How 5G Relates to SDN and NFV Technologies – Part II: Architecture
In Part I of this series, How 5G Relates to SDN and NFV Technologies – Introduction and History, I explained why I believe 2019 is going to be the year of 5G, reviewed the history of wireless communications, and depicted how we got from 1G to 5G. In this blog, I’m going to take a deeper dive and explain 5G architecture and how it relates to SDN and NFV technologies.
5G and its Future Impact
5G is the new generation of radio systems and network architecture that will deliver extreme broadband, ultra-robust low latency connectivity, and massive networking for human beings and the Internet of Things. In a few years, 5G will not only touch technology and applications, but dramatically change the economy, our society and individual lives.
5G will be far more than just a new radio technology: It will combine existing Radio Access Technologies (RATs) in both currently licensed and unlicensed bands, add novel RATs optimized for specific bands and deployments, and scenarios and use cases (some of these use cases have yet to be imagined). It will also demand a different new network architecture based on Network Function Virtualization (NFV) and Software Defined Networking (SDN) technologies. (I covered these technologies in detail in previous blogs so I won’t extend myself here).
5G and the New Concept of Programmability
Programmability will be central to achieving the super-flexibility that network providers and operators will need to support the new communications demands that will come to them from a wide array of devices, users, and companies across many different industries, as well as from other organizations such as government agencies (think: cities/municipalities, army, etc.). To sufficiently manage all these demands, 5G networks will have to be programmable, flexible, modular, software-driven and managed in a holistic fashion in order to enable a diverse and profitable range of services.
But let’s leave the 30,000-foot view and dive deep down into technical architecture whereby I’ll explain (1) what is behind the definition of 5G architecture (3GPP) and (2) compare the 5G Mobile network architecture to a model well-known by computer and networking engineers worldwide.
The OSI Model
The 3GPP stands for the Third Generation Partnership Project and it consists of many standard documents that cover all aspects of the 5G technology (Radio, Access, Core, Security, etc.). These documents are developed by the 3GPP Technical Specifications Groups (TSGs) and Working Groups (WGs) on the main releases. The latest Releases (15 and 16) are the main targets of the 5G architecture, consisting of 5G core and the New Generation radio network (NG-RAN). After these two, yeah, you guessed it, we’ll have release 17 and beyond (in the works) which will add more features and enhancements. There are also other organizations and consortiums of companies working on developing the standards of the 5G including: IETF, ITU, etc.
Now let’s look at the OSI model:
- Starting from the bottom, we see that Layers 1 and 2 of the OSI are replaced by the OWA (Open Wireless Architecture Layer).
- One level up, we have the Network Layer which is divided in two subsections (Upper for the mobile terminal and Lower for each interface). The Network Layer is used to route data from the source IP device to the destination IP device/system.
- On the higher level, we have the OTP (Open Transport Layer) which combines functionality of both the transport layer and session layer.
- Finally, we have the Application Layer which marks the data the proper format required and facilitates encryption and decryption. It also selects the best wireless connection for the given service and takes care of providing the Quality of Service (QoS).
This key new concept in 5G will enable tenants to get different levels of connectivity from their service provider to accommodate use cases. So yes, QoS is built in on the design of 5G.
To better utilize resources—streaming a video in 4K with HDR will be a reality with the bandwidth of 5G and require minimal latency. If you were to use Voice over IP (VoIP) or send an email, however, you wouldn’t need that much. With slicing, we divide the network into ‘virtual slices’ of the underlaying physical network and each one of them supports specific performance guarantees.
You are probably seeing the link between 5G and the SDN and NFV technologies we have been covering. To achieve this network slicing, 5G will be an all-cloud architecture. The specifications provided by 3GPP dictate that the network will be based on a central cloud that’s connected via a backhaul network to many edge computing clouds that are kilometers away from the user and move many services from the core to the Edge. If services can be executed in the network edge, it will take traffic away from the Cloud RAN.
5G architecture will boast cloud-native access, transport, and core networks. Network slicing will be enabled through all of them (edge to edge), allowing diversified 5G services and different QoS. As we can see, the virtualization technology of NFV and the centralized control of SDN are embedded in the DNA of a 5G system. Also, whereas 4G-LTE networks tend to integrate both the user and control planes, 5G architecture will separate these planes, improve flexibility, facilitate centralized control, and ensure easy network slicing.
And do you remember what was one of the tenants for SDN architecture from the previous SDN posts?
Yes, that one, Centralized Network Management!
…To be continued soon in Part III.