6/17/ 2002EMIS 8392 Maya Petkova1 Chapter 4 Traffic and Cost Generators.

6/17/ 2002 EMIS 8392 Maya Petkova 1

Chapter 4Chapter 4

Traffic and Cost Traffic and Cost GeneratorsGenerators


OverviewOverview• Interest to people doing practical network Interest to people doing practical network

design.design.

• Can not get the actual tariffs and traffic.Can not get the actual tariffs and traffic.

• “ “ Practice, practice ! Practice, practice ! ” ”

• Sample scenarios.Sample scenarios.


War StoriesWar Stories War story 4.1War story 4.1 – a company has to make a decision – a company has to make a decision

about converting to routers at the next board about converting to routers at the next board meeting.meeting.

- meeting is in 3 months- meeting is in 3 months - need to have a capital budget and network - need to have a capital budget and network operating costoperating cost - no actual traffic has been measured- no actual traffic has been measured - there are only user populations by job category- there are only user populations by job category PROBLEM: How to take actual measurements andPROBLEM: How to take actual measurements and meet the schedule.meet the schedule. SOLUTION: Network generator.SOLUTION: Network generator.


War Stories – Cont’dWar Stories – Cont’d War story 4.2War story 4.2 - a large company runs a network - a large company runs a network

composed of hundreds of subnetworks with a composed of hundreds of subnetworks with a major restructuring to be done.major restructuring to be done.

- the amount of traffic entering and exiting the- the amount of traffic entering and exiting the network at any connection is known network at any connection is known - no traffic matrix available- no traffic matrix availablePROBLEM: Build a traffic matrix that matches the PROBLEM: Build a traffic matrix that matches the given totals for each site. given totals for each site. SOLUTION: Traffic generator.SOLUTION: Traffic generator.


More on the Advantages of More on the Advantages of Synthesizing Data Synthesizing Data

• Real network design is 90% file preparation, Real network design is 90% file preparation, traffic and cost generators can reduce that traffic and cost generators can reduce that proportion to 50% or even 25%.proportion to 50% or even 25%.

• Creation of unlimited number of interesting Creation of unlimited number of interesting design problems.design problems.

• Reduces the mass of these problems from MB Reduces the mass of these problems from MB to KB – big structures created parametrically.to KB – big structures created parametrically.

• No worries about the security and privacy No worries about the security and privacy issues associated with real data.issues associated with real data.

• Can study a broader range of problems than Can study a broader range of problems than those that can be gathered off of real systems.those that can be gathered off of real systems.


The Structure of a Network The Structure of a Network

Design Problem Design Problem• Need to place switches or multiplexers at the Need to place switches or multiplexers at the

nodes.nodes.• Interconnect the above equipment with links.Interconnect the above equipment with links.• Traffic needs to be routed.Traffic needs to be routed.• Performance needs to be evaluatedPerformance needs to be evaluated..

Above requires data structures and file formats.Above requires data structures and file formats.Almost all tools have proprietary standards.Almost all tools have proprietary standards.


A Small Complete Design A Small Complete Design Problem Problem

• A simple extensible formatA simple extensible format

• 5 nodes to be interconnected.5 nodes to be interconnected.

• All the necessary information to build a All the necessary information to build a network model is provided.network model is provided.

• Organized as a set of tables.Organized as a set of tables.


The SITES TableThe SITES Table• Contains basic information about each site. Contains basic information about each site. • Note: importance of the PARENT column.Note: importance of the PARENT column.


The LINETYPES TableThe LINETYPES TableContains the list of the links to be used in the Contains the list of the links to be used in the

design.design.


The TRAFFIC TableThe TRAFFIC Table• Contains the flows to be carried by the Contains the flows to be carried by the

network.network.• Flow units:Flow units: - voice traffic – Erlangs- voice traffic – Erlangs - data traffic - bits/second- data traffic - bits/second - multiplexer design – virtual- multiplexer design – virtual circuit capacitycircuit capacity


The TARIFF TableThe TARIFF Table• Contains the costs of all the links available for Contains the costs of all the links available for

the design. the design. • By convention missing costs are set to a very By convention missing costs are set to a very

large constant LINFINITYlarge constant LINFINITY


The EQUIPMENT TableThe EQUIPMENT Table• Specifies the cost and capacity of the equipment to be Specifies the cost and capacity of the equipment to be

placed at each site.placed at each site.• Specifies the throughput of the equipment and the Specifies the throughput of the equipment and the

number of lines that can be terminated.number of lines that can be terminated.


