Representative bandwidth calculation systems and methods in a network

• US 10,015,057 B2
• Filed: 01/26/2015
• Issued: 07/03/2018
• Est. Priority Date: 01/26/2015
• Status: Active Grant

• Current Assignee: Ciena Corporation
• Original Assignee: Ciena Corporation
• Inventors: Djukic, Petar, Morris, Todd, Armolavicius, Romualdas
• Primary Examiner(s): Rahman, SM
• Application Number:
  US14/604,875
• Publication Number:
  US 20160218943A1
• Time in Patent Office: 1,254 Days
• Field of Search: 709224
• CPC Class Codes: H04L 41/142 using statistical or mathematical methods
• CPC Class Codes: H04L 43/04 Processing captured monitoring data, e.g. for logfile generation
• CPC Class Codes: H04L 43/0894 Packet rate

Abstract

Systems and methods for a representative bandwidth calculation in a network include receiving a request for representative bandwidth information; obtaining network measurements; combining, filtering, and retiming network measurements based on the request to determine the representative bandwidth information. The systems and methods convert the combined, filtered, and retimed network measurements into the representative bandwidth information which has an additive property in an algebraic domain. The additive property enables the representative bandwidth information to be used for admission control, overbooking determinations, network planning, and/or dynamic pricing in the network.

Claims

1. A method for a representative bandwidth calculation in a network, the method comprising:

receiving a request for representative bandwidth information;

obtaining network measurements;

combining, filtering, and retiming the network measurements based on the request, to determine inputs for a determination of the representative bandwidth information, wherein the representative bandwidth information has an additive property in an algebraic domain based on the combining, filtering, and retiming, wherein the additive property comprises associated representative bandwidth information for traffic being additive on a link with their cumulative traffic having a known probability of exceeding link capacity of the link, wherein the representative bandwidth information is an approximation of historical traffic and its forecast to provide the additive property, and wherein the algebraic domain is one of real-number addition using traffic envelopes as input and calculating the statistical bound considering the number of flows traversing a link;

an effective bandwidth domain calculating a statistical bound on flows traversing a link;

or min-plus algebra calculating a statistical bound on flows traversing a link using network calculus;

determining the representative bandwidth information using the inputs through one of i) a traffic envelope representation which determines the representative bandwidth information from a traffic envelope based on the inputs over time and a probability value that aggregate traffic envelope traffic does not exceed capacity of the link, ii) an effective bandwidth representation which determines the representative bandwidth information through trending, de-trending, finding a moment generating function, and reapplying the trend based on the inputs, and iii) a service curve representation using the inputs to derive curves and using the curves to determine the representative bandwidth information; and

utilizing the representative bandwidth information for admission control of an incoming connection.

2. The method of claim 1, wherein the combining, filtering, and retiming comprises combining and filtering the network measurements at a granularity of the network, a sub-network of the network, a link of the network, a virtual network of the network, an end-to-end flow in the network, a customer in the network, or a class-of-service in the network, wherein the granularity is based on the request, and synchronizing sampling times across the network measurements.

3. The method of claim 1, wherein the combining, filtering, and retiming comprises combining and filtering the network measurements to obtain historical time samples corresponding to busy periods in the past, based on the request, and synchronizing sampling times across the network measurements.

4. The method of claim 1, wherein the network measurements are merged for the retiming by combining ordered lists from different sources to a requested granularity level.

5. The method of claim 1, wherein the additive property enables the representative bandwidth information to be used for the admission control, overbooking determinations, network planning, and/or dynamic pricing in the network.

6. The method of claim 1, wherein the representative bandwidth information is for packet data with variable bandwidth and/or connection-based services with fixed bandwidth.

7. The method of claim 1, wherein the representative bandwidth information is based on a traffic envelope representation which is a value that a traffic does not exceed with some probability.

8. The method of claim 1, wherein the representative bandwidth information is based on an effective bandwidth representation which has a probabilistic additive property in an effective bandwidth domain.

9. The method of claim 1, wherein the representative bandwidth information is based on a service curve representation that uses a min-plus algebra domain.

10. The method of claim 1, further comprising:

utilizing the representative bandwidth information to auto-tune overbooking parameters in the network.

11. The method of claim 1, further comprising:

utilizing the representative bandwidth information for admission control to admit or reject a connection in the network.

12. The method of claim 1, further comprising:

utilizing the representative bandwidth information in network planning to reflect true network usage in the network.

13. The method of claim 1, further comprising:

utilizing the representative bandwidth information to determine a price of a connection in the network.

14. The method of claim 1, wherein the representative bandwidth information is determined at various granularity levels in the network from historically observed measurements and future forecasts to automatically adjust overbooking parameters in the network.

15. A server configured to perform a representative bandwidth calculation in a network, the server comprising:

a network interface;

a processor coupled to the network interface; and

memory storing instructions that, when executed, cause the processor to

receive, via the network interface, a request for representative bandwidth information,

obtain network measurements,

combine, filter, and retimed the network measurements based on the request to determine inputs for a determination of the representative bandwidth information, wherein the representative bandwidth information has an additive property in an algebraic domain based on the combined, filtered, and retimed network measurements, wherein the additive property comprises associated representative bandwidth information for traffic being additive on a link with their cumulative traffic having a known probability of exceeding link capacity of the link, wherein the representative bandwidth information is an approximation of historical traffic and its forecast to provide the additive property, and wherein the algebraic domain is one of real-number addition using traffic envelopes as input and calculating the statistical bound considering the number of flows traversing a link;

an effective bandwidth domain calculating a statistical bound on flows traversing a link;

or min-plus algebra calculating a statistical bound on flows traversing a link using network calculus,

determine the representative bandwidth information using the inputs through one of i) a traffic envelope representation which determines the representative bandwidth information from a traffic envelope based on the inputs, ii) an effective bandwidth representation which determines the representative bandwidth information through trending, de-trending, finding a moment generating function, and reapplying the trend based on the inputs, and iii) a service curve representation using the inputs to derive curves and using the curves to determine the representative bandwidth information; and

utilize the representative bandwidth information for admission control of an incoming connection.

16. The server of claim 15, wherein the representative bandwidth information is determined at various granularity levels in the network from historically observed measurements and future forecasts to automatically adjust overbooking parameters in the network.

17. The server of claim 15, wherein the server is either a Software Defined Networking (SDN) controller or coupled to an SDN controller.