Collaborative joint caching and transcoding in mobile edge networks

• Published in: Journal of network and computer applications Vol. 136; pp. 86 - 99
• Main Authors: Bilal, Kashif, Baccour, Emna, Erbad, Aiman, Mohamed, Amr, Guizani, Mohsen
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 15.06.2019
• Subjects: Adaptive bitrate video transcoding; Backhaul traffic; Collaborative caching and processing; Mobile Edge Computing; The X2 interface; The X2 interface; Collaborative caching and processing; Adaptive bitrate video transcoding; Backhaul traffic; Mobile Edge Computing
• ISSN: 1084-8045; 1095-8592
• DOI: 10.1016/j.jnca.2019.02.004

Video streaming has become a leading consumer of network resources in the last decade. Despite considerable developments, video content providers still face major challenges, which include minimizing data transfer from Content Delivery Network (CDN) or origin servers, CDN cost, and video startup delays. Recent edge computing technologies, such as Mobile Edge Computing (MEC) introduces new opportunities for Radio Access Networks (RANs) by providing computing and storage resources at the Mobile Base Stations (MBSs). Caching and processing videos at the edge networks relieve excessive data transfers over the backhaul links and minimize the viewers perceived delay. Collaborative caching and processing strategies have been proposed to efficiently utilize the edge resources, where neighboring MEC servers share the cached videos. However, such strategies introduce new challenges due to excessive backhaul links utilization for video sharing and limited resources. We propose a collaborative joint caching and processing strategy using the X2 network interface for sharing video data among multiple caches. Our design aims to minimize: (a) backhaul links usage for sharing video data, (b) network usage in transferring data from the CDN, (c) the viewer perceived delay, and (d) CDN cost. We also propose to fetch the higher bitrate version video from the origin/CDN servers and transcode it to the requested version on the fly to effectively use the Adaptive Bit Rate (ABR) streaming and online transcoding. This joint caching and processing approach is formulated as a minimization problem, subject to storage, processing, and bandwidth constraints. We also propose an online greedy algorithm that controls video transcoding, sharing using the X2 or backhaul links, and manages video caching and removing at the edge caches. Simulation results prove a better performance of our proposed algorithm compared to the recent edge caching approaches in terms of cost, average delay, cache removal, and cache hit ratio for different configurations.