Error Resilient Internet Video Transmission

Presenters : Ciro Noronha - Cobalt Digital

Year : 2019

dlp_document_download : C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 Error Resilient Internet Video Transmission Ciro A. Noronha, Ph.D. Director of Technology, Compression Systems Cobalt Digital Juliana W. Noronha University of California, Davis C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 Motivation • There are a number of protocols in use today to transport Video over IP. • Since the "I" in IP stands for "Internet", the Internet can (potentially) be used to transport Video over IP.  Low - cost contribution links!! • However, not all Video over IP protocols are suitable for transporting Video on the Internet because:  The Internet drops packets  Video over IP is compressed and needs every bit  Video over IP cannot take packet drops  The Video over IP protocol has to handle this issue C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 So, where are packets really lost? Router Congestion! C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 What is an "acceptable" packet loss? • Video compression works by removing redundancy from the content — Every bit of compressed video is very important • There is a simple way to look at the effect of packet loss: — Assume that every packet that is dropped by the network causes a noticeable glitch in the video  A block of packets dropped together causes one glitch — Decide how many glitches per (day/hour/minute) is acceptable to you C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 Some numbers Dropping one packet in Produces a glitch every 1,000 2.6 seconds 10,000 26 seconds 100,000 4 minutes 23 seconds 1,000,000 44 minutes 10,000,000 7 hours 19 minutes Assume a 4 Mb/s stream, with 1316 - byte packets In order to achieve reliable operation on the Internet, a network protocol is needed to "recover" in some way the packets that have been lost. C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 Protocols Considered • SMPTE - 2022 FEC — Transmit redundant information with the packets — Losses may be recovered from received packets and redundant information • Retransmission (ARQ) — If a packet is lost, receiver will request a retransmission C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 RTP plus SMPTE - 2022 FEC • Basic idea: — Transmit the video using RTP  That gets you timestamps and sequence numbers  Sequence numbers let you know when packets were dropped — Transmit "extra" FEC packets — If packets are lost in the network, it may be possible to rebuild them from the received packets and FEC packets:  For each N packets send 1 FEC packets  If there is one loss in this set of N+1 packets, it can be corrected — Use a matrix arrangement to deal with burst losses C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 Some FEC Numbers Columns Rows Recovery Capability Overhead Latency @ 2 Mb/s Latency @ 10 Mb/s 5 5 5 pkts every 25 20% 263 ms 53 ms 10 5 10 pkts every 50 20% 526 ms 105 ms 20 5 20 pkts every 100 20% 1052 ms 211 ms 10 10 10 pkts every 100 10% 1052 ms 211 ms C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 ARQ • ARQ stands for: — A utomatic R epeat re Q uest — A utomatic R epeat Q uery • This is the generic name for a number of retransmission strategies in the face of packet loss — Standard TCP uses a couple of ARQ variants • In video transmission, the most useful variant is "Selective Retransmission" (NACK - based) — If you don’t hear from me, everything is OK — If I miss anything, I let you know and you resend just that C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 ARQ Illustration Sender Internet Receiver Network Round - trip Delay Transmitted packets are saved for possible retransmission If the buffers are big enough, multiple retransmissions of the same packet can be supported C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 Comparison of FEC and ARQ • FEC and ARQ have "decent" latency (typically 1 second or less) — May be acceptable for some forms of live contribution • How do these two protocols compare? — Statistical models — Testing on a simulated network — Measurement data C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 A little probability and statistics… • Assume independent loss probability for each transmitted packet (binomial distribution) • Calculate the rate of packets still lost after correction with statistical analysis • This allows us to theoretically compare the performance of the various protocols and settings • Our variables are: R = number of requests (ARQ) N = number of packets per row (FEC) M = number of packets per column (FEC) IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 Network Simulator • Windows - based network simulator custom - built for this test • Random drops, random burst drop size • Test scenario: — End - to - end real - time video — Select max burst loss — Increase loss percentage until video is "not watchable" (subjective) Encoder Decoder Network Simulator C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 Simulator Results C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 Field Test Data • Locations: • Santa Clara, CA • Champaign, IL • ISP: Comcast • Network Round Trip Time: 75 ms • Number of hops: 12 • Target bit rate: 3 Mb/s • Equipment: • 9223 Encoder • 9990 - DEC Decoder C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 RTP/SMPTE - 2022 Test Data Test Duration 65 hours Test Start Date 05/19/17, 3:50PM Network Packet Loss 0.0158% Corrected Packet Loss 0.0027% Correction Ratio 83% Bandwidth Overhead 25% Network Glitch Interval 1 minute 13 seconds Corrected Glitch Interval 7 minutes 12 seconds Protocol Latency 702 ms Parameters: 20x5 matrix, row and column C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 RTP/ARQ Test Data Test Duration 169 hours Test Start Date 05/24/17, 12:30PM Network Packet Loss 0.0257% Corrected Packet Loss 0.000078% Correction Ratio 99.7% Bandwidth Overhead 0.027% Network Glitch Interval 46 seconds Corrected Glitch Interval 4 hours 7 minutes Protocol Latency 400 ms Parameters: up to 4 retries allowed C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 FEC/ARQ Comparison Scaling: • Latency — ARQ latency is constant — FEC latency decreases with increasing bit rate • Overhead — ARQ overhead will increase with packet loss — FEC overhead is constant Parameter 2022 FEC ARQ Network Packet Loss 0.0158% 0.0257% Corrected Packet Loss 0.0027% 0.000078% Correction Ratio 83% 99.7% Bandwidth Overhead 25% 0.027% Network Glitch Interval 1 minute 13 seconds 46 seconds Corrected Glitch Interval 7 minutes 12 seconds 4 hours 7 minutes Protocol Latency 702 ms 400 ms C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 ARQ Standardization Status • The Video Services Forum (VSF) started a group around NAB 2017 to standardize a low - latency video transport protocol over the Internet • RIST : R eliable I nternet S tream T ransport • ARQ has been selected as the base protocol • VSF TR - 06 - 1 was published October 2018 C U R AT E D BY IP SHOWCASE THEATER AT NAB – APRIL 8 - 11, 2019 Thank You Ciro A. Noronha, Ph.D. Cobalt Digital ciro.noronha@cobaltdigital.com +1 650 208 - 0605