MODULE 5: MULTIMEDIA NETWORKING
5.1 Multimedia Networking Applications
• A multimedia network application can be defined as any application that employs audio or video.
5.1.1 Properties of Video
1) High Bit Rate
• Video distributed over the Internet use
→ 100 kbps for low-quality video conferencing.
→ 3 Mbps for streaming high-definition (HD) movies.
• The higher the bit-rate,
→ better the image quality and
→ better the overall user viewing experience.
2) Video Compression
• A video can be compressed, thereby trading off video-quality with bit-rate.
• A video is a sequence of images, displayed at a constant rate.
For example: 24 or 30 images per second.
• An uncompressed digital image consists of an array of pixels.
• Each pixel is encoded into a number of bits to represent luminance and color.
• There are two types of redundancy in video:
1) Spatial Redundancy
➢ An image that consists of mostly white space has a high degree of redundancy.
➢ These images can be efficiently compressed without sacrificing image quality.
2) Temporal Redundancy
➢ Temporal redundancy reflects repetition from image to subsequent image.
➢ For exanple:
If image & subsequent image are same, re-encoding of subsequent image can be avoided.
5.1.2 Properties of Audio
• PCM (Pulse Code Modulation) is a technique used to change an analog signal to digital data
(digitization).
• PCM consists of 1) Encoder at the sender and 2) Decoder at the receiver.
[Link] PCM Encoder
• Digital audio has lower bandwidth requirements than video.
• Consider how analog audio is converted to a digital-signal:
• The analog audio-signal is sampled at some fixed rate. This operation is referred to as sampling.
• For example: 8000 samples per second.
• The value of each sample is an arbitrary real number.
• Each sample is then rounded to one of a finite number of values. This process is called quantization.
• The number of such finite values is called as quantization-values.
• The number of quantization-values is typically a power of 2. For ex: 256(28) quantization-values.
• Each of the quantization-values is represented by a fixed number of bits.
• For example:
If there are 256(28) quantization-values, then each value is represented by 8 bits.
• Bit representations of all values are then concatenated to form digital representation of the signal. This
process is called encoding.
• For example:
If an analog-signal is sampled at 8000 samples per second & each sample is represented by 8 bits, then
the digital-signal will have a rate of 64000 bits per second (8000*8=64000).
[Link] PCM Decoder
• For playback through audio speakers, the digital-signal can be converted back to an analog-signal. This
process is called decoding.
• However, the decoded analog-signal is only an approximation of the original signal.
• The sound quality may be noticeably degraded.
• The decoded signal can better approximate the original analog-signal by increasing
i) sampling rate and
ii) number of quantization-values,
• Thus, there is a trade-off between
→ quality of the decoded signal and
→ bit-rate & storage requirements of the digital-signal.
5.1.3 Types of Multimedia Network Applications
• Three broad categories of multimedia applications:
1) Streaming stored audio/video
2) Conversational voice/video-over-IP and
3) Streaming live audio/video.
[Link] Streaming Stored Audio & Video
• The underlying medium is prerecorded video. For example: a movie.
• These prerecorded videos are placed on servers.
• The users send requests to the servers to view the videos on-demand.
• Nowadays, many Internet companies provide streaming video. For example: YouTube.
• Three key distinguishing features of streaming stored video:
1) Streaming
• The client begins video playout within few seconds after it begins receiving the video from the server.
• At the same time,
i) The client will be playing out from one location in the video.
ii) The client will be receiving later parts of the video from the server.
• This technique avoids having to download the entire video-file before playout begins.
2) Interactivity
• The media is prerecorded, so the user may pause, reposition or fast-forward through video-content.
• The response time should be less than a few seconds.
3) Continuous Playout
• Once playout of the video begins, it should proceed according to the original timing of the recording.
• The data must be received from the server in time for its playout at the client.
Otherwise, users experience video-frame skipping (or freezing).
[Link] Conversational Voice- and Video-over-IP
• Real-time conversational voice over the Internet is often referred to as Internet telephony.
• It is also commonly called Voice-over-IP (VoIP).
• Conversational video includes the video of the participants as well as their voices.
• Most of today‟s voice applications allow users to create conferences with three or more participants.
• Nowadays, many Internet companies provide voice application. For example: Skype & Google Talk.
• Two parameters are particularly important for voice applications:
1) Timing considerations and
2) Tolerance of data loss
• Timing considerations are important because voice applications are highly delay-sensitive.
• Loss-tolerant means
Occasional loss only causes occasional glitches in audio playback & these losses can be partially/fully
hidden.
[Link] Streaming Live Audio & Video
• These applications are similar to broadcast radio, except that transmission takes place over Internet.
• These applications allow a user to receive a live radio transmitted from any corner of the world.
