霧運算:CSP (通訊服務供應商) 的機會與課題
Fog Computing: Opportunities & Challenges for CSPs
|出版日期||內容資訊||英文 25 Pages
|霧運算:CSP (通訊服務供應商) 的機會與課題 Fog Computing: Opportunities & Challenges for CSPs|
|出版日期: 2016年12月13日||內容資訊: 英文 25 Pages||
本報告提供霧運算 (Fog Computing)概要和今後的普及預測，提供其與各種網路和IoT的關聯性等相關分析，再考察通訊服務供應商可從該領域所得的市場機會，及為此應該解決的課題。
The certainty of change -- along with the uncertainty that change brings -- is now painfully common for communications service providers (CSPs). While CSPs and the networks and services they deliver are at the heart of the global economy, the now-constant barrage of technology innovations, business disruptions, and regulatory shifts has put CSPs under enormous pressure that shows no sign of easing.
Fog computing (a.k.a. fog networking) is the latest potential disruptor for CSPs and their technology suppliers. The term itself raises the specter of fear, uncertainty, and doubt that CSPs have faced since the dawn of this century. If virtualization and the cloud weren't murky enough concepts, fog comes rolling in to add to the degree of difficulty for developing and implementing long-term business and technology plans.
Despite its vaguely menacing name, the underlying concept of fog computing is fairly clear. Very simply put, fog computing pushes the concept of virtualization out to the very edges of the network, with the goal of getting network latency as close to zero as possible. In this context, fog is a logical extension of the network virtualization that is taking place deeper inside the network.
Mobile edge computing (MEC) can be viewed as one aspect of the fog model, specific to mobile networks. The MEC architecture uses virtualization to push applications and processing tasks closer to the user, potentially as far as the base station -- which not only greatly reduces the distance from the network to the end user, but also reduces backhaul requirements. ETSI is now working on standards for MEC, which is seen as essential to the successful deployment of 5G networks and services.
But fog computing's potential benefits can be extended to wireline networks, as well. The most obvious application for fog in the wireline network is the extension of cloud services out to the network edge, again with lower latency in mind. Extending the fog model to all types of networks will also be critical for applications that will run over different types of networks, such as Internet of Things (IoT) systems. And taken to its logical conclusion, fog will eventually extend to envelop the end devices themselves.
While the basic concept of fog computing is relatively straightforward, implementation and deployment issues are far from clear. One obvious reason is that fog is still in a fairly early stage of development. The OpenFog Consortium, an industry/university group, was launched in November 2015, and to this point its output consists of a handful of white papers. Another cause is the very real possibility that different implementations will require variations on the overall fog theme: Yes, fog computing involves pushing more functions out to the edge, but which functions are pushed and where they end up will vary.
This report presents Heavy Reading's current thinking on fog computing and its likely role in the development of next-generation networks and services. For this initial look at fog and its potential impact on CSP networks and services, we've focused on five basic aspects of fog deployment: the conventional core network, mobile networks, CSP data centers, IoT systems, and security. Subsequent reports will provide deeper dives into these areas and other aspects of fog computing and networking.
Fog Computing: Opportunities & Challenges for CSPs is organized as follows: