Multimedia :Multimedia 15-441 Computer Networks 10/02/02 Xavier Appé


Outlines :Outlines Difference with classic applications Classes of multimedia applications Requirements/Constraints Problems with today’s Internet and solutions Common multimedia protocols RTP, RTCP Accessing multimedia data through a web server Conclusion


Difference with classic applications :Difference with classic applications Highly delay-sensitive Packets are useless if they arrive too late Loss-tolerant (for the most part) Packet loss can be concealed


Outlines :Outlines Difference with classic applications Classes of multimedia applications Requirements/Constraints Problems with today’s Internet and solutions Common multimedia protocols RTP, RTCP Accessing multimedia data through a web server Conclusion


Classes of multimedia Applications :Classes of multimedia Applications Streaming Stored Audio and Video Streaming Live Audio and Video Real-Time Interactive Audio and Video Others


Class: Streaming Stored Audio and Video :Class: Streaming Stored Audio and Video The multimedia content has been prerecorded and stored on a server User may pause, rewind, forward, etc… The time between the initial request and display start can be 1 to 10 seconds Constraint: after display start, the playout must be continuous


Class: Streaming Live Audio and Video :Class: Streaming Live Audio and Video Similar to traditional broadcast TV/radio, but delivery on the Internet Non-interactive just view/listen Can not pause or rewind Often combined with multicast The time between the initial request and display start can be up to 10 seconds Constraint: like stored streaming, after display start, the playout must be continuous


Class: Real-Time Interactive Audio and Video :Class: Real-Time Interactive Audio and Video Phone conversation/Video conferencing Constraint: delay between initial request and display start must be small Video: <150 ms acceptable Audio: <150 ms not perceived, <400 ms acceptable Constraint: after display start, the playout must be continuous


Class: Others :Class: Others Multimedia sharing applications Download-and-then-play applications E.g. Napster, Gnutella, Freenet Distance learning applications Coordinate video, audio and data Typically distributed on CDs


Outlines :Outlines Difference with classic applications Classes of multimedia applications Requirements/Constraints Problems with today’s Internet and solutions Common multimedia protocols RTP, RTCP Accessing multimedia data through a web server Conclusion


Challenge :Challenge TCP/UDP/IP suite provides best-effort, no guarantees on expectation or variance of packet delay Performance deteriorate if links are congested (transoceanic) Most router implementations use only First-Come-First-Serve (FCFS) packet processing and transmission scheduling


Problems and solutions :Problems and solutions Limited bandwidth Solution: Compression Packet Jitter Solution: Fixed/adaptive playout delay for Audio (example: phone over IP) Packet loss Solution: FEC, Interleaving


Problem: Limited bandwidth Intro: Digitalization :Problem: Limited bandwidth Intro: Digitalization Audio x samples every second (x=frequency) The value of each sample is rounded to a finite number of values (for example 256). This is called quantization Video Each pixel has a color Each color has a value


Problem: Limited bandwidthNeed for compression :Problem: Limited bandwidthNeed for compression Audio CD quality: 44100 samples per seconds with 16 bits per sample, stereo sound 44100*16*2 = 1.411 Mbps For a 3-minute song: 1.441 * 180 = 254 Mb = 31.75 MB Video For 320*240 images with 24-bit colors 320*240*24 = 230KB/image 15 frames/sec: 15*230KB = 3.456MB 3 minutes of video: 3.456*180 = 622MB


Audio compression :Audio compression Several techniques GSM (13 kbps), G.729(8 kbps), G723.3(6.4 and 5.3kbps) MPEG 1 layer 3 (also known as MP3) Typical compress rates 96kbps, 128kbps, 160kbps Very little sound degradation If file is broken up, each piece is still playable Complex (psychoacoustic masking, redundancy reduction, and bit reservoir buffering) 3-minute song (128kbps) : 2.8MB


Image compression: JPEG :Image compression: JPEG Divide digitized image in 8x8 pixel blocks Pixel blocks are transformed into frequency blocks using DCT (Discrete Cosine Transform). This is similar to FFT (Fast Fourier Transform) The quantization phase limits the precision of the frequency coefficient. The encoding phase packs this information in a dense fashion


