SlideShare uma empresa Scribd logo
1 de 35
Fundamentals of
Network Performance Engineering
Dr Neil Davies
Predictable Network Solutions Ltd
Peter Thompson
Predictable Network Solutions Ltd
Martin Geddes
Martin Geddes Consulting Ltd
© 2013 All Rights Reserved
PREDICTABLE
NETWORK
SOLUTIONS
Dr Neil Davies
Co-founder, Predictable Network Solutions Ltd
Ex: University of Bristol (23 years).
Former technical head of joint university/research institute (SRF/PACT).
Peter Thompson
CTO, Predictable Network Solutions Ltd
Ex: GoS Networks, U4EA, SGS-Thomson, INMOS & Universities of Bristol,
Warwick and Cambridge.
Authority on technical and commercial issues of converged networking.
Martin Geddes
Founder, Martin Geddes Consulting Ltd
Ex: BT, Telco 2.0, Sprint, Oracle, Oxford University.
Thought leader on future of telecommunications industry.
PREDICTABLE
NETWORK
SOLUTIONS
Dr Neil Davies
Co-founder, Predictable Network Solutions Ltd
Ex: University of Bristol (23 years).
Former technical head of joint university/research institute (SRF/PACT).
Peter Thompson
CTO, Predictable Network Solutions Ltd
Ex: GoS Networks, U4EA, SGS-Thomson, INMOS & Universities of Bristol,
Warwick and Cambridge.
Authority on technical and commercial issues of converged networking.
Martin Geddes
Founder, Martin Geddes Consulting Ltd
Ex: BT, Telco 2.0, Sprint, Oracle, Oxford University.
Thought leader on future of telecommunications industry.
PREDICTABLE
NETWORK
SOLUTIONS
The only ex-ante network performance
engineering company in the world.
• New mathematical performance techniques.
• Performance assessment methodology.
• World’s first network contention management
solution.
Consultancy on the future of
telecoms and the Internet.
• Business model innovation.
• Technology & product ideation.
• Organisation development.
• Public & private workshops.
This presentation is taken from the content for
Fundamentals of Network
Performance Workshop
For information on locations and
timing of public events visit
www.sustainablebroadband.com
PREDICTABLE
NETWORK
SOLUTIONS
Overview
What is
“Network
Performance
Engineering”?
3 Basic
Concepts
G, S and V
Implications:
Broadband,
LTE, SDN, NFV
PREDICTABLE
NETWORK
SOLUTIONS
What is
networking?
• Networking is inter-process
communications
– What matters is enabling computation
• We only care about the effects
visible to the computation processes
– We don’t per se care about
technologies, mechanisms or policies.
• The only visible effect of the
network to the computation
processes is (paradoxically)
to lose and delay data!
Networking is a
statistical
“game of chance”
• We’re sharing a fixed and finite
transmission resource through
statistical multiplexing
• Good outcomes come from
– many “good coincidences”
– few “bad coincidences”
• In the game of chance, networks
have some choices over what to lose
and delay
PREDICTABLE
NETWORK
SOLUTIONS
What is
“Network Performance Engineering?”
• Network performance engineering is about
delivering good enough outcomes…
– Acceptable quality of experience (QoE) to user
– Low cost to network operator
• …and managing the trade-offs in achieving
these…
• …by tipping the odds in the game of chance in
favour of lower cost and higher QoE
PREDICTABLE
NETWORK
SOLUTIONS
Overview
What is
“Network
Performance
Engineering”?
3 Basic
Concepts
G, S and V
Implications:
Broadband,
LTE, SDN, NFV
PREDICTABLE
NETWORK
SOLUTIONS
Three essential concepts of
network performance engineering
1. Loss and delay accumulate along a path
PREDICTABLE
NETWORK
SOLUTIONS
The raw data we want to work with is
end-to-end path delay, because that’s what
the computation processes experience.
Three essential concepts of
network performance engineering
1. Loss and delay accumulates along a path
2. What matters is the distribution of loss and
delay
PREDICTABLE
NETWORK
SOLUTIONS
This is the simplest view
of the probability
distribution, but it is of
limited use for
managing performance
Cumulative view allows
(de)composition of loss
and delay along the path:
can isolate performance
issues to specific network
elements and links.
We are most interested
in the “tails” and their
structure: these are
what cause application
QoE failure, and whose
mitigation drives cost.
Three essential concepts of
network performance engineering
1. Loss and delay accumulates along a path
2. What matters is the distribution of loss and
delay
3. A model of causality: decompose and predict
PREDICTABLE
NETWORK
SOLUTIONS
How can you know what kind of intervention will address your
QoE or cost issue, and what its effect will be?
Is there another way of looking at this data
that will help us to select the right
intervention and predict its effect?
Overview
What is
“Network
Performance
Engineering”?
3 Basic
Concepts
G, S and V
Implications:
Broadband,
LTE, SDN, NFV
PREDICTABLE
NETWORK
SOLUTIONS
Sort by packet size:
a clear structure
emerges
Example packet
delay: what is it
comprised of?
Transit time of
hypothetical zero
length packet
Look at how there is this
boundary line. Packets on the line
experienced a network where all
buffers were empty; those above
had to wait for other traffic in
buffers. Note that the difference
in delay along this dotted line is
related only to packet size.
Geographic delayG
Every packet experienced a
structural delay due to the
speed of light, routing
lookup overheads.
Serialisation delayS
Packets with bigger payloads experience more delay:
as they are being duplicated by each network element it takes longer
to turn the packet into a bitstream, and back again into a packet.
Variable contention
delay
V
The remainder of the delay is
not structural, but is induced
by applying a demand load to
the shared transmission
supply. We have choices over
how we allocate this delay.
Geographic delay
Serialisation delay
Variable contention
delay
G
S
V
All delay is (everywhere and
always) comprised of these
three basic elements.
Network technology
or design
Link rate
Scheduling
G
S
V
Once we understand their
contribution to QoE and cost, we can
measure and manage the right thing!
Ideas like “jitter” conflate delay
from V and S, along with loss.
Measure the wrong thing, and
you manage the wrong thing.
Packets whose delay is on this line
are experiencing no contention,
even though their delays are
varying due to packet size.
Summary (thus far)
1. Measure paths… not points.
2. Analyse distributions… not averages.
3. Extract structure… for understanding
and prediction.
PREDICTABLE
NETWORK
SOLUTIONS
Note that these simple principles are not common practise in
network performance engineering today.
(That’s why you should do business with us.)
Overview
What is
“Network
Performance
Engineering”?
3 Basic
Concepts
G, S and V
Implications:
Broadband,
LTE, SDN, NFV
PREDICTABLE
NETWORK
SOLUTIONS
So what? Broadband
Megabits/second are an insufficient measure: G and V matter too.
The broadband market is not being regulated correctly!
PREDICTABLE
NETWORK
SOLUTIONS
Example: Two different ADSL providers in the same location, with same measured “speed”
Great for gaming Useless for gaming
So what? LTE
• Cellular will never be as good
as low-spec ADSL
– G and V are too high
– Has implications for real-time
media, gaming
• Nothing in 3G/4G standards
and networks supports
consistent loss and delay
– Yet this is needed for real-time
value added services
PREDICTABLE
NETWORK
SOLUTIONS
Downstream delay over a 3G connection
Too much variability for
TCP to work well.
So what? SDN and NFV
Software Defined
Networking (SDN)
resource model is
restricted to the arbitrary
concept of “bandwidth”.
– So can’t ask for the right
G, S and V loss and delay
characteristics.
PREDICTABLE
NETWORK
SOLUTIONS
Delays measured across UK Internet exchange
Bad virtualisation is likely to be the result!
“Bandwidth” is too weak
a proxy for what matters
in network performance
For further information on
network performance engineering
download white papers at
www.pnsol.com/publications.html
PREDICTABLE
NETWORK
SOLUTIONS
For further insight, webinars
and workshops
sign up for
Future of Communications
email newsletter
www.martingeddes.com
Neil Davies
Neil.Davies@pnsol.com
Peter Thompson
Peter.Thompson@pnsol.com
Martin Geddes
mail@martingeddes.com
PREDICTABLE
NETWORK
SOLUTIONS

