5G Could Usher in New Era of Mobile Cloud
In a survey carried out earlier this year by Heavy Reading, 57% of communications service providers (CSPs) said that entering new markets to generate additional revenue streams is one of their top three incentives for deploying 5G.
But what will CSPs offer that customers can't already get elsewhere? There's no point in deploying new 5G networks if CSPs offer only faster 4G-like services. Public cloud providers like AWS and Microsoft Azure already offer all the bandwidth, storage and compute needed for the most common enterprise applications. Competing on connectivity and bandwidth alone will not be enough.
The real opportunity to differentiate in 5G comes from CSPs' ability to take control of the customer experience. Properly addressing a wide range of vertical industry customers requires a nuanced understanding of their specific needs. Depending on the use cases, services and applications they want to run, many verticals require highly varied and deterministic guarantees on availability and throughout. All these add up to very specific quality of experience (QoE) demands, on a per-customer basis.
Ensuring these QoE demands is dependent on 5G's unique architectural features in two important areas.
Owning the mobile cloud mesh
First, 5G will be a highly distributed series of virtualized functions, spread across core network data centers and devolved to local processing nodes at the "edge." These "mobile edge" nodes and access points will communicate with the core network but also between themselves, creating dynamic meshes of connectivity.
Mesh architecture helps give 5G its latency and speed advantages, and the Heavy Reading survey showed that over 70% of CSPs believe ownership of access and edge cloud assets gives them an advantage over public cloud providers. Ownership allows CSPs to guarantee reliability and speed at the edge, access and core network layers. They can therefore add more value to mobile cloud platforms versus the major providers, which can only really affect service quality in the data center. (Cloud providers tend to see network connectivity as nothing more than an Ethernet cable. They underestimate the complexity of operating reliable, global networks.)
Control at the edge is vital for a wide range of specific use cases and applications where minimal latency and maximum output is essential to the customer experience. This could be smart city transport initiatives and autonomous vehicles, virtual and augmented reality apps in retail environments or e-health services for patients.
Slicing up mobile cloud SLAs
Second, the ability to "slice" the network, and offer low-latency, high-speed and high-capacity services on a per-customer basis, will allow CSPs to deliver a range of performance-driven service level agreements (SLAs) for specific customer needs. However, monetization rests on service reliability and quality that customers are willing to pay for. CSPs will need to commit to commercial SLAs that can be measured and delivered.
Taking control of mobile cloud QoE
With control over 5G's slicing and mobile edge capabilities, CSPs will transform into mobile cloud providers (MCPs), take control of customer QoE and displace public cloud providers that cannot offer the same service guarantees on this scale.
The targets are clear, the features to differentiate are there and the opportunity to succeed is within reach. But switching on 5G and network slicing, and hoping it works "out of the box" is a pipe dream. CSPs will control their own 5G technology deployments, but vertical industries will also require a range of applications to be hosted in a complex mix of hybrid public and private clouds.
5G's own highly dynamic and distributed mesh architectures, combined with broader cloud connectivity, introduces significant management complexity; if CSPs are serious about using QoE to differentiate and become MCPs, it is essential to understand the user experience and assure service quality and SLAs for mission-critical or low-latency services from end to end, across virtualized, cloud and physical networks, and especially on a "per slice, per customer" basis.
Practically, this can only be achieved with unified application, service and network visibility and trusted analytics. Heavy Reading's survey showed that 74% of CSPs rated unified performance visibility across all network and application layers as a key capability in 5G networks.
Advanced, fully virtualized performance assurance is needed to deliver end-to-end performance visibility. This could be likened to a "nervous system" for networks that's capable of streaming high-quality performance management data to analytics and machine learning solutions, as well as reporting and orchestration systems, providing full visibility of any issues that may impact customers.
The ability to combine and correlate highly granular active test and passive monitoring metrics and network data from multiple vendors and sources can feed machine learning capabilities to detect anomalies from baseline performance, provide insight into network behavior and issues and predict future network issues before they occur. This real-time, contextual insight can be used to automate the performance of virtual and hybrid networks, particularly for closed-loop automation and self-healing networks.
To fully capitalize on the 5G opportunity in a new era of mobile cloud leaders, CSPs must use these new assurance capabilities to transform into MCPs by seizing control of the customer experience where others have failed.
— Richard Piasentin, Chief Marketing and Chief Strategy Officer, Accedian