The development of 5G networks began to greatly accelerate in 2020. At the 11th Global Mobile Broadband Forum (MBBF) in Shanghai in 2020, Huawei launched "1+N" 5G target networks. In order to embrace the approaching golden decade of 5G, full spectrum evolution and the construction of one high-bandwidth simplified target network capable of ensuring ubiquitous connectivity with on-demand overlay of 'N' capabilities is required. However, many challenges await on the path to "1+N" target network evolution. One such challenge involves effectively deploying BtoB capabilities and maintaining 2G, 3G, 4G, and 5G simultaneously. Evolving towards Autonomous Driving Network (ADN) is one of the best ways to cope with the increasing complexity of mobile network. A network architecture featuring “layered autonomy and vertical collaboration” and scenario –based OPENNESS is required for the “1+N” mobile network moving forwards the intelligent autonomous network.
Architecture of "layered autonomy and vertical collaboration" lays the foundation for wireless autonomous driving networks
The predominant belief within the industry today assumes that fully intelligent autonomous networks will be implemented step by step. To avoid further increasing network complexity, an open architecture featuring hierarchical domain-based autonomy and vertical cross-domain collaboration is required for wireless autonomous driving networks. Huawei launched iMaster MAE (Mobile Automation Engine), based on this open architecture, to help build 5G autonomous network capabilities and ecosystems. Such architecture consists of the cross-domain collaboration layer, single-domain autonomous layer, and NE layer. Open interfaces (APIs and SDKs) can be used between cross-domain closed-loop and single-domain closed-loop to implement mutual collaboration and information exchange. The use of such interfaces to open atomic network capabilities in a more efficient and simplified manner, and to implement collaboration between the layers of mobile networks, is key to the continuous evolution of autonomous driving networks.
Scenario-based openness enables simplified collaboration
How do we define efficient collaboration? Traditional operations support systems (OSSs) for wireless networks have northbound interfaces (NBIs) to connect to upper-layer systems. Interaction between systems involves the delivery of various data such as configuration, performance, fault, and topology etc,. The large amount of data and complex service interfaces are pain points for cross-domain collaboration and the gap that needs to be bridged for moving towards autonomous driving network.
Consequently, a layer of scenario-based interfaces between cross-domain and single-domain systems is built to convert disordered data and complex commands and instructions into atomic capabilities with functional significance. Upper-layer systems can then use these interfaces to invoke the atomic capabilities of lower-layer domains, avoiding large amounts of redundant data processing. We can use Lego blocks as an example: instead of laying out all the small pieces, you can use standardized block modules for different building scenarios. This greatly improves building efficiency and satisfies the personalized requirements of individual builders. Similarly, cross-domain collaboration becomes more efficient and simpler as a result.
Intent-based openness paves the way for full autonomous driving networks
Just as autonomous driving networks cannot be achieved in a single day, openness of network capabilities is also a step by step process. Network capability openness by iMaster MAE can be classified into atomic API openness, scenario-based API openness, and intent-based API openness by automation level of networks and open content. Upper-layer systems can flexibly invoke these APIs to customize scenarios.
Atomic API openness: A lower-layer OSS can open alarm data, traffic statistics, and measurement reports to an upper-layer network management system (NMS) through atomic APIs. Carriers can extract useful information from such data and develop required applications.
Scenario-based API openness: Atomic capabilities in the planning, construction, maintenance, optimization, and operation phases of carrier workflows are open over scenarios-based APIs. Carriers can use DevOps agile development enabled by the scenario-based API ecosystem to customize appropriate O&M automation applications. For example, in a site deployment scenario, atomic capabilities such as data collection, network design, data preparation, data activation, engineering commissioning, and network acceptance can be provided to upper-layer systems through scenario-based APIs. Carriers can invoke these atomic capabilities through upper-layer systems as required, and customize scenario-specific site deployment processes. This eliminates the need for more than ten manual breakpoints and cuts down on frequent manual interactions in the traditional site deployment process.
In country C, iMaster MAE seamlessly interconnects and collaborates with the upper-layer NMS of carriers by taking advantage of scenario-based openness. With automatic detection and identification of hardware, as well as expert global site deployment experience, iMaster MAE improves the deployment efficiency of quality 5G sites by over 50%. Intent-based API openness: The preceding cases of automatic site deployment use network automation capabilities between level 2 and level 3. As the automation level evolves, the generalization of automatic O&M scenarios becomes stronger and the problem handling scope becomes wider. As such, collaboration across systems and domains needs to be simpler and more efficient. With intent-based APIs, upper-layer systems no longer just invoke the atomic capabilities of the lower-layer OSS in planning, deployment, O&M, and operation phases. Instead, they can invoke applications in domains during network construction and operation phases to meet service intent requirements. The ultimate goal of autonomous driving networks is to build an intelligent autonomous network based on such simplified, intent-based openness.
O&M automation in 5GtoB service scenarios is in urgent need of such intent-based openness. If an industry user says "I want to provision machine vision services involving 30 cameras," they expect lower-layer systems to transfer the corresponding configuration details to upper-layer systems and automatically resolve resource and policy conflicts on the network. This also facilitates the deployment of highly reliable 5G 2B networks, perfectly aligning with industry Service Level Agreement (SLA) requirements.
Atomic, scenario-based, and intent-based openness can be encapsulated step by step to reduce the complexity of a single domain or layer, gradually simplifying vertical cross-domain collaboration, and eventually accelerating the development of mobile network autonomous driving.
This content is sponsored by Huawei.