Pub Date : 2021-06-15DOI: 10.1109/MedComNet52149.2021.9501281
Kariem Fahmi, D. Leith, S. Kucera, H. Claussen
In this paper, we discuss the challenges faced by MPTCP when used to aggregate multiple WAN/Internet connections in Multi-WAN Routers (MWR). We observe that the two architectural variants, proxying and tunneling, used to deploy MPTCP in MWR suffer from key performance problems. First, the proxy variant creates one MPTCP connection for each TCP connection, which results in a large number of parallel uncoordinated MPTCP connections, which we explain leads to underutilization of the available capacity, suboptimal multipath scheduling, and increased loss rate. Second, the tunnel variant, which relies on encapsulating TCP over MPTCP, stacks two reliability layers and leads to a large number of spurious retransmissions, an issue known as TCP meltdown. Instead, we propose a new multi-path solution more suited to MWR, called BOOST. This solution eliminates the problems with both the proxy and tunnel approaches by multiplexing TCP connections over a single persistent multi-path connection. BOOST also takes a novel approach to multi-path scheduling that combines multipath load balancing and scheduling. In particular, short flows are transmitted across a single link to avoid HoL blocking while longer flows are opportunistically transmitted across multiple paths, utilizing left-over capacity. Evaluations show that BOOST provides better throughput, lower losses, and retransmissions while requiring less memory compared to both MPTCP variants
{"title":"BOOST: Transport-Layer Multi-Connectivity Solution for Multi-Wan Routers","authors":"Kariem Fahmi, D. Leith, S. Kucera, H. Claussen","doi":"10.1109/MedComNet52149.2021.9501281","DOIUrl":"https://doi.org/10.1109/MedComNet52149.2021.9501281","url":null,"abstract":"In this paper, we discuss the challenges faced by MPTCP when used to aggregate multiple WAN/Internet connections in Multi-WAN Routers (MWR). We observe that the two architectural variants, proxying and tunneling, used to deploy MPTCP in MWR suffer from key performance problems. First, the proxy variant creates one MPTCP connection for each TCP connection, which results in a large number of parallel uncoordinated MPTCP connections, which we explain leads to underutilization of the available capacity, suboptimal multipath scheduling, and increased loss rate. Second, the tunnel variant, which relies on encapsulating TCP over MPTCP, stacks two reliability layers and leads to a large number of spurious retransmissions, an issue known as TCP meltdown. Instead, we propose a new multi-path solution more suited to MWR, called BOOST. This solution eliminates the problems with both the proxy and tunnel approaches by multiplexing TCP connections over a single persistent multi-path connection. BOOST also takes a novel approach to multi-path scheduling that combines multipath load balancing and scheduling. In particular, short flows are transmitted across a single link to avoid HoL blocking while longer flows are opportunistically transmitted across multiple paths, utilizing left-over capacity. Evaluations show that BOOST provides better throughput, lower losses, and retransmissions while requiring less memory compared to both MPTCP variants","PeriodicalId":272937,"journal":{"name":"2021 19th Mediterranean Communication and Computer Networking Conference (MedComNet)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133770825","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 : 2021-06-15DOI: 10.1109/MedComNet52149.2021.9501235
M. Knitter, W. Endemann, R. Kays
The upcoming IEEE 802.11ax standard introduces spatial reuse and enhanced spatial division multiplexing as two measures to improve user throughput by parallel transmitting links and spatial streams. Related work shows that the proposed spatial reuse concept does not fully utilize the spatial reuse potential. The standard limits spatial division multiplexing to one network. This paper investigates a spatial division multiplexing concept across multiple, decentrally organized networks to cancel co-channel interference. The paper combines the concept with an improved version of spatial reuse and introduces a corresponding channel access protocol, CSMA/SDMSR. An IEEE 802.11ax based simulation of reference scenarios demonstrates a CSMA/SDMSR system performance improvement up to 108 % and decreased latency compared to standard CSMA/CA.
