BACKGROUND IMAGE: stock.adobe.com
5G wireless technology clearly enables more phone calls over a broader area range with increased connection reliability,...
but it's also designed with more than those capabilities in mind. However, for 5G to fulfill next-level capabilities, networks will need support from cloud-native infrastructures.
Technology expectations for 5G
Fundamental expectations of 5G technology call for it to deliver massive volumes of data at extremely high data rates over a large number of connections, with high reliability and low latency. 5G is designed to serve as the wireless network technology for IoT and should have countless other uses, including massive arrays of sensing devices, ultrahigh-volume data and streaming media transmission, self-operating vehicles, crucial emergency services and other mission-critical technologies.
A 5G network architecture relies on network slicing to deal with high external demands. These slices are functionally defined virtual regions of the network dedicated to handling the speed, volume, reliability and latency needs of different use categories. Network slicing requires a scalable, software-defined infrastructure -- one equally adept at handling low-volume, high-speed, low-latency use cases as it is at handling conditions involving a high volume of users and traffic, as well as rapidly moving users and hotspots.
While many of the basic standards of 5G implementation understandably center on the choice of transmission bands and radio technology, any working 5G implementation highly depends on the IT infrastructure that supports it, which is made clear by industry standards group Next Generation Mobile Networks (NGMN).
The requirements for 5G, as described by NGMN, include the virtualization of core communication network functions with software-defined, information-centric network technology across the communications pipeline.
5G network architecture needs microservices
Many key players in 5G technology -- including Huawei, in its white paper on 5G network architecture -- emphasize that cloud-native infrastructure is a fundamental requirement for any high-speed, high-volume, virtualized communication network. For 5G to work as intended, a complex ecosystem of services residing in the cloud must support, manage and maintain it.
Microservices is the ideal cloud-based architecture for 5G, rather than a monolithic architecture. Only microservices can properly support a 5G network architecture, because no set of monolithic applications can deliver the same requirements of responsiveness, flexibility, updatability and scalability that 5G demands.
Virtualized network services also must adapt to new technologies and demands on the system as they come along. With a microservices-based architecture, this is a relatively easy task, accomplished via changes to individual microservices rather than the whole system. The technologies included in 5G will likely change rapidly after the initial rollout, so this kind of adaptability is a necessity.
Additionally, signal-related expectations of 5G, such as high availability, require the kind of flexibility that microservices can deliver. According to NGMN, remote-location equipment should be self-healing, which means it requires flexible, built-in, AI-based diagnostic and repair software capable of at least re-establishing lost communication when isolated. For advanced self-repair, remote-location equipment also requires sophisticated cloud-based services that include the ability to deliver frequent updates to the remote system.
In many ways, self-repairing remote equipment is a specialized version of general support for IoT, which is a key expectation of 5G. IoT support requires a high level of scalability in terms of resources, delivering only needed services to each device. IoT support also requires scalability for handling connections, shifting quickly, if necessary, from a relatively small number of devices -- each of which produces a limited amount of data -- to a high volume of connections, producing high volumes of data with rapid response demands. Microservices can handle this kind of scenario.
Core communication service providers and providers of peripheral services should also favor microservices for supporting 5G, due to bottlenecks. The high speeds required by 5G are likely to expose the existence and location of any bottlenecks in current communications pipelines and could put the providers, which are responsible for such slowdowns, under pressure to remedy them. The earlier that providers identify this problem, the earlier they can switch from the monolith to a cloud-based microservices architecture. Get ahead of the game, and prepare to greet 5G with the type of virtualized, microservices-based infrastructure that it needs.