Pub Date : 2020-01-27DOI: 10.1109/6GSUMMIT49458.2020.9083785
Idris Badmus, Abdelquoddouss Laghrissi, Marja Matinmikko-Blue, A. Pouttu
The concept of a softwarized network leveraging technologies such as SDN/NFV, comes with different merits such as decreased Operational Expenses (OPEX) and less dependency on underlying hardware components. With the amount of increased flexibility, reconfigurability and programmability attributed to future technologies (i.e., 5G and beyond), and towards the complete network virtualization and softwarization, a new set of requirements/parameters can be identified affecting the latency in a virtualized network. In this paper, we identify different latency requirements for a virtualized network. These requirements include the Virtual Network Function (VNF) deployment time, establishment/connection time and application instantiation time. We further test how some factors such as VNFs' resource usage, the applications running within the VNF and the shared status of the VNF, coordinately affect the identified latency requirement for a virtualized network. Experimentally, for performance analysis, we deploy a softwarized network based on the ETSI-NFV architecture, using open source tools. The results show that the new set of latency requirements is relevant for consideration in order to achieve an overall ultra-reliable low latency and how different the factors can affect these new requirements, especially in the core network. Furthermore, the result of our performance analysis proves the trade-off between latency of a virtualized network and the resource usage of the VNFs.
{"title":"Identifying Requirements Affecting Latency in a Softwarized Network for Future 5G and Beyond","authors":"Idris Badmus, Abdelquoddouss Laghrissi, Marja Matinmikko-Blue, A. Pouttu","doi":"10.1109/6GSUMMIT49458.2020.9083785","DOIUrl":"https://doi.org/10.1109/6GSUMMIT49458.2020.9083785","url":null,"abstract":"The concept of a softwarized network leveraging technologies such as SDN/NFV, comes with different merits such as decreased Operational Expenses (OPEX) and less dependency on underlying hardware components. With the amount of increased flexibility, reconfigurability and programmability attributed to future technologies (i.e., 5G and beyond), and towards the complete network virtualization and softwarization, a new set of requirements/parameters can be identified affecting the latency in a virtualized network. In this paper, we identify different latency requirements for a virtualized network. These requirements include the Virtual Network Function (VNF) deployment time, establishment/connection time and application instantiation time. We further test how some factors such as VNFs' resource usage, the applications running within the VNF and the shared status of the VNF, coordinately affect the identified latency requirement for a virtualized network. Experimentally, for performance analysis, we deploy a softwarized network based on the ETSI-NFV architecture, using open source tools. The results show that the new set of latency requirements is relevant for consideration in order to achieve an overall ultra-reliable low latency and how different the factors can affect these new requirements, especially in the core network. Furthermore, the result of our performance analysis proves the trade-off between latency of a virtualized network and the resource usage of the VNFs.","PeriodicalId":385212,"journal":{"name":"2020 2nd 6G Wireless Summit (6G SUMMIT)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114253672","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 : 2020-01-26DOI: 10.1109/6GSUMMIT49458.2020.9083767
Benjamin Sliwa, Robert Falkenberg, C. Wietfeld
Machine learning-based data rate prediction is one of the key drivers for anticipatory mobile networking with applications such as dynamic Radio Access Technology (RAT) selection, opportunistic data transfer, and predictive caching. User Equipment (UE)-based prediction approaches that rely on passive measurements of network quality indicators have successfully been applied to forecast the throughput of vehicular data transmissions. However, the achievable prediction accuracy is limited as the UE is unaware of the current network load. To overcome this issue, we propose a cooperative data rate prediction approach which brings together knowledge from the client and network domains. In a real world proof-of-concept evaluation, we utilize the Software Defined Radio (SDR)-based control channel sniffer FALCON in order to mimic the behavior of a possible network-assisted information provisioning within future 6G networks. The results show that the proposed cooperative prediction approach is able to reduce the average prediction error by up to 30%. With respect to the ongoing standardization efforts regarding the implementation of intelligence for network management, we argue that future 6G networks should go beyond network-focused approaches and actively provide load information to the UEs in order to fuel pervasive machine learning and catalyze UE-based network optimization techniques.
