Title: Proactive Video Content Caching
Company: BI2S Ltd.
Where: Nicosia (Cyprus)
When (time-plan): January to June 2020 (6 months)
ICON supports the localized, personalized and timely provision of media resources to users through the adoption of the FLAME’s platform. End users can search and consume content based on their preferences learned by the provided ICON application. At the same, ICON supports the concept of content prosumers, i.e., end users can become content producers, apart from being consumers, having the opportunity to upload media files into the available infrastructure, thus, setting up the basis for novel business models. ICON supports an ‘ecosystem’ of video content objects applied in any domain.
ICON builds on top of the content provided through the FLAME platform. The content is transferred by invoking the appropriate FLAME’s services and is multicasted to various locations in the network. A web application is adopted running in browser in users’ devices and undertakes the responsibility of maintaining the video database while communicating with the available edge access points (i.e., FLAME’s clusters per access point). Content offloading and caching policies are defined while following ML paradigms that evolve dynamically as users makes content requests. In edge access points, information of searches are locally stored becoming the basis for delivering intelligent analytics. A machine learning model is responsible to estimate the future demand, hence, to make the platform able to be adapted to it by employing the appropriate caching strategy. At the back end, a Machine Learning model is responsible to proactively offload video content while providing analytics in multiple dimensions. The experimentation involves the study of a set of KPIs to reveal the performance of both: the FLAME’s platform and the ICON’s content management.
Figure below presents the envisioned generic plan and ICON’s objectives. With the FLAME’s platform, the content is delivered to end users through multicasting according to their personal needs. The multicasting is performed to groups of users requesting the same video files at the same time as the FLAME’s Opportunistic Multicast service design pattern dictates.
ICON employs machine learning algorithms to offload video content to the EDGE before the user request takes place and utilizes content caching strategies to adapt to the current network conditions. Specifically pre-trained feed forward and deep neural network models are deployed that utilize video characteristics such as popularity and size and also the location of the uploaded video in order to choose whether the content should be offloaded to the EDGE. In order to decide the amount of time a video should stay stored at the EDGE the K-means machine learning algorithm is employed that dynamically adapts the caching strategy according to video sizes and the available storage resources at the EDGE.
ICON has conducted experimentation and performance analysis of the employed ML algorithms as well as of the video distribution platform. The trials executed at the FLAME infrastructure at the city of Barcelona have proven the efficiency of the proposed solution.