Streaming of Scalable Multimedia over Content Delivery Cloud Xiaoming Bao, Rongshan Yu, Institute for Infocomm Research, A*STAR, Singapore Email: {baoxm, ryu}@i2r.a-star.edu.sg Abstract—Content Delivery Cloud (CDC) extends Content Delivery Network (CDN) to provide elastic, scalable and low cost services to the customers. For multimedia streaming over CDC, caching the media content onto the edge server from storage cloud is commonly used to minimize the latency of content delivery. It is very important for CDN to balance between the resources being used (storage space, bandwidth, etc) and the performance achieved. Commercial CDNs (such as Akamai, Limelight, Amazon CloudFront) have their proprietary caching algorithms to deal with this issue. In this paper, we propose a method to further improve the efficiency of the caching system for scalable multimedia contents. Specifically, we notice that a scalable multimedia content can be flexibly truncated to lower bit rates on-the-fly based on the available network bandwidth between the edge server to the end users. Therefore, it may not be necessary to cache such a content at its highest quality/rate. Based on this observation, we show that edge server can decide an optimized truncation ratio for the cached scalable multimedia contents to balance between the quality of the media and the resource usage. The proposed optimized truncation algorithm is analyzed and its efficacy in improving the efficiency of the caching system is justified with simulation result. Index Terms—Content delivery network, cloud storage, multi- media streaming, caching algorithm. I. I NTRODUCTION Content Delivery Network (CDN) played an essential role in assisting content providers to deliver multimedia contents to end users efficiently. The objective and internal mechanism of CDN are well explored by [1], [2]. Recently, the cloud computing technology [3] has been widely used in CDN to provide elastic, scalable and low cost services to the customers, which extends the traditional CDN model to so called Content Delivery Cloud (CDC) [4]. Typically, a CDN service cloud includes a number of basic components such as storage, parallel computing engine, controller and user front (Edge/Proxy server). The internal of the cloud architecture is transparent to the users. The globally deployed edge servers respond to the nearby user requests for the contents whose physical locations are totally unaware to the users. The controllers are responsible for retrieving the content from the respective storage cloud and transferring it to the edge server with guaranteed Quality-of-Service (QoS). For cloud-based multimedia computing, including multi- media streaming, QoS requirements in terms of bandwidth, delay and jitter are key factors in designing a multimedia system [5]. Numeric efforts have been done through high level architectural design [5], [6] or Distributed File System (DFS) specific file perfecting [7] to satisfy the QoS requirements for media content delivery service. Practically, caching the media content from storage cloud onto the edge server is a common technique being widely used in CDC to effectively reduce the latency of media content delivery. In this paper, we propose an optimized caching algorithm for streaming scalable encoded multimedia such as MPEG-4 SLS [8] over CDC. The scalable multimedia is encoded in a way such that a low bit rate frame data can be generated by truncating the data from a higher bit rate frame [9]. Therefore, one media content can be encoded into a single source with highest bit rate, while in streaming application the multimedia file can be further truncated on the fly to lower bit rate before sent to end users. The actual bit rate is thus determined by a truncation ratio λ depending on the available network bandwidth from the edge server to the users. We take advantage of this scalable feature to further improve the efficiency of existing caching algorithms in CDC, where the quality of the scalable multimedia file being cached in the edge server is determined based on the network conditions from the edge server to the users. In the proposed algorithm, both the utility function of scalable media as a function of bit-rate and the cost of transmitting and storing the media files from storage cloud onto the edge server are considred to achieve the best balance between the service quality and its associated cost. In this paper, the optimal truncation ratio for the scalable media is formulated as stochastic optimization problem and a solution to this problem is given followed by numerical analysis to the solution. The efficiency of the proposed algorithm is justified with simulation result. The rest of this paper is organized as follows. The for- mulation of the problem of optimal initial truncation ratio determination is given in Section II and a solution based on stochastic optimization is given. The proposed solution are further analyzed in Section III where some interesting properties of the proposed solution are given. Numerical simulation results are given in Section IV. Finally, this paper is concluded in Section V. II. PROBLEM FORMULATION AND SOLUTION We assume that all the scalable multimedia files are stored in storage cloud initially. Upon receiving request by the first end user, a multimedia file will be fetched from the storage cloud to the edge server, and streamed to the end user by the edge server. The multimedia file will be stored in the cache