Software Defined Networking (SDN) is an approach to network management that enables dynamic and efficient network reconfigurations to improve network performance, monitoring and minimizes overall network cost. SDN leverages an open protocol and APIs that enable network elements to be directly programmable, agile, centrally managed and programmatically configured. However, while SDN can be applied to the transport, switching and routing networks layers the underlying physical fiber network has remained static.
The need for software control at the network level was driven by trends such as changing traffic patterns, the rise of cloud services and Big Data. The rise of cloud services and widespread use of broadband mobile devices has increased the diversity of access to applications and data while Big Data technologies such as machine learning or natural language processing have created huge datasets with increasing and variable demand on compute power and storage. While SDN networking has improved the ability of networks to handle these recent trends, demands on the network will only continue to increase. Adding machine-learning-based traffic prediction and dynamic fiber cross connects (DFCCs) can bring significant economic benefits as the network scales to meet increasing data traffic demands.
The optical network diagram shown above identifies the components involved in an IP Link over the optical layer. The components in the IP link include:
R1 and R2, router ports in two different locations
T1 and T2, transponders used for electrical-to-optical and optical-to-electrical conversion
Reconfigurable optical add-drop multiplexers (ROADMs), which today can be colorless and directionless (CD-ROADMs)
RE1, an electrical regenerator as required if the link length is beyond the optical reach of the link
Tail 1 and Tail 2, the combination of a connected router port (R) and transponder (T)
Dynamic Fiber Cross Connects (DFCCs), used to dynamically connect the router port and transponder in the tail.
With traditional traffic engineering, a static data tunnel is created in the network to carry end-to-end traffic in the IP layer. If any component fails along the path or if there is a fiber cut, the entire IP link fails and all the components in the link are out of service. However, by utilizing SDN controllers to manage the data and control plane the active components can be disconnected from the link and reused elsewhere in the network. The DFCC extends this benefit to the router and transponder components in the tail. If one of the components fails, the functional component can be reused and combined with another component to form a new tail. Thus, the SDN controller combined with the DFCC can optimize and overcome traffic fluctuations and network failures in real-time.
Dual-routers configurations are typically used in data centers for 1+1 redundancy, with inter-router optical connections used for communication between the routers. The DFCC adds significant value by passing the inter-router connections through the DFCC. If there is a port failure on a router, the transponder connected to the failed router port can be reconfigured to a working router port. Even with a complete router failure or if one of the routers is taken offline for any reason (failure, software upgrade, etc.…), the router ports on the in-service router can be combined with the transponders through the DFCC to create new tails. The DFCC can also be used to connect a transponder to a different router port in the same office if needed for more efficient routing.
The economic benefit of the DFCC in a network has been analyzed by a group at AT&T and presented at OFC’19 (see reference). Scenario 5 in the table below shows the addition of the DFCC which creates an additional 5% saving on the total network cost, mostly due to savings on the transponder and router port costs. For a $100 M network cost, the savings from implementing dynamic fiber cross connects alone is $5 M.
The Telescent G4 Network Topology Manager (NTM) is an example of an innovative DFCC that allows software control of the physical layer while offering remote diagnostic capability such as optical power and OTDR monitoring. The Telescent NTM uses a robot to route a fiber to any of the ports to offer any-to-any connectivity of the 1,008 ports in each system. The NTM can be initially configured with fewer than the full number of ports and upgraded in a pay-as-you-grow manner. Once made, the connections are equivalent to existing fiber patch panel connections with low loss and are fully latched, allowing traffic to continue uninterrupted as the system is upgraded. The Telescent NTM has passed NEBS Level 3 reliability testing as well as multiple vendor specific qualification tests which have demonstrated a >10 year lifetime. Multiple NTMs systems can be managed through software control offering scaling to 10,000 cross-connects and beyond with machine accurate record keeping and limiting stranded capacity.
By combining the Telescent NTM as a dynamic fiber cross connect in an optical network with other recent advances in software defined networking and machine learning, significant economic benefits can be obtained in the network. Contact Telescent to learn more about how the NTM can improve your network.
Reference: OFC 2019: “Joint Optimization of Packet and Optical layers of a Core Network Using SDN Controller, CD ROADMs and machine-learning-based traffic prediction”, Gagan Choudhury, Gaurav Thakur and Simon Tse, https://ieeexplore.ieee.org/document/8696614