CS 414 Multimedia Systems Design Lecture 19 Multimedia

  • Slides: 24
Download presentation
CS 414 – Multimedia Systems Design Lecture 19 – Multimedia Transport Subsystem (Part 2)

CS 414 – Multimedia Systems Design Lecture 19 – Multimedia Transport Subsystem (Part 2) + Midterm Review Klara Nahrstedt Spring 2014 CS 414 - Spring 2014

Midterm March 7 (Friday), 1 -1: 50 pm, 0216 SC n Closed Book, Closed

Midterm March 7 (Friday), 1 -1: 50 pm, 0216 SC n Closed Book, Closed Notes n You can bring calculator and 1 page cheat sheet n CS 414 - Spring 2014

Covered Material n n n Class Notes (Lectures 1 -15) MP 1 Book Chapters

Covered Material n n n Class Notes (Lectures 1 -15) MP 1 Book Chapters to read/study: ¨ Media Coding and Content processing book Chapter 2, n Chapter 3. 1 -3. 2, 3. 8, n Chapter 4. 1 -4. 2. 2. 1, n Chapter 4. 3 (as discussed in lecture) n Chapter 5, chapter 7. 1 -7. 5, 7. 7 n CS 414 - Spring 2014

Material n Audio Characteristics ¨ Samples, frequency, Nyquist theorem ¨ Perception, psychoacoustic effects, loudness,

Material n Audio Characteristics ¨ Samples, frequency, Nyquist theorem ¨ Perception, psychoacoustic effects, loudness, pitch, decibel, intensity ¨ Sampling rate, quantization n Audio Characteristics ¨ PCM, DPCM, signal-to-noise ratio CS 414 - Spring 2014

Material n Image Characteristics ¨ Sampling, quantization, pixels ¨ Image properties: color CS 414

Material n Image Characteristics ¨ Sampling, quantization, pixels ¨ Image properties: color CS 414 - Spring 2014

Material n Video technology ¨ Color perception: hue, brightness, saturation, ¨ Visual representation: horizontal

Material n Video technology ¨ Color perception: hue, brightness, saturation, ¨ Visual representation: horizontal and vertical resolution, aspect ratio; depth perception, luminance, temporal resolution and motion ¨ Flicker effect ¨ Color coding: YUV, YIQ, RGB ¨ NTSC vs HDTV formats CS 414 - Spring 2014

Material n Basic Coding schemes ¨ Run-length coding ¨ Statistical coding Huffman coding n

Material n Basic Coding schemes ¨ Run-length coding ¨ Statistical coding Huffman coding n Arithmetic coding n n Hybrid codes ¨ JPEG: image preparation, DCT transformation, Quantization, entropy coding, JPEG-2000 characteristics CS 414 - Spring 2014

Material n Hybrid Coding ¨ Video MPEG: image preparation, I, P, B frames characteristics,

Material n Hybrid Coding ¨ Video MPEG: image preparation, I, P, B frames characteristics, quantization, display vs processing/transmission order of frames ¨ Audio MPEG: role of psychoacoustic effect, masking, steps of audio compression ¨ MPEG-4: differences to MPEG-2/MPEG-1 n Audio-visual objects, layering ¨ H. 261, 263, 264, 265 CS 414 - Spring 2014

Sample Problems Consider the following alphabet {C, S, 4, 1}, with probabilities: P(C) =

Sample Problems Consider the following alphabet {C, S, 4, 1}, with probabilities: P(C) = 0. 3, P(S) = 0. 2, P(4)= 0. 25, P(1) = 0. 25. n Encode the word CS 414 using n ¨ Huffman coding and arithmetic coding ¨ Compare which encoding requires less bits CS 414 - Spring 2014

Sample Problems n Describe briefly each step in MPEG-1 audio encoding. Specify the functionality,

Sample Problems n Describe briefly each step in MPEG-1 audio encoding. Specify the functionality, which is performed in each step. You don’t have to provide equations, only a clear explanation of the functionality that is performed inside each step. CS 414 - Spring 2014

Sample Problems What is flicker effect and how to remove it? n Provide five

Sample Problems What is flicker effect and how to remove it? n Provide five differences between MPEG-4 video encoding standard and the previous MPEG video encoding standards n CS 414 - Spring 2014

Covered Aspects of Multimedia Image/Video Capture Audio/Video Perception/ Playback Audio/Video Presentation Playback Image/Video Information

Covered Aspects of Multimedia Image/Video Capture Audio/Video Perception/ Playback Audio/Video Presentation Playback Image/Video Information Representation Transmission Audio Capture Transmission Compression Processing Audio Information Representation Media Server Storage CS 414 - Spring 2014 A/V Playback

We have discussed so far Quality of Service n Multimedia Data Establishment Protocol n

We have discussed so far Quality of Service n Multimedia Data Establishment Protocol n ¨ Negotiation and Translation of Qo. S CS 414 - Spring 2014

What we will talk about today n Multimedia Call Establishment Protocol ¨ Admission and

