new ip and market opportunities · tcp/ip standardized arpanet ceased www, cern tim berners-lee...
TRANSCRIPT
Futurewei Technologies, Inc.
New IP and Market Opportunities
Richard Li
Chief Scientist, Network TechnologiesFuturewei, USA
IEEE International Conference on
High Performance Switching and Routing (HPSR) 2020
May 11-14, 2020
2
Acknowledgement and Disclaimer
• New IP is under research and development by research scientists and professional
engineers across different countries and different organizations
• Some results of this talk may have appeared in ITU, IEEE and ACM publications
Page 3
The Internet has been very successful! But can it sustain all new applications
in decades to come? Can we rest here and take the Internet as it is forever?
Packet Switching Landmark Projects Commercialization
Paul Baran Leonard Kleinrock
Inventor Packet Switching
Inventor Packet Switching
Don Davis, NPL, UK
Frank Heart &Team,
BBN IMP Spec
Queen Elizabeth II
Sends her first email.
IEEE, "A Protocol for Packet Network Intercommunication” , 1974
Vinton Cerf and Bob Kahn
TCP/IP Standardized
ARPANET ceased
WWW, CERN
Tim Berners-LeeCyclades at INRIA,
France, 1971-1979,
Louis Pouzin
Conceptual Experimental Standardization Explooooosion1961
1965
1968
1974
1976
1979
1980
1989
2000
4
Agenda
• Market Opportunities and New Requirements
▪ Need for high-precision communications for industrial machine-type communications
▪ Need for free choice addressing in the age of ManyNets for economy and democracy
▪ Need for qualitative communications for very large volumetric communications
• Current IP
▪ Design decisions: e.g. Fixed Addressing, Inherently Best Effort
▪ Packet Loss, Retransmission, and Host-based Congestion Control
▪ Cerf-Kahn-Mathis Equation
• New IP
▪ Free-Choice Addressing
▪ High-Precision Communications
▪ Qualitative Communications
Page 5
ITU-T Focus Group on Network 2030
https://www.itu.int/en/ITU-T/focusgroups/net2030/Pages/default.aspx
RadiocommunicationITU General Secretariat Standardization Development ITU Telecom Members’ Zone Join ITU
EventsAbout ITU-T Study Groups All Groups Join ITU-T Standards Resources Regional Presence BSG
Study new capabilities
of networks for the year
2030 and beyond
Explore new concepts,
principles, mechanisms,
and architectures
Review Protocol Stack,
and outline
future directions
Identify future use
cases and new
requirements
Establish
July 16 – 27, 2018
Geneva
1st Meeting
October 2 – 4, 2018
New York
2nd Meeting
December 18 – 21, 2018
Hong Kong
3rd Meeting
February 18-20, 2019
London
4th Meeting
May 21-23, 2019
St. Petersburg
5th Meeting
October 14-18, 2019
Geneva
6th Meeting
January 13-15, 2020
Lisbon
6
Driverless Vehicles and Remote Operations
Hazardous environment
Page 7
Latency and Packet Loss vs Safety of Life
Collision-Avoidance Distance
60 km/hour = 16.7m/sec.
distance = 16.7m/sec x 60ms = 1m
30 km/hour = 8.3m/sec.
distance = 8.4m/sec x 60ms = 0.5m
Autonomous Driving Cloud Driving
Local Driving
Decision Point
Local
Sensory Input
Remote
Contextual Advisory
Local
Sensory InputRemote Driving
Decision Point
Latency
Packet Loss Packet LossLatency
Packet Loss is a Serious Issue
Page 8
Industrial Machine-Type Communications and Control
Cloudified
PLC
Major Challenges:
▪ Latency
▪ Availability
▪ Elasticity
▪ Security
▪ Usability
Source: Texas Instruments
Page 9
ManyNets: Embracing Diversity, Variety, Economy, and Autonomy
Non-IP Networks(Growing market segment)
Private Global Backbones(No Need for Internet Transit)
Emerging Satellite Constellations(Global Broadband connectivity for 4 billion people
who are not connected to any network today)
Spread Networks
OneWeb
Starlink
1) Geoff Huston, The Death of Transit and the Future Internet, Keynote Speech at 2nd ITU-T Workshop on Network 2030, Hong Kong, Dec. 2018
2) Mostafa Ammar, Service-Infrastructure Cycle, Ossification, and the Fragmentation of the Internet, Keynote Speech at 3rd ITU-T Workshop on Network 2030, London, UK, Feb. 2019
Page 10
Space Internet
Public Transit Backbone
Private Transit Backbone
Private Transit Backbone
LaserLT
E Groun
d
Statio
n
LEOs or MEOs
Routing in spaceCo. Support Scale
Starlink SpaceX, Google 4K by 2019, then 12K
Oneweb Blue Origin (Bezos), Virgin Orbit
650 by 2019
Boeing Apple (spec) 2956, 1350 in 6 yrs
O3Nb Virgin group, SES 400
CASIC China 300 (54 trial)
Distances Bandwidth delay
(LEO)900-1200 KM
1—200 Gbps 35ms
(MEO)~2000 KM
1-200 Gbps ~60ms
Space to space ~100 KM – ~Tbps~1000 KM ~10 Gbps
How do we integrate terrestrial and space internet?