{"title":"Combined Spatial Division Multiplexing and Spatial Reuse across Decentral Wireless LANs","authors":"M. Knitter, W. Endemann, R. Kays","doi":"10.1109/MedComNet52149.2021.9501235","DOIUrl":"https://doi.org/10.1109/MedComNet52149.2021.9501235","url":null,"abstract":"The upcoming IEEE 802.11ax standard introduces spatial reuse and enhanced spatial division multiplexing as two measures to improve user throughput by parallel transmitting links and spatial streams. Related work shows that the proposed spatial reuse concept does not fully utilize the spatial reuse potential. The standard limits spatial division multiplexing to one network. This paper investigates a spatial division multiplexing concept across multiple, decentrally organized networks to cancel co-channel interference. The paper combines the concept with an improved version of spatial reuse and introduces a corresponding channel access protocol, CSMA/SDMSR. An IEEE 802.11ax based simulation of reference scenarios demonstrates a CSMA/SDMSR system performance improvement up to 108 % and decreased latency compared to standard CSMA/CA.","PeriodicalId":272937,"journal":{"name":"2021 19th Mediterranean Communication and Computer Networking Conference (MedComNet)","volume":"108 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124150087","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 : 2021-05-21DOI: 10.1109/MedComNet52149.2021.9501237
Álvaro López-Raventós, B. Bellalta
The multi-link operation (MLO) is a new feature proposed to be part of the IEEE 802.11be Extremely High Throughput (EHT) amendment. Through MLO, access points and stations will be provided with the capabilities to transmit and receive data from the same traffic flow over multiple radio interfaces. However, the question on how traffic flows should be distributed over the different interfaces to maximize the WLAN performance is still unresolved. To that end, we evaluate in this article different traffic allocation policies, under a wide variety of scenarios and traffic loads, in order to shed some light on that question. The obtained results confirm that congestion-aware policies outperform static ones. However, and more importantly, the results also reveal that traffic flows become highly vulnerable to the activity of neighboring networks when they are distributed across multiple links. As a result, the best performance is obtained when a new arriving flow is simply assigned entirely to the emptiest interface.
{"title":"IEEE 802.11be Multi-Link Operation: When the Best Could Be to Use Only a Single Interface","authors":"Álvaro López-Raventós, B. Bellalta","doi":"10.1109/MedComNet52149.2021.9501237","DOIUrl":"https://doi.org/10.1109/MedComNet52149.2021.9501237","url":null,"abstract":"The multi-link operation (MLO) is a new feature proposed to be part of the IEEE 802.11be Extremely High Throughput (EHT) amendment. Through MLO, access points and stations will be provided with the capabilities to transmit and receive data from the same traffic flow over multiple radio interfaces. However, the question on how traffic flows should be distributed over the different interfaces to maximize the WLAN performance is still unresolved. To that end, we evaluate in this article different traffic allocation policies, under a wide variety of scenarios and traffic loads, in order to shed some light on that question. The obtained results confirm that congestion-aware policies outperform static ones. However, and more importantly, the results also reveal that traffic flows become highly vulnerable to the activity of neighboring networks when they are distributed across multiple links. As a result, the best performance is obtained when a new arriving flow is simply assigned entirely to the emptiest interface.","PeriodicalId":272937,"journal":{"name":"2021 19th Mediterranean Communication and Computer Networking Conference (MedComNet)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129121502","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 : 2021-05-06DOI: 10.1109/MedComNet52149.2021.9501272
T. Si Salem, Gabriele Castellano, G. Neglia, Fabio Pianese, Andrea Araldo
We present the novel idea of inference delivery networks (IDN), networks of computing nodes that coordinate to satisfy inference requests achieving the best trade-off between latency and accuracy. IDNs bridge the dichotomy between device and cloud execution by integrating inference delivery at the various tiers of the infrastructure continuum (access, edge, regional data center, cloud). We propose a distributed dynamic policy for ML model allocation in an IDN by which each node periodically updates its local set of inference models based on requests observed during the recent past plus limited information exchange with its neighbor nodes. Our policy offers strong performance guarantees in an adversarial setting and shows improvements over greedy heuristics with similar complexity in realistic scenarios.
{"title":"Towards Inference Delivery Networks: Distributing Machine Learning with Optimality Guarantees","authors":"T. Si Salem, Gabriele Castellano, G. Neglia, Fabio Pianese, Andrea Araldo","doi":"10.1109/MedComNet52149.2021.9501272","DOIUrl":"https://doi.org/10.1109/MedComNet52149.2021.9501272","url":null,"abstract":"We present the novel idea of inference delivery networks (IDN), networks of computing nodes that coordinate to satisfy inference requests achieving the best trade-off between latency and accuracy. IDNs bridge the dichotomy between device and cloud execution by integrating inference delivery at the various tiers of the infrastructure continuum (access, edge, regional data center, cloud). We propose a distributed dynamic policy for ML model allocation in an IDN by which each node periodically updates its local set of inference models based on requests observed during the recent past plus limited information exchange with its neighbor nodes. Our policy offers strong performance guarantees in an adversarial setting and shows improvements over greedy heuristics with similar complexity in realistic scenarios.","PeriodicalId":272937,"journal":{"name":"2021 19th Mediterranean Communication and Computer Networking Conference (MedComNet)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-05-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128445425","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 : 2021-02-24DOI: 10.1109/MedComNet52149.2021.9501274
Martino Trevisan, Danilo Giordano, I. Drago, Ali Safari Khatouni
The third version of the Hypertext Transfer Protocol (HTTP) is in its final standardization phase by the IETF. Besides better security and increased flexibility, it promises benefits in terms of performance. HTTP/3 adopts a more efficient header compression schema and replaces TCP with QUIC, a transport protocol carried over UDP, originally proposed by Google and currently under standardization too. Although HTTP/3 early implementations already exist and some websites announce its support, it has been subject to few studies. We provide a first measurement study on HTTP/3 adoption and performance. We testify how it has been adopted by some of the leading Internet companies such as Google, Facebook and Cloudflare in 2020. We run a large-scale measurement campaign towards thousands of websites adopting HTTP/3, aiming at understanding to what extent it achieves better performance than HTTP/2. We find that adopting websites often host most web page objects on third-party servers, which support only HTTP/2 or even HTTP/1.1. As excepted, websites loading objects from a limited set of third-party domains (avoiding legacy protocols) are those experiencing larger performance gains. Our experiments however show that HTTP/3 provides sizable benefits only in scenarios with high latency or poor bandwidth.
{"title":"Measuring HTTP/3: Adoption and Performance","authors":"Martino Trevisan, Danilo Giordano, I. Drago, Ali Safari Khatouni","doi":"10.1109/MedComNet52149.2021.9501274","DOIUrl":"https://doi.org/10.1109/MedComNet52149.2021.9501274","url":null,"abstract":"The third version of the Hypertext Transfer Protocol (HTTP) is in its final standardization phase by the IETF. Besides better security and increased flexibility, it promises benefits in terms of performance. HTTP/3 adopts a more efficient header compression schema and replaces TCP with QUIC, a transport protocol carried over UDP, originally proposed by Google and currently under standardization too. Although HTTP/3 early implementations already exist and some websites announce its support, it has been subject to few studies. We provide a first measurement study on HTTP/3 adoption and performance. We testify how it has been adopted by some of the leading Internet companies such as Google, Facebook and Cloudflare in 2020. We run a large-scale measurement campaign towards thousands of websites adopting HTTP/3, aiming at understanding to what extent it achieves better performance than HTTP/2. We find that adopting websites often host most web page objects on third-party servers, which support only HTTP/2 or even HTTP/1.1. As excepted, websites loading objects from a limited set of third-party domains (avoiding legacy protocols) are those experiencing larger performance gains. Our experiments however show that HTTP/3 provides sizable benefits only in scenarios with high latency or poor bandwidth.","PeriodicalId":272937,"journal":{"name":"2021 19th Mediterranean Communication and Computer Networking Conference (MedComNet)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131920344","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 : 2020-11-12DOI: 10.1109/MedComNet52149.2021.9501246