Mais conteúdo relacionado

Mais procurados

Cisco Study: State of Web Security
Cisco Study: State of Web Security Cisco Study: State of Web Security
Cisco Study: State of Web Security Cisco Canada
 
Ch 7: Programming for Security Professionals
Ch 7: Programming for Security ProfessionalsCh 7: Programming for Security Professionals
Ch 7: Programming for Security ProfessionalsSam Bowne
 
Wpa2 psk security measure
Wpa2 psk security measureWpa2 psk security measure
Wpa2 psk security measureShivam Singh
 
Cloud Vulnerabilities and Its Threats
Cloud Vulnerabilities and Its ThreatsCloud Vulnerabilities and Its Threats
Cloud Vulnerabilities and Its ThreatsAri Apridana
 
Low level design template (1)
Low level design template (1)Low level design template (1)
Low level design template (1)anosha jamshed
 
5. message authentication and hash function
5. message authentication and hash function5. message authentication and hash function
5. message authentication and hash functionChirag Patel
 
Network Monitoring System
Network Monitoring SystemNetwork Monitoring System
Network Monitoring SystemRofiq Fauzi
 
Free and open cloud security posture monitoring
Free and open cloud security posture monitoringFree and open cloud security posture monitoring
Free and open cloud security posture monitoringElasticsearch
 
An introduction to SOC (Security Operation Center)
An introduction to SOC (Security Operation Center)An introduction to SOC (Security Operation Center)
An introduction to SOC (Security Operation Center)Ahmad Haghighi
 
Complete Endpoint protection
Complete Endpoint protectionComplete Endpoint protection
Complete Endpoint protectionxband
 
Secure Socket Layer
Secure Socket LayerSecure Socket Layer
Secure Socket LayerNaveen Kumar
 
CS8792 - Cryptography and Network Security
CS8792 - Cryptography and Network SecurityCS8792 - Cryptography and Network Security
CS8792 - Cryptography and Network Securityvishnukp34
 

Mais procurados (20)

Network security
Network security Network security
Network security
 
Cisco Study: State of Web Security
Cisco Study: State of Web Security Cisco Study: State of Web Security
Cisco Study: State of Web Security
 
Ch 7: Programming for Security Professionals
Ch 7: Programming for Security ProfessionalsCh 7: Programming for Security Professionals
Ch 7: Programming for Security Professionals
 
Wpa2 psk security measure
Wpa2 psk security measureWpa2 psk security measure
Wpa2 psk security measure
 
IP security
IP securityIP security
IP security
 
Cloud Vulnerabilities and Its Threats
Cloud Vulnerabilities and Its ThreatsCloud Vulnerabilities and Its Threats
Cloud Vulnerabilities and Its Threats
 
IOT Security
IOT SecurityIOT Security
IOT Security
 
Low level design template (1)
Low level design template (1)Low level design template (1)
Low level design template (1)
 
5. message authentication and hash function
5. message authentication and hash function5. message authentication and hash function
5. message authentication and hash function
 
Network Monitoring System
Network Monitoring SystemNetwork Monitoring System
Network Monitoring System
 
