1. © 1999, Third Generation Partnership Project 2. All Rights Reserved. Permission is granted for copying, reproducing, or duplicating this document only for the legitimate purposes of 3GPP2. No other copying, reproduction, duplication, or distribution is permitted. 1 3GPP2-C21-19990614-004 3GPP2-C12-19990614-004 Video Conferencing on CDMA2000 SOURCE : Jeong-Hoon Park, Dong-Seek Park Samsung Electronics Co. Ltd., +82-331-200-3674, +82-331-200-3195 (fax) celp@mmrnd.sec.samsung.co.kr dspark@mmrnd.sec.samsung.co.kr Matthieu Tisserand, John Villasenor UCLA Electrical Engineering +1-310-825-0228, +1-310-825-7928 (fax) mattiss@icsl.ucla.edu, villa@icsl.ucla.edu DATE : June 14, 1999 ABSTRACT : This contribution presents simulation results of video conferencing with cdma2000 (phase1) RLP RECOMMENDATION : Review and Discuss Notice © 1999, Samsung Electronics. All rights reserved. The information contained in this contribution is provided for the sole purpose of promoting discussion within the Technical Specification Groups of 3GPP2 and is not binding on the contributor. The contributor reserves the right to add to, amend or withdraw the statements contained herein. Samsung Electronics grants a free, irrevocable license to 3GPP2 and its Organizational Partners to incorporate text or other copyrightable material contained in the contribution and any modifications thereof in the creation of 3GPP2 publications; to copyright and sell in Organizational Partner's name any Organizational Partner's standards publication even though it may include portions of the contribution; and at the Organizational Partner's sole discretion to permit others to reproduce in whole or in part such contributions or the resulting Organizational Partner's standards publication. The contributor must also be willing to grant licenses under such contributor copyrights to third parties on reasonable, non-discriminatory terms and conditions, as appropriate.

2. © 1999, Third Generation Partnership Project 2. All Rights Reserved. Permission is granted for copying, reproducing, or duplicating this document only for the legitimate purposes of 3GPP2. No other copying, reproduction, duplication, or distribution is permitted. 2 Table of Contents n Simulation Scenario –Scope of Simulation –Generation of H.324M Stream –Use of existing RLP Framing and MUXing n Simulation Results –Overhead Analysis –Subjective Picture Quality Comparison n Caveats –More Possible Sets of Simulations

3. © 1999, Third Generation Partnership Project 2. All Rights Reserved. Permission is granted for copying, reproducing, or duplicating this document only for the legitimate purposes of 3GPP2. No other copying, reproduction, duplication, or distribution is permitted. 3 Simulation Scenario n Scope of Simulation –Do the following loop-back video conferencing scenario in WCDMA channel User Application (H.324M) SCH RLP Layer including MUX (with the simplest set of parameters) Null Data for Speech 76.8/153.6 kbps 64 kbps/128kbps Physical Layer (with 1/3 convolution code) and error-prone WCDMA channel

4. © 1999, Third Generation Partnership Project 2. All Rights Reserved. Permission is granted for copying, reproducing, or duplicating this document only for the legitimate purposes of 3GPP2. No other copying, reproduction, duplication, or distribution is permitted. 4 Simulation Scenario H.263 Codec Null Data for Speech H.223/Annex B MUX HDLC/Header etc. Payload (Usually Video and Speech. Since speech data has to meet the condition of 24 Bytes/30msec video-only payload can be made in case that video data generation is faster than 30 msec.) RLP Header (based on 4.2 in IS-707) RLP frame data Video Application Layer RLP Header RLP frame data RLP Header RLP frame data RLP Layer MUX Header RLP frame data CRC MUX Header RLP frame data CRC Physical Layer SDU (4 or 8 MUX_PDUs) 1522/3014 bits MUX Layer Do XOR with WCDMA error file to simulate Physical Layer and Channel n Simulation Scenario of RLP/MUX Layer in Supplemental Channel (SCH)

5. © 1999, Third Generation Partnership Project 2. All Rights Reserved. Permission is granted for copying, reproducing, or duplicating this document only for the legitimate purposes of 3GPP2. No other copying, reproduction, duplication, or distribution is permitted. 5 H.263/MPEG4 Video Codec G.723.1 Audio Codec H.223 = Level 0 MUX/ DEMUX Wireless Wired Modem Control V.14 etc. Data Protocols PSTN Network Video In/Out Audio In/Out Data In/Out System Control H.324M Level 3, Level 2, Level 1 Wireless Network Mobile Multiplexer Control Protocol H.245 H.324 Simulation Scenario n Video Conferencing Application: Use H.324M - Mobile Version of H.324 (Video Conferencing Standard for Circuit Switched Networks)

6. © 1999, Third Generation Partnership Project 2. All Rights Reserved. Permission is granted for copying, reproducing, or duplicating this document only for the legitimate purposes of 3GPP2. No other copying, reproduction, duplication, or distribution is permitted. 6 H.263 Video Codec Null Data for Speech Mobile MUX H.223/B MUX Header (16-bit) Payload (video only or video and speech (variable length, upto 254 Bytes) Golay Code (12-bit) Payload Length (8-bit) MC (4-bit) Simulation Scenario n Generation of H.324M Stream n Try to satisfy the condition*: –Length of H.324M stream <= (Physical Layer SDU - RLP overhead) * In case of realistic environments, in general, generating fixed length of video packet at each time index and it requires highly complicated buffer control and sometimes it might cause unwanted picture quality.

