This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read our privacy policy>Close

If you need help, please click here:

Unleashing IT Values, Keeping Telco DNA

Telecom Operators Adopt IT Technologies

As the mobile Internet becomes increasingly widespread throughout the digital economy, traditional communications services such as voice and text messaging are continuously affected by Over-The-Top (OTT) applications such as WeChat, WhatsApp, and Skype. The smartphone market continues to spawn thousands and thousands of OTT applications, each of which increases the competitive pressure faced by telecommunications (telecom) operators to produce high-quality user experience and more attractive new applications.

New competition and new user experiences and behaviors require that operators make changes to their business, marketing, R&D, operation, and service models. These changes require the support of agile IT telecom networks that are substantially similar to enterprise data centers.

Historically, service innovations in carrier networks have been time consuming and expensive. Traditional telecom networks are silo networks and Capital Expenditure (CAPEX) and Operation Expenditures (OPEX) are excessively high. In other words, upfront cost has been a major bottleneck for operators’ development in the mobile Internet era.

New Enterprise cloud computing technologies are providing the technical basis for restructuring communication networks, and bringing new opportunities for telecom network operators. Within a traditional telecom network plant, the Network Elements (NEs), the software and hardware, are decoupled. However, the paradigm shift of cloud computing technologies is brought about by Network Functions Virtualization (NFV) that both shares and flexibly reallocates resources. For telecoms, the first step in cloudification is to upgrade the core network.

Huawei’s CloudCore solution is an NFV-based core network upgrade.

Ensuring Core Network Capability

First and foremost, the switching and control center of a cloud-based carrier-grade network must satisfy the highest service level requirements for reliability, performance, and self-healing faults.

Following Huawei’s ROADS (Real Time, On Demand, All Online, DIY, and Social) predicate, cloud-based core networks must be “open” and available for customer modification. This requires that core network processes be deployed using the most efficient software architecture and design practices.

Building a Cloud-aware Architecture

Cloud-aware software architecture is essential to CloudCore network capabilities. Cloud software architecture must ensure that telecom software is deployed layer-by-layer — service distribution, service processing, data storage, etc. — in order to keep service logic separate from session data. A stateless design for service processing units also is necessary.

  • The CloudCore system architecture specifies that communications session data be separated from service logic such as configuration data, which is stored by dedicated distributed database modules.
  • There is no need for service logic processing programs to store session data.
  • On the front end, service distribution modules perform load distribution.
  • Since service logic processing units do not store session data, there is no need for service distribution modules to distribute messages to specific processing units by user or session; therefore, they can support reconfigurable distribution strategies. For example, fault-tolerance alert messages can be distributed according to the load and health conditions of each processing unit.

A direct benefit of a cloud-aware architecture is the ability to flexibly scale the number of service processing instances. Virtual machine resources for traffic rebalancing are added or released according to traffic fluctuations within a matter of minutes, independent of the number of active sessions.

Maintaining Carrier-Grade Service Levels

A number of common technical methods are used to ensure Service Level Agreements (SLAs):

  • Multiple-layer reliability.
  • Key Performance Indicator (KPI)-based self-healing.
  • Service queue control.
  • Virtual resource optimization.
  • Heterogeneous resource pools.

Software and hardware are decoupled in CloudCore networks, and the software is separated by layers. Reliability strategies are deployed from the bottom — servers, storage, and networks — to top-layer applications. We employ multiple network adapter redundancy, network adapter bundling, multiple-disk array networking, and a High Availability (HA) applications layer to ensure 99.999 percent reliability over the entire system.

KPI-based health checks and fault self-healing are essential to ensure carrier-grade SLAs. In traditional system fault processes, entire systems are reset after sufficient degradation, risking dropped sessions and data loss. The KPI-based health monitoring systems service KPIs in real time and compare current statistics with historical norms. When key KPIs deteriorate to preset thresholds, alarms are generated and automatic recoveries are rapidly invoked based only on specific fault cases.