Ch06 Wireless Network Security
Ch06 Wireless Network SecurityCh06 Wireless Network Security
Ch06 Wireless Network Security
 
Web security
Web securityWeb security
Web security
 
Free and open cloud security posture monitoring
Free and open cloud security posture monitoringFree and open cloud security posture monitoring
Free and open cloud security posture monitoring
 
An introduction to SOC (Security Operation Center)
An introduction to SOC (Security Operation Center)An introduction to SOC (Security Operation Center)
An introduction to SOC (Security Operation Center)
 
Wireshark
Wireshark Wireshark
Wireshark
 
Complete Endpoint protection
Complete Endpoint protectionComplete Endpoint protection
Complete Endpoint protection
 
Secure Socket Layer
Secure Socket LayerSecure Socket Layer
Secure Socket Layer
 
F5 Web Application Security
F5 Web Application SecurityF5 Web Application Security
F5 Web Application Security
 
CS8792 - Cryptography and Network Security
CS8792 - Cryptography and Network SecurityCS8792 - Cryptography and Network Security
CS8792 - Cryptography and Network Security
 
Wifi Security
Wifi SecurityWifi Security
Wifi Security
 

Destaque

Why ∆Q is the ideal network metric
Why ∆Q is the ideal network metricWhy ∆Q is the ideal network metric
Why ∆Q is the ideal network metricMartin Geddes
 
The Properties and Mathematics of Data Transport Quality
The Properties and Mathematics of Data Transport QualityThe Properties and Mathematics of Data Transport Quality
The Properties and Mathematics of Data Transport QualityMartin Geddes
 
Response to DCMS Digital Communications Infrastructure Strategy consultation
Response to DCMS Digital Communications Infrastructure Strategy consultationResponse to DCMS Digital Communications Infrastructure Strategy consultation
Response to DCMS Digital Communications Infrastructure Strategy consultationMartin Geddes
 
Advanced network performance measurement techniques
Advanced network performance measurement techniquesAdvanced network performance measurement techniques
Advanced network performance measurement techniquesMartin Geddes
 
Introduction to ΔQ and Network Performance Science (extracts)
Introduction to ΔQ and Network Performance Science (extracts)Introduction to ΔQ and Network Performance Science (extracts)
Introduction to ΔQ and Network Performance Science (extracts)Martin Geddes
 
Network performance optimisation using high-fidelity measures
Network performance optimisation using high-fidelity measuresNetwork performance optimisation using high-fidelity measures
Network performance optimisation using high-fidelity measuresMartin Geddes
 
Evaluating the internet end-user experience in the Russian Federation
Evaluating the internet end-user experience in the Russian FederationEvaluating the internet end-user experience in the Russian Federation
Evaluating the internet end-user experience in the Russian FederationMartin Geddes
 
The Ladder: How money and multiplexing are connected
The Ladder: How money and multiplexing are connectedThe Ladder: How money and multiplexing are connected
The Ladder: How money and multiplexing are connectedMartin Geddes
 

Destaque (8)

Why ∆Q is the ideal network metric
Why ∆Q is the ideal network metricWhy ∆Q is the ideal network metric
Why ∆Q is the ideal network metric
 
The Properties and Mathematics of Data Transport Quality
The Properties and Mathematics of Data Transport QualityThe Properties and Mathematics of Data Transport Quality
The Properties and Mathematics of Data Transport Quality
 
Response to DCMS Digital Communications Infrastructure Strategy consultation
Response to DCMS Digital Communications Infrastructure Strategy consultationResponse to DCMS Digital Communications Infrastructure Strategy consultation
Response to DCMS Digital Communications Infrastructure Strategy consultation
 
Advanced network performance measurement techniques
Advanced network performance measurement techniquesAdvanced network performance measurement techniques
Advanced network performance measurement techniques
 
Introduction to ΔQ and Network Performance Science (extracts)
Introduction to ΔQ and Network Performance Science (extracts)Introduction to ΔQ and Network Performance Science (extracts)
Introduction to ΔQ and Network Performance Science (extracts)
 
Network performance optimisation using high-fidelity measures
Network performance optimisation using high-fidelity measuresNetwork performance optimisation using high-fidelity measures
Network performance optimisation using high-fidelity measures
 
Evaluating the internet end-user experience in the Russian Federation
Evaluating the internet end-user experience in the Russian FederationEvaluating the internet end-user experience in the Russian Federation
Evaluating the internet end-user experience in the Russian Federation
 
The Ladder: How money and multiplexing are connected
The Ladder: How money and multiplexing are connectedThe Ladder: How money and multiplexing are connected
The Ladder: How money and multiplexing are connected
 

Semelhante a Fundamentals of network performance engineering

Future of Broadband workshop presentation - ITU Telecom World 2013
Future of Broadband workshop presentation - ITU Telecom World 2013Future of Broadband workshop presentation - ITU Telecom World 2013
Future of Broadband workshop presentation - ITU Telecom World 2013Martin Geddes
 
Research Challenges and Characteristic Features in Wireless Sensor Networks
Research Challenges and Characteristic Features in Wireless Sensor NetworksResearch Challenges and Characteristic Features in Wireless Sensor Networks
Research Challenges and Characteristic Features in Wireless Sensor NetworksEswar Publications
 
Telecommunications network design
Telecommunications network designTelecommunications network design
Telecommunications network designnerdic
 
Big Data and Next Generation Network Challenges - Phdassistance
Big Data and Next Generation Network Challenges - PhdassistanceBig Data and Next Generation Network Challenges - Phdassistance
Big Data and Next Generation Network Challenges - PhdassistancePhD Assistance
 
