Load Distribution Strategies in Cluster-based Transcoding Servers for Mobile Clients

Dongmahn Seo, Joahyoung Lee, Yoon Kim, Changyeol Choi, Hwangkyu Choi, Inbum Jung, “Load Distribution Strategies in Cluster-based Transcoding Servers for Mobile Clients,” ICCSA 2006, LNCS 3983, pp. 1156.1165, 2006.

Load Distribution Strategies in Cluster-Based
Transcoding Servers for Mobile Clients
Dongmahn Seo, Joahyoung Lee, Yoon Kim, Changyeol Choi,
Hwangkyu Choi, and Inbum Jung
Department of Computer Science and Engineering,
Kangwon National University, Chuncheon, 200-701, Korea
{sarum, jinnie4u, yooni, cychoi, hkchoi, ibjung}@kangwon.ac.kr
Abstract. The recent advance in wireless network technologies has enabled
the streaming media service on the mobile devices such as PDAs
and cellular phones. Since the wireless network has low bandwidth channels
and mobile devices are actually composed of limited hardware specifications,
the transcoding technology is needed to adapt streaming media
to the given mobile devices. When large scale mobile clients demand the
streaming service, load distribution strategies among transcoding servers
highly impact on the total number of QoS streams. In this paper, the
resource weighted load distribution strategy is proposed for the fair load
balancing and the more scalable performance in cluster-based transcoding
servers. Our proposed strategy is based on the weight of resources
consumed for transcoding to classified client grades and the maximum
number of QoS streams actually measured in transcoding servers. The
proposed policy is implemented on cluster-based transcoding system. In
experiments, we evaluate its fair load distribution and scalable performance
according to the increase of transcoding servers.