• For example: live cricket commentary.
• Today, thousands of radio stations around the world are broadcasting content over the Internet.
• Live broadcast applications have many users who receive the same audio program at the same time.
• The network must provide an average throughput that is larger than the video consumption rate.
5.2 Streaming Stored Video
• Prerecorded videos are placed on servers.
• Users send requests to these servers to view the videos on-demand.
• The media is prerecorded, so the user may pause, reposition or fast-forward through video-content.
• Three categories of applications:
1) UDP streaming
2) HTTP streaming and
3) Adaptive HTTP streaming.
• A main characteristic of video-streaming is the extensive use of client-side buffering.
• Two advantages of client-side buffering:
1) Client-side buffering can mitigate effects of varying end-to-end delays
2) This can mitigate effects of varying amounts of available bandwidth b/w server & client.
5.2.1 UDP Streaming
• The server transmits video at a rate that matches the client‟s video consumption rate.
• The server transmits the video-chunks over UDP at a steady rate.
• UDP does not employ a congestion-control mechanism.
• Therefore, the server can push packets into the network at the video consumption rate.
• Typically, UDP streaming uses a small client-side buffer. (RTP → Real-Time Transport Protocol).
• Using RTP, the server encapsulates the video-chunks within transport packets.
• The client & server also maintain a control-connection over which the client sends commands (such as
pause, resume and reposition).
• The RTSP (Real-Time Streaming Protocol) is a popular open protocol for a control-connection.
• Disadvantages:
1) Unreliability
➢ UDP streaming can fail to provide continuous playout „.‟ of varying amt of available bandwidth
2) Costly & Complex
➢ A media control server (RTSP) is required
→ to process client-to-server interactivity requests and
→ to track client-state for each ongoing client-session.
➢ This increases the overall cost and complexity of deploying a large-scale application.
3) Firewall Problem
➢ Many firewalls are configured to block UDP traffic.
➢ This prevents the users behind the firewalls from receiving the video.
5.2.2 HTTP Streaming
• The video is stored in an HTTP server as an ordinary file with a specific URL.
• Here is how it works:
1) When a user wants to see the video, the client
→ establishes a TCP connection with the server and
→ issues an HTTP GET request for that URL.
2) Then, the server responds with the video file, within an HTTP response message.
3) On client side, the bytes are collected in a client application buffer.
4) Once no. of bytes in this buffer exceeds a specific threshold, the client begins playback.
• Advantages:
1) Not Costly & Complex
➢ Streaming over HTTP avoids the need for a media control server (RTSP).
➢ This reduces the cost of deploying a large-scale application.
2) No Firewall Problem
➢ The use of HTTP over TCP also allows the video to traverse firewalls and NATs more easily.
3) Prefetching Video
➢ The client downloads the video at a rate higher than the consumption rate.
➢ Thus, prefetching video-frames that are to be consumed in the future.
➢ This prefetched video is stored in the client application buffer
• Nowadays, most video-streaming applications use HTTP streaming. For example: YouTube
[Link] Client Application Buffer & TCP Buffers
• Figure 5.1 illustrates the interaction between client and server for HTTP streaming.
• On the server side,
1) The bytes of the video file are sent into the server‟s socket.
2) Then, the bytes are placed in the TCP send buffer before.
3) Finally, the bytes are transmitted into the Internet.
• On the client side,
1) The application (media-player) reads bytes from the TCP receive-buffer (thro client-socket)
2) Then, the application places the bytes into the client-buffer.
3) At the same time, the application periodically
→ grabs video-frames from the client-buffer
→ decompresses the frames and
→ displays the frames on the user‟s screen.
Figure 5.1: Streaming stored video over HTTP/TCP
[Link] Early Termination & Repositioning the Video
• HTTP streaming systems make use of the byte-range header in the HTTP GET request message.
• Byte-range header specifies the range of bytes the client currently wants to retrieve from the video.
• This is particularly useful when the user wants to reposition to a future point in the video.
• When the user repositions to a new position, the client sends a new HTTP request.
• When server receives new HTTP request, the server sends the requested-bytes.
• Disadvantage:
When a user repositions to a future point in the video, some prefetched-but-not-yet-viewed data will
go unwatched. This results in a waste of bandwidth and server-resources.
5.2.3 Adaptive Streaming & DASH
• Problem with HTTP streaming:
All clients receive the same encoding of video, despite the large variations in the amount of bandwidth
available to different clients.
Solution: Use DASH (Dynamic Adaptive Streaming over HTTP).
[Link] DASH
• The video is encoded into several different versions.
• Each version has a different bit-rate and a different quality level.