Houssam ElBouanani, C. Barakat, W. Dabbous, T. Turletti
Emulation has become a popular approach for the validation and evaluation of network research. It provides researchers with a contained, customizable, and scalable testing environment, which can be easily packaged and published for potential readers to reproduce their results. However, as the network components are only virtual, emulation lacks the inherent realism of physical testbeds. In light of this, monitoring specific metrics of the emulated network has been proposed as a solution to mitigate to some degree inaccuracies caused by emulation. While this is not difficult to implement in a single-machine setting (e.g. with Mininet), monitoring is limited by the lack of time synchronization in scenarios where the emulation is distributed over multiple physical machines (e.g., Distrinet). In this paper we tackle the case of packet delay monitoring, to which we propose a methodology for passively measuring one-way delays with underlying assumptions about time synchronization, and round-trip delays otherwise. For an efficient implementation of our methodology, we propose an eBPF-based packet measurement tool that performs better than current packet sniffers under emulation-specific assumptions. We implement and evaluate our system in an open testbed and show that it can reach results within few microseconds of perfect accuracy and precision.
{"title":"Passive Delay Measurement for Fidelity Monitoring of Distributed Network Emulation","authors":"Houssam ElBouanani, C. Barakat, W. Dabbous, T. Turletti","doi":"10.1109/MedComNet52149.2021.9501246","DOIUrl":"https://doi.org/10.1109/MedComNet52149.2021.9501246","url":null,"abstract":"Emulation has become a popular approach for the validation and evaluation of network research. It provides researchers with a contained, customizable, and scalable testing environment, which can be easily packaged and published for potential readers to reproduce their results. However, as the network components are only virtual, emulation lacks the inherent realism of physical testbeds. In light of this, monitoring specific metrics of the emulated network has been proposed as a solution to mitigate to some degree inaccuracies caused by emulation. While this is not difficult to implement in a single-machine setting (e.g. with Mininet), monitoring is limited by the lack of time synchronization in scenarios where the emulation is distributed over multiple physical machines (e.g., Distrinet). In this paper we tackle the case of packet delay monitoring, to which we propose a methodology for passively measuring one-way delays with underlying assumptions about time synchronization, and round-trip delays otherwise. For an efficient implementation of our methodology, we propose an eBPF-based packet measurement tool that performs better than current packet sniffers under emulation-specific assumptions. We implement and evaluate our system in an open testbed and show that it can reach results within few microseconds of perfect accuracy and precision.","PeriodicalId":272937,"journal":{"name":"2021 19th Mediterranean Communication and Computer Networking Conference (MedComNet)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-11-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128824781","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 : 2020-11-10DOI: 10.1109/MedComNet52149.2021.9501236
Mattia Lecci, Matteo Drago, A. Zanella, M. Zorzi
Many current and future multimedia and industrial applications, like video streaming, eXtended Reality or remote robot control, are characterized by periodic data transmissions with strict latency and reliability constraints. In an effort to meet the stringent demand of such traffic sources, the WiGig standards support a contention-free channel access mechanism, named Service Period, that makes it possible to allocate dedicated time intervals to certain wireless stations. However, the standard only covers the fundamental aspects that ensure interoperability, while the actual schedule logic is left to vendors. In this paper, we propose two algorithms for joint admission control and scheduling of periodic traffic streams with contrasting performance objectives, specifically a simple scheduler and a max-min fair scheduler. The schemes are compared in two different scenarios, in order to characterize and highlight some fundamental trade-offs. As expected from their design principles, the simple scheduler tends to accept more homogeneous resource allocation requests, while the max-min scheduler can efficiently handle more diversified requests, at the cost of a small loss in terms of total resource utilization.
