International telecommunication union
In-network data processing
Download 0.91 Mb.
|
- Navigate this page:
- 11 Resource usage optimization
- 12 Resource abstraction
10 In-network data processingIn-network data processing is a system that provides with network wide data processing and application services by network nodes. Increase of video traffic, the expansion of IoT, and shorter response time requirement need a basic structural change of current ICT system configuration where the data processing is done at the remote data center and the network just functions as a data pipe. In 5G network, it is required that the network node will provide with data processing and application services with the aim of reducing the network congestion and also shortening response time, when appropriate. ICN and the edge computing are typical examples. Edge computing works very well to shorten the response time and reduce the network congestion when the target data for computing is closed to an edge node area. In IoT, however, there are cases that the target data for processing span to many edge node areas, therefore the inner node of a network is more appropriate for processing. Another example can be the on-path data processing, which applies a series of data processing in tandem manner on the transmission path, and is frequently used in big data processing. There also exist the service provisioning cases that inner network node is better suited to perform, where the service user is sparse and distributed to several edge nodes. Gap analysis One use case scenario of in-network data processing is included in ITU-T SG13 that deals with requirements and architecture with in-network data processing. However, only a limited number of use case scenarios are described for In Network Data Processing. Further discussion for viable in-network data processing for 5G mobile network is necessary. 11 Resource usage optimizationA collaborative redundancy reduction methodology in an end-to-end path of softwarized network for resource usage optimization is a new problem to tackle with. Taking advantage of SDN control, a collaborative redundancy reduction methodology is a new virtualized network functionality that dynamically offloads computational operations and memory management tasks of de-duplication to the group of the software designed network virtual functions. As this methodology efficiently chains storage de-duplication and network redundancy elimination functions and virtualizes de-duplication processes, it achieves effective performance without introducing high processing and memory overhead. Gap analysis: A large portion of digital data is transferred repeatedly across networks and duplicated in storage systems, which costs excessive bandwidth, storage, energy, and operations. Thus, great effort has been made in both areas of mobile and fixed networks and storage systems to mitigate the redundancies. However, due to the lack of the coordination capabilities, expensive procedures of C-H-I (Chunking, Hashing, and Indexing) are incurring recursively on the end-to-end path of data processing. Redundancy reduction methodology in an end-to-end path of softwarized network may be needed for resource usage optimization. Some identified gaps in 5G network end-to-end path resource optimization are:
12 Resource abstractionAs 5G services provided on a slice utilizes end-to-end resources to implement functional components, it is necessary to define unified abstraction of resource to adopt for the best practice and performance of network. The detailed information of the physical resource can be abstracted so that other systems, applications, services, or users can access the capabilities of the virtual resource by using abstracted interfaces. Therefore, it is necessary to define unified abstraction of resource to facilitate API, as followings:
In short, resource abstraction purpose is to ensure transparent network technologies and architecture for user. In other words, this 5G resource abstraction allows end-user to handle and use underlying resources in 5G network infrastructure easily (enable and maintain simplification for network users). Gap analysis This is no common model that can provide abstraction of various capabilities supported by physical resources that constitute end-to-end scope and are not covered by existing networks, including, physical radio interfaces, packet forwarding and routing in access networks. The granularity of current abstraction model may not be sufficient to support various approaches to satisfy end-to-end quality of application, while minimizing impact on utilization of networks. Directory: ITU-T -> focusgroups focusgroups -> International Telecommunication Union focusgroups -> Innovation of ict in Developing Countries focusgroups -> International Telecommunication Union focusgroups -> Telecommunication focusgroups -> Focus Group on Smart Grid focusgroups -> Telecommunication focusgroups -> Telecommunication focusgroups -> Telecommunication focusgroups -> Innovation of ict in Developing Countries focusgroups -> Innovation of ict in Developing Countries Download 0.91 Mb. Share with your friends:
|