Survey Article

Quality of Service (QoS) Issues in Multimedia Wireless Network (181-202)  
        Sulata Mitra and Mosa A. Abu-Rgheff
The advances in multimedia applications over a wide area network have directed considerable research into the quality of service. A comprehensive exposition of the specifications and management of quality of service (QoS) in wireless networks and in distributed computing systems, supporting multimedia applications, are important for both service providers and end users.   This article is a survey to explore issues concerning the quality of service in the current and future multimedia services in wireless networks.   The survey is done in two phases. In the first phase, a survey of various issues in third generation multimedia applications is presented. The second phase highlights research on QoS issues in the developing technology of the fourth generation wireless networks.

Research Articles

Mesh-based P2P Live Video Streaming  with StreamComplete (203-237)  
        Federico Covino and Massimo Mecella 
This paper introduces StreamComplete, a new architecture and prototype system for mesh-based P2P live video streaming; it realizes a new concept of overlay network's management as well as merges the best practices of tree- and mesh-based approaches. \StreamComplete creates a dynamic overlay network that optimizes itself on the basis of local conditions of peers. We have extensively evaluated the performance and the behavior of StreamComplete over the PlanetLab. Our experiments demonstrate the ability of peers in managing the overlay network with autonomic behavior in case of network changes. Furthermore the system is scalable w.r.t. the cardinality of the overlay network, by requiring very few control traffic.

Distributed Scheduling in a Time-Varying Ad Hoc Network (238-254)  
        Tina Heikkinen and Ari Hottinen 
This paper studies opportunistic distributed scheduling in an ad hoc wireless network, assuming partial orthogonality among multiple transmitters. The approach is based on game theory. A fair distributed scheduling scheme in a time-correlated channel is defined using a synchronous game for highly orthogonal transmitters and using an asynchronous game based on one-at-a-time transmission otherwise. Distributed game heuristics only require local node level information but still achieve a significant portion (at least 80 % in example cases) of the sum of rates obtained using coalitional uplink proportional fair scheduling for a wide range of orthogonality factors. An asynchronous game based on one-at-a-time transmission performs well relative to PFS for non-orthogonal transmitters. In addition to noncooperative game models, a cooperative game model for threshold-based scheduling is studied.

A Scheduling Method to Reduce Waiting Time for P2P Streaming Systems (255-270)  
        Yusuke Gotoh, Kentaro Susuki, Tomoki Yoshihisa, and Masanori Kanazawa
Recently, live streaming systems with peer-to-peer (P2P) technology have attracted much attention and are changing how we watch movies. In P2P streaming systems, a peer that plays the movie receives data from other peers. By sequentially playing the received data, users can watch the entire movie from beginning to the end. We previously proposed a method to reduce waiting time in P2P streaming. In conventional methods, by receiving the first chunk of data sequentially from a peer with large bandwidth and making a delivery schedule that considers the finishing time to delivery each content, waiting time is reduced effectively. However, these methods do not consider the case where heterogeneous peers deliver data to multiple peers. Also, since selected peers deliver the data to a peer using all the bandwidth, the number of available peers that deliver data decreases. In this paper, we propose a scheduling method to reduce the waiting time for selecting peers in P2P streaming by selecting peers and considering the available bandwidth. By designing and implementing P2P streaming systems, we consider situations in which our proposed system is effective. Our evaluation shows that our proposed method reduces the average waiting time 62.5\% more than conventional methods at maximum.

Back to JMM Online Front Page