iMaster NCE: Driving IP Network Automation with Path Computing Element

At MPLS SD & AI Net World Congress 2022, Huawei introduced Path Computing Element, a key capability that empowers the iMaster NCE IP network automation solution, and demonstrated the capability in commercial cases.

With rapidly emerging new services like 5G and cloud, operators urgently require an automated network that can satisfy stringent SLAs. According to a survey by TM Forum, 43% of global operators believe that current network automation capabilities fall short of providing agile, real-time, and differentiated services.

In particular, automatic scheduling of network traffic is a major challenge.

The Automation You Need

Huawei launched Path Computing Element to automatically schedule network traffic throughout the lifecycle, helping customers build next-generation automated IP networks.

Path Computing Element has three highlights:

• Visualized digital map: displays multi-dimensional indicators, such as latency and bandwidth utilization, in real time. The latency circle function helps users with accurate service planning.
• Navigation-based path computation: Network paths are automatically planned, while also satisfying differentiated service SLAs.
• Minutes-level automatic optimization: continuously satisfies service SLAs.

How Path Computing Element Works

Built on three core technologies — a combination of 15+ path computation factors, flexible multi-vendor device control, and the management of millions of tunnels — Path Computing Element truly implements network-wide control and E2E path management, unleashing the value of automated networks.

A Combination of 15+ path computation factors

New services, such as HD live broadcast, cloud gaming, and AR/VR, have led to increasingly diverse requirements on network bearer quality.

To meet diverse service requirements, operators must provide differentiated products, as path computation based on a single factor cannot meet complex service quality requirements. By using a Huawei-proprietary cloud-map algorithm, Path Computing Element offers more than 15 path computation factors such as bandwidth, latency, traffic, and availability. These can be flexibly combined to compute the optimal path that best matches customers’ service intent.

Operators can perform more refined service classification, so that services are carried on network paths with different SLAs, including bandwidth, latency, and packet loss rate. Service SLAs are satisfied with optimal resource utilization, enabling operators to provide diversified network products.

Flexible multi-vendor device control

Operator networks usually consist of devices from multiple vendors. To implement E2E path control on the entire network, multi-vendor device control is a must. The significant differences in the configurations of multi-vendor devices mean that conventional script configuration incurs a prohibitively heavy workload subject to frequent configuration changes.

To address this issue, Path Computing Element uses standard protocols and an open multi-vendor framework to quickly adapt to multi-vendor devices, construct a complete path computation topology, and implement E2E path control network-wide.

It complies with more than 70 standard protocols through which it collects topology information and delivers paths. The protocols standardize multi-vendor interoperability, reducing integration costs by 50%. And the open multi-vendor framework makes device driver development fast and cost-effective, loading driver packages online and proactively adapting to the differences between multi-vendor devices.

Owing to its global commercial use, Path Computing Element has accumulated an extensive set of multi-vendor control capabilities. At European Advanced Networking Test Center (EANTC), the solution successfully completed interoperability tests with devices from different vendors.

Management of millions of tunnels

In the 5G and cloud era, integrated transport networks are becoming increasingly complex in structure. To satisfy the differentiated SLAs of tenants, each service exclusively occupies a tunnel, increasing the number of tunnels by several hundred times. To implement E2E control on network-wide paths, the controller must manage a massive number of tunnels. Path Computing Element supports millions of tunnels and takes just moments to compute paths network-wide, which can even meet future requirements of large-scale network management.

E2E management is truly achieved through just one controller.

Knowing that network automation is where the industry is heading, global operators are building their own next-generation automated networks. Telecom Malaysia (TM) used Huawei iMaster NCE to build its Next Generation Transport (NGT) network with E2E automation.

“Huawei iMaster NCE integrates multi-vendor IP devices and offers whole network control capabilities, so it can provide a real global network view and real end-to-end management. NCE supports an optimal path computation algorithm to meet different service quality requirements. NCE also supports service agility to quickly meet changing business requirements.”

Nariman Razali, Vice President, Network Planning Department, Telecom Malaysia

To date, Path Computing Element has been deployed at more than 100 commercial sites around the world, enabling top operators in Asia Pacific, South Africa, and West Europe to build their next-generation automated networks.

Click here for more information about the Huawei iMaster NCE-IP solution.

Further Reading

• Huawei iMaster NCE: Accelerating Network Automation & Agility for Carriers
• The “DoubleS” Approach to Intelligent Optical Networks

Disclaimer: Any views and/or opinions expressed in this post by individual authors or contributors are their personal views and/or opinions and do not necessarily reflect the views and/or opinions of Huawei Technologies.