JPEG Compression :JPEG Compression


Video compression :Video compression Popular techniques MPEG 1 for CD-ROM quality video (1.5Mbps) MPEG 2 for high quality DVD video (3-6 Mbps) MPEG 4 for object-oriented video compression


Video Compression: MPEG :Video Compression: MPEG MPEG uses inter-frame encoding Exploits the similarity between consecutive frames Three frame types I frame: independent encoding of the frame (JPEG) P frame: encodes difference relative to I-frame (predicted) B frame: encodes difference relative to interpolated frame Note that frames will have different sizes Complex encoding, e.g. motion of pixel blocks, scene changes, … Decoding is easier then encoding MPEG often uses fixed-rate encoding I P B B B B B B P P I B B B B


MPEG Compression (cont.) :MPEG Compression (cont.)


MPEG System Streams :MPEG System Streams Combine MPEG video and audio streams in a single synchronized stream Consists of a hierarchy with meta data at every level describing the data System level contains synchronization information Video level is organized as a stream of group of pictures Group of pictures consists of pictures Pictures are organized in slices …


MPEG System Streams (cont.) :MPEG System Streams (cont.)


MPEG System Streams (cont.) :MPEG System Streams (cont.)


Problem: Packet Jitter :Problem: Packet Jitter Jitter: Variation in delay Example 1 3 5 4 3 2 Sender No jitter 1 2 5 4 6 6 Receiver Jitter pkt 6 pkt 5


Dealing with packet jitter :Dealing with packet jitter How does Phone over IP applications limit the effect of jitter? A sequence number is added to each packet A timestamp is added to each packet Playout is delayed


Dealing with packet jitterFixed playout delay :Dealing with packet jitterFixed playout delay Fixed playout delay


Dealing with packet jitterAdaptive playout delay :Dealing with packet jitterAdaptive playout delay Objective is to use a value for p-r that tracks the network delay performance as it varies during a transfer. The following formulas are used: di = (1-u)di-1 + u(ri – ti) u=0.01 for example i = (1-u)i-1 + u|ri-ti-di| Where ti is the timestamp of the ith packet (the time pkt i is sent) ri is the time packet i is received pi is the time packet i is played di is an estimate of the average network delay i is an estimate of the average deviation of the delay from the estimated average delay


Problem: Packet loss :Problem: Packet loss Loss is in a broader sense: packet never arrives or arrives later than its scheduled playout time Since retransmission is inappropriate for Real Time applications, FEC or Interleaving are used to reduce loss impact.


Recovering from packet lossForward Error Correction :Recovering from packet lossForward Error Correction Send redundant encoded chunk every n chunks (XOR original n chunks) If 1 packet in this group lost, can reconstruct If >1 packets lost, cannot recover Disadvantages The smaller the group size, the larger the overhead Playout delay increased


Recovering from packet loss Piggybacking Lo-fi stream :Recovering from packet loss Piggybacking Lo-fi stream With one redundant low quality chunk per chunk, scheme can recover from single packet losses


Recovering from packet loss Interleaving :Recovering from packet loss Interleaving Divide 20 msec of audio data into smaller units of 5 msec each and interleave Upon loss, have a set of partially filled chunks


Recovering from packet lossReceiver-based Repair :Recovering from packet lossReceiver-based Repair The simplest form: Packet repetition Replaces lost packets with copies of the packets that arrived immediately before the loss A more computationally intensive form: Interpolation Uses Audio before and after the loss to interpolate a suitable packet to cover the loss


Movie Time :Movie Time


Outlines :Outlines Difference with classic applications Classes of multimedia applications Requirements/Constraints Problems with today’s Internet and solutions Common multimedia protocols RTP, RTCP Accessing multimedia data through a web server Conclusion


Real Time Protocol (RTP) :Real Time Protocol (RTP) RTP logically extends UDP Sits between UDP and application Implemented as an application library What does it do? Framing Multiplexing Synchronization Feedback (RTCP)


