Resilient Semi-Passive Optical Link Failure Protection

Introduction

A recent Cisco VNI Global Mobile Data Traffic Forecast disclosed that in five years there will be 4 billion more mobile-ready devices and connections and an average mobile connection speed will increase by 2.4-fold [1]. The Citrix mobile consumer survey in 2014 revealed that 54% of mobile subscribers would abandon a slow-loading page in less than 10 seconds and 63% of millennials were more likely to blame the mobile network for mobile video stalling [2]. With the advent of smart devices, cloud services, newer technologies for fixed and wireless connectivity and very impatient consumers, there is tremendous pressure to strengthen the access and mobile backhaul segment of the network.

The copper connections to the enterprise and towers can no longer suffice the exploding capacity requirements leading to increased fiber penetration. The best way, as already verified by numerous installations worldwide, is to leverage the existing fiber infrastructure such as fiber to the node or cabinet to provide cost-effective interconnection for financial to business enterprises and additionally for small to macro cellular networks [3].  Together with the multiplexing technology the fiber-based network can be made timeless and cheap improving the power and space requirements.  

The fiber-to-the-x (x: cabinet, antenna, enterprise, etc.), like any other network, is also subject to failures which may arise due to link cuts or equipment malfunctions. Protection planning against such failures and ensuring a high level of network performance is a crucial part of network design, especially for networks where every second of the interruption in the data traffic leads to economic dismay and also exhaustion of valuable resources (time and expert personnel) that would be much appreciated elsewhere. The fronthauling or backhauling network and enterprise access networks can thus benefit from the classical technique of multipath transmission that provides redundant path or backup equipment in the case of failure.

In this paper, the vastly implemented classical optical line protection approach is briefly introduced. A novel semi-passive layout especially appropriate for power and latency constrained fixed and wireless application is introduced. The semi-passive approach is illustrated in detail with application scenarios in enterprise and fronthauling architectures.

Optical Line Protection

It is important for operators that the designed network recovers quickly in an event of failure and keeps disruption of the traffic to its minimum. In practice, implementing protection against fiber failures and link recovery is possible via higher layer protocols however it becomes very slow. Low latency protection is acritical factor to financial and wireless application.

Optical line protection which protects the link by switching at optical layer has therefore received growing attention especially since it is fast and data rate, wavelength and protocol transparent. This protection scheme, referred to as optical protect switching (OPS) is an automatic system used for failure recovery, where the connection needs to be restored within seconds after failure.

There are numerous optical protection switching layouts. In the classical scheme, an active OPS module contains a 3-dB splitter and a switch. The active OPS modules, as depicted in Figure 1, are deployed symmetrically at the two ends of a dual fiber link after the client equipment to protect the client signal. The splitter in the OPS module duplicates the client signal and transmits it on both the working and backup lines. At the receiver the received optical power in the working or primary line is measured and compared to a certain threshold. The OPS switches to the backup line if the detected optical power in the working line is below this pre-set threshold preventing from network disruption.



The OPS can also be used in manual switching mode where the switching occurs when an appropriate command from the user is received. The classical OPS can be implemented for both point-to-point and ring architecture and is independent of data rate, wavelength and protocol. This type of OPS however requires power and management interface on both the ends. For applications where there is a tight space and power constraints at the customer premises, a semi-passive solution is desired.

A semi-passive OPS, as the name describes is a novel scheme where the passive 3-dB coupler is deployed at the customer edge and an active switch is implemented at the central office. The design differs in the amount of capacity and the operational complexity requirement of the application. A single semi-passive OPS module can be de-signed to contain many splitters to increase the density and capacity. In the following sections we illustrate the application of semi-passive OPS for fronthauling and enterprise networks.

Protecting Enterprise Network

All the localized traffic like video, voice and sensitive data from financial institutions to health, insurance and business enterprises to data centers is linked to other locations via fiber networks. Therefore it is of utmost importance that the network is resilient to fiber cuts with less than a second time to recover. Network operators nowadays provide fiber based enterprise services to their customers where the end customers usually own the ONT device at customer premises. In such cases, it becomes difficult for operators to monitor the fiber and to restore a service quickly when a fault occurs.

For the smooth functioning of the network, in this case, the operator would need a complete passive solution at the customer premises so that the fiber fault can be restored with minimal disruption not only to the traffic but also to the customer.  
In Figure 2, an example of a semi-passive OPS deployment for a single fiber bi-directional transmission between an enterprise and central office is depicted. The WDM signal from the CPE is multiplexed to a single fiber. A passive coupler is deployed at the remote customer end and an active switch is implemented at the central office to accomplish optical line protection. At the central office the WDM signal after the multiplexer is connected to the appropriate OLT.

With such a semi-passive OPS architecture, the operational cost is minimized with no power and management requirements for the remote customer node. The latency in the system is also kept to the minimum. The semi-OPS is transparent to wavelength, data rate and protocols giving the operator the flexibility of single solution for all. 



Protecting Front/Backhaul Network

Since the revenue per bit for mobile data service is very low, it is imperative to deploy a solution that is cost effective in both CAPEX and OPEX and that also represents a future proof technology for seamless communication. While cost can be an important deciding factor, optical fiber line protection for small cell to macro cell deployments also needs to be resilient to harsh environmental conditions and insensitive to electrical power outages.

Figure 3 depicts the application of semi-passive OPS in a mobile fronthauling application. This example assumes the use of dual fiber for bi-direction transmission between the central office and the remote radio head (RRH) node. A passive coupler is deployed at RRH node and an active switch is implemented at the central office. A RRH node typically has three or more RRHs. In this example three RRHs are considered which are independently operating in three different wavelengths.

The appropriate wave-length is achieved by inserting the corresponding transceiver at the RRH. With the aid of a multiplexer the three different wavelengths in three fibers are combined into a single fiber. The 3-dB coupler after the multiplexer splits the WDM signal into two working and backup line. At the central office, the active switch detects the signal in the working like and switches to the backup line when fault occurs.



Because of the use of purely passive coupler at the RRH end, with such a semi-passive OPS architecture, no power supply is required at the remote node thus minimizing the operational cost and making the design resilient to electrical power surges. There is also a possibility of passive remote monitoring and management via the implementation of a reflector at the RRH node when desired. This design is appropriate for harsh environment and extended temperature range and can support WDM or Optical Add/Drop (OADM) integration for point-to-point and ring topologies. The baseband unit (BBU) unit at the central office has management network access to the switch to monitor and user defined threshold values.

Conclusion

Since fiber convergence to enterprise and mobile network is a demanding application with changing requirements, resilience in network is clearly an interesting subject matter for all the service operators. While classical OPS provides a good protection against fiber failures, there are applications where a remote customer or RRH node end is desired to be completely passive. This can be either to reduce the cost and power consumption or to make the architecture simple with no management requirements.

Semi-passive OPS accomplishes this by implementing a passive coupler at the remote end and active switch at the central office. Semi-passive OPS is indifferent to wavelength, data rate and protocols and clearly provides an edge to line protection with its economic, robust, low latency and flexible design to decrease the outage time and provide a superior network connection.