{"title":"Exploiting Scheduled Access Features of mmWave WLANs for Periodic Traffic Sources","authors":"Mattia Lecci, Matteo Drago, A. Zanella, M. Zorzi","doi":"10.1109/MedComNet52149.2021.9501236","DOIUrl":"https://doi.org/10.1109/MedComNet52149.2021.9501236","url":null,"abstract":"Many current and future multimedia and industrial applications, like video streaming, eXtended Reality or remote robot control, are characterized by periodic data transmissions with strict latency and reliability constraints. In an effort to meet the stringent demand of such traffic sources, the WiGig standards support a contention-free channel access mechanism, named Service Period, that makes it possible to allocate dedicated time intervals to certain wireless stations. However, the standard only covers the fundamental aspects that ensure interoperability, while the actual schedule logic is left to vendors. In this paper, we propose two algorithms for joint admission control and scheduling of periodic traffic streams with contrasting performance objectives, specifically a simple scheduler and a max-min fair scheduler. The schemes are compared in two different scenarios, in order to characterize and highlight some fundamental trade-offs. As expected from their design principles, the simple scheduler tends to accept more homogeneous resource allocation requests, while the max-min scheduler can efficiently handle more diversified requests, at the cost of a small loss in terms of total resource utilization.","PeriodicalId":272937,"journal":{"name":"2021 19th Mediterranean Communication and Computer Networking Conference (MedComNet)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116629163","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 : 2020-10-13DOI: 10.1109/MedComNet52149.2021.9501238
Felipe Gómez-Cuba, Tommaso Zugno, Junseok Kim, Michele Polese, S. Bahk, M. Zorzi
This paper analyzes Hybrid Beamforming (HBF) and Multi-User Multiple-Input Multiple-Output (MU-MIMO) in millimeter wave (mmWave) 5th generation (5G) cellular networks considering the full protocol stack with TCP/IP traffic and MAC scheduling. Prior work on HBF and MU-MIMO has assumed full-buffer transmissions and studied link-level performance. We report non-trivial interactions between the HBF technique, the front-loaded channel estimation pilot scheme in NR, and the constraints of MU-MIMO scheduling. We also report that joint multi-user beamforming design is imperative, in the sense that the MU-MIMO system cannot be fully exploited when implemented as a mere collection of single-user analog beams working in parallel. By addressing these issues, throughput can be dramatically increased in mmWave 5G networks by means of Spatial Division Multiple Access (SDMA).
{"title":"Full-stack Hybrid Beamforming in mmWave 5G Networks","authors":"Felipe Gómez-Cuba, Tommaso Zugno, Junseok Kim, Michele Polese, S. Bahk, M. Zorzi","doi":"10.1109/MedComNet52149.2021.9501238","DOIUrl":"https://doi.org/10.1109/MedComNet52149.2021.9501238","url":null,"abstract":"This paper analyzes Hybrid Beamforming (HBF) and Multi-User Multiple-Input Multiple-Output (MU-MIMO) in millimeter wave (mmWave) 5th generation (5G) cellular networks considering the full protocol stack with TCP/IP traffic and MAC scheduling. Prior work on HBF and MU-MIMO has assumed full-buffer transmissions and studied link-level performance. We report non-trivial interactions between the HBF technique, the front-loaded channel estimation pilot scheme in NR, and the constraints of MU-MIMO scheduling. We also report that joint multi-user beamforming design is imperative, in the sense that the MU-MIMO system cannot be fully exploited when implemented as a mere collection of single-user analog beams working in parallel. By addressing these issues, throughput can be dramatically increased in mmWave 5G networks by means of Spatial Division Multiple Access (SDMA).","PeriodicalId":272937,"journal":{"name":"2021 19th Mediterranean Communication and Computer Networking Conference (MedComNet)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126398804","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：