Dynamic policy engine for multi-layer network management

Patent metadata

Field	Value
Publication number	US-9819565-B2
Application number	US-201514605161-A
Country	US
Kind code	B2
Filing date	Jan 26, 2015
Priority date	Jan 26, 2015
Publication date	Nov 14, 2017
Grant date	Nov 14, 2017

Abstract

Dynamic policy systems and methods for implementing network management of a network include pre-calculating management policies for the network, in an off-line manner, based on past observed network load taking into consideration seasonality of network load, forecasts of the network load, and planned events on the network load; selecting a policy from the pre-calculated management policies, to manage the network in an on-line manner, wherein the selecting is either time-based or performance-based; and performing network management functions in the network utilizing the selected policy.

First claim

What is claimed is:

1. A dynamic policy method for implementing network management of a network, the dynamic policy method comprising:
pre-calculating management policies for the network, in an off-line manner, based on past observed network load taking into consideration seasonality of network load, forecasts of the network load, and planned events on the network load;
selecting a policy from the pre-calculated management policies, to manage the network in an on-line manner, wherein the selecting is either time-based or performance-based; and
performing network management functions in the network utilizing the selected policy to adjust one or more parameters in the network to change operation thereof,
wherein the network management functions are provided through a Software Defined Networking (SDN) management plane, and
wherein the pre-calculated management policies comprise administrative weight policies, routing and wavelength assignment policies, and channel assignment policies.

2. The dynamic policy method of claim 1 , wherein the selected policy is reevaluated over time or based on detecting the network load has changed significantly enough to warrant a policy change.

3. The dynamic policy method of claim 1 , wherein the pre-calculating is performed before real-time deadlines are needed to be met to perform the network management functions.

4. The dynamic policy method of claim 1 , wherein the selected policy is maintained using a longest-hold approach where the selected policy is kept for as long as possible to keep network performance acceptable.

5. The dynamic policy method of claim 1 , wherein the management policies further comprise dynamic pricing policies and zone assignment policies.

6. The dynamic policy method of claim 1 , wherein the selecting is determined using a utility function for the management policies, which is determined through one of a formula, optimization technique, and simulation.

7. The dynamic policy method of claim 1 , wherein the selecting is first time-based, selecting the selected policy based on a time-based policy look up, and second performance-based, adjusting to a different policy if the selected policy is not acceptable based on the network load.

8. The dynamic policy method of claim 1 , wherein the management policies relate to performing Operations, Administration, Maintenance, and Provisioning (OAM&P) in the network, in an automated manner.

9. A management system configured to provide dynamic policy management for implementing network management of a network, the management system comprising:
a processor; and
memory storing instructions that, when executed, cause the processor to
pre-calculate management policies for the network, in an off-line manner, based on past observed network load taking into consideration seasonality of network load, forecasts of the network load, and planned events on the network load,
select a policy of the pre-calculated management policies, to manage the network in an on-line manner, wherein the selecting is either time-based or performance-based, and
perform network management functions in the network utilizing the selected policy to adjust one or more parameters in the network to change operation thereof,
wherein the network management functions are provided through a Software Defined Networking (SDN) management plane, and
wherein the pre-calculated management policies comprise administrative weight policies, routing and wavelength assignment policies, and channel assignment policies.

10. The management system of claim 9 , wherein the selected policy is reevaluated over time or based on detecting the network load has changed significantly enough to warrant a policy change.

11. The management system of claim 9 , wherein the off-line manner is performed before real-time deadlines are needed to be met to perform the network management functions.

12. The management system of claim 9 , wherein the selected policy is maintained using a longest-hold approach where the selected policy is kept for as long as possible to keep network performance acceptable.

13. The management system of claim 9 , wherein the management policies further comprise dynamic pricing policies and zone assignment policies.

14. The management system of claim 9 , wherein the selecting is determined using a utility function for the management policies, which is determined through one of a formula, optimization technique, or simulation.

15. The management system of claim 9 , wherein the selecting is first time-based, selecting the selected policy based on a time-based policy look up, and second performance-based, adjusting to a different policy if the selected policy is not acceptable based on the network load.

16. The management system of claim 9 , wherein the management policies relate to performing Operations, Administration, Maintenance, and Provisioning (OAM&P) in the network, in an automated manner.

17. A network configured to provide dynamic policy management for implementing network management, the network comprising:
a plurality of nodes that are interconnected; and
a controller communicatively coupled to the plurality of nodes, wherein the controller comprises memory storing instructions that, when executed, cause the controller to
pre-calculate management policies for the network, in an off-line manner, based on past observed network load taking into consideration seasonality of network load, forecasts of the network load, and planned events on the network load,
select a policy of the pre-calculated management policies, to manage the network in an on-line manner, wherein the selecting is either time-based or performance-based, and
perform network management functions in the network utilizing the selected policy to adjust one or more parameters in the network to change operation thereof,
wherein the controller is a Software Defined Networking (SDN) controller, and
wherein the pre-calculated management policies comprise administrative weight policies, routing and wavelength assignment policies, and channel assignment policies.

Assignees

• Djukic Petar
• Morris Todd
• Miedema David Jeffrey
• Ciena Corp

Inventors

• Djukic Petar
• Morris Todd
• Miedema David Jeffrey

Classifications

• H04L47/127

  by using congestion prediction · CPC title

• H04L47/12

  Avoiding congestion; Recovering from congestion · CPC title

• H04L41/0893

  Assignment of logical groups to network elements · CPC title

• H04L43/0876 Primary

  Network utilisation, e.g. volume of load or congestion level · CPC title

• H04L41/0894

  Policy-based network configuration management · CPC title