Transcript Evolution of Mobile Video Delivery in the US

Evolution of Mobile
Video Delivery in the
US
Jeff Van Dyke
Director of Technology
Office of the CTO
Dialogic, Inc.
3
Agenda
• Video Quality Factors
• Characteristics of Video Services
• Video Delivery Methods
– 3G-324M
– Video over IP
– Broadcast Video
• Where are We Headed?
4
Video Quality Factors
• Video quality is determined by:
–
–
–
–
Display resolution and size
Codec and level of compression
Frame rate
Network quality of service (QOS)
• Given a specific codec improving the video quality requires
more network bandwidth and better QOS
• High quality video delivery requires high bandwidth and high
QOS from the network.
– None of today’s networks were originally designed for video
• PSTN – low bandwidth, high QOS for voice
• Cellular – low bandwidth, low/medium QOS for voice and low bit rate
data (SMS)
• Internet – medium bandwidth, low inherent QOS for telnet, SMTP and
FTP
5
Display Considerations
• Common screen resolutions
• What can today’s devices
support?
– iPhone - 480 x 320 pixels
– LG Dare - 400 x 240 pixels
• Standard YouTube videos
are QVGA scaled to 480 x
360 pixels
Size
Width
Height
QCIF
QVGA
CIF
VGA
176
320
352
640
144
240
288
480
6
Frame Rate
• The perception of smooth motion requires
approximately 18 frames per second
– In practice mobile video is delivered at 10 – 15 FPS
7
Characteristics of Video Services
• Interactive communication (e.g. video call)
–
–
–
–
–
–
Low complexity
Usually small display sizes (QCIF) with low frame rates (~10)
Compared by users to dedicated video phones or soft phones
Lowest user expectations for quality
Camera placement on handsets is problematic for this application
Bandwidth: < 64 Kb/s
• User and content provider generated streaming video (e.g. YouTube)
–
–
–
–
Low to medium complexity
Larger display sizes (CIF or QVGA) and medium frame rates (15 – 20)
User expectation set by YouTube and similar sites
Bandwidth: 128 Kb/s – 300 Kb/s
• Entertainment video (e.g. broadcast TV shows)
–
–
–
–
Medium to high complexity (e.g. sports broadcasts)
Larger displays (>= CIF and QVGA) and high frame rates (>= 20)
High user expectation based on movie and TV viewing
Bandwidth: > 300 Kb/s
8
Video Delivery - 3G-324M
• Multiplex of control, audio and video in a single 64
Kb/s channel
– Video bandwidth limited to about 45 Kb/s
• Advantages
– Proven, deployed technology
– Low, deterministic transport delay
– Infrastructure available
• Disadvantages
– Limited bandwidth makes it difficult to deliver a
compelling user experience
9
3G-324M Protocol Diagram
3G-324M protocol
(3G TS 26.111)
Video Codec:
H.263
MPEG-4
H.264
Audio Codec:
H.223
Multiplexer/
Demultiplexer
Network
Interface
Video I/O
Equipment
AMR-NB*
G.723.1
AMR-WB*
H.223 Annex A
H.223 Annex B
H.223 Annex C
H.223 Annex D
Audio I/O
Equipment
Data Codec
V.14
LAPM
NSRP
CCSRL
User Data
H.245
Call Control
System
Control
Call Control (via ISUP, ISDN, BICC)
--3G TS 26.112--
10
Video Delivery - IP
• Purely packet based
– Call control via SIP
– Media transport via RTP
• Advantages
– Sufficient bandwidth for delivering video that meets user
expectations
– Flexibility to support various video services
• Disadvantages
– Infrastructure issues
•
•
•
•
IMS networks not fully in place
Quality of service for video transport
IP address limitations (IPv4 vs. IPv6)
Confusion and competition among transport technologies (e.g. HSPA,
WIMAX, LTE)
11
Video Delivery - Broadcast
• Separate network for broadcast TV content
– MediaFLO, DVB-H
• Advantages
– Doesn’t saturate current voice/data spectrum with video
– Can deliver good quality (high frame rates and good QOS)
• Disadvantages
– Completely separate infrastructure
– Can’t be used for interactive communications or user generated web
content
• Do mobile users really want broadcast content?
– Streaming model gives users more control and flexibility
12
Where are We Heading?
• 3G-324M not sufficient to deliver rich video services
• Video over IP breaks through the bandwidth limitation of 3G324M but:
– There are technical issues (e.g. QOS)
– There are economic issues (e.g. infrastructure costs)
– There are many transport alternatives currently in competition with
each other (e.g. HSPA, WIMAX, LTE)
• Lots of noise around broadcast TV but will consumers want
it?
• Opinion: Video over IP will win eventually but it will be a
bumpy ride
– “May you live in interesting times.” – reputed ancient Chinese curse