1

Dr. Lawrence RobertsFounder, Chairman,Anagran

What will follow the InternetWhat will follow the Internet

2

The Beginning of the InternetARPANET became the Internet

• 1965 – MIT- 1st Packet Experiment -Roberts• 1967 - Roberts to ARPA – Designs ARPANET• 1969 – ARPANET Starts – 1st Packet Network• 1971 – ARPANET Grows to 18 nodes

• 1983 – TCP/IP installed on ARPANET – Kahn/Cerf• 1986 – NSF takes over network - NSFNET• 1991 – Internet opened to commercial use

Roberts at MIT Computer

ARPANET1971

3

Internet Early History

1

10

100

1,000

10,000

100,000

1969 1971 1973 1975 1977 1979 1981 1983 1985 1987

Ho

sts

or

Tra

ffic

inb

ps

/10

Hosts

TrafficTCP/IPNCP

EMAILFTP

ICCC Demo

Aloha-Packet Radio

SATNET - Satellite to UK

Spans US

Ethernet

DNSPacketRadioNET

“Internet”Name first used- RFC 675

TCP/IP Design

X.25 – Virtual Circuit standard

Roberts term at ARPA Kahn term at ARPA

Cerf term at ARPA

4

Original Internet DesignIt was designed for Data

File Transfer and Email main activities

Constrained by high cost of memory

– Only Packet Destination Examined

– No Source Checks

– No QoS

– No Security

– Best Effort Only

– Voice Considered

– Video not feasible

ARPANET1971

Not much change since then

5

Voice Totally moving to packets

– Low loss, low delay required

Video Totally moving to packets

– Low loss, low delay jitter required

Emergency Services No Preference Priority

Broadband Edge Must control Edge Traffic

P2P utilizes TCP unfairness – multiple flows– Congests network – 5% of users take 80% of capacity

Changing Use of InternetMajor changes in Network Use

6

Network Change Required

Video Quality Poor

– Packet Loss, Delay Jitter, VoD can be too slow

– Cannot determine if quality path exists

– Separate Networks needed for Voice, Data, and Video

Fairness Missing – Multi-flow applications overload Internet

– Currently P2P is reducing throughput of normal users

High Cost, Power, Size – Major reduction possible

Emergency Services - Need Preference Priorities

Security – Need User Authorization and Source Checking

– Must know who you are connected to

7

Where should we be in the next decade

Everyone has mobile wireless Internet device

– Voice and video person-to-person communication

– Knowledge and Entertainment access – much as video

– Location knowledgeable (GPS) – locate anything

– Payment capable, Access capable (keys), ID capable

– Remote control for anything – TV, computer, etc.

– Secured for individual – can identify individual to net

Larger computers secured from wireless device

8

Where should Internet be in the next decade

2008 2018

% World Population On-Line 22% 99%

Total Traffic PB/month 3,200 191,000

Traffic per User GB/month 2.2 26

GB/mo/user Developed areas 2.7 156

GB/mo/user Less Dev. areas 0.5 3

This means people in less developed areas will have morecapacity than is available in developed areas today!

Also, users in developed areas could see 3-10 hours of video perday (HD or SD)

Requires a 60 times increase in capacity (Moore’s Law increase)

9

TCP - Network StabilityTCP has Allowed the Network to Scale

TCP and Network Equipment keep a balance

This balance keeps the network stable– TCP speeds up until a packet lost, then slows down

– Network drops packets if overloaded

Result:– TCP grows to fill network

– Network then loses random packets

– All traffic impacted by packet losses, random rate changes

– However, system is basically stable

TCP Network

10

A New Alternative - Flow ManagementTCP or the Network need to Change

Network Equipment now drops random packets– All traffic suffers delay and jitter if there is any congestion

– Voice & Video do not slow down but still lose packets

– Data flows often lose several packets and stall

Flow Behavior Management - a new control alternative– Control the rate of each TCP flow individually

– Smoothly adjust the TCP rates to fill the desired capacity

– Flows can be put into virtual segments called classes

– Priority can adjust the relative class rates as desired

– Assured Rate option allows bandwidth guarantees for classes

Replacing random drops with rate control:– Network Stability is maintained

– All traffic moves smoothly without random loss

– Voice & Video flows cleanly with no loss or delay jitter

– Unfairness can be eliminated

11

Multi-Flow Unfairness Solution

12

Fairness in the beginning

A flow was a file transfer, or a voice call

The voice network had 1 flow per user

– All flows were equal (except for 911)

– Early networking was mainly terminal tocomputer

– Again we had 1 flow (each way) per user

– No long term analysis was done on fairness

It was obvious that under congestion:

Users are equalthus

Equal Capacity per Flowwas the design

13

Internet Traffic Grown 1012 since 1970

In 1999 P2P applications discovered using multiple flows could give themmore capacity and their traffic moved up to 80% of the network capacity

World Internet Traffic - History

0.000000001

0.00000001

0.0000001

0.000001

0.00001

0.0001

0.001

0.01

0.1

1

10

100

1000

10000

1970 1975 1980 1985 1990 1995 2000 2005

Pe

taB

yte

sp

er

mo

nth

Normal Traffic

P2P

TCP

ARPANET NSFNET COMMERICAL

Double each year

Electronics – Double every 18 months

14

Where is the Internet now?

