System and method for network performance monitoring

• US 8,213,323 B1
• Filed: 12/18/2003
• Issued: 07/03/2012
• Est. Priority Date: 12/18/2003
• Status: Active Grant

• Current Assignee: T-Mobile Innovations LLC (Deutsche Telekom AG)
• Original Assignee: Sprint Communications Company LP (Deutsche Telekom AG)
• Inventors: Papagiannaki, Konstantina
• Primary Examiner(s): Ngo, Ricky
• Assistant Examiner(s): Kao, Eric
• Application Number:
  US10/737,922
• Time in Patent Office: 3,120 Days
• Field of Search: 370/229, 370/230, 370/231, 370/233, 370/234, 370/235, 370/241, 370/247, 370/252, 370/253, 370/232, 370/236, 370/236.2, 370/241.1, 370/251
• US Class Current: 370/252
• CPC Class Codes: H04L 43/0882 Utilisation of link capacity
• CPC Class Codes: H04L 43/16 Threshold monitoring

Abstract

A method and system are provided for monitoring network performance in order to reveal micro-behavior of network traffic. At least two distinct durations are set for monitoring link utilization. At least one link utilization threshold is set for each duration. Link utilization is monitored at a router using a counter. In addition, a number of durations is provided during which the at least one link utilization threshold has been exceeded.

Claims

1. A method for monitoring network performance in order to reveal micro-behavior of network traffic, the method comprising:

monitoring traffic through a router;

setting at least two distinct time durations over a unidirectional IP link that traverses through the router for monitoring link utilization in the router, wherein the link utilization is a measurement of a use of the unidirectional IP link, wherein the link utilization is monitored simultaneously during each of the at least two distinct time durations, wherein the at least two distinct time durations are based on the size of the network to avoid overloading polled network elements, wherein Global Positioning System time stamps are utilized to monitor a transmission of data through the unidirectional IP link, wherein the router includes a monitoring system that has a processor and a memory, wherein the memory includes timers, statistics calculation tools, and stored statistics, wherein the timers implement different durations of a time scale, and wherein the statistics calculation tool computes link utilization by dividing a collected number of bytes by a capacity of the unidirectional IP link;

adjusting the at least two distinct time durations over the unidirectional IP link so that congestion can be detected;

setting at least one link utilization threshold for each distinct time duration in the router;

monitoring link utilization at the router using a counter;

identifying intervals of time durations where the at least one link utilization threshold has been exceeded; and

providing a number of intervals of time durations during which the at least one link utilization threshold has been exceeded.

2. The method of claim 1, further comprising analyzing collected data for a specific time scale across different utilization thresholds in order to determine an amplitude of a micro-congestion episode.

3. The method of claim 2, further comprising analyzing collected data in order to determine frequency of the micro-congestion episode and analyzing collected data for specific values of the at least one link utilization threshold across multiple time scales to determine duration of a micro-congestion episode.

4. The method of claim 1, further comprising setting the distinct durations to 1 ms and 10 ms.

5. The method of claim 1, further comprising collecting monitored data at a pre-set interval.

6. The method of claim 5, further comprising using a 5 minute pre-set interval.

7. The method of claim 1, further comprising setting the link utilization threshold based on network characteristics.

8. A method for monitoring network performance in order to reveal micro-behavior of network traffic, the method comprising:

monitoring traffic through a router;

setting at least one distinct duration over a unidirectional IP link that traverses through the router for monitoring link utilization at the router wherein the link utilization is a measurement of a use of the unidirectional IP link where data packets are monitored at the input of the unidirectional IP link and the output of the unidirectional IP link, wherein the size of at least one distinct time duration is based on the size of the network to avoid overloading polled network elements, wherein Global Positioning System time stamps are utilized to monitor a transmission of data packets through the unidirectional IP link, wherein the router includes a monitoring system that has a processor and a memory, and wherein the memory includes timers, statistics calculation tools, and stored statistics, wherein the timers implement different durations of a time scale, and wherein the statistics calculation tool computes link utilization by dividing a collected number of bytes by a capacity of the unidirectional IP link;

adjusting the at least one distinct duration so that congestion can be detected;

setting a collection interval for collecting monitored data at the router, wherein the collection interval is larger than the at least one distinct duration; and

analyzing the collected data.

9. The method of claim 8, further comprising setting at least two distinct durations for monitoring link utilization over the unidirectional IP link.

10. The method of claim 9, further comprising setting at least one link utilization threshold for each duration.

11. The method of claim 10, further comprising monitoring link utilization at a the router using a counter.

12. The method of claim 11, further comprising providing a number of durations during which the at least one link utilization threshold has been exceeded.

13. The method of claim 8, further comprising analyzing collected data for a specific time scale across different utilization thresholds in order to determine an amplitude of a micro-congestion episode.

14. The method of claim 8, further comprising analyzing collected data in order to determine frequency of the micro-congestion episode and analyzing collected data from specific values of the at least one link utilization threshold across multiple time scales to determine duration of a the micro-congestion episode.

15. A non-transitory computer-readable medium having computer-executable instructions for performing the method recited in claim 8.

16. A system for monitoring network performance in order to reveal micro-behavior of network traffic, the system comprising:

a processing unit and a memory wherein the memory includes timers, statistics calculation tools, stored statistics, and a data collection apparatus;

the timers implemented by a router for monitoring link utilization during at least two distinct durations over a unidirectional IP link wherein the link utilization is a measurement of a use of the unidirectional IP link, wherein the link utilization is monitored simultaneously during each of the at least two distinct time durations, wherein the at least two distinct time durations are based on the size of the network to avoid overloading polled network elements, and wherein Global Positioning System time stamps are utilized to monitor a transmission of data through the unidirectional IP link;

the statistics calculation tools for implementing a threshold monitor for determining whether link utilization during the at least two distinct durations exceeds the at least two distinct thresholds wherein the statistics calculation tool computes link utilization by dividing a collected number of bytes by a capacity of the unidirectional IP link;

the stored statistics accessible to a network operator;

the data collection apparatus for enabling collection of the link utilization data from the router and for holding data used by the timers and statistics calculation tools;

an operator interface for receiving inputs into the system from a user using an input device such as a keyboard or a pointing device;

a network interface implemented to collect utilization metrics at different time scales; and

a system bus implemented to couple the processing unit, the memory, the operator interface, and the network interface.

17. The system of claim 16, wherein the statistics calculation tools determine link throughput based on collected data.

18. One or more non-transitory computer-readable storage media having computer-useable instructions embodied thereon to perform a method for determining parameters for monitoring micro-congestion episodes in a network, the method comprising:

collecting data over at least one time interval having a specified duration, the data including at least a number of packets entering a unidirectional link of a router and a number of packets exiting the unidirectional link of the router wherein the size of at least one distinct time duration is based on the size of the network to avoid overloading polled network elements, wherein Global Positioning System time stamps are utilized to monitor a transmission of data packets through the unidirectional IP link;

determining link throughput based on the collected data;

adjusting the at least one time interval to detect a delay;

calculating delays for each time interval that is adjusted;

relating link throughput with calculated delays;

selecting at least one monitoring time interval based on the relationship between throughput and calculated delays; and

analyzing the amplitude, frequency and duration of the micro-congestion episodes based on the collected data

wherein the collected data is analyzed for a specific time scale across different utilization thresholds and the collected data is analyzed for specific values of at least one link utilization threshold across multiple time scales.