The EQUIPMENT The EQUIPMENT Table – Cont’d Table – Cont’d

• Throughput measurementThroughput measurement - for data switches - packets/second- for data switches - packets/second - for voice switches – call setups/second- for voice switches – call setups/second - multiplexers usually set up static circuits and- multiplexers usually set up static circuits and have only a degree constrainthave only a degree constraint• Most of the cost of the network 90% - 95% is in Most of the cost of the network 90% - 95% is in

the links rather then in the nodes. Link errors the links rather then in the nodes. Link errors could be very expensive.could be very expensive.

• Detailed description in Appendix D.Detailed description in Appendix D.


The PARMS TableThe PARMS TableContains information to guide the design processContains information to guide the design process


Changing the Problem SizeChanging the Problem Size• Suppose instead of 5 we have 100 nodes.Suppose instead of 5 we have 100 nodes.• What is the effect on the size of the tables?What is the effect on the size of the tables? - LINETYPES, EQUIPMENT, PARMS tables do - LINETYPES, EQUIPMENT, PARMS tables do

not changenot change - SITES table grows linearly- SITES table grows linearly - TRAFFIC table may contain entries- TRAFFIC table may contain entries - TARIFF table – number of link costs is ( )- TARIFF table – number of link costs is ( )• May have 10,000 traffic records, 4950 link costs.May have 10,000 traffic records, 4950 link costs.• If created by hand will contain thousands and If created by hand will contain thousands and

thousands of errors.thousands of errors.


The Most Important Lesson fromThe Most Important Lesson from Ch.4 - According to the Author Ch.4 - According to the Author • All the tables with over a few hundred entries All the tables with over a few hundred entries

must be produced automatically.must be produced automatically.

• Traffic and cost generators will facilitate the Traffic and cost generators will facilitate the process of producing complete design process of producing complete design problems provided that tables SITES, problems provided that tables SITES, LINETYPES , EQUIPMENT and PARMS are put LINETYPES , EQUIPMENT and PARMS are put together.together.


The SITES Table for The SITES Table for Network Generation Network Generation

• Squareworld – sites chosen at random.Squareworld – sites chosen at random.• Real life – assembled by interviews, reviews of Real life – assembled by interviews, reviews of

personnel records, real estate records, personnel records, real estate records, examining existing networks, management examining existing networks, management reports.reports.

• Design Principle 4.1.Design Principle 4.1. The first thing a network designer needs to The first thing a network designer needs to

know is the location of the sites to be know is the location of the sites to be connected, just as builder needs a survey of a connected, just as builder needs a survey of a building site.building site.


The SITES Table for The SITES Table for Network Generation – Network Generation – Cont’dCont’d Four additional columns to the previously Four additional columns to the previously

discussed SITES table.discussed SITES table.1.1. POPULATIONPOPULATION – The number of users at a site – The number of users at a site

or the census population.or the census population.2.2. TRAFOUTTRAFOUT – The traffic leaving the site for – The traffic leaving the site for

other sites expressed in the appropriate other sites expressed in the appropriate units.units.

3.3. TRAFINTRAFIN – The traffic entering the site from – The traffic entering the site from other sites.other sites.

4.4. LEVELLEVEL – The level of the site in the hierarchy. – The level of the site in the hierarchy.


Sample SITES TableSample SITES Table


Traffic GeneratorsTraffic Generators• The job of the traffic generator is to spread the The job of the traffic generator is to spread the

traffic around the network.traffic around the network.• Various types of network have different units of Various types of network have different units of

traffic.traffic.

• Being able to fill in the traffic matrix for data Being able to fill in the traffic matrix for data

networks is more important than for voice and networks is more important than for voice and multiplexer networks since the quantum of traffic multiplexer networks since the quantum of traffic is much smaller and the accounting is sketchier.is much smaller and the accounting is sketchier.