The Internet is still equal capacity per flow under congestion

Computers, not users, now generate flows

– Any process can use any number of flows

– P2P takes advantage of this using 10-1000 flows

Congestion typically occurs at the Internet edge

– Here, many users share a common capacity pool

– TCP generally expands until congestion occurs

– This forces equal capacity per flow

– Then the number of flows determines each users capacity

The result is therefore unfair to users who paid the same

P2P FTP

15

Internet Traffic Recently

Since 2004, total traffic has increased 64% per year, about Moore’s Law– P2P has increased 91% per year – Consuming most of the capacity growth

– Normal traffic has only increased 22% per year –Significantly slowdown from past

Since P2P slows other traffic 5:1, users can only do 1/5 as much

This may account for the normal traffic being about 1/5 what it should bewith normal growth

World Internet Traffic - History

0

500

1000

1500

2000

2500

3000

3500

2000 2002 2004 2006 2008

Pe

taB

yte

sp

er

mo

nth

Normal Traffic

P2P

16

A New Fairness Rule is Needed

Inequity in TCP/IP – Currently equal capacity per flow– P2P has taken advantage of this, using 10-500 flows– This gives the 5% P2P users 80-95% of the capacity– P2P does not know when to stop until it sees congestion

Instead we should give equal capacity for equal pay– This is simply a revised equality rule – similar users get equal capacity– This tracks with what we pay– If network assures all similar users get equal service, file sharing will find the

best equitable method – perhaps slack time and local hosts

This is a major worldwide problem– P2P is not bad, it can be quite effective– But, without revised fairness, multi-flow applications can take capacity away

from other users, dramatically slowing their network use– It then becomes an arms race – who can use the most flows

17

ISP Problem - Dow nstream Capacity Use

6 Mbps Dow nstream , 3 Mbps upstream

0

20,000

40,000

60,000

80,000

100,000

5.0% 4.5% 4.0% 3.5% 3.0% 2.5% 2.0% 1.5% 1.0% 0.5% 0.0%% Users with P2P

Tra

ffic

Mb

ps

The Impact on ISP’s

Uncontrolled P2P growth slows normal traffic, users upset

ISP’s cannot win by adding capacity – P2P applications use it

DPI has been tried but is less than effective– Also targeting only P2P without fairness censured by FCC

Unless a fair solution is deployed, service will deteriorate– All applications will consider using multi-flows

– This leads to a major collapse of service levels world wide

Asymmetric service – Upstream is congested by P2P and downstream limited by slow ACK’s

No Regulation DPI

Equalize

Average Users

P2P Users

Wasted Capacity

18

Average User Data Rate Downstream1 Gbps Internet Feed, 100 Mbps User Acces

0

100

200

300

400

500

600

700

800

900

1,000

0.0%0.5%1.0%1.5%2.0%2.5%3.0%3.5%4.0%4.5%5.0%% Users with P2P

User

Data

Rate

-K

bp

s

The Impact on Average Users

Uncontrolled P2P impacts normal users if any congestion

– P2P generally expands until it sees congestion

– Congestion is not normally that visible – 50% average is sufficient

– The short overload peaks cause delay, loss and flow slowdown

Average Users flows run at speed of P2P flows

– P2P can use lots of very low speed flows

No Regulation DPI

Equalization

Result – Average user only gets 30 Kbps per flow – 3% of expected

19

We Must Resolve Unfairness Soon

Any application that uses more flows gets more capacity

This capacity is often taken at the expense of other users

P2P started the trend to use multiple flows - now up to 500

However, to compete, other applications will use multi-flows

– FTP and HTTP could go to using 1000 flows to compete

– Soon all ports will be used and NAT will fail

Also, UDP without flow control could be used to compete

In either case the Internet will be in trouble

20

Revised Fairness Criteria

We need not stick with “Equal Capacity per Flow”

That was our concept in the early days – one flow per person

Now, computers generate the flows, not humans

We provide or sell peak capacity to each user

We should provide “Equal Capacity per User” when overloaded– In general: Equal Capacity for Equal Pay

Today’s technology allows rate control of every flow– For multi-flow unfairness – divide capacity equally between users

– P2P still works, privacy maintained, and most users run 5-10 times faster

P2P FTP

Equal Capacity per User

21

Why is it Important to Change Fairness Rule?

P2P is attractive and growing rapidly

It cannot determine its fair share itself

The network must provide the fair boundary

Without fairness, normal users will slow down and stall

Multi-flow applications will be misled on economics

– Today most P2P users believe their peak capacity is theirs

– They do not realize they are slowing down other users

– The economics of file transfer are thus badly misjudged

– This leads to globally un-economic product decisions

User equality will lead to economic use of communications

22

Flow Behavior Management – Why now?

Routers, WAN Optimizers, & DPI depend on processing

– They do packet classification – processing is expensive

Memory cost has come down faster than processing cost

Flow Behavior Manager is memory based

– Keeping knowledge about 4 M flows and 100 K users

Allows it to support four 10 Gbps trunks in 1 RU

23

Summary – Flow Management

Flow Management can easily be added to any network

Allows fairness change to “equal capacity for equal pay”

– This will keep multi-flow applications from drowning the net

Eliminates delay and jitter

– TCP data throughput improved

– Voice and video quality moved from fair to great

Allows virtual separation of traffic into rate controlled classes

Eliminates packet loss and delay with overload

Provides complete flow records for all traffic

– Greatly improves understanding of network traffic & problems