Optimization evaluation system, optimization evaluation device, optimization evaluation method, and optimization evaluation program

• US 8,493,882 B2
• Filed: 04/09/2009
• Issued: 07/23/2013
• Est. Priority Date: 05/29/2008
• Status: Active Grant

• Current Assignee: NEC Corporation
• Original Assignee: NEC Corporation
• Inventors: Miyoshi, Kazunori
• Primary Examiner(s): Lai, Andrew
• Assistant Examiner(s): Nguyen, Chuong M
• Application Number:
  US12/991,779
• Publication Number:
  US 20110058498A1
• Time in Patent Office: 1,566 Days
• Field of Search: 370230-236, 370250-252, 370412-470, 709220-224
• US Class Current: 370/253
• CPC Class Codes: H04L 41/14 Network analysis or design
• CPC Class Codes: H04L 43/045 for graphical visualisation of monitoring data
• CPC Class Codes: H04L 43/0829 Packet loss
• CPC Class Codes: H04L 45/00 Routing or path finding of packets in data switching networks routing or path finding in wireless networks H04W40/00
• CPC Class Codes: H04L 45/12 Shortest path evaluation
• CPC Class Codes: H04L 45/123 Evaluation of link metrics techniques for monitoring network metrics H04L43/08
• CPC Class Codes: H04L 45/125 based on throughput or bandwidth
• CPC Class Codes: H04L 45/70 Routing based on monitoring results

Abstract

Provided is an optimization evaluation system, wherein the effect of the optimization of traffic characteristic in a communication network is quantitatively evaluated. An optimization evaluation system which evaluates effects of optimization done by an optimization function in a communication network that has the optimization function for optimizing a communication traffic characteristic, including: a communication traffic analyzing module which acquires a communication traffic variation distribution based on measured communication traffic data; and a communication traffic evaluation module which executes processing for quantitatively evaluating the effects of the optimization of the communication traffic characteristic executed by the optimization function based on a fact whether or not the variation distribution calculated by the communication traffic analyzing module is a power function.

Claims

1. An optimization evaluation system which evaluates effects of optimization done by an optimization function in a communication network that has the optimization function for optimizing a communication traffic characteristic, the system comprising a traffic monitor that is disposed at a router located at two hops or more stages before a band control device to observe a total communication traffic of outlet line of the router, wherein

the traffic monitor includes:

a communication traffic measuring module which observes the communication traffic and measures packet size or packet arrival interval of a communication packet;

a communication traffic analyzing module which rebuilds observation data for each of the traffic characteristics based on a measurement result acquired by the communication traffic measuring module, supplies the traffic characteristic to be optimized that can be expressed by the traffic variation distribution and a variable, and calculates the variation cumulative probability density distribution of the traffic characteristic which satisfies a first condition in which the variable is less than or equal to a value or a complementary distribution of the variation cumulative probability density distribution of the traffic characteristic which satisfies a second condition in which the variable is greater than or equal to the value;

a memory module which stores threshold values of each of the traffic characteristics, the traffic characteristics, and observation start time as well as observation end time of each packet;

a communication traffic evaluation module which compares the observation data of each traffic characteristic to a power function with an exponent of −

1 to −

1.3 when an origin is taken as 10 to power of 0 by using an analyzed result acquired by the communication traffic analyzing module and information stored in the memory module, judges whether the observed communication traffic is in a state having a margin with respect to limits of the buffer of the router or the communication band or in a state having congestion generated by exceeding the limits of the buffer or the communication band, and acquires the extent of the margin and the extent of the congestion as an evaluation result;

a feedback module which performs processing for feeding back the evaluation result judged or acquired by the communication traffic evaluation section to a device or a system which has an optimization function; and

a center node, wherein;

the traffic monitor is one of a plurality of traffic monitors;

the traffic monitor transmits information showing the variation cumulative probability density of the communication traffic characteristic measured by the communication traffic measuring module to the center node; and

the center node detects errors among the traffic monitors by comparing the information showing the cumulative probability density distribution received from each traffic monitor.

2. The optimization evaluation system as claimed in claim 1, wherein the traffic evaluation module:

judges whether or not the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 fits the variation cumulative probability density distribution of the communication traffic characteristic measured by the communication traffic measuring module within a range of 1 decade;

judges as being in a state showing an optimum traffic efficiency by taking a fact that the power function fits the cumulative probability density distribution within the range of 1 decade as a condition that the variation cumulative probability density distribution of the observed communication traffic characteristic exhibits a power law; and

acquires, as a scaling region, a fitting range of the power function and the cumulative probability density distribution as a comparison region when executing comparison processing with respect to a variation cumulative probability distribution of a communication traffic characteristic that is acquired by another measurement.

3. The optimization evaluation system as claimed in claim 2, wherein

the traffic evaluation module calculates the extent of the margin or the extent of the congestion of the communication traffic by acquiring a difference between the power function and the variation cumulative probability density distribution of the communication traffic characteristic acquired by the measurement in the scaling region where the cumulative probability density distribution is the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 by using an average value on a horizontal axis of the scaling region.