What we will talk about today n Multimedia Call Establishment Protocol ¨ Admission and Reservation Operations Bandwidth Admission n Processing Admission n n Data Streaming/Transmission Operations ¨ Traffic Shaping CS 414 - Spring 2014

Bandwidth Admission Test n Consider bi – reserved bandwidth for the ‘i’ connection ¨

Bandwidth Admission Test n Consider bi – reserved bandwidth for the ‘i’ connection ¨ Bmax – maximal bandwidth at the network interface ¨ n n Admission test (if all connections declare their bandwidth requirements bi at the same time): ¨ ∑ b ≤ Bmax (i=1, …n) i Example: n n n ¨ Bmax = 100 Mbps, Bandwidth requirement of connection 1 b 1 = 10 Mbps Bandwidth requirement of connection 2 b 2 = 20 Mbps Admission Control Condition: b 1 + b 2 < Bmax n n Step 1: if b 1 < Bmax then admit b 1, reserve b 1, adjust Bmax to Bavail = Bmax – b 1 Step 2: if b 2 < Bavail then admit b 2, reserve b 2, adjust Bavail to Bavail = Bavail – b 2 CS 414 - Spring 2014

Packet/Frame Scheduling Admission Systems have queues n We need packet/frame scheduling policies for admitting

Packet/Frame Scheduling Admission Systems have queues n We need packet/frame scheduling policies for admitting new streams n We need frame/packet schedulability tests n ¨ Note that in networking only NONPREEMPTIVE scheduling exists!!! CS 414 - Spring 2014

Packet/Frame Scheduling Admission Control ei – processing of a packet ‘i’ in network node

Packet/Frame Scheduling Admission Control ei – processing of a packet ‘i’ in network node Admission Test: ei ≤ deadline ∑ servei/ (1/r) ≤ 1 (i=1, …, n) serve – packet/frame service time at the processors – constant time due to hardware implementation q_in and q_out are queuing times N – number of packets in queue λ – service rate q = N/λ (Little Theorem) r – service rate of the switch 1/r – packet/frame period on CS 414 - Spring 2014 processor

Resource Reservation/Allocation n Bandwidth reservation ¨ Pessimistic reservation with maximal bandwidth allocation: Given (MN,

Resource Reservation/Allocation n Bandwidth reservation ¨ Pessimistic reservation with maximal bandwidth allocation: Given (MN, RA, and MA) n if then CS 414 - Spring 2014

Pessimistic Resource Reservation (Example) n n Example: Consider sequence of MPEG video frames of

Pessimistic Resource Reservation (Example) n n Example: Consider sequence of MPEG video frames of size 80 KB, 60 KB, 20 KB, 60 KB, 20 KB (Group of Pictures I, P, B, B ), Pessimistic frame size calculation: ¨ n n MA = max(80, 60, 20, 20) = 80 KB Given video frame rate RA = 20 fps If Given MN = 10 KB (network packet size, e. g. , packet size for the transport layer like TCP/UDP), then need to consider bandwidth/ throughput reservation for ¨ BN = 10 KB x (8 network packets per application frame) x 20 application frames per second= 1600 KB/second = 12800 Kbps CS 414 - Spring 2014

Optimistic Resource Reservation/Allocation Optimistic reservation considers average bandwidth allocation n Given MA, RA, MN,

Optimistic Resource Reservation/Allocation Optimistic reservation considers average bandwidth allocation n Given MA, RA, MN, where n n Then CS 414 - Spring 2014

Optimistic Resource Reservation (Example) n n Example: Consider sequence of MPEG video frames of

Optimistic Resource Reservation (Example) n n Example: Consider sequence of MPEG video frames of size 80 KB, 60 KB, 20 KB, 60 KB, 20 KB (Group of Pictures I, P, B, B, ), Optimistic frame size calculation: ¨ n n MA = 280/7 = 40 KB Given video frame rate RA = 20 fps If Given MN = 10 KB (network packet size, e. g. , packet size for the transport layer like TCP/UDP), then need to consider bandwidth/ throughput reservation for ¨ BN = 10 KB x (4 network packets per application frame) x 20 application frames per second= 800 KB/second = 6400 Kbps CS 414 - Spring 2014

Sender-Oriented Reservation Protocol CS 414 - Spring 2014

Sender-Oriented Reservation Protocol CS 414 - Spring 2014

Receiver-Oriented Reservation Protocol CS 414 - Spring 2014

Receiver-Oriented Reservation Protocol CS 414 - Spring 2014

Conclusion n Multimedia Call Establishment Protocol requires ¨ Qo. S Parameter negotiation (exchange) ¨

Conclusion n Multimedia Call Establishment Protocol requires ¨ Qo. S Parameter negotiation (exchange) ¨ Qo. S Parameter translation ¨ Admission Control of resources needed to provide Qo. S requirements Bandwidth admission n Frame/Packet scheduling admission n ¨ Reservation of resources for admitted multimedia streams CS 414 - Spring 2014