Network Automation (2.0) Takes on Service Assurance Challenges

Summary Bullets:

Automated network service provisioning (via SDN technologies) is an established industry norm; however, dealing with service quality assurance requires advanced capabilities such as real-time network visualization, multi-factor path computation, and the ability to dynamically handle high-scale network variations.
Service assurance benefits include optimization of network resources, increased revenue potential, and reduced operational costs, among others; achieving these objectives can yield significant operational efficiency and improved quality of experience for end users.

Operators continue to invest and scale their IP and transport networks to meet growing capacity and new use cases demands, and the need to expand automated network management beyond network element and service provisioning, as well as begin to address service assurance, has become a top priority. Experience has shown that today’s highly complex and dynamic networks rely on automation as the key to successfully delivering high-quality services. In the two most recent networking conferences (i.e., MWC 22 and MPLS SD & AI Net World), network automation demonstrations and proof of concept (PoC) presentations by vendors and network operators alike were front and center and show promise when addressing service quality assurance.

The foundation for network automation has evolved of the past decade, and according to the recent EANTC-Interop Test 2022, significant software-defined networking (SDN) evolution and robustness across vendors was evident, noting that protocols for managing the routing policies are well-instrumented. Network automation, to date, has been primarily focused on network resource configuration and service provisioning, thanks to the technical shift from command-line interfaces and scripts to application programming interfaces and workflows; however, to ensure service quality, solutions must be capable of visualizing network-wide resources and have the capability to dynamically adjust to changing traffic patterns, network anomalies, and provide visibility into network performance and trends.

Today’s IP networks are transitioning quickly to leverage SRv6, network slicing, cloud services, and SDN to deliver advanced services. This level of complexity requires automation to establish and maintain network service level agreements (SLAs). For simplicity, the first phase of network automation can be characterized as Network Automation 1.0; the market is now entering the Network Automation 2.0 phase, which extends addresses service assurance.

Network Automation 2.0 presents the greatest challenges for vendors and operators; solutions should include the following characteristics to adequately address end-to-end service assurance:

Real-time network mapping to enable multi-vendor device control, large-scale network management, and real-time visualization of network-wide resources to expose indicators such as latency, bandwidth utilization, and network anomalies.
Differentiated path computation capabilities and complete path computation algorithms to meet service and product quality requirements, provide mass-scale tunnel management, and support network-wide path management.
Automatic traffic scheduling, real-time service quality monitoring, automatic adjustment to service quality changes, and minutes-level recovery. Path computation algorithms need to combine multiple factors to establish optimal network paths to meet SLA requirements.
The ability to support network slicing, provide real-time control and monitoring of the network resources, and support automated O&M operations to reduce the time to repair.

Vendor solutions have emerged, such as the Path Computing Element newly launched in Huawei’s iMaster Network Cloud Engine IP (NCE-IP), which delivers the following benefits to network operators:

Source: Huawei

Cross-domain network traffic automatic scheduling.
Multi-vendor support using standard network protocols such as BGP-LS and SRv6 to reduce integration time by 50%.
Closed-loop automation to improve the overall service experience.
Centralized traffic engineering to provide a network-wide perspective to achieve resource optimization.
Network digital map to provide real-time network visualization, which displays multi-dimensional indicators such as latency and bandwidth utilization in real time to better control the service lifecycle and maintain optimum network paths.
Optimal path computation algorithm based on more than 15 factors to automatically plan network paths required by services.

Typical use case examples for Network Automation 2.0 using Huawei’s NCE-IP:

Provided automatic optimization for MTN Nigeria to relieve network congestion in minutes caused by fiber cuts and to guarantee user experience.
Enabled SLA on-demand services for TM Malaysia by leveraging one centralized network controller for end-to-end network traffic management.

The use cases noted above demonstrate the network automation can be leveraged to deliver services faster, ensure SLAs are maintained, and drastically reduce time-to-service.