SURVEY PAPER

You are expected to write a survey paper on a recent research topic related to this class. This will constitute 25% of your final grade. A survey paper, as described by ACM Computing Surveys, is "a paper that summarizes and organizes recent research results in a novel way that integrates and add understanding to work in the field. A survey article assumes a general knowledge of the area; it emphasizes the classification of the existing literature, developing a perspective on the area, and evaluating trends." For the purpose of this class, "general knowledge of the area" includes the topics we cover in class. You may ignore "evaluating trends".

You are expected to read at least 4 papers on one of the selected topics (see below), and produce a survey on the topic. The survey paper must be 6-8 pages in length, using the LaTeX/Word template provided. Ideally, it is of publication quality (you may pretend you are submitting it to ACM Computing Surveys). It should not be just a concatenation of paper reviews. You will be graded based on your writing, presentation, and how it enhance the understanding of the research topic.

You may select a research topic related (or the same as) your research project. Be careful when you choose the papers to read. You would want to read good papers solving interesting aspect of the research problems. Citeseer citation listings may be helpful in deciding which paper is important and which are not (This rule of thumb does not apply to recent papers). You may also want to restrict your readings to papers published in rank 1 and rank 2 journals and conferences (ranking list available here).

Due Date

The survey paper is due on 11th Oct.

Review Format

For survey, you should follow the ACM proceeding style guidelines. Please use the given LaTeX (recommended) or Word template so that to ensure that there are little variations in margin and font size.

• ACM Proceedings Template

Useful Links

• ACM Computing Surveys

Examples

• X. Wang, H. Schulzrinne, "Comparison of adaptive Internet multimedia applications," Accepted to The IEICE Transactions on Communications.
• Yang Richard Yang and Simon S. Lam. Internet Multicast Congestion Control: A Survey. In Proceedings of ICT 2000, Acapulco, Mexico, May 2000
• S. W. Carter, D. D. E Long and J.-F. Paris, Video-on-Demand Broadcasting Protocols, In Multimedia Communications: Directions and Innovations (J. D. Gibson, Ed.), Academic Press, San Diego, 2000, pages 179--189

Topics

You may propose other survey topics that are of interest to you. Citations in the description are there to help you get started with your survey. You should not limit your survey to the work cited below.

• Single Source Multicast (SSM) Many media applications do not require many-to-many communications supported by IP multicast. Applications such as webcast or VoD only requires 1-to-many communications. SSM schemes such as EXPRESS and Simple Multicast support 1-to-many communications in a simpler and more efficient way, compare to IP Multicast.
• Streaming with TCP Several recent work by Buck Krasic (OGI), Sam Liang (Stanford) etc. showed that TCP can be used for streaming as well, in constrast to the traditional view that UDP should be used for streaming media applications.
• On-line Smoothing Algorithms Do a survey on on-line smoothing algorithms for live video (I will cover off-line smooth for pre-recorded video if time permits).
• Modeling MPEG Traffic Researchers often need to generate VBR video traffic to evaluate a new system or a new protocol. Since real video database is not always available, a model for such type of traffic is used instead. Focus your survey on modelling of MPEG video instead of general VBR traffic.
• Characterizing Media Server Workload Understanding the behavior of media server workload can help researchers in designing realisitic systems. There are several recent work (Veloso, Almeida, Chesire etc.) in this area.
• Predicting MPEG Execution Time Predicting the time taken to decode an MPEG frame is important for various scheduling algorithms. For example, on a battery-powered device, we want to be able to predict how much power will be consumed when a particular frame is decoded.
• Power-aware Video Decoding Video decoding is processor intensive. As video playback becoming popular on battery-powered device (cellphone, PDA), decoding video while efficiently consume power becomes an important issue.
• Adaptive Media Playout A receiver typically buffers some data before playing it back to absorb jitter. Adaptive media playout allows the client to adjust its playback rate (increase video frame-rate, time-compress audio) to reduce buffer size and latency.
• Path Diversity Streaming media packets normally follows a single path in the network. Path diversity allows a stream to follow multiple paths from a source to a destination to reduce effects of network variability such as bursty error.
• Peer-to-peer Streaming In one-to-many communication, a receiver might help the sender by relaying the received streams to a peer (i.e., another receiver)
• Many-to-one Streaming Instead of downloading stream from a server, a client might download from multiple servers and combined the data from different servers to form a stream. (Note that even though P2P streaming can be considered M-to-1 streaming as well, here we consider reliable servers as sources here, not transient peers.)
• Other Possible Topics: Authentication of Streams, Streaming over Wireless Network, Transcoding Proxies, etc etc. See papers in related conferences/journals for more ideas.