Source: Internet
11
We are in the dawn of a big change from digital society to holographic society
Holographic Society
Healthcare/ Surgery Agriculture
Entertainment
Shopping
Industry
Education
Tourism
HapticTechnology
Tactile Internet
Digital Senses
Non-contact Force
feedback
Virtual Movement
Remote Touch
Remote Control
Fast Response High Precision
Sight Hearing Touch Smell Taste
Page 12
Holograms and Holographic Type Communications
Throughput
4K/8K HD AR/VR Hologram
35Mbps~140Mbps 25Mbps~20Gbps 10 Gbps~10 Tbps
band
width
band
width
Synchronization of parallel streams
4K/8K HD VR/AR Hologram
Audio/Video(2) Multiple tiles (12)~thousands
(view-angles)
streams streams
20” wide
6’0
” t
all
Raw data; no optimization or compression. color, FP
(full parallax), 30 fps
4”
4”
Dimensions Raw Data
Tile 4 x 4 inches 30 Gbps
Human 72 x 20 inch 4.32 Tbps
(reference: 3D Holographic Display and Its Data
Transmission Requirement,
10.1109/IPOC.2011.6122872), derived from for
‘Holographic three-dimensional telepresence’; N.
Peyghambarian, University of Arizona)
360 degrees of view
6 degrees of freedom
/VR
Image
Capture
Framing
Streaming
DecodingDisplay
Encoding
Network
Transport
Motion-to-Photon Time: Total 20 ms
5-7 ms
How do we transport very large volumetric data?
Page 13
Attaching Digital Senses to Holograms
Media Evolution
Hologram
AR/VR
Video
Audio
Image
Text
1T/s
D
1ms
1G/s
D
17ms
100M/s 33ms
64k/s 50ms
IEEE Digital Senses Initiative
Coverage Model
Privacy
Identity
Security
Trust
Ethics
Transforming Industries
Capture
Reproduce
Synthesize
Undersized
Respond
Public Awareness
User Acceptance
Content Richness
Ecosystem Readiness
Smart Robots Wearables
and more… Consumer
Healthcare
Human or
Machine
Sight
Touch
Other
Senses
TasteSmell
Hearing
Augmented Reality
Virtual
Reality
Human
Augmentation
Page 14
IP/MPLS in Mobile Backhaul Networks
App (user)
TCP (user)
IP (user)
PDCP
RLC
MAC
PHY
App (user)
TCP (user)
IP (user)
PDCP
RLC
MAC
PHY
App (user)
TCP (user)
IP (user)
GTP-U (S1)
UDP (Nwk)
IP (Nwk)
IP/MPLS
Backhaul
Eth/Nwk
App (user)
TCP (user)
IP (user)
GTP-U (S1)
UDP (Nwk)
IP (Nwk)
IP/MPLS
Backhaul
Eth/Nwk
App (user)
TCP (user)
IP (user)
No guarantee on E2E throughput
and latency by current TCP/IP
Inefficient
retransmission
Inefficient use of protocols
Tunnels over tunnels
Repeating header fields
Radio retransmissions
are not synchronized
with TCP flow control
Retransmit wasteful
packets
Cellular network Fixed, IP based wireline network
Not suitable for mMTC and uRLLC
Low efficient user payload, unsuitable for
mMTC and short messages
No E2E QoS, unsuitable for uRLLC
How do we provide end-to-end service guarantee for 5G/B5G/6G-enabled applications?
