Improving Video-on-Demand Performance by Multi-channel/ Multicast Approach in Cellular Networks Reza Shokri Kalan International Computer Institute, Ege University, Izmir, Turkey
[email protected] Abstract- On-demand-video streaming has been applied in wired networks with success. However, it remains a challenging task in wireless networks due to bandwidth problems and the effect of wireless interference and client mobility. To address these problems, we unified cache server with multi-channel/multicast technique and improves speed up by limiting the number of video request from the remote server. The proposed solution is an efficient when more clients request the same video at the same or closer time. Also, video locality based on cache server, which located in the base station controller reduce core network traffics by elimination number of request to remote media services. The simulation results show that the suggested methods improve response time and speedup that lead to better quality of service. Keywords-Video-on-demand, Caching
I.
multicast,
Frequency
band,
INTRODUCTION
In recent years broadcasting video over wireless networks has become more popular. Video-on-demand (VoD) is a multimedia system which enables clients to access and play video from the remote media server through wired or wireless networks. Wireless networks enable users to request and display the video anytime anywhere. The VoD systems can be categorized into two main approaches: (i) client/server and (ii) distributed or Peer-to-Peer (P2P) [1, 2]. In the client/server technique, the video files are stored on multiple servers. Each client independently connects to the server and plays the requested video from the server. Although the client/server model is the simplest, by increasing number of clients, bottleneck (or I/O) problems reduce server performance. Peerto-Peer as a decentralized and multi point communication technique reduces server loads and tries to avoid bottleneck problems on the server side by saving bandwidth. Content delivery latency as a part of the quality of service (QoS) is an important issue in real time communication. A caching system on the client side improves data access latency. In this manner clients can download video from suitable neighbors. Due to limited bandwidth, each client can forward cached segment to one client at the same time. The client switches to active mode during cache forwarding and stays in this mode until cache forwarding is finished. Although recently mobile devices come with more memory size which can be used for caching media streams, channel allocation is still a big problem. Multicast as a stream sharing mechanism can
improve the frequency spectrum utilization of wireless communication in multipoint services. The simplest approach to stream media data to a client is to create a unicast connection from the server to each requesting user, where the media server is situated behind the Base Station (BS). In this case, we need individual free channels for each user’s communication. The communication between Mobile Station (MS) and BS is started if there is a free channel, otherwise MS must be waiting until channel allocation. Unfortunately, this approach greedy and will quickly consume all available bandwidth of the downlink and lead to bottleneck problems in core network and server side when multiple users concurrently request the media services. The aim of our approach is minimizing individual unicast connections between client and media server. In summary, our contribution in this paper organized as follows: We begin the paper with related works. In Section 3, we focus on scheduled multicast and patching technique. Section 4, includes the basic idea of VoD in cellular networks. Architecture of our proposed system discussed in section 5. Section 6, explain the simulation results. Finally, we draw to a conclusion. II.
RELATED WORKS
Providing on-demand services to a large number of clients in ‘‘real-time’’ imposes a high resource requirement on the underlying network and server. Therefore, traditional client/server architecture cannot provide a scalable solution, as it requires a dedicated communication channel per user. VoD servers can be arranged as a centralized, distributed, and a proxy or content delivery network (CDN) architecture. There is a good background of multimedia services in wired networks [3, 4, 5, 6], however less research reported on multimedia services in wireless networks. In Cache-and-Relay approach, peers store downloaded video in their memory, ready to relay the stored video to other peers in future, leading to an asynchronous P2P video sharing [4]. Proxy caching improves startup latency and network traffic, but it is costly. VoD eliminates the need for proxy server or a centralized coordinator, and it improves performance by both multithreading and peer-to-peer cooperation based on the locality principle.