The Road towards Wireless Dense & Heterogeneous Networks: The CROWD Perspective
The Road towards Wireless Dense & Heterogeneous Networks: The CROWD PerspectiveThe Road towards Wireless Dense & Heterogeneous Networks: The CROWD Perspective
The Road towards Wireless Dense & Heterogeneous Networks: The CROWD PerspectiveIIT CNR
 
PhD proposal in KTH, By Amin Azari
PhD proposal in KTH, By Amin AzariPhD proposal in KTH, By Amin Azari
PhD proposal in KTH, By Amin Azariamin azari
 
Optical Networks Automation Overview: A Survey
Optical Networks Automation Overview: A SurveyOptical Networks Automation Overview: A Survey
Optical Networks Automation Overview: A SurveySergio Cruzes
 
Designing network topology.pptx
Designing network topology.pptxDesigning network topology.pptx
Designing network topology.pptxKISHOYIANKISH
 
Bonneau - Complex Networks Foundations of Information Systems - Spring Review...
Bonneau - Complex Networks Foundations of Information Systems - Spring Review...Bonneau - Complex Networks Foundations of Information Systems - Spring Review...
Bonneau - Complex Networks Foundations of Information Systems - Spring Review...The Air Force Office of Scientific Research
 
Fiber Optic Network Topologies Ring Star and Mesh.pptx
Fiber Optic Network Topologies Ring Star and Mesh.pptxFiber Optic Network Topologies Ring Star and Mesh.pptx
Fiber Optic Network Topologies Ring Star and Mesh.pptxVERSITRONINC
 
Proactive Population-Risk Based Defense Against Denial of Cyber-Physical Serv...
Proactive Population-Risk Based Defense Against Denial of Cyber-Physical Serv...Proactive Population-Risk Based Defense Against Denial of Cyber-Physical Serv...
Proactive Population-Risk Based Defense Against Denial of Cyber-Physical Serv...IRJET Journal
 
Data Communications and Net-Centric computing
Data Communications and Net-Centric computingData Communications and Net-Centric computing
Data Communications and Net-Centric computingKomalah Nair
 
Network Planning & Design: An Art or a Science?
Network Planning & Design: An Art or a Science?Network Planning & Design: An Art or a Science?
Network Planning & Design: An Art or a Science?Vishal Sharma, Ph.D.
 
Fiber Optic Network Topologies Ring Star and Mesh.pptx
Fiber Optic Network Topologies Ring Star and Mesh.pptxFiber Optic Network Topologies Ring Star and Mesh.pptx
Fiber Optic Network Topologies Ring Star and Mesh.pptxVERSITRONINC
 
Study on reliability optimization problem of computer By Dharmendra Singh[Srm...
Study on reliability optimization problem of computer By Dharmendra Singh[Srm...Study on reliability optimization problem of computer By Dharmendra Singh[Srm...
Study on reliability optimization problem of computer By Dharmendra Singh[Srm...Dharmendrasingh417
 
IRJET- Cost Effective Scheme for Delay Tolerant Data Transmission
IRJET- Cost Effective Scheme for Delay Tolerant Data TransmissionIRJET- Cost Effective Scheme for Delay Tolerant Data Transmission
IRJET- Cost Effective Scheme for Delay Tolerant Data TransmissionIRJET Journal
 
Geddes/PNSol - Broadband market evolution
Geddes/PNSol - Broadband market evolutionGeddes/PNSol - Broadband market evolution
Geddes/PNSol - Broadband market evolutionMartin Geddes
 
OPTIMIZED ROUTING AND DENIAL OF SERVICE FOR ROBUST TRANSMISSION IN WIRELESS N...
OPTIMIZED ROUTING AND DENIAL OF SERVICE FOR ROBUST TRANSMISSION IN WIRELESS N...OPTIMIZED ROUTING AND DENIAL OF SERVICE FOR ROBUST TRANSMISSION IN WIRELESS N...
OPTIMIZED ROUTING AND DENIAL OF SERVICE FOR ROBUST TRANSMISSION IN WIRELESS N...IRJET Journal
 
Taming limits with approximate networking
Taming limits with approximate networkingTaming limits with approximate networking
Taming limits with approximate networkingJunaid Qadir
 

Semelhante a Fundamentals of network performance engineering (20)

Future of Broadband workshop presentation - ITU Telecom World 2013
Future of Broadband workshop presentation - ITU Telecom World 2013Future of Broadband workshop presentation - ITU Telecom World 2013
Future of Broadband workshop presentation - ITU Telecom World 2013
 
Research Challenges and Characteristic Features in Wireless Sensor Networks
Research Challenges and Characteristic Features in Wireless Sensor NetworksResearch Challenges and Characteristic Features in Wireless Sensor Networks
Research Challenges and Characteristic Features in Wireless Sensor Networks
 
Telecommunications network design
Telecommunications network designTelecommunications network design
Telecommunications network design
 
Big Data and Next Generation Network Challenges - Phdassistance
Big Data and Next Generation Network Challenges - PhdassistanceBig Data and Next Generation Network Challenges - Phdassistance
Big Data and Next Generation Network Challenges - Phdassistance
 
Black_Box_Cabling_Guide.pdf
Black_Box_Cabling_Guide.pdfBlack_Box_Cabling_Guide.pdf
Black_Box_Cabling_Guide.pdf
 
The Road towards Wireless Dense & Heterogeneous Networks: The CROWD Perspective
The Road towards Wireless Dense & Heterogeneous Networks: The CROWD PerspectiveThe Road towards Wireless Dense & Heterogeneous Networks: The CROWD Perspective
The Road towards Wireless Dense & Heterogeneous Networks: The CROWD Perspective
 