15
Market Opportunity and New Requirements
▪ Entropy-Based Communications▪ Semantics-Based Communications▪ Knowledge-Based Communications
▪ Industrial Machine-type Communications▪ Industrial Control and Manufacturing▪ Industrial Internet▪ Tactile Internet▪ Autonomous Driving▪ Cloud Driving▪ Digital Twin▪ Holographic Twin▪ Holographic Society▪ Network and Computing Convergence▪ Smart City▪ Smart Agriculture▪ Smart healthcare▪ ManyNets▪ Space-Terrestrial Integrated Network ▪ Private Internet▪ Non-IP Networks
▪ Premium services▪ Adaptable Responsive▪ User Programmable▪ Business Easing▪ Compute Power Network
▪ In-Time Guaranteed Transport▪ On-Time Guaranteed Transport▪ Lossless Transport
Free-Choice Addressing
Qualitative Communications
Moving Beyond Native Multiplexing
High-Precision Communications
▪ Mix and Match▪ Industrial Machine-Type Addressing▪ Machine ID located in a Workshop▪ Power-Efficient Addressing▪ ManyNets
16
Agenda
• Market Drivers and New Requirements
▪ Need for high-precision communications for industrial machine-type communications
▪ Need for free choice addressing in the age of ManyNets for economy and democracy
▪ Need for qualitative communications for very large volumetric communications
• Current IP
▪ Design decisions: e.g. Fixed Addressing, Inherently Best Effort
▪ Packet Loss, Retransmission, and Host-based Congestion Control
▪ Cerf-Kahn-Mathis Equation
• New IP
▪ Free-Choice Addressing
▪ High-Precision Communications
▪ Qualitative Communications
▪ Better Traffic Engineering
17
Router and Packet Statistical Multiplexing
Egress CardIngress Card Switch Fabric
Core
DDR接口
Serd
es IO
Serd
es IO
DDR FIB
MAC NP TM/FIC
FIB
Burst
Maximal Resource Utilization Inherently Best Effort Forwarding
Header Processing Delay Queueing DelayTransmission
Delay
Packets are dropped when
buffer is fullPackets are dropped when
buffer is full
Page 18Futurewei Technologies, Inc.
IP is an artifact (made up of a few design decisions)
One Size Fits All(Fixed Addressing)
Statistical Multiplexing Capabilities and Services
Best Effort. Default and Most Popular
DiffServ (on a per-hop basis)
Traffic Engineering
▪ Traffic Steering (Explicit Path)
▪ Minimal Bandwidth Guarantee
▪ Fast Re-Route
Maximize network utilization: Matching
traffic demand to available capacity
One common network layer to
connect everything globally
Packet switching: 59 years (2020-1961)
TCP/IP (Cerf’s paper): 46 years (2020-1974)
IPv4 (RFC 791): 39 years (2020-1981)
IPv6 (RFC 1883): 25 years (2020-1995)
MPLS (RFC 3031): 19 years (2020-2001)
As of now, total RFCs: 8778
(If an engineer studies 5RFCs a week, it takes 33.76 years to finish all them)
Page 19
Transport of Traffic: Packet Loss and Retransmission
𝐓𝐡𝐫𝐨𝐮𝐠𝐡𝐩𝐮𝐭 ≤ 𝐦𝐢𝐧(𝐁𝐖,𝐖𝐢𝐧𝐝𝐨𝐰𝐒𝐢𝐳𝐞
𝐑𝐓𝐓,𝐌𝐒𝐒
𝐑𝐓𝐓×𝐂
𝛒 )
Packet Loss
Congestion ControlRetransmissionTransmission
Page 20
Transport of Traffic: Improvements and Enhancements
See ACM SIGCOMM 2019 Keynote (Mark Handley) for his personal achievements
Congestion Control
In Host
Rate Limiting
Multi-Path
Pacing
Scheduling
In Network
Load Balancing ECMP
Flow LB
Packet LB
Flowlet LBRerouting
Priority Queuing
Back Pressure
It is an eternal and never-stopping topic in ACM Sigcomm
Google Scholar returns 13,200 publications on “TCP
Congestion Control”
No one size fits all Time to read: 36 years
Read one paper a day: 13200/365
Page 21
Transport of Traffic: Extrapolation from Cerf-Kahn-Mathis Equation
The result may vary with CPUs, Links, Buffers, etc, but throughput, latency and packet loss are coupled closely together
You can’t make omelet without breaking eggs!