7. © 1999, Third Generation Partnership Project 2. All Rights Reserved. Permission is granted for copying, reproducing, or duplicating this document only for the legitimate purposes of 3GPP2. No other copying, reproduction, duplication, or distribution is permitted. 7 104 bits 40/40 bits 368/560 bits 0000.... 744/1128 bits 38.4/57.6 kbps 76.8/115.2 kbps 368 bits....0000.... 1512/2280 bits 153.6 kbps 3048 bits 0000....MCMC MCMCMCMC MCMCMC 8/8 bits Simulation Scenario n Used Physical Layer SDU and MUX_PDU (Simulation with 76.8 and 153.6 kbps) n 1 format E + 3/7 Format D Data RLP frames per Physical Layer SDU

8. © 1999, Third Generation Partnership Project 2. All Rights Reserved. Permission is granted for copying, reproducing, or duplicating this document only for the legitimate purposes of 3GPP2. No other copying, reproduction, duplication, or distribution is permitted. 8 Simulation Results n Simulation Condition: –Format E/D with 41/42 Bytes filled with information –No control RLP frame, Non-LTU mode (no use of CRC) n Overhead per Physical Layer SDU –For 76.8 kbps Total Information: 41 x 8 x 1 + 42 x 8 x 3 in each SDU = 1336 bits for 76.8 kbpsTotal Information: 41 x 8 x 1 + 42 x 8 x 3 in each SDU = 1336 bits for 76.8 kbps Overhead = (1512 - 1336)/1512  11.6%Overhead = (1512 - 1336)/1512  11.6% –For 153.6 kbps Total Information: 41 x 8 x 1 + 42 x 8 x 7 in each SDU = 2680 bits for 153.6 kbpsTotal Information: 41 x 8 x 1 + 42 x 8 x 7 in each SDU = 2680 bits for 153.6 kbps Overhead = (3048 - 2680)/3048  12.1%Overhead = (3048 - 2680)/3048  12.1% 20 msec 40 bits 368 bits 0000.... 76.8 kbps 1512 bits MCMCMCMC HeaderInformationHeaderInformation 18/10 bits 328/336 bits without CRC

9. © 1999, Third Generation Partnership Project 2. All Rights Reserved. Permission is granted for copying, reproducing, or duplicating this document only for the legitimate purposes of 3GPP2. No other copying, reproduction, duplication, or distribution is permitted. 9 Simulation Results n Problems in Using RLP Layer –Overhead –Number of lost H.223 MUX_PDUs n Simulation Results –Comparison of available video bandwidth –Number of lost H.223 MUX_PDUs BitrateWith RLPComparison ItemsWithout RLP 76.3 153.6 kbps 51 99 Available Video Rate (kbps) 57 108 Without RLP 89.18 82.91 Video Length (sec) 86.97 82.22 4482 4129 Number of SDU Windows 4354 4125 89.64 82.59 Total Time Elapsed (sec) 87.09 82.50

10. © 1999, Third Generation Partnership Project 2. All Rights Reserved. Permission is granted for copying, reproducing, or duplicating this document only for the legitimate purposes of 3GPP2. No other copying, reproduction, duplication, or distribution is permitted. 10 Simulation Results n Picture Quality Comparison in WCDMA Channel –Error Resilience in H.263: Data Partitioning Slice Mode (Annex K, V) –Channel Characteristics Bit Error Rate: 1.35 x 10 -3 and 1.26 x 10 -3Bit Error Rate: 1.35 x 10 -3 and 1.26 x 10 -3 Average Length of Burst Error: 16 and 17 bitsAverage Length of Burst Error: 16 and 17 bits H.324M @153.6 kbps 108 kbps Video Physical Layer SDUs @153.6 kbps 99 kbps Video H.324M @76.3 kbps 57 kbps Video Physical Layer SDUs @76.3 kbps 51 kbps Video

11. © 1999, Third Generation Partnership Project 2. All Rights Reserved. Permission is granted for copying, reproducing, or duplicating this document only for the legitimate purposes of 3GPP2. No other copying, reproduction, duplication, or distribution is permitted. 11 Conclusive Remarks and Caveats n The H.324 streams have been investigated in WCDMA channels with and without RLP frames n It turned out that: –The smallest overhed due to RLP layer is about 20% of bandwidth of Physical Layer SDU –If every Physical Layer SDU does not contain exactly one H.324 PDU, use of RLP causes increased number of lost H.223 MUX_PDU n Future works include: –At Video Codec Layer: Find out optimum video bandwidth that can incorporate with fixed length window of Physical Layer SDUFind out optimum video bandwidth that can incorporate with fixed length window of Physical Layer SDU –RLP Related Issues: No use of RLP Layer: In-depth investigation of tansparent RLP mode (including bypass mode) in order maximize video quality and minimize number of lost video packets. Investigation of more error robustness in Physical Layer SDU with minimum overhead is also ‘must’No use of RLP Layer: In-depth investigation of tansparent RLP mode (including bypass mode) in order maximize video quality and minimize number of lost video packets. Investigation of more error robustness in Physical Layer SDU with minimum overhead is also ‘must’ Use of RLP Layer: Comparison between fixed and variable size of RLPUse of RLP Layer: Comparison between fixed and variable size of RLP –More realistic simulation environment?