Chapter one Introduction

Chapter One
Introduction
1.1 Motivation
Information in computer science might be represented in multiple forms. In the beginning the first used forms of information were text and images, after the huge increase of using computers and internet another type of information appears, it is the hyper text which is the combination of text and images. Hyper text is text pages that connected by hyper links (hyper link provide non linear access to the text pages no need to view all the pages sequentially to access a specific page) and then new forms are used like Audio and motion pictures or videos, nowadays a combination of these representations as well as the traditional forms (text and images) is called multimedia [LIZ04].
Multimediais a data type may contain multiple types of information like: audio, video, and animation in addition to traditional media (text, graphics, drawings and images), Digital multimedia is a data type represented by series of bits and manipulated by some multimedia applications and stored on some storage units like on CD or DVD or in multimedia servers.
Digital multimedia can be interactive (non linear access) that can be manipulated with computer programs (create, edit, transfer), or non interactive (linear access) where the viewer is unable to control the multimedia like the presentation of movie in the cinema [LIZ04].
Hypermedia is a collection of many types of information like text, images even videos in one website or more that are connected together by some hyper links that provide non linear access to the media like encyclopedia which user can jump from article to the other or jump to the next mark in DVD. The figure 1.1 below shows the structure of Hypermedia.

Figure 1.1 Hypermedia
Multimedia applications are a computer based programs that are designed to manipulate multimedia data with different sources like text, graphics, audio and video [NIE07].
Multimedia applications can be found in many areas such as:
Education, nowadays computer based training applications that designed to simulate the scientific experiments by creating the reactions environment and show the hypothetical results that should be produced from the row materials that will react in the reaction environment, these applications are safer for students from some experiments in real world as well they use some visual effects to explain some phenomenon like physical, chemical and geographic learning applications.
Advertising, multimedia applications can be found in advertising and journalism like the advertising of products and places like restaurants and companies, in journalism nowadays electronic copies of magazines and journals.
Entertainment, in this field multimedia is used to design video games and visual effects of games and movies, interactive multimedia is used heavily in video games that the user participates alone or with others on the same computer or over the internet not just sit and watch.
Industrial sector, multimedia is used as a way to help present information to shareholders, superiors and coworkers. Multimedia is also helpful for providing employee training, advertising and selling products all over the world via virtually unlimited web-based technology.
Mathematical and scientific research, multimedia is mainly used for modeling and simulation. For example, a scientist can look at a molecular model of a particular substance and manipulate it to arrive at a new substance. Representative research can be found in journals such as the Journal of Multimedia, in engineering multimedia applications now used to design buildings and machines before implementing them in real world [LIZ04].
Since multimedia represents some motion captured and stored on some remote place there must be some kind of delivery mechanism to transfer multimedia files from its source (producer) to the destination (consumer) [LIZ04].
Multimedia delivery can be divided into two categories offline and online delivery, offline delivery means transferring multimedia, without using internet, by using removable storage medium like CD-ROM to the hard disk on one station, multimedia was spread widely after the increase using of CD’s DVD’s and now Blue ray which are large storage media can hold the large size of multimedia files, online delivery (also called Streaming) uses the internet for transferring the multimedia from the producer which is usually a server with bulky storage to the consumer which is the client who can be any terminal mobile phone, personal computer or television and takes into consideration the statues of the network the client and server are connected to.
Online delivery transmits video files for events happened at the same time like video conferences, football games or video calls; or already processed and stored videos like video clips and movie trailers, online delivery multimedia and websites should be written in an efficient way to transmit or deliver the multimedia files and they are able to be read by the client, also there must be enough resources for the network and the client (which are available bandwidth, network throughput, congestion, and for the client the available buffer and client’s device speed, etc..) in order to transmit the multimedia file efficiently and the client has a successful experience [NIE07][LIZ04].
The adaptive streaming is the process of transmitting videos from the server to the client over internet in an efficient way depending on the network and client’s statues; to deliver videos faster and fewer pauses different qualities are saved to the same video and the server switches between these qualities according to the network and client’s statues.
In this thesis, a proposed protocol for streaming videos in adaptive way is introduced; the protocol has two sides the server or the sending parity which is responsible of transmitting the video files to the client in an adaptive way and the receiving party known as the client who receives the videos and displays them on his screen, adaptive means taking into consideration the statues of network like the available bandwidth in order to transmit the video file to the client smoothly, the processing of the video files is done on line, in other words the video frames are processed during the transmission according to the available bandwidth measured at the moment, dislike the used technologies nowadays which they save more than one copy of the video file with different bitrates and transmit the proper bitrate version of the video to the client according to the available bandwidth.