RTP packet format :RTP packet format Payload Type: 7 bits, providing 128 possible different types of encoding; eg PCM, MPEG2 video, etc. Sequence Number: 16 bits; used to detect packet loss


RTP packet format (cont) :RTP packet format (cont) Timestamp: 32 bytes; gives the sampling instant of the first audio/video byte in the packet; used to remove jitter introduced by the network Synchronization Source identifier (SSRC): 32 bits; an id for the source of a stream; assigned randomly by the source


Timestamp vs. Sequence No :Timestamp vs. Sequence No Timestamps relates packets to real time Timestamp value sampled from a media specific clock Sequence number relates packets to other packets


Audio silence example :Audio silence example Consider audio data type What do you want to send during silence? Not sending anything Why might this cause problems? Other side needs to distinguish between loss and silence Receiver uses Timestamps and sequence No. to figure out what happened


RTP Control Protocol (RTCP) :RTP Control Protocol (RTCP) Used in conjunction with RTP. Used to exchange control information between the sender and the receiver. Three reports are defined: Receiver reception, Sender, and Source description Reports contain statistics such as the number of packets sent, numberof packets lost, inter-arrival jitter Typically, limit the RTCP bandwidth to 5%. Approximately one sender report for threereceiver reports


Outlines :Outlines Difference with classic applications Classes of multimedia applications Requirements/Constraints Problems with today’s Internet and solutions Common multimedia protocols RTP, RTCP Accessing multimedia data through a web server Conclusion


Streaming Stored Multimedia Example :Streaming Stored Multimedia Example Audio/Video file is segmented and sent over either TCP or UDP, public segmentation protocol: Real-Time Protocol (RTP) User interactive control is provided, e.g. the public protocol Real Time Streaming Protocol (RTSP)


Streaming Stored Multimedia Example :Streaming Stored Multimedia Example Helper Application: displays content, which is typically requested via a Web browser; e.g. RealPlayer; typical functions: Decompression Jitter removal Error correction: use redundant packets to be used for reconstruction of original stream GUI for user control


Streaming from Web Servers :Streaming from Web Servers Audio: in files sent as HTTP objects Video (interleaved audio and images in one file, or two separate files and client synchronizes the display) sent as HTTP object(s) A simple architecture is to have the Browser request the object(s) and after their reception pass them to the player for display - No pipelining


Streaming from a Web Server (cont) :Streaming from a Web Server (cont) Alternative: set up connection between server and player, then download Web browser requests and receives a Meta File (a file describing the object) instead of receiving the file itself; Browser launches the appropriate Player and passes it the Meta File; Player sets up a TCP connection with a streaming server Server and downloads the file


Using a Streaming Server :Using a Streaming Server


Options when using a streaming server :Options when using a streaming server Use UDP, and Server sends at a rate (Compression and Transmission) appropriate for client; to reduce jitter, Player buffers initially for 2-5 seconds, then starts display Use TCP, and sender sends at maximum possible rate under TCP; retransmit when error is encountered; Player uses a much large buffer to smooth delivery rate of TCP


Real Time Streaming Protocol (RTSP) :Real Time Streaming Protocol (RTSP) For user to control display: rewind, fast forward, pause, resume, etc… Out-of-band protocol (uses two connections, one for control messages (Port 554) and one for media stream) RFC 2326 permits use of either TCP or UDP for the control messages connection, sometimes called the RTSP Channel As before, meta file is communicated to web browser which then launches the Player; Player sets up an RTSP connection for control messages in addition to the connection for the streaming media


Meta File Example :Meta File Example Twister


RTSP Operations :RTSP Operations


Outlines :Outlines Difference with classic applications Classes of multimedia applications Requirements/Constraints Problems with today’s Internet and solutions Common multimedia protocols RTP, RTCP Accessing multimedia data through a web server Conclusion


Conclusion :Conclusion None of the proposed solutions give a real guarantee to the user that multimedia data will arrive on time. Couldn’t we reserve some bandwidth for our multimedia transfer?