PhD proposal in KTH, By Amin Azari
PhD proposal in KTH, By Amin AzariPhD proposal in KTH, By Amin Azari
PhD proposal in KTH, By Amin Azari
 
Optical Networks Automation Overview: A Survey
Optical Networks Automation Overview: A SurveyOptical Networks Automation Overview: A Survey
Optical Networks Automation Overview: A Survey
 
Designing network topology.pptx
Designing network topology.pptxDesigning network topology.pptx
Designing network topology.pptx
 
Bonneau - Complex Networks Foundations of Information Systems - Spring Review...
Bonneau - Complex Networks Foundations of Information Systems - Spring Review...Bonneau - Complex Networks Foundations of Information Systems - Spring Review...
Bonneau - Complex Networks Foundations of Information Systems - Spring Review...
 
Fiber Optic Network Topologies Ring Star and Mesh.pptx
Fiber Optic Network Topologies Ring Star and Mesh.pptxFiber Optic Network Topologies Ring Star and Mesh.pptx
Fiber Optic Network Topologies Ring Star and Mesh.pptx
 
Proactive Population-Risk Based Defense Against Denial of Cyber-Physical Serv...
Proactive Population-Risk Based Defense Against Denial of Cyber-Physical Serv...Proactive Population-Risk Based Defense Against Denial of Cyber-Physical Serv...
Proactive Population-Risk Based Defense Against Denial of Cyber-Physical Serv...
 
Data Communications and Net-Centric computing
Data Communications and Net-Centric computingData Communications and Net-Centric computing
Data Communications and Net-Centric computing
 
Network Planning & Design: An Art or a Science?
Network Planning & Design: An Art or a Science?Network Planning & Design: An Art or a Science?
Network Planning & Design: An Art or a Science?
 
Fiber Optic Network Topologies Ring Star and Mesh.pptx
Fiber Optic Network Topologies Ring Star and Mesh.pptxFiber Optic Network Topologies Ring Star and Mesh.pptx
Fiber Optic Network Topologies Ring Star and Mesh.pptx
 
Study on reliability optimization problem of computer By Dharmendra Singh[Srm...
Study on reliability optimization problem of computer By Dharmendra Singh[Srm...Study on reliability optimization problem of computer By Dharmendra Singh[Srm...
Study on reliability optimization problem of computer By Dharmendra Singh[Srm...
 
IRJET- Cost Effective Scheme for Delay Tolerant Data Transmission
IRJET- Cost Effective Scheme for Delay Tolerant Data TransmissionIRJET- Cost Effective Scheme for Delay Tolerant Data Transmission
IRJET- Cost Effective Scheme for Delay Tolerant Data Transmission
 
Geddes/PNSol - Broadband market evolution
Geddes/PNSol - Broadband market evolutionGeddes/PNSol - Broadband market evolution
Geddes/PNSol - Broadband market evolution
 
OPTIMIZED ROUTING AND DENIAL OF SERVICE FOR ROBUST TRANSMISSION IN WIRELESS N...
OPTIMIZED ROUTING AND DENIAL OF SERVICE FOR ROBUST TRANSMISSION IN WIRELESS N...OPTIMIZED ROUTING AND DENIAL OF SERVICE FOR ROBUST TRANSMISSION IN WIRELESS N...
OPTIMIZED ROUTING AND DENIAL OF SERVICE FOR ROBUST TRANSMISSION IN WIRELESS N...
 
Taming limits with approximate networking
Taming limits with approximate networkingTaming limits with approximate networking
Taming limits with approximate networking
 

Mais de Martin Geddes

CARE initiative technical prospectus
CARE initiative technical prospectusCARE initiative technical prospectus
CARE initiative technical prospectusMartin Geddes
 
Martin Geddes Livestream - March 2018
Martin Geddes Livestream - March 2018Martin Geddes Livestream - March 2018
Martin Geddes Livestream - March 2018Martin Geddes
 
Digital supply chain quality management
Digital supply chain quality managementDigital supply chain quality management
Digital supply chain quality managementMartin Geddes
 
Stationarity is the new speed
Stationarity is the new speedStationarity is the new speed
Stationarity is the new speedMartin Geddes
 
Example high-fidelity network measures using ∆Q metrics
Example high-fidelity network measures using ∆Q metricsExample high-fidelity network measures using ∆Q metrics
Example high-fidelity network measures using ∆Q metricsMartin Geddes
 
Superfast or superfit? The case for UK broadband policy reform
Superfast or superfit? The case for UK broadband policy reformSuperfast or superfit? The case for UK broadband policy reform
Superfast or superfit? The case for UK broadband policy reformMartin Geddes
 
Broadband service quality - rationing or markets?
Broadband service quality - rationing or markets?Broadband service quality - rationing or markets?
Broadband service quality - rationing or markets?Martin Geddes
 
Introduction to network quality arbitrage
Introduction to network quality arbitrageIntroduction to network quality arbitrage
Introduction to network quality arbitrageMartin Geddes
 
The End of Information Technology: Introducing Hypersense & Human Technology
The End of Information Technology: Introducing Hypersense & Human TechnologyThe End of Information Technology: Introducing Hypersense & Human Technology
The End of Information Technology: Introducing Hypersense & Human TechnologyMartin Geddes
 
Sample proposal summary for quality arbitrage business unit
Sample proposal summary for quality arbitrage business unitSample proposal summary for quality arbitrage business unit
Sample proposal summary for quality arbitrage business unitMartin Geddes
 