Literature Review

In this section reviews to some related work for this thesis about the dynamic video streaming over the internet

“A Buffer-Driven Approach to Adaptively Stream Stored Video over Internet”, by Dejian Ye [YED02], In dynamic video streaming technique many parameters can affect the streaming operation of video from a server to client one of these parameters is the buffer size of the sending and receiving parts, authors focused on the server and client buffer size and according to their statues server decide the quality of transmitted video, the buffer size and amount of available buffer size in the sender and receiver sides decide the quality of the transmitted video from the server to the client.
“ENet – Dynamic QoS Controller for Video Streaming Application”, by Qingyang Wang [WAN10], the popularity of video on demand and streaming over the internet increased and the limitations between the server and the client that may appear during streaming like supporting data type or bandwidth or computer speed and throughput and how to maintain a good QOS the authors design Enet which is a tool added to the server. Enet is a simple Fuzzy System used to switch between multiple bit rate versions of video, it picks the most appropriate bit rate to the client’s properties, this method takes a large storage space since it stores multiple versions of video with different bit rate for each video copy.
“Adaptation Algorithm for Adaptive Streaming over HTTP”, by Konstantin Miller [MIL12]. Internet videos (live or on demand) take a large amount of internet traffic, in this paper authors build an algorithm for the client side which tries to improve the delivery of video to the client by making client selects the appropriate bit rate depending on the client and network conditions like bandwidth available buffer and network throughput. They integrated the algorithm with a prototype implementation of a streaming client based on the MPEG DASH (Dynamic Adaptive Streaming over HTTP) standard which divide the video file into segments and each segment is saved in a number of different bit rates and client take the responsibility of selecting the most appropriate bit rate depending on the clients and network state. The design of the algorithm download the video file as segments one after the other and it uses one buffer that results client cannot receive the segment N unless it complete receiving segment N-1.
“Confused, Timid, and Unstable: Picking a Video Streaming Rate is Hard”, by Te-Yuan Huang [HUN12], Hulu, Vudu and Net Flux are three video streaming service providers that provide videos with high quality to the users with some fee. Videos provided by these services are divided to chunks with multiple bit rates for each chunk (when bit rate is slow chunk size is small and high bit rate the chunk size is big) and server switches between these chunks with different bit rates according to the available bandwidth. The problem is when requesting a video and start watching it and after some time another video requested a downward spiral effect appears which is bit rate goes down to the lowest level despite there is enough bandwidth for higher bit rate.

“USING SCALABLE VIDEO CODING FOR DYNAMIC ADAPTIVE TREAMING OVER HTTP IN MOBILE ENVIRONMENTS”, by Christopher Muller [MUL12], DASH is the approach that is used to transfer video over the internet in dynamic way to the user, it’s done by saving multiple copies of the video each copy with different bit rate and the server starts switching between these copies or bitrates according to client’s specifications. In this paper authors uses scalable video coding (SVC) instead of advanced video coding (AVC or H264) which provide more flexibility to the system with the available bandwidth and the buffer size, But if the client switches from one bit rate to the other (higher or lower) and he already requested a chunk before, this chunk will be canceled. The server will save multiple copies of video as a group of chunks. SVC experiment and code depend on windows so it’s not available to other operating systems.
“OPTIMIZING HTTP-BASED ADAPTIVE VIDEO STREAMING FOR WIRELESS ACCESS NETWORKS” by Xiaoling Qiu [Qui 13], Authors in this paper propose new approach for dynamic streaming videos over HTTP; their approach measure the available bandwidth of the network and client’s CPU speed and old measurements into account to select the best video quality using an optimization algorithm. This method contains optimization algorithm to select quality of segments to be sent to the client, but this method suffer from high computation rate.

Aim of thesis

The aim of this thesis is to build an efficient streaming protocol for transmitting video files from the server to the client with accepted waiting time at the client side.
The work in the thesis focuses on video files transmission to provide the accepted delivery time, videos usually are large size files this needs compression method to be used to reduce the video file size when the available bandwidth is below the required level for transmitting videos without compression, also this thesis takes into consideration the quality of the delivered video, the quality of the received video is maintained since the used compression technique in the proposed protocol is lossless technique so there is no loss in the transmitted video file quality at the client side.
Another important side in video streaming technique is the buffer management; the proposed streaming protocol uses a variable size for the buffers at the server side and the client side. This approach guarantees the efficient delivery of the required video file without take more than the required memory space.
The limitations that affect the proposed protocol execution are:

The execution on the internet is hard because it needs to reserve domain in a server over the internet, so the proposed protocol implemented on Local Area Network.
The bandwidth values are not real and they were generated randomly to simulate the change on the available bandwidth in real world but the limit of bandwidth was selected depending on used LAN defined bandwidth.

Outlines of the thesis

This thesis is organized in five chapters as follows:
Chapter one: “Introduction”
This chapter begins with explanation of the meaning of multimedia and its applications in real world and also contains a list of reviews works related to this work.
Chapter two: “Theoretical Background”
This chapter contains description of the main concepts of video file, video streaming, compression and quality.
Chapter Three: “Video Streaming Protocol”
This chapter contains the proposed streaming protocol between the server and the client that satisfy the aim of this thesis of delivering video from the server to the client with little or no pauses and with acceptable quality.
Chapter Four: “Experimental Results”
This chapter contains the results of the video streaming protocol introduced in the thesis.
Chapter Five “Conclusion and Future work”
This chapter contains the conclusions and the future work for the thesis.
1