Cement your C3 with network automation
C3 stands for command, control and communications and is a key aspect of defense operations. Along with AI, cloud and cybersecurity, these are the four priorities identified by the US Department of Defense in its 2019 Digital Modernization Strategy. The report also presents a vision of “a more secure, coordinated, seamless, transparent and cost-effective IT architecture that transforms data into actionable information and ensures dependable mission execution in the face of a persistent cyber threat,” in order to support these four priorities.
There are many parts to this foundational IT architecture, but at its base it will depend on connectivity and communications to be able to leverage data across its entire operations. Along with the key technologies identified by the DoD, such as cloud and AI, defense operations will also adopt 5G networking as the critical “last mile.”
One of the most important aspects of 5G is its ability to provide “slices” for specific users and their applications. A slice is essentially a virtual network partition that contains dedicated resources to constantly meet quality of service (QoS) requirements such as bandwidth, security and latency. Each slice can be configured dynamically to support whatever operational requirements the use case requires, from delay-sensitive virtual reality and drone control communications, to IoT sensors and video streaming from 8k CCTV cameras.
5G slicing opens up a new service paradigm where slices can be quickly created and deleted as missions are launched and completed. Also, the massive bandwidth that 5G enables means that the WAN needs to rely more on optical transport. In order to fully support 5G, the WAN transport layer has to address new challenges.
The first challenge is having the agility to respond to the lifecycle of 5G slices. Can a corresponding transport slice with the right QoS resources be setup and canceled as 5G services come and go? With todays’ IP/MPLS VPN technology, an end-to-end 5G slice would take hours or even days to configure.
The second challenge is how to manage the end-to-end transport across technology domains of the WAN network. Backhauling traffic from the 5G radios would usually require microwave transmission at the edge and optical transport in the core. These layers are normally managed separately and apart from IP/MPLS. With the adoption of cloud computing, AI/ML, IoT and data centers, the nodes are increasing in capacity. The topology is becoming more complex. This increases operational challenges such as multi-path topology discovery and path diversity analysis.
Beyond 5G, an additional challenge is how to accommodate legacy communications needs, especially from those low-speed TDM-based systems, when TDM network equipment is beyond end-of-life and, soon, end-of-support. IP/MPLS service routers can emulate TDM circuits but the migration from older equipment and continuing operations again takes time and relies on TDM expertise, which is dwindling.
In the past, network challenges were met with enhancements or new innovations. For instance, the addition of MPLS to IP networking met the QoS challenge. 5G increases wireless bandwidth and reduces latency. But the nature of these three challenges dictates that we need to change the network operation paradigm — bringing in network automation to help.
Figure 1: WAN architecture for 5G backhaul
Figure 1 shows a typical WAN architecture for 5G backhaul. The WAN network manager plays the role of transport slice controller. The end-to-end network slice orchestrator orders the provisioning of a transport slice through the WAN manager, a RAN slice and a core slice via the other two controllers. It provides the service requirements (also known as service intents) including RAN and core endpoints, as well as transport QoS requirements to the WAN manager. The WAN manager uses its intent-based network automation capability to build a custom transport slice that can meet the QoS requirements of the 5G slice.
The first challenge we identified — provisioning the transport slice at the speed of service — is thus met. Furthermore, as 5G slices are created and user traffic changes in real time, the transport network automatically responds and optimizes to consistently meet the QoS objectives. The WAN manager continually collects telemetry data for network statistics and OAM results, such as delay. It then proactively monitors the delay performance and optimizes the path in the WAN if the performance falls short of the objective.
The cross-domain challenge of managing the service across IP and optical layers is met by the cross-domain network manager, which provides an intuitive visualization of the multi-layer topology by leveraging protocols such as LLDP snooping. It can also compare traffic counts to gather cross-layer connection information to build out a full multi-layer topology. This provides full visualization of both IP and optical topologies in a fully correlated way.
Another highlight of this approach is the ability to analyze service path diversity in a multi-layer IP/optical network, which is pivotal to reliable service delivery. Does an alternate or backup service share the same optical path as the principal service, for instance? Network automation can perform cross-layer path diversity analysis to detect the shared risk. Then the unified network manager can re-reroute the alternate service to offer true path diversity, thereby improving service reliability.
For TDM traffic, the network manager also computes and establishes, if necessary, a network path connecting the two service endpoints. This hides the TDM technology complexities as well as eliminating network and traffic engineering efforts, thus allowing operators to confidently carry out the TDM migration. Network analytics can automate the monitoring of key TDM circuit parameters such as network delay and jitter buffer depth at TDM endpoints. Alarms can be raised when the measurement crosses a pre-configured threshold. This allows operators to take remedial measures, either in automated or manual manner, before the TDM circuits experience service degradation.
As defense forces modernize to strengthen C3 and embrace cloud and AI for data-informed, data-driven decisions, the WAN is critical for the transport of data at speed and scale. But to harness the true benefits of 5G, cloud, AI, IoT and big data, the WAN will also have to undergo its own transformation through automation to provide the additional capabilities that advanced defense applications will need. In this way it will strengthen defense competitiveness and resilience in the face of external threats, as well as boost efficiencies and effectiveness in daily operations.
To learn more read our Automating the WAN for digital defense modernization white paper.