- Supports industry’s widest range of access technologies that can be used separately or combined to improve service quality and expand footprint
- Cloud-native architecture delivers greater flexibility in resource deployment and efficient utilization to drive better economics
- Nokia’s Cloud Packet Core is 5G ready to support expanded service and network capabilities, new QoE mechanisms, network slicing and evolved connectionless services
There are many new demand drivers for re-architecting the core of the network. Both public and private networks need multi-access for ultra-broadband and IoT. Enterprise, public safety and government needs are evolving to embrace a wide array of industry access technologies, including licensed, shared and unlicensed wireless spectrum and fixed access. This is not just for worker mobility in the field, but is especially needed to support IoT applications and machine-type communications (MTC).
Meanwhile, consumers continue to demand ubiquitous network support for more bandwidth-intensive applications, with mobile video and cloud-based applications leading the way. And many governments have committed to deliver national-level broadband speeds of 30 mbs and above, which are driving the need for hybrid fixed-mobile access solutions for remote and rural areas.
A cloud-based converged packet core is emerging as one of the critical pieces of technology for supporting these demands. It needs to support a more flexible, multi-access network architecture. It should enable fixed and wireless technologies to be used individually or together to ensure the delivery of a seamless service experience. It should also provide the foundation for the evolution to 5G. And it should be built around cloud-native principles, so that it can scale from the needs of isolated enterprise field-area networks to the largest service provider networks supporting ultra-broadband business and IoT/MTC services.
While service providers might once have dreamed of simplifying the range of access technologies with which they must deal, the reality is that with access sometimes being the bandwidth bottleneck, they need to exploit as many access types as possible. This means using licensed wireless spectrum in conjunction with unlicensed and shared wireless spectrum, such as Wi-Fi, MulteFire and Citizens Band Radio Service (CBRS). Service providers will also extend fiber and copper fixed networks as close as possible to users and supplement existing macro cells with small cells. Add to all of this, support for the unique, often-connectionless demands of IoT sensors and devices.
The good news is that this massive proliferation of access types doesn’t imply siloed services and a fragmented user experience. One of the keys is to converge these functions on a common packet core that can ensure seamless services across various access types. By abstracting the service from the access technologies using standards-based IP networking capabilities such as Multi-Path TCP (MP-TCP), it is now also possible to offer concurrent service delivery across multi-access technologies. For instance, a consumer or business with heavy bandwidth needs might use a DSL service supplemented by LTE to get the bandwidth or service reliability they require.
IoT and MTC
A significant part of the new demand being placed on the network is coming from IoT devices and sensors that transmit information or data using machine-type communications (MTC). MTC is widely used in many industrial and utility networks. They often require only narrowband services, but conversely put heavy loads on the signaling plane. A converged packet core needs to incorporate the latest 3GPP IoT/MTC feature sets in order to support these narrowband IoT connectionless services, but must also rely on its underlying architecture to ensure the realization of the economics of the wide range of IoT/MTC services.
Some of the most mission-critical services — characteristic of power utilities, intelligent transportation systems (ITS) or resource management/protection systems, for instance — require services that have minimum tolerances for jitter and delay. In order to minimize latency, packet core functions need to be able to be distributed and made available in close proximity to where they are needed.
The networking industry is embracing network function virtualization (NFV) and software-defined networking (SDN). There is a corresponding rush by vendors to virtualize their product offerings. Unfortunately, there are also many examples of network functions being ported to NFV with minimal re-architecting. Cloud-native refers to re-architecting a traditional network function to avoid the limits of vertically integrated network functions. It requires building it from the ground up to take true advantage of the flexibility and scalability made possible by the cloud. This includes a common data layer, stateless functional software elements with state-efficient processing, centralized and distributed deployment architectures and automated lifecycle management.
One of the key drivers for this move to the cloud is, of course, economics. But its real impact on TCO lies in its efficient use of compute resources; in other words, its ability to allocate compute resources to only those tasks that have an immediate need, moving dormant state information into a common data layer for retrieval at the appropriate moment. This makes it highly efficient.
Additional efficiencies are achieved by a cloud-native packet core that incorporates software disaggregation principles. Functional tasks are broken down into smaller elements and separated into independently scalable processes. This disaggregation of tasks and functions means that network compute and processing resources can be managed independently. High levels of automation are required to provide the agility to support both event- and service-driven change across multiple access technologies. For example, large one-time events like outdoor concerts that spike network demand can have additional resources quickly deployed, and then reallocated when and where they're needed.
By building the packet core from the ground up to be cloud-native, it is also possible to make it ready for some of the most advanced aspects made possible by 5G. These include expanded service and network capabilities, such as new QoE mechanisms, expanded network slicing and evolved connectionless services.
Enterprise and industry verticals will especially be able to take advantage of the expanded network slicing capability of 5G and a cloud-native packet core. Analogous to multi-access services, discussed above, network slicing abstracts the network service from the underlying network infrastructure. This will help re-define the enterprise LAN to include wide area networking services. Wherever the worker roams, he or she is always on the appropriate network, which may occupy its own dedicated slice of the service provider’s fixed or wireless network.
Several recent examples of IoT device hacking, from HVAC equipment to consumer PVRs, has highlighted the security issues associated with widespread IoT. Devices can often be released to market with insufficient attention to security. When we look at public utilities and infrastructure services, the issues become critical to national security. Using network slicing, enterprises and governments will be capable of isolating IoT and MTC devices on dedicated network slices, to help secure these network solutions.
Nokia Cloud Packet Core
In this article, I have tried to set out what I see as the principle requirements for the new converged packet core. In undertaking to re-architect the Nokia Cloud Packet Core we took these demands into consideration. We were constantly mindful of multi-access driven broadband evolution, IoT/MTC, the move to 5G and the special needs of enterprises, vertical industries and governments. Consequently, we embraced cloud and web-scale principles within our Cloud Packet Core.
For enterprises, vertical industries and governments, and service providers, building field-area networks using LTE or other unlicensed or shared wireless spectrum technologies, we have pre-integrated AirFrame-based packet core appliances that can be scaled to meet varying business and operational requirements. And for our largest service providers, we can provide a massively scalable packet core that can handle anything they throw at it, from millions of ultra-broadband users to billions of IoT/MTC devices and sensors. And we can do this with much greater efficiency as well as provide a solid foundation for their evolution to 5G in the coming years.
For more information visit our website on the Nokia Cloud Packet Core. We also have a white paper available Embracing an Industry Inflection Point with the Cloud-Native Packet Core. Or, watch this video: Nokia Cloud Packet Core: Profit from opportunity, evolve with confidence.