738.7 (Mbps)
522.3
233.6165.2
73.9 52.2 23.4 (Mbps)
0.0
100.0
200.0
300.0
400.0
500.0
600.0
700.0
800.0
0.001% 0.002% 0.010% 0.020% 0.100% 0.200% 1.000%
TCP
Th
rou
ghp
ut
(Mb
ps)
Packet Loss Ratio
TCP Throughput (Mbps) drops as Packet Loss Rate increases, Guaranteed RTT= 5ms
307.80 (us)
217.64
97.3368.83
30.78 21.76 9.73 (us)
0.00
50.00
100.00
150.00
200.00
250.00
300.00
350.00
0.001% 0.002% 0.010% 0.020% 0.100% 0.200% 1.000%
RTT
(u
s)
Packet Loss Ratio
Ultra-low Latency (us) demands as Packet Loss Ratio increases, Guaranteed Throughput = 12Gbps
If you lose 1 packet per 10,000 packets, your latency is 0.1 ms in order to yield a throughput of 12 Gbps
• improvements exist, but the nature of the correlation between throughput, packet loss and latency keeps similar.
22
Agenda
• Market Drivers and New Requirements
▪ Need for high-precision communications for industrial machine-type communications
▪ Need for free choice addressing in the age of ManyNets for economy and democracy
▪ Need for qualitative communications for very large volumetric communications
• Current IP
▪ Design decisions: e.g. Fixed Addressing, Inherently Best Effort
▪ Packet Loss, Retransmission, and Host-based Congestion Control
▪ Cerf-Kahn-Mathis Equation
• New IP
▪ Free-Choice Addressing
▪ High-Precision Communications
▪ Qualitative Communications
▪ Better Traffic Engineering
Page 23
What are new in New IP?
▪ Fixed Addressing
▪ Best-Effort
▪ DiffServ
▪ Traffic Engineering
▪ Quantitative Communications
IPv4 and IPv6
▪ High-Precision Communications
▪ In-Time Guaranteed Transport
▪ On-Time Guaranteed Transport
▪ Lossless Networking
▪ Free-Choice Addressing
▪ Mix-and-Match
▪ Industrial Machine-Type Addressing
▪ Power-Efficient Addressing
▪ Domain-Specific Addressing in ManyNets
▪ Qualitative Communications▪ Entropy-Based Communications
▪ Semantics-Based Communications
▪ Knowledge-Based Communications
▪ Inherited and backward compatible
▪ Best-Effort
▪ DiffServ
▪ Traffic Engineering
▪ Quantitative Communications
New IP
Page 24
What Can We Learn from Postal Services?
TrackableMeasurement and Telemetry
AssurableGuaranteed
BillableNew Source of Revenue
IP datagram used to be called “letter-gram”, and it enjoys many analogies to postal letters.
Today’s postal services are no longer the postal services of 50 years ago.
Postal services have greatly evolved, but IP hasn’t!
CustomizableProgrammability
Customize Delivery Time
Deliver to Another Address
Hold at FedEx Location
Sign for a Package
Provide Delivery Instructions
Sign for a Package
Page 25
Imagine a New IP Packet as a FedEx-like Datagram
IP Header Contract User Payload
A packet carries a contract between an application and the network.
The network and routers fulfill the contract.
2 The packet arrives at 35ms sharp, no sooner no later
3 It requires a throughput of 12 Gbps
4 No packet loss. If lost, you get a compensation.
5 Track it
1 The packet arrives in 35ms
FedEx-like IP Datagram
Ref: Richard Li, et al, A New Framework and Protocol for Future Networking Applications, ACM Sigcomm 2018 NEAT Workshop, Budapest, Hungary, August 2018
2 The package arrives at 9:00am next day
3 The weight is 12kg
4 No package loss. If lost, you get a refund of $$$.
5 Status track
1 The package arrives in 1 day
FedEx Package
Page 26
What Can a Contract Do in New IP?
High-Precision
Communications
❖ Lossless networking
❖ Throughout guarantee
❖ Latency Guarantee
(in-time, on-time,
coordinated)
User Network
Interface (UNI)
In-Band
Signaling
High-Precision
Telemetry
User-Defined
Networking
Application-Specific
Programmability
Preferred Path Routing
(PPR). . .
