Migration towards newly emerging networks

13 Migration towards newly emerging networks

The starting network phase utilizes current and state-of-the-art network technologies (existing technologies): LTE, IP-based network, etc.

The benefit of this deployment is during the migration intermediate phase, all end-to-end resources can still maintain the conventional communication means to communicate with each other. As a result, this mechanism enables migrated end-to-end resources are deployed in conjunction with existing devices. It enhances the migration process feasibility by enabling both the gradual 5G deployment and maintains current communication models simultaneously during intermediate period.

The requirements for this 5g network migration are stated as followings:

• 
  The 5G network is a foundation of the future network and having a mechanism to smoothly evolve to the future network which is under discussion in ITU-T SG13/Q15:
  
• 
  The migration scenario from the early stage of 5G network to the future network:
  
• 
  A locality based service provisioning mechanism and architecture: Mobile Edge Computing, which is one of the hot topics of 5G mobile system discussion, and local area computing are examples.
  
• 
  Possibility of the in-network data processing/service provisioning capability, where each network node carries out some data processing and service provisioning: This feature is especially efficient to handle IoT and big data.
  
• 
  Adoption of emerging network technology:
  

• 
  The application of network softwarization concept as a core technology to make 5G network: SDN and NFV are the examples.
  
• 
  Adoption of multiple logical networks (Slice), each having different architecture that fits to the major services provided on the slice: IP slice, ICN slice, IoT slice, and low latency slice, etc. will be candidates.
  
• 
  Having clear API for the development of a variety of applications and services developments and their provisioning:
  

In the final phase, the target network is formed and the 5G deployment is fully achieved (the migration process is finished), i.e. all the parts meet the requirements of network softwarization for IMT-2020 networks. Furthermore, in order to guarantee high QoS, low latency and high reliability of future network, the 5G deployment should be based on the state-of-the-art network framework, such as ICN which is referred as Data Aware Networking (DAN) in ITU documents (ITU-T Y.3033) and widely-recognized for its high performance and low latency.

Network virtualization described in [ITU-T Y.3011] allows the network providers to integrate legacy support and keep backward compatibility by allocating the existing networks to LINPs (i.e., slices) for deploying new network technologies and services or migrating to new network architecture. )

It is expected that network softwarization, especially slices can provide migration paths to newly emerging network architecture since it may be possible to accommodate multiple network architectures in slices concurrently. However, there is yet no activity observed for discussing the detailed migration scenario.

14 RAN virtualization and slicing under software control

The 5G mobile network needs to support flexible network capabilities with software defined radio access technologies in terms of frequency band, transmission schemes, antenna configuration, multiplex access attributes, for example, in order to achieve network optimization for a variety of services in dynamic manner in various environments with a reasonable cost, In such circumstances, a flexible programming scheme with software controlled virtualization will be required for gaining benefits of network capabilities and performances for network operators, service providers and end users.

Figure 9. Virtual RAN physical model - example

An overall picture around possible RAN virtualization is illustrated in Figure 10. It enables software based control of data transport from user device of access network, fronthaul, backhaul to core network.

Figure 10. RAN Virtualization model - example

RAN virtualization may own essential capabilities as follows for increasing network gain:

• 
  RAN controller integrates overall network control, scheduling, and data transport control throughout the user devices, remote radio unit (RRU), radio access schemes, fronthaul, backhaul, and radio resources such as transport bandwidth, RAT attributes, signaling on BBU,
  
• 
  Depending on the requirements for service application, the network slices are flexibly arranged and scaled up or down in a configuration set with appropriate network resources, virtual network functionalities (VNFs) in the virtual network topology in dynamic way.
  
• 
  The RAN has a capability of orchestrating the VNF chain by arranging and scheduling the virtual machines, storage memories, processing units, and so on. In consequence, all the data processing functions like the vEPC and the transport lines become programmable in software.
  
• 
  Transport SDN is also a softwarized capability for the data transport control in the path of X-hauls through the appropriate virtual network functions on the slice per the necessary data services and performances. The fronthaul and backhaul network path, the bandwidth, QoS, wavelength, time-slots, etc. are controlled by the TSDN in the routing between the RRUs and the core network.
  
• 
  On each slice, the network resources are flexibly allocated in a scalable manner under the RAN controller. Network resources are pooled, and idle resources are re-locatable among some network slices.
  

As a result of the expected capabilities as above, the network can provide comfortable quality of experience (QoE) for a variety of services in a reasonable cost (CAPEX and OPEX) with higher flexibility and agility. However, in order to realize it, some gaps need to be studied for overcome.

Virtualization of RAN domain in conjunction with software control is expected as effective solution to provide appropriate QoE for diversified service requirements in dynamic way with a reasonable cost. RAN resources and the functionalities are mapped onto the network slices in association with the service profiles.

Following elements should be defined.

1. 
   Slice management and the arrangement of VNFs, virtual topology, software based transport control on the slice.
   

• 
  Control of resources mapping to slices.
  
• 
  Computation of resource to slice assignments, trade-offs.
  
• 
  Decision of the timing when to turn on/off a slice.
  
• 
  Control of slicing within the UE, OS etc.
  
• 
  Management of the end to end resources on a given slice.
  
• 
  Inter-slice management (moving resources between slices)
  

2. 
   Activation of slice attributes such as the application drive, resiliency, OAM, and security on each slice and inter-slice.
   

• 
  Start/Stop/Management of applications residing within a slice.
  
• 
  Resiliency of slice control
  
• 
  OAM within a slice, between slices.
  
• 
  Security within slice, between slices.
  

3. 
   Appropriate APIs in some network elements such as follows:
   

---------------------------------------------------------------------------------

Reference documents on C-RAN

[ITU-R REP M.2320] Report ITU-R M.2320-0 (11/2014), “Future technology trends of terrestrial IMT systems”, 5.6.4 Cloud-RAN

[NGMN] “SUGGESTIONS ON POTENTIAL SOLUTIONS TO C-RAN BY NGMN ALLIANCE”, DATE: 03-JANUARY-2013, VERSION 4.0

[CMRI] China Mobile Research Institute, “C-RAN, The Road Towards Green RAN (Version 3.0)”, White Paper, Version 3.0 (Dec, 2013)

[ARIB] ARIB 2020 and Beyond Ad Hoc Group White Paper, “Mobile Communications Systems for 2020 and beyond”, Version 1.0.0, October 8, 2014, A.7.4 Cloud-RAN (C-RAN)