Service queue traffic control is independent of CPU usage. The resource load — CPU, dynamic memory, and message packages — is a latency factor that delays clearing each message queue. When resource use is high, the number of traffic control queue service messages are reduced, which in turn increases service message delay and the size of the queue. When the service message quantity or delay exceeds a defined threshold, messages will overflow the queue and trigger traffic control, resulting in an improved CPU utilization.

NFV performance deteriorates because scheduling management and virtual machine interrupts consume system resources — by 20 to 30 percent in extreme scenarios. Optimal software design is the only solution. Virtual machine scheduling (computation and storage) and memory access efficiency are the key factors for improving run-time efficiency. Virtual storage performance is improved through Raw Device Mapping (RDM), a VMware resource that enables the mapping between a virtual machine and a Storage Area Network (SAN). Virtual network performance is improved by reducing I/O interrupts and reducing data copies between the “OS kernel” and the “User modes” of the Intel® Data Plane Development Kit (DPDK).

Openness

The Huawei CloudCore system is designed to support heterogeneous mixtures of Commercial Off-The-Shelf (COTS) hardware and open operating systems such as FusionSphere, VMware, and OpenStack.

Because in the real world traditional core networks and cloud-based networks must coexist, Huawei has built CloudCore to operate side-by-side with legacy telecom switching networks as a virtual instance within a unified resource pool under the control of a single Operations and Maintenance (O&M) process.

In all cases, the goal is to enable traditional telecom networks to compete effectively with OTT players. First and foremost, this is achieved by the delivery of pre-integrated hardware and software products from multiple vendors wrapped in an innovative open-source software solution that preserves and aggregates longstanding industry values.

Huawei CloudCore Architecture

Core Network Cloudification

Leading Architecture and Design

The Huawei CloudCore solution encompasses all core network NEs and, therefore, fully complies with the design requirements of a cloud-based core network.

This is due in no small part to Huawei’s continuing, substantial contribution to the NFV standards process, primarily through the European Telecommunications Standards Institute (ETSI).

Huawei has joined all nine possible NFV research topics within the ETSI standards organization — with the top position among 348 accepted proposals — and occupies two vice-chairman positions.

Industry Acknowledgement

Huawei dedication to the CloudCore effort was acknowledged at the 2014 and 2015 IMS World Forum for “Most Innovative Virtualized IMS Solution,” due to our leading technology, successful pilots, and commercial experience.

NFV Open Lab

In January, 2015, to satisfy customer demand for pre-integrated solutions, Huawei launched an NFV Open Lab in Xi’an, China with three primary responsibilities:

  • Develop a multi-scenario, multi-vendor integration and verification capability. Build a flexible multi-vendor integration and verification platform that includes Big Data analysis, continuous integration, and NFV planning practices to provide reliable data support and decision-making by operators.
  • Perform joint development and quick mutual qualification innovations with industrial organizations, operators, and partners.
  • Build open and cooperative NFV industrial ecosystems with commitments, so far, from more than 150 partners.

Mature Commercial Capabilities

Huawei CloudCore solution has delivered multiple successful commercial cases, including Vodafone-Italy and the successful delivery of the first commercial cloud IP Multimedia Subsystem (IMS) project in the industry. Huawei provided a cloud-IMS based on third-party servers and a Cloud OS. As primary integrator, Huawei was responsible for the vertical integration of the cloud platform from the bottom layer (COTS hardware and Cloud OS) to top-layer IMS applications software. Huawei completed delivery in short order by leveraging a professional delivery team using efficient integration tools and their extensive integration experience.

With leading architecture and design, great industrial contributions, industry acknowledgement, and mature commercial capabilities, the Huawei CloudCore solution is providing strong support to operators for building new, agile telecom networks that lead the cloudification journey for the core network.

By Liu Hao

Sr. Manager of Core Network Marketing, Huawei Technologies, Inc.