Cloud Computing in ICT’s World of Duality
Cloud computing enhances the value of IT for enterprises. Instead of simply using IT for productivity gains and cost reduction, many companies are merging communications and network technologies into a unified cloud architecture that, underpinned by Infrastructure-as-a-Service (IaaS), drives network virtualization.
Companies with a deep and precise understanding of the changing environment can adapt well and establish alliances that will help them thrive amid ICT transformation in the new ecosystem.
Although IT and Communications Technology (CT) follow separate paths of development, they have come to be intertwined and interacting in ways that have synthesized a variety of innovative concepts, such as virtualization, Software-Defined-X, and the Internet of Things. When it is difficult to determine if a particular platform is an information technology or a computer technology, it is possible that only the term “ICT” is applicable.
Information Industry Model
Today’s information industry model classifies ICT into three areas, as shown in Figure 1. Conceptually, the ICT industry is best represented by Big Data and cloud computing. Technically, it is based on IT and CT. For implementation, Internet Data Center (IDC) and Enterprise Data Center (EDC) are the two major factors.
Figure 1. Information industry model
Big Data and cloud computing are interdependent concepts resulting from IT development. Big Data describes a virtualized objective world, while cloud computing comprises virtualized ICT resources. When cloud computing was conceived, virtualization empowered it to meet the needs of Big Data applications. Consequently, cloud computing has gradually penetrated into the production areas of various industries.
As a system responsible for organizing the production activities of an enterprise, an EDC is designed to improve the operational efficiency of its production units. Implementing its functions requires simplifying the paths through which information is disseminated throughout the enterprise, the integration of business processes, use of Big Data applications, adoption of flatter organizational structures, and extension of digitalization to the furthest nodes of the organizational structure.
A virtualized EDC is a cloud computing platform presented as a “private cloud” that allows a high degree of information sharing within the enterprise. Although this cloud computing platform also disseminates information, its primary functions are information gathering, analysis, and decision-making; therefore, an EDC more closely resembles an IT system that adopts computer technologies than a CT system.
An IDC is an information dissemination system designed to improve information exchange and communication experiences, for social networking and other media services. Implementing its functions requires simplifying the paths through which information is disseminated.
A virtualized IDC is a cloud computing platform presented as a “public cloud.” Although it is also used to process information, its primary functions are information correlation, sharing, and dissemination; therefore, an IDC more closely resembles a CT system that adopts information technologies than an IT system.
Duality: Cloud Computing and Virtualization
Today's ICT systems are built upon both cloud computing and virtualization technologies: with cloud computing manifesting external to ICT systems, and virtualization being inherently internal. The contrast between the two technologies illustrates a duality at the core of the ICT platform, as shown in Figure 2 (see overleaf).
Figure 2. The duality of an ICT system in the Big Data era
Cloud computing provides agile services. Outside, it provides comprehensive services to ensure service availability. The technical services inside the cloud support continuous process optimizations — inheritable and iterative — that enable a smooth evolution of services over the service life from system-launch to system-maturity.
Cloud computing is a platform for delivering intelligent services. Being able to receive information and understand requirements from external resources; the cloud platform can then apply its internal processes to process the request and automatically respond.
Cloud computing supports automated responses to requirements, imbuing them with enhanced mobility and free customization potential. Examples of intelligent services include Software-Defined Radio (SDR) and Software-Defined Networking (SDN). Cloud computing is platform built to understand external service requests, and grow because of the ability to manage the cost of resources between peak events. Cloud platforms are built to anticipate a defined range of resource requests, and also to anticipate a growing range of applications-on-demand.
From the inside view of an ICT system, virtualization is the method for implementing cloud computing. Cloud services provided on cloud platforms constitute cloud computing. The intention is that cloud platforms are extremely and inherently agile, designed to provide end-to-end services for various applications. The fundamental idea of X-as-a-Service (XaaS) is the availability of accessing packaged resources on-demand. Service resources within the each cloud plant are built on a foundation of three basic features:
- Isolation: Cloud platforms provide comprehensive services to all applications that meet specified requirements. Service applications use cloud services at layers of abstraction above the internal details of the cloud platform.
- Interaction: Cloud platforms exchange information with service applications through Application Programming Interfaces (APIs). This implementation of ICT intelligence supports dynamic service customization.
- Ease of accessibility: Since cloud platforms are highly mobile, and are able to deliver on-demand, ubiquitous access to services for distribution over Wide Area Networks (WANs).
Restructuring for Effective Cloud Operations
The job of cloud services is to organize and allocate resources. Unlike conventional ICT services, cloud computing resources are not deployed from scratch, Instead, by upgrading and restructuring existing ICT processes to accommodate cloud services. In all cases, the goal for restructuring is to calculate optimized cloud platform configurations.
Restructuring improves operational efficiency and the service application usability, the results of which are agile and intelligent cloud services. Open APIs provide access for external services to reach the virtualization engines internal to the cloud.
A critical issue in all cases is if the volume of data in an ICT system is allowed to grow without constraint. Without a timely intervention, the system would soon collapse — for which either system component optimization or system restructuring will solve the crisis.
There are four key considerations in the solution: Decoupling, Convergence, Open Access, and Intelligence.
- Decoupling simplifies processes by establishing direct paths for sharing information flows, which creates the benefit of shrinking device count and the floor space required to deploy services at scale. Aggressive measures are taken to minimize redundancy, reduce dirty data, and eliminate ineffective processes.
- Convergence extricates and normalizes the base functionality of resource and service interoperation — the success for which is expressed as consistent services, uniformly scheduled and with managed levels. The result is the foundation for an ordered expansion of the ICT system.
- Open Access is a precondition for terminal and service interfaces that must be updated from the SDN host.
SDN provides an example paradigm for trading monolithic devices with limited access to root level activities for CT systems that divide the Application, Control, and Device layers — a physical and logical separation of network resource management from data packet processing and packet forwarding. Among other unique devices, Huawei provides a programmable Ethernet Network Processor — ENP Transformer 1.0 — for users to access service application systems.
- Intelligence is expressed by two critical cloud service characteristics: First, external requirements information, exchanged between systems, enables the seamless integration of services delivered by the cloud platform. Second, internal requirements information can be processed in coordination between platforms. SDN is required because the physical world is dynamic and networks must be modifiable. Only by being modifiable can networks provide optimal on-demand services to users and maximally utilize ICT resources. In the Big Data era, ICT tasks follow a long-tailed distribution pattern characterized by irregular demand bursts and variable patterns of mobility.
Huawei has rich experience in Component Restructuring, including the SingleRAN solution based on the horizontal convergence of GSM/UMTS/LTE (GUL) and the integrated U2000 Unified Network Management System that has combined the N2000 and I2000 solutions.
Process Restructuring involves vertical convergence of cloud system services. Among the solutions is the uniform planning of multiple layered Optical Transport Network Interfaces (ONI) to minimize redundancy. Optical overhead Channels (QoCs) can be uniformly implemented when the transmission modes of IPv4, Synchronous Digital Hierarchy (SDH), and Wavelength-Division Multiplexing (WDM) complete their evolution to IPv6 and OTN.
ICT Drives to Meet Unchanging Needs
ICT technologies continue to evolve to meet the needs of a growing market. The demand for greater efficiency and flexibility is an unending need that focuses the conversation on the evolution of open, pragmatic services that are responsive to customer needs. As new technologies emerge, cloud computing is poised to play a strong continuing role in maximizing the benefits of ICT’s dual origins from the IT and CT sectors.