• Two main tasks:
1) The client dynamically requests video-chunks from the different versions: low & high.
i) When the available bandwidth is high, the client selects chunks from a high-rate version.
For ex: Fiber connections can receive a high-quality version.
ii) When the available bandwidth is low, the client naturally selects from a low-rate version.
For ex: 3G connections can receive a low-quality version.
2) The client adapts to the available bandwidth if end-to-end bandwidth changes during session.
➢ This feature is particularly important for mobile-users.
➢ The mobile-users see their bandwidth fluctuate as they move with respect to base-stations.
• HTTP server stores following files:
1) Each video version with a different URL.
2) Manifest file provides a URL for each version along with its bit-rate.
• Here is how it works:
1) First, the client requests the manifest file and learns about the various versions.
2) Then, the client selects one chunk at a time by specifying
→ URL and
→ byte range in an HTTP GET request message.
3) While downloading chunks, the client
→ measures the received bandwidth and
→ runs a rate determination-algorithm.
i) If measured-bandwidth is high, client will choose chunk from high-rate version.
ii) If measured-bandwidth is low, client will choose chunk from low-rate version
4) Therefore, DASH allows the client to freely switch among different quality-levels.
• Advantages:
1) DASH can achieve continuous playout at the best possible quality level w/o frame freezing.
2) Server-side scalability is improved: Because
→ the client maintains the intelligence to determine which chunk to send next.
3) Client can use HTTP byte-range request to precisely control the amount of prefetched video.
5.2.4 CDN
[Link] Motivation for CDN
• The streaming video service can be provided is as follows:
1) Build a single massive data-center.
2) Store all videos in the data-center and
3) Stream the videos directly from the data-center to clients worldwide.
• Three major problems with the above approach:
1) More Delay
➢ If links provides a throughput lesser than consumption-rate, the end-to-end throughput will also
be below the consumption-rate.
➢ This results in freezing delays for the user.
2) Network Bandwidth is wasted
➢ A popular video may be sent many times over the same links.
3) Single Point of Failure:
➢ If the data-center goes down, it cannot distribute any video streams.
Problem Solution: Use CDN (Content Distribution Network).
[Link] CDN Types
• A CDN
→ manages servers in multiple geographically distributed locations
→ stores copies of the videos in its servers, and
→ attempts to direct each user-request to a CDN that provides the best user experience.
• The CDN may be a private CDN or a third-party CDN.
1) Private CDN
➢ A private CDN is owned by the content provider itself.
➢ For example:
Google‟s CDN distributes YouTube videos
2) Third Party CDN
➢A third-party CDN distributes content on behalf of multiple content providers CDNs.
➢ Two approaches for server placement:
i) Enter Deep
¤ The first approach is to enter deep into the access networks of ISPs.
¤ Server-clusters are deployed in access networks of ISPs all over the world.
¤ The goal is to get close to end users.
¤ This improves delay/throughput by decreasing no. of links b/w end user & CDN cluster
ii) Bring Home
¤ The second approach is to bring the ISPs home.
¤ Large clusters are built at a smaller number of key locations.
¤ These clusters are connected using a private high-speed network.
¤ Typically, clusters are placed at a location that is near the PoPs of many tier-1 ISPs. For example:
within a few miles of both Airtel and BSNL PoPs in a major city.
¤ Advantage:
Lower maintenance and management overhead.
¤ Disadvantage:
Higher delay and lower throughput to end users.
[Link] CDN Operation
• When a browser wants to retrieve a specific video, the CDN intercepts the request.
• Then, the CDN
1) determines a suitable server-cluster for the client and
2) redirects the client‟s request to the desired server.
• Most CDNs take advantage of DNS to intercept and redirect requests.
• CDN operation is illustrated in Figure 5.2.
Figure 5.2: DNS redirects a user‟s request to a CDN server
• Suppose a content provider “NetCinema” employs the CDN company ”KingCDN” to distribute videos.
• Let URL = [Link]
• Six events occur as shown in Figure 5.2:
1) The user visits the Web page at NetCinema.
2) The user clicks on the following link:
[Link]
➢ Then, the user‟s host sends a DNS query for “[Link]”.
3) The user‟s local-DNS-server (LDNS) forwards the DNS-query to an authoritative-DNS-server
“NetCinema”.
➢ The server “NetCinema” returns to the LDNS a hostname in the KingCDN‟s domain.
➢ For example: “[Link]”.
4) The user‟s LDNS then sends a second query, now for “[Link]”.
➢ Eventually, KingCDN‟s DNS system returns the IP addresses of a “KingCDN” server to LDNS.
5) The LDNS forwards the IP address of the “KingCDN" server to the user‟s host.
6) Finally, the client
→ establishes a TCP connection with the server
→ issues an HTTP GET request for the video.