{"title":"Towards Cooperative Data Rate Prediction for Future Mobile and Vehicular 6G Networks","authors":"Benjamin Sliwa, Robert Falkenberg, C. Wietfeld","doi":"10.1109/6GSUMMIT49458.2020.9083767","DOIUrl":"https://doi.org/10.1109/6GSUMMIT49458.2020.9083767","url":null,"abstract":"Machine learning-based data rate prediction is one of the key drivers for anticipatory mobile networking with applications such as dynamic Radio Access Technology (RAT) selection, opportunistic data transfer, and predictive caching. User Equipment (UE)-based prediction approaches that rely on passive measurements of network quality indicators have successfully been applied to forecast the throughput of vehicular data transmissions. However, the achievable prediction accuracy is limited as the UE is unaware of the current network load. To overcome this issue, we propose a cooperative data rate prediction approach which brings together knowledge from the client and network domains. In a real world proof-of-concept evaluation, we utilize the Software Defined Radio (SDR)-based control channel sniffer FALCON in order to mimic the behavior of a possible network-assisted information provisioning within future 6G networks. The results show that the proposed cooperative prediction approach is able to reduce the average prediction error by up to 30%. With respect to the ongoing standardization efforts regarding the implementation of intelligence for network management, we argue that future 6G networks should go beyond network-focused approaches and actively provide load information to the UEs in order to fuel pervasive machine learning and catalyze UE-based network optimization techniques.","PeriodicalId":385212,"journal":{"name":"2020 2nd 6G Wireless Summit (6G SUMMIT)","volume":"126 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122489889","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 : 2020-01-25DOI: 10.1109/6GSUMMIT49458.2020.9083751
Juuso Haavisto, J. Riekki
Mixed reality (MR) applications are expected to become common when 5G goes mainstream. However, the latency requirements are challenging to meet due to the resources required by video-based remoting of graphics, that is, decoding video codecs. We propose an approach towards tackling this challenge: a client-server implementation for transacting intermediate representation (IR) between a mobile UE and a MEC server instead of video codecs and this way avoiding video decoding. We demonstrate the ability to address latency bottlenecks on edge computing workloads that transact graphics. We select SPIR-V compatible GPU kernels as the intermediate representation. Our approach requires know-how in GPU architecture and GPU domain-specific languages (DSLs), but compared to video-based edge graphics, it decreases UE device delay by sevenfold. Further, we find that due to low cold-start times on both UEs and MEC servers, application migration can happen in milliseconds. We imply that graphics-based location-aware applications, such as MR, can benefit from this kind of approach.
{"title":"Interoperable GPU Kernels as Latency Improver for MEC","authors":"Juuso Haavisto, J. Riekki","doi":"10.1109/6GSUMMIT49458.2020.9083751","DOIUrl":"https://doi.org/10.1109/6GSUMMIT49458.2020.9083751","url":null,"abstract":"Mixed reality (MR) applications are expected to become common when 5G goes mainstream. However, the latency requirements are challenging to meet due to the resources required by video-based remoting of graphics, that is, decoding video codecs. We propose an approach towards tackling this challenge: a client-server implementation for transacting intermediate representation (IR) between a mobile UE and a MEC server instead of video codecs and this way avoiding video decoding. We demonstrate the ability to address latency bottlenecks on edge computing workloads that transact graphics. We select SPIR-V compatible GPU kernels as the intermediate representation. Our approach requires know-how in GPU architecture and GPU domain-specific languages (DSLs), but compared to video-based edge graphics, it decreases UE device delay by sevenfold. Further, we find that due to low cold-start times on both UEs and MEC servers, application migration can happen in milliseconds. We imply that graphics-based location-aware applications, such as MR, can benefit from this kind of approach.","PeriodicalId":385212,"journal":{"name":"2020 2nd 6G Wireless Summit (6G SUMMIT)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130889523","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 : 2020-01-07DOI: 10.1109/6GSUMMIT49458.2020.9083828
O. Z. Alsulami, A. Alahmadi, Sarah O. M. Saeed, S. Mohamed, T. El-Gorashi, M. Alresheedi, J. Elmirghani
Optical wireless communication (OWC) systems are a promising communication technology that can provide high data rates into the tens of Tb/s and can support multiple users at the same time. This paper investigates the optimum allocation of resources in wavelength division multiple access (WDMA) OWC systems to support multiple users. A mixed-integer linear programming (MILP) model is developed to optimise the resource allocation. Two types of receivers are examined, an angle diversity receiver (ADR) and an imaging receiver (ImR). The ImR can support high data rates up to 14 Gbps for each user with a higher signal to interference plus noise ratio (SINR). The ImR receiver provides a better result compared to the ADR in term of channel bandwidth, SINR and data rate. Given the highly directional nature of light, the space dimension can be exploited to enable the co-existence of multiple, spatially separated, links and thus aggregate data rates into the Tb/s. We have considered a visible light communication (VLC) setting with four wavelengths per access point (red, green, yellow and blue). In the infrared spectrum, commercial sources exist that can support up to 100 wavelengths, significantly increasing the system aggregate capacity. Other orthogonal domains can be exploited to lead to higher capacities in these future systems in 6G and beyond.
{"title":"Optimum Resource Allocation in 6G Optical Wireless Communication Systems","authors":"O. Z. Alsulami, A. Alahmadi, Sarah O. M. Saeed, S. Mohamed, T. El-Gorashi, M. Alresheedi, J. Elmirghani","doi":"10.1109/6GSUMMIT49458.2020.9083828","DOIUrl":"https://doi.org/10.1109/6GSUMMIT49458.2020.9083828","url":null,"abstract":"Optical wireless communication (OWC) systems are a promising communication technology that can provide high data rates into the tens of Tb/s and can support multiple users at the same time. This paper investigates the optimum allocation of resources in wavelength division multiple access (WDMA) OWC systems to support multiple users. A mixed-integer linear programming (MILP) model is developed to optimise the resource allocation. Two types of receivers are examined, an angle diversity receiver (ADR) and an imaging receiver (ImR). The ImR can support high data rates up to 14 Gbps for each user with a higher signal to interference plus noise ratio (SINR). The ImR receiver provides a better result compared to the ADR in term of channel bandwidth, SINR and data rate. Given the highly directional nature of light, the space dimension can be exploited to enable the co-existence of multiple, spatially separated, links and thus aggregate data rates into the Tb/s. We have considered a visible light communication (VLC) setting with four wavelengths per access point (red, green, yellow and blue). In the infrared spectrum, commercial sources exist that can support up to 100 wavelengths, significantly increasing the system aggregate capacity. Other orthogonal domains can be exploited to lead to higher capacities in these future systems in 6G and beyond.","PeriodicalId":385212,"journal":{"name":"2020 2nd 6G Wireless Summit (6G SUMMIT)","volume":"56 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133887714","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 : 2020-01-05DOI: 10.1109/6GSUMMIT49458.2020.9083856
O. Simeone, Sangwoo Park, Joonhyuk Kang
Machine learning methods adapt the parameters of a model, constrained to lie in a given model class, by using a fixed learning procedure based on data or active observations. Adaptation is done on a per-task basis, and retraining is needed when the system configuration changes. The resulting inefficiency in terms of data and training time requirements can be mitigated, if domain knowledge is available, by selecting a suitable model class and learning procedure, collectively known as inductive bias. However, it is generally difficult to encode prior knowledge into an inductive bias, particularly with black-box model classes such as neural networks. Meta-learning provides a way to automatize the selection of an inductive bias. Meta-learning leverages data or active observations from tasks that are expected to be related to future, and a priori unknown, tasks of interest. With a meta-trained inductive bias, training of a machine learning model can be potentially carried out with reduced training data and/or time complexity. This paper provides a high-level introduction to meta-learning with applications to communication systems.
{"title":"From Learning to Meta-Learning: Reduced Training Overhead and Complexity for Communication Systems","authors":"O. Simeone, Sangwoo Park, Joonhyuk Kang","doi":"10.1109/6GSUMMIT49458.2020.9083856","DOIUrl":"https://doi.org/10.1109/6GSUMMIT49458.2020.9083856","url":null,"abstract":"Machine learning methods adapt the parameters of a model, constrained to lie in a given model class, by using a fixed learning procedure based on data or active observations. Adaptation is done on a per-task basis, and retraining is needed when the system configuration changes. The resulting inefficiency in terms of data and training time requirements can be mitigated, if domain knowledge is available, by selecting a suitable model class and learning procedure, collectively known as inductive bias. However, it is generally difficult to encode prior knowledge into an inductive bias, particularly with black-box model classes such as neural networks. Meta-learning provides a way to automatize the selection of an inductive bias. Meta-learning leverages data or active observations from tasks that are expected to be related to future, and a priori unknown, tasks of interest. With a meta-trained inductive bias, training of a machine learning model can be potentially carried out with reduced training data and/or time complexity. This paper provides a high-level introduction to meta-learning with applications to communication systems.","PeriodicalId":385212,"journal":{"name":"2020 2nd 6G Wireless Summit (6G SUMMIT)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130706812","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 : 2019-12-20DOI: 10.1109/6GSUMMIT49458.2020.9083779
M. Salehi, Antti Tölli, S. P. Shariatpanahi
We study the joint effect of beamformer structure and subpacketization level on the achievable rate of cache-enabled multi-antenna communications. We use appropriate low-SNR approximations, to show that using simple zero-forcing (ZF) beamformers, increasing subpacketization degrades the achievable rate; in contrast to what has been shown in the literature for more complex, optimized beamformers. We also numerically analyze the probability distribution of symmetric rate terms, in order to confirm the validity of mathematical outputs. The results suggest that for improving the content delivery rate at low-SNR, subpacketization level and beamformer complexity should be jointly increased.
{"title":"Subpacketization - Beamformer Interaction in Multi-Antenna Coded Caching","authors":"M. Salehi, Antti Tölli, S. P. Shariatpanahi","doi":"10.1109/6GSUMMIT49458.2020.9083779","DOIUrl":"https://doi.org/10.1109/6GSUMMIT49458.2020.9083779","url":null,"abstract":"We study the joint effect of beamformer structure and subpacketization level on the achievable rate of cache-enabled multi-antenna communications. We use appropriate low-SNR approximations, to show that using simple zero-forcing (ZF) beamformers, increasing subpacketization degrades the achievable rate; in contrast to what has been shown in the literature for more complex, optimized beamformers. We also numerically analyze the probability distribution of symmetric rate terms, in order to confirm the validity of mathematical outputs. The results suggest that for improving the content delivery rate at low-SNR, subpacketization level and beamformer complexity should be jointly increased.","PeriodicalId":385212,"journal":{"name":"2020 2nd 6G Wireless Summit (6G SUMMIT)","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-12-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124074273","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：