Traffic Generators - Traffic Generators - Uniform Traffic Uniform Traffic

• Simplest traffic – for all and we defineSimplest traffic – for all and we define • Can give rise to interesting problems - CST Can give rise to interesting problems - CST

(communication spanning tree) problem – find the (communication spanning tree) problem – find the lowest cost tree to connect n sites, where the cost lowest cost tree to connect n sites, where the cost of each link is proportional to the flow across it. In of each link is proportional to the flow across it. In 1994 C.Palmer and A. Kershenbaum produced a 1994 C.Palmer and A. Kershenbaum produced a creditable algorithm for solving the CST problem.creditable algorithm for solving the CST problem.

• Not realistic for many networks.Not realistic for many networks.• Better fits the traffic inside the switches.Better fits the traffic inside the switches.• Used to test the behavior of algorithms.Used to test the behavior of algorithms.


Traffic Generators – Traffic Generators – RandomRandom Traffic (Voice&Data) Traffic (Voice&Data)

• Suppose we have n sites.Suppose we have n sites.• Suppose we specify the max and the min for Suppose we specify the max and the min for

the traffic in either Erlangs or bits/second.the traffic in either Erlangs or bits/second.• Randreq.cRandreq.c generates random traffic in the generates random traffic in the

(min, max) range between each pair of sites.(min, max) range between each pair of sites. for (i=0; i<n; ++i) {for (i=0; i<n; ++i) { for (j=0; j<n; ++j) {for (j=0; j<n; ++j) { traffic[i][j]=min_req+((double)rand())*(max_req- traffic[i][j]=min_req+((double)rand())*(max_req-

min_req)/RAND_MAX;min_req)/RAND_MAX; } /* endfor */} /* endfor */ } /* endfor */} /* endfor */


Traffic Generators - Random Traffic Generators - Random Traffic (MUX traffic) Traffic (MUX traffic)

• Randreq2.cRandreq2.c generates random set of circuit generates random set of circuit requirements for the multiplexer design problem.requirements for the multiplexer design problem.

• Suppose we have n nodes and nreq=m requirements Suppose we have n nodes and nreq=m requirements for circuits of type D56 – 56,000 bps links. for circuits of type D56 – 56,000 bps links.

for (i=0; i < nreq; ++i) {for (i=0; i < nreq; ++i) { end1 = rand()%n+1;end1 = rand()%n+1; end2 = rand()%n+1;end2 = rand()%n+1; if (end1==end2) {if (end1==end2) { --i;--i; continue;continue; } /* end if */} /* end if */ fprintf(fp,”N%d N%d D56\n”, end1, end2);fprintf(fp,”N%d N%d D56\n”, end1, end2); }/* end for */}/* end for */


More Realistic Traffic: More Realistic Traffic: A Simple Generator A Simple Generator

• Uniform and random traffic generators produce Uniform and random traffic generators produce nonrealistic traffic (does not match what is nonrealistic traffic (does not match what is actually observed by traffic surveys).actually observed by traffic surveys).

• How to produce realistic traffic?How to produce realistic traffic? - determine the factors that could influence- determine the factors that could influence the total traffic between 2 sites.the total traffic between 2 sites. - assume we have 2 nodes and that- assume we have 2 nodes and that have populations and and are have populations and and are

apart.apart.


More Realistic Traffic: More Realistic Traffic: A Simple Generator – Cont’d A Simple Generator – Cont’d

• Simplest notion of population is the human Simplest notion of population is the human population. It could be the portion that uses population. It could be the portion that uses telecommunications (different for voice&data).telecommunications (different for voice&data).

• Network designer chooses how to model the Network designer chooses how to model the site population.site population.

• If population is homogeneous then we can use If population is homogeneous then we can use a model based on physics.a model based on physics.


More Realistic Traffic: More Realistic Traffic: A Simple Generator – Cont’d A Simple Generator – Cont’d

• If we chose model based on gravitation If we chose model based on gravitation attraction: Pop_Power =1, Dist_Power = 2.attraction: Pop_Power =1, Dist_Power = 2.

• If we chose model based on magnetism:If we chose model based on magnetism: Pop_Power =1, Dist_Power = 3.Pop_Power =1, Dist_Power = 3.• If site traffic is independent of population: If site traffic is independent of population:

Pop_Power = 0.Pop_Power = 0.• In all cases is used as a scaling factor to put In all cases is used as a scaling factor to put

the traffic into the range we wish.the traffic into the range we wish.