Page 27
E Pluribus Unum: Mix-and-Match, Flexible Addressing System
OAM from IPv4 hosts
to non-IPv4 nodes
IPv4 host directly talks with
IPv6 server, and vice versa
Satellites of different
companies talk with each other
New IP
LAN
New IP
Backbone Network
IoT
Device
#1
IoT Device #2
GW3Prefix P, 104 bits Server
31
Raw
Data
DIP: y (24 bit)
SIP: x (24 bit)
DIP:D (128 bit)
SIP: x (24 bit) DIP: D (128 bit)
SIP: P + x (128 bit)
2
2
Raw Data4
5
Comparison
In current IP, the source and destination addresses are of the same type, but cannot be mixed
Page 28
Moving SatellitesTopology-based & Geography-based Addressing
(10, 10)
(12, 10)
(12, 8)
(14, 10)
(16, 8)
(18, 10)
(20, 10)
AS 2
New IP Packet Header
IP ADDR 2:2::2
Geo ADDR (20, 10)
payload
New IP Packet Header
IP ADDR 2:2::2
payload
AS 1
Sender
AS 3
Receiver
New IP Packet Header
IP ADDR 2:2::2
payload
IP PrefixGeo
Address
3:3::3/32 (30, 15)
2:2::2/32 (20, 10)
2:2::2/32
Page 29
Receiver
Current IP: Quantitative Communications
Sender
Packet Packet
Packet Corrupted Packet
What is received What is sent
Every bit and byte has the same significance to
routers/switches
Good for
• File/Document Transfer
• Banking, Shopping
Overkill for some applications
• Holograms
• Disaster Environment
Syntax
=
Page 30
New IP: Qualitative Communications
Sender Receiver
Qualitative Packet
Ref: A Framework for Qualitative Communications using Big Packet Protocol, ACM Sigcomm 2019 NEAT
Workshop, Beijing, August 19, 2019. Available at: https://dl.acm.org/citation.cfm?id=3342201
Noisy Link
Congested Node
Congested Node
Congested Node
What is received What is maximally meant
In payload, bits and bytes are not equally
significant. Instead, they are differential
in their entropies
Good for
• Large volume of image-like data
• Holographic type communications
• Media with digital senses
• Disaster Environment
Less significant bits and bytes may be dropped
Partial or degraded, yet useful, packets may be
repaired and recovered before being rendered
Semantics
=
31
Progression and Evolution
Header Contract User Payload
Header User PayloadIPv4/IPv6
Enhancement with “Contract”
Shipping Directive Contract User Payload
Header Evolution Payload Evolution
New IP
ManyNets
Free-Choice AddressingVLV&TIC
Qualitative Communications
BBE-Beyond Best Effort
HPC- High Precision Communications
VLV&TIC
User-Defined Networking
Better Traffic Engineering
Page 32
New IP
Industrial Use(Stringent KPI)
+
Consumer Use
Autonomous System
IPv4/IPv6
Some
Industrial Use(Non-Stringent KPI)
+
Consumer Use
Autonomous System
IPv4/IPv6
Existing
and
Most Networks
Autonomous System
Deployment under Consideration
❖ New IP can be deployed within an autonomous
system (AS) in the existing framework.
❖ New IP is especially good for industrial machine-
type communications that often require high-
precision KPI, ultra-low latency, and whose
devices may have addresses other than
IPv4/IPv6.
❖ Examples of New IP Trial Deployment:
❖ Metropolitan Networks
❖ Mobile Backhaul for 5G/B5G/6G
❖ Industrial Manufacturing Facility
❖ Industrial Park and Campus
❖ Autonomous Driving and Remote Cloud Driving
interconnected
33
Concluding Remarks▪ While the existing IP is pretty much like classical postal letters, New IP is more like modern
FedEx-style packages.
▪ New IP is motivated by
• Industrial machine-type communications (Industry 4.0, Industrial Internet, Smart World)
• Mobile Backhaul Transport for 5G/B5G/6G
• Emerging Industry Verticals (driverless vehicles)
• ITU-T Network 2030
▪ Research and empirical results have been published in ITU, IEEE, and ACM. Some industrial
manufacturing-related companies, service providers and network operators have shown their
interest in New IP
▪ Some prototypes have been implemented by different organizations in different countries.
However, Standards Developing Organizations (SDO) have yet to standardize it.
▪ New IP has been, unfortunately, politicized in media, and misinformation about it has been
spread
34
Thank You!