FCC Open Internet Transparency - a review by Martin Geddes
FCC Open Internet Transparency - a review by Martin GeddesFCC Open Internet Transparency - a review by Martin Geddes
FCC Open Internet Transparency - a review by Martin GeddesMartin Geddes
 
Beasts to superheroes to gods
Beasts to superheroes to godsBeasts to superheroes to gods
Beasts to superheroes to godsMartin Geddes
 
Beyond 'neutrality' - how to reconnect regulation to reality?
Beyond 'neutrality' - how to reconnect regulation to reality?Beyond 'neutrality' - how to reconnect regulation to reality?
Beyond 'neutrality' - how to reconnect regulation to reality?Martin Geddes
 
What is a 'polyservice' network?
What is a 'polyservice' network?What is a 'polyservice' network?
What is a 'polyservice' network?Martin Geddes
 
The perception gap: the barrier to disruptive innovation in telecoms
The perception gap: the barrier to disruptive innovation in telecomsThe perception gap: the barrier to disruptive innovation in telecoms
The perception gap: the barrier to disruptive innovation in telecomsMartin Geddes
 
Lastingham - Parish mast report
Lastingham - Parish mast reportLastingham - Parish mast report
Lastingham - Parish mast reportMartin Geddes
 
BT Operate Case Study
BT Operate Case StudyBT Operate Case Study
BT Operate Case StudyMartin Geddes
 
Essential science for broadband regulation
Essential science for broadband regulationEssential science for broadband regulation
Essential science for broadband regulationMartin Geddes
 

Mais de Martin Geddes (20)

CARE initiative technical prospectus
CARE initiative technical prospectusCARE initiative technical prospectus
CARE initiative technical prospectus
 
Martin Geddes Livestream - March 2018
Martin Geddes Livestream - March 2018Martin Geddes Livestream - March 2018
Martin Geddes Livestream - March 2018
 
Digital supply chain quality management
Digital supply chain quality managementDigital supply chain quality management
Digital supply chain quality management
 
Stationarity is the new speed
Stationarity is the new speedStationarity is the new speed
Stationarity is the new speed
 
Example high-fidelity network measures using ∆Q metrics
Example high-fidelity network measures using ∆Q metricsExample high-fidelity network measures using ∆Q metrics
Example high-fidelity network measures using ∆Q metrics
 
Superfast or superfit? The case for UK broadband policy reform
Superfast or superfit? The case for UK broadband policy reformSuperfast or superfit? The case for UK broadband policy reform
Superfast or superfit? The case for UK broadband policy reform
 
Broadband service quality - rationing or markets?
Broadband service quality - rationing or markets?Broadband service quality - rationing or markets?
Broadband service quality - rationing or markets?
 
Introduction to network quality arbitrage
Introduction to network quality arbitrageIntroduction to network quality arbitrage
Introduction to network quality arbitrage
 
The End of Information Technology: Introducing Hypersense & Human Technology
The End of Information Technology: Introducing Hypersense & Human TechnologyThe End of Information Technology: Introducing Hypersense & Human Technology
The End of Information Technology: Introducing Hypersense & Human Technology
 
Sample proposal summary for quality arbitrage business unit
Sample proposal summary for quality arbitrage business unitSample proposal summary for quality arbitrage business unit
Sample proposal summary for quality arbitrage business unit
 
The Guardian Avatar
The Guardian AvatarThe Guardian Avatar
The Guardian Avatar
 
FCC Open Internet Transparency - a review by Martin Geddes
FCC Open Internet Transparency - a review by Martin GeddesFCC Open Internet Transparency - a review by Martin Geddes
FCC Open Internet Transparency - a review by Martin Geddes
 
Beasts to superheroes to gods
Beasts to superheroes to godsBeasts to superheroes to gods
Beasts to superheroes to gods
 
Beyond 'neutrality' - how to reconnect regulation to reality?
Beyond 'neutrality' - how to reconnect regulation to reality?Beyond 'neutrality' - how to reconnect regulation to reality?
Beyond 'neutrality' - how to reconnect regulation to reality?
 
What is a 'polyservice' network?
What is a 'polyservice' network?What is a 'polyservice' network?
What is a 'polyservice' network?
 
Superfit broadband
Superfit broadbandSuperfit broadband
Superfit broadband
 
The perception gap: the barrier to disruptive innovation in telecoms
The perception gap: the barrier to disruptive innovation in telecomsThe perception gap: the barrier to disruptive innovation in telecoms
The perception gap: the barrier to disruptive innovation in telecoms
 
Lastingham - Parish mast report
Lastingham - Parish mast reportLastingham - Parish mast report
Lastingham - Parish mast report
 
BT Operate Case Study
BT Operate Case StudyBT Operate Case Study
BT Operate Case Study
 
Essential science for broadband regulation
Essential science for broadband regulationEssential science for broadband regulation
Essential science for broadband regulation
 

Último

"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks..."LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...Fwdays
 
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 3652toLead Limited
 
Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?Mattias Andersson
 
APIForce Zurich 5 April Automation LPDG
APIForce Zurich 5 April  Automation LPDGAPIForce Zurich 5 April  Automation LPDG
APIForce Zurich 5 April Automation LPDGMarianaLemus7
 
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | DelhiFULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhisoniya singh
 
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr LapshynFwdays
 
Benefits Of Flutter Compared To Other Frameworks
Benefits Of Flutter Compared To Other FrameworksBenefits Of Flutter Compared To Other Frameworks
Benefits Of Flutter Compared To Other FrameworksSoftradix Technologies
 
Understanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitectureUnderstanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitecturePixlogix Infotech
 
Connect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck PresentationConnect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck PresentationSlibray Presentation
 
Unlocking the Potential of the Cloud for IBM Power Systems
Unlocking the Potential of the Cloud for IBM Power SystemsUnlocking the Potential of the Cloud for IBM Power Systems
Unlocking the Potential of the Cloud for IBM Power SystemsPrecisely
 
My Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 PresentationMy Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 PresentationRidwan Fadjar
 
Install Stable Diffusion in windows machine
Install Stable Diffusion in windows machineInstall Stable Diffusion in windows machine
Install Stable Diffusion in windows machinePadma Pradeep
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersThousandEyes
 
Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)Allon Mureinik
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxMalak Abu Hammad
 
SIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge GraphSIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge GraphNeo4j
 
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024BookNet Canada
 

Último (20)

"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks..."LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
"LLMs for Python Engineers: Advanced Data Analysis and Semantic Kernel",Oleks...
 