4. The optimization evaluation system as claimed in claim 3, wherein

the traffic evaluation module calculates the extent of the margin or the extent of the congestion of the communication traffic by acquiring a difference between the power function and the variation cumulative probability density distribution of the communication traffic characteristic measured by the communication traffic measuring module in the scaling region where the cumulative probability density distribution is the power function with the exponent of −

1 when the origin is taken 10 to the power of 0 by using the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 and some of the values on the horizontal axis of the scaling region in the cumulative probability density distribution of the data measured by the communication traffic measuring module.

5. The optimization evaluation system as claimed in claim 3, wherein

the traffic evaluation module calculates the extent of the margin or the extent of the congestion of the communication traffic by acquiring a difference between the power function and the variation cumulative probability density distribution of the communication traffic characteristic measured by the communication traffic measuring module in the scaling region where the cumulative probability density distribution is the power function with the exponent of −

1 by using the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 and a greatest value on the horizontal axis of the scaling region in the cumulative probability density distribution of the data measured by the communication traffic measuring module.

6. The optimization evaluation system as claimed in claim 3, wherein

the traffic evaluation module calculates the extent of the margin or the extent of the congestion of the communication traffic by acquiring a difference between the power function and the variation cumulative probability density distribution of the communication traffic characteristic measured by the communication traffic measuring module in the scaling region where the cumulative probability density distribution is the power function with the exponent of −

1 as an integration value of differences between the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 and the cumulative probability density distribution of the data measured by the communication traffic measuring module within the scaling region.

7. The optimization evaluation system as claimed in claim 2, wherein the traffic evaluation module performs fitting of the variation cumulative probability density distribution of the communication traffic characteristic measured by the communication traffic measuring module and the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 through a curve of a least-square or a linear approximation method by using the value acquired based on the traffic data measured by the communication traffic measuring module, a data result calculated in bilogarithm used for showing a normal power law, or a data result calculated by taking either the vertical axis or the horizontal axis as semilogarithm.

8. The optimization evaluation system as claimed in claim 1, which selectively includes the traffic monitor in advance to a server connected to the router that is selected randomly from the routers as nodes with maximum number of outdegrees by using an estimation result of structure of the communication network searched by using a traceroot command in a case where topology of the measurement-target network is unknown.

9. The optimization evaluation system as claimed in claim 1, which includes the traffic monitor in advance on a server that is connected to the router with a large value of Betweenness function in a case where topology of the measurement-target network is known and the outdegree distribution of the node exhibits the power law.

10. An optimization evaluation device which evaluates effects of optimization done by an optimization function in a communication network that has the optimization function for optimizing a communication traffic characteristic, which is disposed at a router located at two hops or more stages before a band control device, the optimization evaluation device comprising:

a communication traffic measuring module which observes a total communication traffic of outlet line of the router, and measures packet size or packet arrival interval of a communication packet;

a communication traffic analyzing module which rebuilds observation data for each of the traffic characteristics based on a measurement result acquired by the communication traffic measuring module, supplies the traffic characteristic to be optimized that can be expressed by the traffic variation distribution and a variable, and calculates the variation cumulative probability density distribution of the traffic characteristic which satisfies a first condition in which the variable is less than or equal to a value or a complementary distribution of the variation cumulative probability density distribution of the traffic characteristic which satisfies a second condition in which the variable is greater than or equal to the value;

a memory module which stores threshold values of each of the traffic characteristics, the traffic characteristics, and observation start time as well as observation end time of each packet;

a communication traffic evaluation module which compares the observation data of each traffic characteristic to a power function with an exponent of −

1 when an origin is taken as 10 to the power of 0 by using an analyzed result acquired by the communication traffic analyzing module and information stored in the memory module, judges whether the observed communication traffic is in a state having a margin with respect to limits of the buffer of the router or the communication band or in a state with congestion generated by exceeding the limits of the buffer or the communication band, and acquires the extent of the margin and the extent of the congestion as an evaluation result; and

a feedback module which performs processing for feeding back the evaluation result judged or acquired by the communication traffic evaluation section to a device or a system which has an optimization function, wherein;

the optimization evaluation device transmits information showing the variation cumulative probability density of the communication traffic characteristic measured by the communication traffic measuring module to a center node; and

the center node detects errors among a plurality of traffic monitors by comparing the information showing the cumulative probability density distribution received from each traffic monitor.

11. The optimization evaluation device as claimed in claim 10, wherein the traffic evaluation module:

judges whether or not the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 fits the variation cumulative probability density distribution of the communication traffic characteristic measured by the communication traffic measuring module within a range of 1 decade;

judges as being in a state showing an optimum traffic efficiency by taking a fact that the power function fits the cumulative probability density distribution within the range of 1 decade as a condition that the variation cumulative probability density distribution of the measured communication traffic characteristic exhibits a power law; and

acquires, as a scaling region, a fitting range of the power function and the cumulative probability density distribution as a comparison region when executing comparison processing with respect to a variation cumulative probability distribution of a communication traffic characteristic that is acquired by another measurement.

