提供程序对资源提供程序的基础设施不具有可视性，因此它必须查询资源提供程序，以确定可能的分配情况及其成本。我们表明，当有许多资源提供者和许多vnf组成SFC时，为发现最低成本的SFC嵌入而进行的查询数量会迅速增加，从而导致部署时间过长。为了减少后一种数量，我们建议策略性地查询资源提供者，而不是一次收集所有可能分配的信息。我们提供了在这种方法中要进行的查询数量的限制，并建议利用最短路径发现算法来减少查询数量，从而减少SFC部署时间。我们的数值结果表明，该算法是相当有效的，并且可以显著缩短部署时间，特别是当分片提供商可以提供分配成本的初始估计时。
{"title":"Shortening the Deployment Time of SFCs by Adaptively Querying Resource Providers","authors":"Ali El Amine, O. Brun, Slim Abdellatif, Pascal Berthou","doi":"10.1109/GLOBECOM46510.2021.9685356","DOIUrl":"https://doi.org/10.1109/GLOBECOM46510.2021.9685356","url":null,"abstract":"We consider the SFC embedding (SFCE) problem in the Slice as a Service (SlaaS) model. In this model, a slice provider leases resources from multiple cloud and network providers in order to instantiate the Service Function Chain (SFC) requested by a slice tenant. As the slice provider has no visibility on the infrastructures of the resource providers, in which resources may be purchased and released quite rapidly, it has to query them to determine what are the possible allocations and their costs. We show that when there are many resource providers and many VNFs composing the SFC, the number of queries to be made for discovering a minimum cost SFC embedding grows quickly, leading to excessively long deployment times. In order to reduce the latter quantity, we propose to query resource providers strategically, rather than collecting the information on all possible allocations at once. We provide bounds on the number of queries to be made in this approach, and propose to exploit a Shortest Path Discovery algorithm in order to reduce this number of queries and thus the SFC deployment time. Our numerical results suggest that this algorithm is fairly efficient, and that the deployment times can be significantly shortened, in particular when initial estimates of allocation costs can be provided by the slice provider.","PeriodicalId":200641,"journal":{"name":"2021 IEEE Global Communications Conference (GLOBECOM)","volume":"153 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133949155","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 : 2021-12-01DOI: 10.1109/GLOBECOM46510.2021.9685685
Yuyan Wu, Runzhou Li, P. Hong
Due to the characteristic of large-radix routers, the Dragonfly topology can achieve low diameter, high performance/cost ratio. However, in the Dragonfly networks deployed with Remote Direct Memory Access (RDMA), existing packet-level routing algorithms which are mostly based on queue length information, are neither good enough to achieve load balancing nor meet the requirement of in order. To tackle the above issues, we first analyze the drawbacks of flow-level source routing in RDMA-enabled Dragonfly networks. Then, a flow-level rerouting scheme that can estimate traffic distribution and link load based on the routers' history information is proposed. Finally, the simulation results show that our scheme can obtain significant performance gains over existing algorithms in both average flow completion time (AFCT) and saturation throughput. In particular, under the adversarial traffic pattern, our scheme can greatly reduce the AFCT of flow-level UGAL by 25% and improve the saturation throughput by 13% while avoiding disorder.
由于大基数路由器的特性，蜻蜓拓扑可以实现低直径、高性能/成本比。然而，在部署RDMA (Remote Direct Memory Access)的蜻蜓网络中，现有的分组级路由算法大多基于队列长度信息，既不能很好地实现负载均衡，也不能满足有序的要求。为了解决上述问题，我们首先分析了支持rdma的蜻蜓网络中流级源路由的缺点。然后，提出了一种基于路由器历史信息估计流量分布和链路负载的流级重路由方案。仿真结果表明，该方案在平均流完井时间(AFCT)和饱和吞吐量方面都比现有算法有显著的性能提升。特别是在对抗流量模式下，我们的方案可以在避免混乱的同时，将流级UGAL的AFCT大大降低25%，将饱和吞吐量提高13%。
{"title":"Flow-Level Rerouting in RDMA-Enabled Dragonfly Networks","authors":"Yuyan Wu, Runzhou Li, P. Hong","doi":"10.1109/GLOBECOM46510.2021.9685685","DOIUrl":"https://doi.org/10.1109/GLOBECOM46510.2021.9685685","url":null,"abstract":"Due to the characteristic of large-radix routers, the Dragonfly topology can achieve low diameter, high performance/cost ratio. However, in the Dragonfly networks deployed with Remote Direct Memory Access (RDMA), existing packet-level routing algorithms which are mostly based on queue length information, are neither good enough to achieve load balancing nor meet the requirement of in order. To tackle the above issues, we first analyze the drawbacks of flow-level source routing in RDMA-enabled Dragonfly networks. Then, a flow-level rerouting scheme that can estimate traffic distribution and link load based on the routers' history information is proposed. Finally, the simulation results show that our scheme can obtain significant performance gains over existing algorithms in both average flow completion time (AFCT) and saturation throughput. In particular, under the adversarial traffic pattern, our scheme can greatly reduce the AFCT of flow-level UGAL by 25% and improve the saturation throughput by 13% while avoiding disorder.","PeriodicalId":200641,"journal":{"name":"2021 IEEE Global Communications Conference (GLOBECOM)","volume":"118 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134161938","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：