This blog is by Volker Held at Nokia Solutions and Networks.
Cloud computing has been a mega-trend in the IT industry for several years, and the telecommunications industry is now moving to the cloud as well. Rather than quickly jumping on the bandwagon, it’s important to consider the very stringent requirements needed to build your telco cloud. When done right, operators can reduce service deployment times from months to hours and even minutes – and can achieve drastic TCO savings, as calculations have shown.
Many network operators are currently defining their cloud strategies. For several reasons, it makes sense to start the cloud journey at the “heart” of the network in the Mobile Core. Here the cost-benefit ratio is clearly the best since payback time is fast, the technology is mature, and the implementation is relatively riskless if certain requirements are met.
So what is important for operators at this first stage in order to take full advantage of the telco cloud?
1) Make telco-specific adaptations of the cloud stack
The basic components of the cloud platform are virtualized computation and storage resources that are connected via virtualized LANs. An infrastructure manager manages these virtual resources and provides APIs that allow automated access to these resources. However, adjustments to the cloud platform are necessary to meet the high availability and latency demands of network functions, especially for signalling. Therefore, NSN works closely with its partners on cloud stack extensions that meet these requirements, and NSN contributes adjustments to the OpenStack Forum so that they become part of its defined open standards – thereby avoiding proprietary solutions.
2) Ensure independency of hardware and cloud stack
Operators should not bind themselves to a particular cloud provider API. The industry must either agree on a limited number of API versions, or better yet, operators should choose open solutions to be independent of specific hardware or cloud stacks.
3) Add orchestration and application management
Virtualization is not nearly enough. A fully functional telco cloud also requires orchestration and application management capabilities in order to provide automated network functions and to achieve the maximum elasticity and scalability gains. These capabilities allow the continuous monitoring of the runtime status of the software in the telco cloud and distribute the software on different clouds to flexibly cope with fluctuations in capacity demand.
In addition, the network functions can be adapted to the volume and type of traffic on the network. Faults in the management of the cloud infrastructure are the biggest risk regarding telco cloud availability. Only a high degree of automation by suitable orchestration and applications management systems can minimize this risk. Thus, the application deployment should be controlled by pre-set templates that define exactly which virtual resources are required. Orchestration and application management should be separated from the underlying IT infrastructure by clearly defined APIs.
4) Avoid complexity by integrating into legacy FCAPS system
The new cloud infrastructure should also be integrated into the existing FCAPS (Fault, Configuration, Accounting, Performance, Security) system in order to avoid additional complexity. This is especially important in the transition phase, when classical network elements are run in parallel with virtual network functions.
5) Adjust processes
The telco cloud is set to change the way operators run their networks and will affect many parts of their organizations. In order to reap its full rewards, the technological transformation must be accompanied by a fundamental shift in an operator’s working processes and organization.
Exploiting the full benefits of cloud technology
NSN has made detailed calculations of the benefits of cloud technology based on a realistic reference mobile network serving approximately 10 million subscribers. The study shows that fewer site locations produces savings in site costs of up to 65%, while IT administration improvements through automation generate 25% cost savings. Additional TCO savings are achieved by on-demand resource utilization. In addition, operators will be able to create networks for particular services and/or specific customers quickly and automatically. In other words, a service such as Voice over LTE (VoLTE) could be deployed from scratch in a number of hours.
NSN telco cloud is ready for commercial deployment
NSN has been a leading proponent of the telco cloud from an early stage, with the launch of Liquid Core in 2011, which incorporated virtualized MSS and TAS network elements. Since then, NSN has run proof-of-concepts and live cloud trials to provide the technology foundation for the telco cloud. These projects have demonstrated that core network software can be run on virtualized infrastructure with the virtualization of network functions such as IMS and TAS, which are required for VoLTE, HSS front-end as well as MME and S/P Gateway.
Additional projects have demonstrated the telco cloud readiness of end-to-end VoLTE and verified the automated deployment and elastic scaling of virtualized network elements, live migration of virtual machines from one server to another, and recovery from hardware failures. NSN was first to deploy a commercial telco cloud platform in a live operator multi-vendor network.
Check out our recent webinar: Reinventing telcos for the cloud where we share our approach and hands-on experience from several cloud trials with global operators who are ushering networks into the cloud era. You’ll also hear about our latest innovation, cloud-ready application management, which enables operators to freely choose the most competitive hardware and cloud platform for their needs.
For more information about NSN’s telco cloud approach, please see our recent press releases: Mobile broadband leader NSN to unveil #1GBperday$ portfolio enhancements and NSN builds on telco cloud leadership with multi-stack cloud approach.