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
Tech-Forward - Achieving Business Readiness For Copilot in Microsoft 365
 
Vulnerability_Management_GRC_by Sohang Sengupta.pptx
Vulnerability_Management_GRC_by Sohang Sengupta.pptxVulnerability_Management_GRC_by Sohang Sengupta.pptx
Vulnerability_Management_GRC_by Sohang Sengupta.pptx
 
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptxE-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
E-Vehicle_Hacking_by_Parul Sharma_null_owasp.pptx
 
Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?
 
APIForce Zurich 5 April Automation LPDG
APIForce Zurich 5 April  Automation LPDGAPIForce Zurich 5 April  Automation LPDG
APIForce Zurich 5 April Automation LPDG
 
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | DelhiFULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
FULL ENJOY 🔝 8264348440 🔝 Call Girls in Diplomatic Enclave | Delhi
 
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
"Federated learning: out of reach no matter how close",Oleksandr Lapshyn
 
Benefits Of Flutter Compared To Other Frameworks
Benefits Of Flutter Compared To Other FrameworksBenefits Of Flutter Compared To Other Frameworks
Benefits Of Flutter Compared To Other Frameworks
 
DMCC Future of Trade Web3 - Special Edition
DMCC Future of Trade Web3 - Special EditionDMCC Future of Trade Web3 - Special Edition
DMCC Future of Trade Web3 - Special Edition
 
Understanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitectureUnderstanding the Laravel MVC Architecture
Understanding the Laravel MVC Architecture
 
Connect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck PresentationConnect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck Presentation
 
Unlocking the Potential of the Cloud for IBM Power Systems
Unlocking the Potential of the Cloud for IBM Power SystemsUnlocking the Potential of the Cloud for IBM Power Systems
Unlocking the Potential of the Cloud for IBM Power Systems
 
My Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 PresentationMy Hashitalk Indonesia April 2024 Presentation
My Hashitalk Indonesia April 2024 Presentation
 
Install Stable Diffusion in windows machine
Install Stable Diffusion in windows machineInstall Stable Diffusion in windows machine
Install Stable Diffusion in windows machine
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
 
Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)Injustice - Developers Among Us (SciFiDevCon 2024)
Injustice - Developers Among Us (SciFiDevCon 2024)
 
The Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptxThe Codex of Business Writing Software for Real-World Solutions 2.pptx
The Codex of Business Writing Software for Real-World Solutions 2.pptx
 
SIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge GraphSIEMENS: RAPUNZEL – A Tale About Knowledge Graph
SIEMENS: RAPUNZEL – A Tale About Knowledge Graph
 
