Cloud NF: The Way to Efficiency and Simplicity

With the rapid development of mobile internet and the widespread use of smart terminals, global data traffic has increased exponentially. Even though mobile networks are evolving to 4G, legacy 2G and 3G networks will still be around for some time. In a traditional network, equipment is usually designed with a dedicated architecture and specific function. Therefore, there are a variety of network elements (NEs) in the network. This complicates network OAM and also increases OAM costs. In addition, the ARPU of telecom operators is being eaten away by OTT players. Challenges facing operators include:

●    costly expansion and replacement. To cope with the exponential increase of signaling and data traffic, it is necessary to expand or even replace NEs located in different areas. One of the foremost concerns of an operator is smooth, flexible expansion in response to unpredictable traffic fluctuation.

●    difficult OAM. Every function or NE requires dedicated hardware and software. This increases OAM complexity. The capacity of a single system is also limited. The only way to increase capacity is to replace hardware or add extra NEs. This makes OAM difficult and increases the workload of OAM personnel.

●    high power consumption. Operators usually run all their NEs at maximum capacity to ensure QoS. However, for direct communication and dynamic capacity scaling between these NEs, a lot of power is needed.

●    closely coupled hardware and software that restricts innovation. The equipment used for a service might have latent capabilities that are unrealized but locked because of the close coupling with software. It is difficult to protect investment because additional hardware and software needs to be deployed for each new services and application.

●    long R&D cycle and high customization cost. Network evolution and standardization of new functions is a long process. The R&D cycle is long, and frequent service customization is costly.

Exponential growth in data traffic increases the requirements on network architecture in terms of capacity, delay, power consumption, and new service deployment. If we look at the history of mobile network architecture, from layered vertical architecture in the 2G era to flat, open architecture in the 4G or even 5G era, the goal of mobile network evolution has been simplicity and efficiency.

The evolution of mobile networks to flat, cloud-based networks is inevitable and will improve operator competitiveness and simplify OAM. In response to this trend, ZTE has developed cloud network function (Cloud NF).

Cloud NF is based on network function virtualization (NFV) and integrates mobile core networks and controllers. It uses virtualization technology to separate software architecture from hardware architecture and combines all NE functions into an organic whole. Cloud NF also uses software defined networking (SDN) to separate the forwarding plane of packet gateways from the core network. This improves forwarding efficiency.

Based on the universal infrastructure platform, Cloud NF uses a cloud management system to integrate computing, storage, and network resources and enables all NE functions to be deployed as one function on the unified underlying platform. Cloud NF can function as the radio network controller, core network, and eAPP. Controllers comprise 2G/3G RNC and BSC, small cell AG (Femto GW), security access gateway (SeGW), and radio resource policy controller (RRPC). Core network contains control plane NEs and the control and complex data processing functions of user plane NEs. A user plane NE can separate the forward part, based on the SDN architecture, to form a forwarding plane. eAPP refers to wireless-related applications.

Cloud NF architecture can be flexibly deployed according to network scale and coverage, and it manages cross-regional resources in a unified way. Traditionally, NEs were independently planned and deployed, but now they are horizontally planned and deployed. The entire mobile network is planned uniformly and can dynamically adapt to changes in services. In other words, the number and capacity of NEs is determined by actual network capacity and planning parameters, such as traffic model. The internal processing model of NEs determines the demand for virtual machines (VMs), storage, and network resources. VM resources are allocated by each NE. Once the software of an NE has been loaded, resources are automatically fed to the NE. Cloud NF architecture is a solution for efficient 2G, 3G and 4G deployment and takes into account the demands of both legacy and future networks. It has open capabilities that enable third-party applications to be deployed on the telecom elastic cloud platform and that enable virtual network function (NFV) to be deployed on a third-party platform.

Cloud NF has the following features that help operators cope with their challenges:

●    universal hardware platform. Cloud NF uses an industry-leading universal hardware platform that has distinct advantages in terms of extensibility, flexibility, evolution, and decoupling of software and hardware.

●    virtual network function. Software and hardware can be decoupled. Virtualization makes software deployment more convenient. Cloud NF enables mobile network equipment to be quickly deployed, and this reduces costs during network commissioning. Cloud NF also supports frequent version changes, and this greatly lowers the risk associated with equipment upgrade. It also improves network stability.

●    automatic OAM. Cloud NF supports automatic shelf installation, software installation, and service configuration. This facilitates network OAM and greatly improves deployment efficiency and quality. Cloud NF also has remarkable advantages in terms of plug and play and flexible capacity configuration.

●    flexible service scalability. Services can be provisioned only through remote software configuration, and provisioning time is greatly shortened. Cloud NF supports network-based sharing so that different dedicated services can be deployed on the same network. It can also provide greater openness so that innovative third-party applications can be quickly deployed.

●    carrier-class reliability. To meet the need for highly reliable telecom equipment, Cloud NF provides backup modes for multiple layers, including node, VM and service layers. It also supports 1+1, N+1, and pool backup modes.

●    energy conservation. Cloud NF uses an energy-conserving multicore CPU.  Each computing node can automatically power-off in order to conserve energy. The energy-conservation mechanism involves shutting down unused processing cores, decreasing the number of occupied physical machines (by concentrating VMs onto less physical machines), and shutting down unused physical machines.

As the NFV and SDN industry chain matures, future mobile networks will become more open, flexible, controllable, and intelligent. As networks evolve to 4G, existing 2G, 3G and 4G networks will continue to coexist for quite some time. As a key device in 2G and 3G networks, controllers must be evolvable. To maximize profits, operators need to provide LTE networks while addressing the issue of 2G and 3G coexistence.  A cloud solution in which the controllers and core network share the same platform is therefore inevitable. ZTE’s Cloud NF is such a solution.