The SDN-based IP+optical solution provides multi-layer protection, which protects services without wasting resources and reduces TCO. IP+optical synergy enables carriers to fully utilize any means available, such as optical fiber redundancy, router port redundancy, and router redundancy, to improve service availability.

Optical Fiber Redundancy Protection

Multi-Layer Restoration by Optical ASON (MLR-O) attempts to restore service paths by optical layer ASON.

Compared with IP layer protection, optical layer protection is more economic and effective. GMPLS/ASON enables the control plane of the optical network to dynamically change optical paths.

When a fault occurs on the network, GMPLS/ASON restores services by rerouting. Compared with the traditional protection mode, rerouting protects services without wasting resources.

As shown in the following figure, an LSP traverses A, D, G, and I. When a fault occurs on the link between D and G, the rerouting process starts:

1.         D notifies its GMPLS module of the fault after detecting the fault.

2.         The GMPLS module on D detects the affected service and sends a notification to A.

3.         Upon receipt of the notification, As GMPLS module calculates an end-to-end path and sends a PATH message over the new path to set a restoration path for I.

4.         Upon receipt of the PATH message, Is GMPLS module returns an RESV message over the new path to A. Upon receipt of the RESV message, A enables the alarm function and sends a PATH message to downstream devices, requesting these devices to enable their alarm functions for the new path.

5.         After all devices on the new path have their alarm functions enabled, the original LSP in non-revertive mode is deleted. The rerouting process is then complete.

NOTE:

Comparatively speaking, IP layer 1:1 protection is faster than ASON rerouting protection. Therefore, use IP layer 1:1 protection for high-priority services and ASON rerouting protection for low-priority services.

As shown in the following figure, two service links exist between routers A and B, one for high-priority services (marked red), and the other for low-priority services (marked blue). In the traditional solution, two protection IP links are required to ensure that the service switching time is less than 50 ms and packet loss duration is also less than 50 ms.

In real-world situations, only the high-priority IP link needs rapid service switching, and the low-priority IP link can use ASON rerouting protection.

Compared with complete IP link protection, IP+optical protection reduces required network investments. For example, in this scenario, only six IP ports are needed to implement the functions implemented by eight IP ports in the traditional protection solution.

Multi-layer protection offers the following benefits to carriers:

• Rapid recovery for high-priority services: 1:1 TE HSB ensures that high-priority services can be switched to the protection link within 50 ms when the working link fails.
• Reduced protection costs for low-priority services: Low-priority services can be restored through ASON rerouting without using too many IP ports.

Interface Redundancy Protection

Multi-layer Restoration by Router Port (MLR-P) restores service paths by sharing IP protection ports.

Generally, IP links are deployed in 1:1 redundancy mode for protection, wasting link resources. In real-world situations, not all IP links require 1:1 protection.

Can multiple working IP links share the same protection IP link? The answer is yes.

As shown in the following figure, two IP links (marked yellow and blue respectively at both ends) share the same protection link (marked green at both ends).

The SDN-based IP+optical solution provides the IP port redundancy architecture, which allows a unidirectional protection link to be shared by N working IP links. If a working IP link fails, the SDN controller switches services from this link to the protection link. Specifically, the SDN controller completely replicates parameter settings of the failed IP link to the shared protection IP link and binds this link to the corresponding optical path. Compared with 1:1 protection, this implementation saves (N-1) x 2 router and optical ports.

 

The following describes another MLR-P scenario.

The preceding figure shows a typical dual-plane IP core network where PEs are dual-homed to Ps. Two services (with ports marked yellow and blue respectively) are deployed on the network. Port redundancy protection is provided for each service in addition to P-level protection. The SDN-based IP+optical solution provides global protection for the two services:

• Optical-layer fault: ASON rerouting.
• PE1 -> P1 link fault: PE1 switches services to P2.
• P1 -> N1 link fault: P1 switches services to the shared protection link.
• P3 -> N3 link fault: P3 switches services to the shared protection link.
• P1 fault: P2 opens a new port for services switched from P1.
• P3 fault: P4 opens a new port for services switched from P3.

Port redundancy protection offers the following benefits to carriers:

• Simplified network planning: For a dedicated router, unidirectional IP link protection only needs an additional IP port, and network complexity does not need to be considered.
• Reduced investments: The traditional solution provides 1:1 protection at both the IP and optical layers. In some scenarios, services are overprotected, wasting resources. The IP+optical solution allows you to use optical layer resources to protect IP services, reducing network investments.

Router Redundancy

Multi-layer Restoration by Router Aggregation (MLR-A) restores service paths by redundant aggregation routers.

In some scenarios, due to reasons such as lack of optical-layer resources and node faults, some IP links cannot be restored and need node redundancy for service protection. As shown in the following figure, multiple paths exist between R1 and R5. The working path traverses R2 and the protection path traverses R4. If the IP link between R1 and R2 cannot be restored and the working path takes a detour to R3, this is called router redundancy protection.

In the preceding scenario, three levels of protection are available:

• Level 1: TE FRR

High-priority services from R1 to R4 use TE link protection. R4 provides service protection, not restoration.

The working path is R1 -> N1 -> N4 -> R2 -> R5, and the protection path is R1 -> N1 -> N2 -> R4 -> R5.

• Level 2: ASON rerouting

If the optical fiber between N1 and N4 fails, the optical layer performs ASON rerouting. After the ASON finds an optical path between N3 and N4, the IP link from R1 to R2 recovers.

The working path is N1 -> N4, and the protection path is N1 -> N3 -> N4.

• Level 3: router redundancy

An additional router R3 serves as the redundant router. If the optical fiber between N3 and N4 fails, the IP link from R1 to R2 cannot be restored with level-1 or level-2 protection, and R3 becomes necessary for service recovery.

The working path is R1 -> N1 -> N4 -> R2 -> R5, and the protection path is R1 -> N1 -> N3 -> R3 -> R5.

In normal situations, the IP layer uses TE FRR to protect high-priority services and the optical layer uses ASON rerouting to protect low-priority services. If TE FRR or ASON rerouting fails, router redundancy provides another service recovery method.