12. The optimization evaluation device as claimed in claim 11, wherein

the traffic evaluation module calculates the extent of the margin or the extent of the congestion of the communication traffic by acquiring a difference between the power function and the variation cumulative probability density distribution of the communication traffic characteristic acquired by the measurement in the scaling region where the cumulative probability density distribution is the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 by using an average value on a horizontal axis of the scaling region.

13. The optimization evaluation device as claimed in claim 12, wherein

the traffic evaluation module calculates the extent of the margin or the extent of the congestion of the communication traffic by acquiring a difference between the power function and the variation cumulative probability density distribution of the communication traffic characteristic measured by the communication traffic measuring module in the scaling region where the cumulative probability density distribution is the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 by using the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 and some of the values on the horizontal axis of the scaling region in the cumulative probability density distribution of the data measured by the communication traffic measuring module.

14. The optimization evaluation device as claimed in claim 12, wherein

the traffic evaluation module calculates the extent of the margin or the extent of the congestion of the communication traffic by acquiring a difference between the power function and the variation cumulative probability density distribution of the communication traffic characteristic measured by the communication traffic measuring module in the scaling region where the cumulative probability density distribution is the power function with the exponent of −

1 by using the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 and a greatest value of the scaling region in the cumulative probability density distribution of the data measured by the communication traffic measuring module.

15. The optimization evaluation device as claimed in claim 12, wherein

the traffic evaluation module calculates the extent of the margin or the extent of the congestion of the communication traffic by acquiring a difference between the power function and the variation cumulative probability density distribution of the communication traffic characteristic measured by the communication traffic measuring module in the scaling region where the cumulative probability density distribution is the power function with the exponent of −

1 as an integration value of differences between the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 and the cumulative probability density distribution of the data measured by the communication traffic measuring module within the scaling region .

16. The optimization evaluation device as claimed in claim 11, wherein the traffic evaluation module performs fitting of the variation cumulative probability density distribution of the communication traffic characteristic measured by the communication traffic measuring module and the power function with the exponent of −

1 when the origin is taken as 10 to the power of 0 through a curve of a least-square or a linear approximation method by using the value acquired based on the traffic data measured by the communication traffic measuring module, a data result calculated in bilogarithm used for showing a normal power law, or a data result calculated by taking either the vertical axis or the horizontal axis as semilogarithm.

17. The optimization evaluation device as claimed in claim 10, which is selectively disposed in advance to a server connected to the router that is selected randomly from the routers as nodes with maximum number of outdegrees by using an estimation result of structure of the communication network searched by using a traceroot command in a case where topology of the measurement-target network is unknown.

18. The optimization evaluation device as claimed in claim 10, which is disposed in advance on a server that is connected to the router with a large value of Betweenness function in a case where topology of the measurement-target network is known and the outdegree distribution of the node exhibits the power law.

19. An optimization evaluation method which evaluates effects of optimization done by an optimization function in a communication network that has the optimization function for optimizing a communication traffic characteristic, the method implemented in a traffic monitor that is disposed at a router located at two hops or more stages before a band control device to observe a total communication traffic of outlet line of the router, wherein

the traffic monitor executes:

processing that observes the communication traffic and measures packet size or packet arrival interval of a communication packet;

processing that rebuilds observation data for each of the traffic characteristics based on the measurement result, supplies the traffic characteristic to be optimized that can be expressed by the traffic variation distribution and a variable, and calculates the variation cumulative probability density distribution of the traffic characteristic which satisfies a first condition in which the variable is less than or equal to a value or a complementary distribution of the variation cumulative probability density distribution of the traffic characteristic which satisfies a second condition in which the variable is greater than or equal to the value;

processing that compares the observation data of each traffic characteristic to a power function with an exponent of −

1 when an origin is taken as 10 to power of 0 by using stored threshold values of each traffic characteristic, the traffic characteristics, observation start time as well as observation end time of each packet, and a calculated result of the complementary distribution, judges whether the observed communication traffic is in a state having a margin with respect to limits of the buffer of the router or the communication band or in a state having congestion generated by exceeding the limits of the buffer or the communication band, and acquires the extent of the margin and the extent of the congestion as an evaluation result; and

processing that feeds back the judged or acquired evaluation result to a device or a system which has an optimization function, wherein;

the traffic monitor is one of a plurality of traffic monitors;

the traffic monitor transmits information showing the variation cumulative probability density of the communication traffic characteristic measured by the communication traffic measuring module to a center node; and

the center node detects errors among the traffic monitors by comparing the information showing the cumulative probability density distribution received from each traffic monitor.