Practical, Real-time Centralized Control for CDN-based Live Video Delivery

Mukerjee, Matthew K.; Naylor, David; Jiang, Junchen; Han, Dongsu; Seshan, Srinivasan; Zhang, Hui

doi:10.1145/2829988.2787475

Cited by 51 publications

(36 citation statements)

References 31 publications

(42 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The root controller periodically creates a general model of the streaming system, which is then used by the children controllers (on a coarser time-scale) to take decisions on the best QoE strategy for the video clients. A similar approach is also employed by Mukerjee et al to optimize the performance of live video streams [13]. The use of an SDN controller to optimize the behavior of HAS clients has been studied by Egilmez et al [14].…”

Section: Related Workmentioning

confidence: 99%

A machine learning-based framework for preventing video freezes in HTTP adaptive streaming

Petrangeli

Wauters

et al. 2017

Journal of Network and Computer Applications

View full text Add to dashboard Cite

HTTP Adaptive Streaming (HAS) represents the dominant technology to deliver videos over the Internet, due to its ability to adapt the video quality to the available bandwidth. Despite that, HAS clients can still suffer from freezes in the video playout, the main factor influencing users' Quality of Experience (QoE). To reduce video freezes, we propose a network-based framework, where a network controller prioritizes the delivery of particular video segments to prevent freezes at the clients. This framework is based on OpenFlow, a widely adopted protocol to implement the SoftwareDefined Networking principle. The main element of the controller is a Machine Learning (ML) engine based on the Random Undersampling Boosting algorithm and fuzzy logic, which can detect when a client is close to a freeze and drive the network prioritization to avoid it. This decision is based on measurements collected from the network nodes only, without any knowledge on the streamed video or on the client's characteristics.In this paper, we detail the design of the proposed ML-based framework and compare its performance with other benchmarking HAS solutions, under various video streaming scenarios. Particularly, we show through extensive experimentation that the proposed approach can reduce video freezes and freeze time with about 65% and 45% respectively, when compared to benchmarking algorithms. These results represent a major improvement for the QoE of the users watching multimedia content online.

show abstract

Section: Related Workmentioning

confidence: 99%

A machine learning-based framework for preventing video freezes in HTTP adaptive streaming

Petrangeli

Wauters

et al. 2017

Journal of Network and Computer Applications

View full text Add to dashboard Cite

show abstract

“…Constraints (18)- (19) deal with the capacity of the links between leaf cache nodes and metro areas. Constraint (18) computes the amount of connections supporting each link and constraint (19) assures that the requested bitrate does not exceeds a given maximum.…”

Section: B Cdn User Reallocation (Cdn_usher)mentioning

confidence: 99%

“…Analyzing video sessions, authors in [17] concluded that a centralized controller could improve user experience, while authors in [18] introduced presented a centralized algorithm for live video optimization providing real-time, fine-grained control. In addition, placing new cache nodes to accommodate spikes in demand and consolidate workload in few cache nodes when the load decreases can also bring benefits.…”

mentioning

confidence: 99%

Big Data-backed video distribution in the telecom cloud

Ruiz

Germán

Contreras

et al. 2016

Computer Communications

View full text Add to dashboard Cite

Telecom operators are starting the deployment of Content Delivery Networks (CDN) to better control and manage video contents injected into the network. Cache nodes placed close to end users can manage contents and adapt them to users' devices, while reducing video traffic in the core. By adopting the standardized MPEG-DASH technique, video contents can be delivered over HTTP. Thus, HTTP servers can be used to serve contents, while packagers running as software can prepare live contents. This paves the way for virtualizing the CDN function. In this paper, a CDN manager is proposed to adapt the virtualized CDN function to current and future demand. A Big Data architecture, fulfilling the ETSI NFV guide lines, allows controlling virtualized components while collecting and pre-processing data. Optimization problems minimize CDN costs while ensuring the highest quality. Re-optimization is triggered based on threshold violations; data stream mining sketches transform collected into modeled data and statistical linear regression and machine learning techniques are proposed to produce estimation of future scenarios. Exhaustive simulation over a realistic scenario reveals remarkable costs reduction by dynamically reconfiguring the CDN.Peer ReviewedPostprint (author's final draft

show abstract

“…Finally, while this paper focuses on a VoD service, proactive content placement has recently been investigated in the context of live video streaming (Mukerjee et al, 2015), which focuses on the end-to-end optimization of the stream delivery. In contrast to our approach, this targets very short re-configuration intervals (in the order of milliseconds) and involves changing different parameters (e.g., video stream placement and bitrate encoding selection), which are features specific to this type of service.…”

Section: Related Workmentioning

confidence: 99%

Hybrid multi-tenant cache management for virtualized ISP networks

Claeys

Tuncer

Famaey

et al. 2016

Journal of Network and Computer Applications

View full text Add to dashboard Cite

In recent years, Internet Service Providers (ISPs) have started to deploy Telco Content Delivery Networks (Telco CDNs) to reduce the pressure on their network resources. The deployment of Telco CDNs results in reduced ISP bandwidth utilization and improved service quality by bringing the content closer to the end-users. Furthermore, virtualization of storage and networking resources can open up new business models by enabling the ISP to simultaneously lease its Telco CDN infrastructure to multiple third parties. Previous work has shown that multi-tenant proactive resource allocation and content placement can significantly reduce the load on the ISP network. However, the performance of this approach strongly depends on the prediction accuracy for future content requests. In this paper, a hybrid cache management approach is proposed where proactive content placement and traditional reactive caching strategies are combined. In this way, content placement and server selection can be optimized across tenants and users, based on predicted content popularity and the geographical distribution of requests, while simultaneously providing reactivity to unexpected changes in the request pattern. Based on a Video-on-Demand (VoD) production request trace, it is shown that the total hit ratio can be increased by 43% while using 5% less bandwidth compared to the traditional Least Recently Used (LRU) caching strategy. Furthermore, the proposed approach requires 39% less migration overhead compared to the proactive placement approach we previously proposed in Claeys et al. (2014b) and achieves a hit ratio increase of 19% and bandwidth usage reduction of 7% in the evaluated VoD scenarios and topology.

show abstract

Practical, Real-time Centralized Control for CDN-based Live Video Delivery

Cited by 51 publications

References 31 publications

A machine learning-based framework for preventing video freezes in HTTP adaptive streaming

A machine learning-based framework for preventing video freezes in HTTP adaptive streaming

Big Data-backed video distribution in the telecom cloud

Hybrid multi-tenant cache management for virtualized ISP networks

Contact Info

Product

Resources

About