Even More Realistic Traffic Even More Realistic Traffic

Models ModelsQ. What happens if the 2 nodes in the previous Q. What happens if the 2 nodes in the previous

models are in the same location?models are in the same location?A.A. Division by 0.Division by 0.Q.Q. What happens if the 2 nodes have What happens if the 2 nodes have

populations of 1,000,000 each and the populations of 1,000,000 each and the distance between them is 1000 miles?distance between them is 1000 miles?

A. The effect of distance is lost.A. The effect of distance is lost.Goal: Make calculation more unit independent.Goal: Make calculation more unit independent.


Even More Realistic TrafficEven More Realistic Traffic Models – Cont’d Models – Cont’d

• To normalize the calculation we define: To normalize the calculation we define:

• Define to avoid division by 0.Define to avoid division by 0.• Define to avoid all the traffic to and from Define to avoid all the traffic to and from

small nodes to be 0.small nodes to be 0.• Define traffic by Define traffic by


Even More Realistic TrafficEven More Realistic Traffic Models – Cont’d Models – Cont’d

• Setting the scale factorSetting the scale factor - Set Pop_max =1000- Set Pop_max =1000 - Set Pop_Power = 1- Set Pop_Power = 1 - Set dist_max = 500- Set dist_max = 500 - Set Dist_Power = 1- Set Dist_Power = 1• We want to get 500 bps of traffic between 2 We want to get 500 bps of traffic between 2

sites of population 500 with distance of 250.sites of population 500 with distance of 250.• Select = 1000.Select = 1000.


Traffic Normalization:Traffic Normalization: Uniform Normalization Uniform Normalization Example 4.1.Example 4.1. A A company has 50 sites linked by 85 E1 company has 50 sites linked by 85 E1

lines. The average number of hops in a route is 2.75. Links lines. The average number of hops in a route is 2.75. Links have an average utilization of 55%. What value of have an average utilization of 55%. What value of should be chosen to generate the traffic?should be chosen to generate the traffic?

Total traffic is Total traffic is (85x2x2,048,000x0.55)/2.75 = 69.632 Mbps(85x2x2,048,000x0.55)/2.75 = 69.632 Mbps Let where is the traffic to Let where is the traffic to be carried from node to nodebe carried from node to node Choose = 69,632,000/TChoose = 69,632,000/T Works for all values of Pop_Power and Dist_Power.Works for all values of Pop_Power and Dist_Power.


Traffic Normalization: Row Traffic Normalization: Row

Normalization Normalization Example 4.2.Example 4.2. A company has 50 sites linked by a A company has 50 sites linked by a

network. At each site we observe the total flow of traffic network. At each site we observe the total flow of traffic out to the network, . We wish to normalize the out to the network, . We wish to normalize the requirements so that the traffic out to the network for requirements so that the traffic out to the network for each site matches the observations. each site matches the observations.

The traffic from node to all other nodes isThe traffic from node to all other nodes is

We can define We can define


Traffic Normalization: RowTraffic Normalization: Row Normalization – Cont’d Normalization – Cont’d

• Normalization will be different at each node.Normalization will be different at each node.• Row normalization allows us to match the row Row normalization allows us to match the row

sums to the observed flows.sums to the observed flows.• Row normalization is as the matrix multiplication Row normalization is as the matrix multiplication

of the diagonal matrix with the matrix of the diagonal matrix with the matrix

• If measurements available only at some of the If measurements available only at some of the sites it is reasonable to use the average of these sites it is reasonable to use the average of these values for the rest of the nodes in the values for the rest of the nodes in the network.network.


Traffic Normalization: Row Traffic Normalization: Row

and Column and Column NormalizationNormalization Example 4.3. Example 4.3. A company has 50 sites linked by a A company has 50 sites linked by a

networknetwork. . At each site we observe the total flow of traffic At each site we observe the total flow of traffic out to the network, .We also observe the traffic out to the network, .We also observe the traffic flow in from the network We wish to normalize flow in from the network We wish to normalize the requirements so that the traffic out to the network and the requirements so that the traffic out to the network and in from network at each node matches the observations.in from network at each node matches the observations.

Theorem 4.1. Theorem 4.1. If , If , then the normalization problem has no solution.then the normalization problem has no solution.

Proof:Proof: Obvious, argue using sources and sinks. Obvious, argue using sources and sinks.