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
 

Fundamentals of network performance engineering

  • 1. Fundamentals of Network Performance Engineering Dr Neil Davies Predictable Network Solutions Ltd Peter Thompson Predictable Network Solutions Ltd Martin Geddes Martin Geddes Consulting Ltd © 2013 All Rights Reserved PREDICTABLE NETWORK SOLUTIONS
  • 2. Dr Neil Davies Co-founder, Predictable Network Solutions Ltd Ex: University of Bristol (23 years). Former technical head of joint university/research institute (SRF/PACT). Peter Thompson CTO, Predictable Network Solutions Ltd Ex: GoS Networks, U4EA, SGS-Thomson, INMOS & Universities of Bristol, Warwick and Cambridge. Authority on technical and commercial issues of converged networking. Martin Geddes Founder, Martin Geddes Consulting Ltd Ex: BT, Telco 2.0, Sprint, Oracle, Oxford University. Thought leader on future of telecommunications industry. PREDICTABLE NETWORK SOLUTIONS
  • 3. Dr Neil Davies Co-founder, Predictable Network Solutions Ltd Ex: University of Bristol (23 years). Former technical head of joint university/research institute (SRF/PACT). Peter Thompson CTO, Predictable Network Solutions Ltd Ex: GoS Networks, U4EA, SGS-Thomson, INMOS & Universities of Bristol, Warwick and Cambridge. Authority on technical and commercial issues of converged networking. Martin Geddes Founder, Martin Geddes Consulting Ltd Ex: BT, Telco 2.0, Sprint, Oracle, Oxford University. Thought leader on future of telecommunications industry. PREDICTABLE NETWORK SOLUTIONS The only ex-ante network performance engineering company in the world. • New mathematical performance techniques. • Performance assessment methodology. • World’s first network contention management solution. Consultancy on the future of telecoms and the Internet. • Business model innovation. • Technology & product ideation. • Organisation development. • Public & private workshops.
  • 4. This presentation is taken from the content for Fundamentals of Network Performance Workshop For information on locations and timing of public events visit www.sustainablebroadband.com PREDICTABLE NETWORK SOLUTIONS
  • 5. Overview What is “Network Performance Engineering”? 3 Basic Concepts G, S and V Implications: Broadband, LTE, SDN, NFV PREDICTABLE NETWORK SOLUTIONS
  • 6. What is networking? • Networking is inter-process communications – What matters is enabling computation • We only care about the effects visible to the computation processes – We don’t per se care about technologies, mechanisms or policies. • The only visible effect of the network to the computation processes is (paradoxically) to lose and delay data!
  • 7. Networking is a statistical “game of chance” • We’re sharing a fixed and finite transmission resource through statistical multiplexing • Good outcomes come from – many “good coincidences” – few “bad coincidences” • In the game of chance, networks have some choices over what to lose and delay PREDICTABLE NETWORK SOLUTIONS
  • 8. What is “Network Performance Engineering?” • Network performance engineering is about delivering good enough outcomes… – Acceptable quality of experience (QoE) to user – Low cost to network operator • …and managing the trade-offs in achieving these… • …by tipping the odds in the game of chance in favour of lower cost and higher QoE PREDICTABLE NETWORK SOLUTIONS
  • 9. Overview What is “Network Performance Engineering”? 3 Basic Concepts G, S and V Implications: Broadband, LTE, SDN, NFV PREDICTABLE NETWORK SOLUTIONS
  • 10. Three essential concepts of network performance engineering 1. Loss and delay accumulate along a path PREDICTABLE NETWORK SOLUTIONS
  • 11. The raw data we want to work with is end-to-end path delay, because that’s what the computation processes experience.
  • 12. Three essential concepts of network performance engineering 1. Loss and delay accumulates along a path 2. What matters is the distribution of loss and delay PREDICTABLE NETWORK SOLUTIONS
  • 13. This is the simplest view of the probability distribution, but it is of limited use for managing performance
  • 14. Cumulative view allows (de)composition of loss and delay along the path: can isolate performance issues to specific network elements and links.
  • 15. We are most interested in the “tails” and their structure: these are what cause application QoE failure, and whose mitigation drives cost.
  • 16. Three essential concepts of network performance engineering 1. Loss and delay accumulates along a path 2. What matters is the distribution of loss and delay 3. A model of causality: decompose and predict PREDICTABLE NETWORK SOLUTIONS How can you know what kind of intervention will address your QoE or cost issue, and what its effect will be?
  • 17. Is there another way of looking at this data that will help us to select the right intervention and predict its effect?
  • 18. Overview What is “Network Performance Engineering”? 3 Basic Concepts G, S and V Implications: Broadband, LTE, SDN, NFV PREDICTABLE NETWORK SOLUTIONS
  • 19. Sort by packet size: a clear structure emerges
  • 20. Example packet delay: what is it comprised of?
  • 21. Transit time of hypothetical zero length packet Look at how there is this boundary line. Packets on the line experienced a network where all buffers were empty; those above had to wait for other traffic in buffers. Note that the difference in delay along this dotted line is related only to packet size.
  • 22. Geographic delayG Every packet experienced a structural delay due to the speed of light, routing lookup overheads.
  • 23. Serialisation delayS Packets with bigger payloads experience more delay: as they are being duplicated by each network element it takes longer to turn the packet into a bitstream, and back again into a packet.
  • 24. Variable contention delay V The remainder of the delay is not structural, but is induced by applying a demand load to the shared transmission supply. We have choices over how we allocate this delay.
  • 25. Geographic delay Serialisation delay Variable contention delay G S V All delay is (everywhere and always) comprised of these three basic elements.
  • 26. Network technology or design Link rate Scheduling G S V Once we understand their contribution to QoE and cost, we can measure and manage the right thing!
  • 27. Ideas like “jitter” conflate delay from V and S, along with loss. Measure the wrong thing, and you manage the wrong thing. Packets whose delay is on this line are experiencing no contention, even though their delays are varying due to packet size.
  • 28. Summary (thus far) 1. Measure paths… not points. 2. Analyse distributions… not averages. 3. Extract structure… for understanding and prediction. PREDICTABLE NETWORK SOLUTIONS Note that these simple principles are not common practise in network performance engineering today. (That’s why you should do business with us.)
  • 29. Overview What is “Network Performance Engineering”? 3 Basic Concepts G, S and V Implications: Broadband, LTE, SDN, NFV PREDICTABLE NETWORK SOLUTIONS
  • 30. So what? Broadband Megabits/second are an insufficient measure: G and V matter too. The broadband market is not being regulated correctly! PREDICTABLE NETWORK SOLUTIONS Example: Two different ADSL providers in the same location, with same measured “speed” Great for gaming Useless for gaming
  • 31. So what? LTE • Cellular will never be as good as low-spec ADSL – G and V are too high – Has implications for real-time media, gaming • Nothing in 3G/4G standards and networks supports consistent loss and delay – Yet this is needed for real-time value added services PREDICTABLE NETWORK SOLUTIONS Downstream delay over a 3G connection Too much variability for TCP to work well.
  • 32. So what? SDN and NFV Software Defined Networking (SDN) resource model is restricted to the arbitrary concept of “bandwidth”. – So can’t ask for the right G, S and V loss and delay characteristics. PREDICTABLE NETWORK SOLUTIONS Delays measured across UK Internet exchange Bad virtualisation is likely to be the result! “Bandwidth” is too weak a proxy for what matters in network performance
  • 33. For further information on network performance engineering download white papers at www.pnsol.com/publications.html PREDICTABLE NETWORK SOLUTIONS
  • 34. For further insight, webinars and workshops sign up for Future of Communications email newsletter www.martingeddes.com
  • 35. Neil Davies Neil.Davies@pnsol.com Peter Thompson Peter.Thompson@pnsol.com Martin Geddes mail@martingeddes.com PREDICTABLE NETWORK SOLUTIONS