Traffic Normalization: Row Traffic Normalization: Row and and Column Normalization - Column Normalization - Cont’dCont’d

• Rowcol.c normalizes the requirements matrix to Rowcol.c normalizes the requirements matrix to match row and column sums.match row and column sums.

• No guaranteed success, especially when there are No guaranteed success, especially when there are too many zeros in the traffic matrix too many zeros in the traffic matrix

• Works well with the traffic generated by our basic Works well with the traffic generated by our basic population and distance nodes.population and distance nodes.

• Uses row_scale and col_scale to measure how far Uses row_scale and col_scale to measure how far each row and column are from the desired totals.each row and column are from the desired totals.

• Variable max_scale gives the max value of any of Variable max_scale gives the max value of any of the scaling factorsthe scaling factors

• Iterates until have converged by having max_scale Iterates until have converged by having max_scale betweenbetween 0.999 and 1.001 0.999 and 1.001


The Final Traffic The Final Traffic Generator:Generator: The Level Matrix The Level Matrix

• Introduced so far traffic generators produced Introduced so far traffic generators produced symmetric traffic – symmetric traffic –

• Often traffic is asymmetric – terminal traffic, Often traffic is asymmetric – terminal traffic, Web browsing, database query.Web browsing, database query.

• To allow for asymmetric traffic we introduce To allow for asymmetric traffic we introduce the notion of levels and level matrix. the notion of levels and level matrix.

Definition 4.1.Definition 4.1. A level matrix will be an A level matrix will be an matrix with entries in .matrix with entries in .


The Final Traffic The Final Traffic Generator:Generator: The Level Matrix – Cont’d The Level Matrix – Cont’d Example:Example: - suppose type 1 nodes are requesting Web - suppose type 1 nodes are requesting Web

pagespages - suppose type 2 nodes are supplying Web - suppose type 2 nodes are supplying Web

pages pages - suppose each incoming page is 3 times larger - suppose each incoming page is 3 times larger

than the requestthan the request - the level matrix is - the level matrix is


The Final Traffic The Final Traffic Generator:Generator: The Level Matrix – Cont’d The Level Matrix – Cont’d

• Further modification of the basic network Further modification of the basic network generation formulageneration formula

• If is of level 1 and is of level 2 If is of level 1 and is of level 2

• If is at level 1, thenIf is at level 1, then


The Final Traffic The Final Traffic Generator:Generator: The Level Matrix – Cont’d The Level Matrix – Cont’d

• Realistic network traffic is event driven.Realistic network traffic is event driven.• There is certain degree of randomness (unexpected There is certain degree of randomness (unexpected

forces).forces).• Let rand() is a pseudo-random generator that produces Let rand() is a pseudo-random generator that produces

strings of numbers in the range [0, RAND_MAX].strings of numbers in the range [0, RAND_MAX].• We model randomness by adding random factor We model randomness by adding random factor • Absolutely, positivelyAbsolutely, positively the final traffic generatorthe final traffic generator::


The Final Traffic The Final Traffic Generator: Generator: The Level Matrix – Cont’d The Level Matrix – Cont’d• The final traffic generator can be combined The final traffic generator can be combined

with any of the 3 normalizations.with any of the 3 normalizations.• ROW/COL normalization may have some ROW/COL normalization may have some

problems with convergence.problems with convergence.• Example – use the level matrix from slide 36 Example – use the level matrix from slide 36

for the following data:for the following data:


Traffic Generators and Traffic Generators and Sensitivity Analysis Sensitivity Analysis

• Useful to create traffic suites rather than just a Useful to create traffic suites rather than just a single set of traffic.single set of traffic.

• Allow us to assess the capability of the network to Allow us to assess the capability of the network to respond to a change.respond to a change.

• With network generator easy – add additional load With network generator easy – add additional load and see if the network still performs correctly and see if the network still performs correctly

• Sensing the capacity is important if there is a Sensing the capacity is important if there is a requirement not to redesign for a period of timerequirement not to redesign for a period of time

• If traffic grows 20% per year and if the network If traffic grows 20% per year and if the network begins to falter at 35% - not a good candidate begins to falter at 35% - not a good candidate network.network.

6/17/ 2002EMIS 8392 Maya Petkova1 Chapter 4 Traffic and Cost Generators.

Documents