The Role of Optimization in 5G Networks

F5 Miniatura
Published February 16, 2017

As we all know, 4G and LTE are designed to improve capacity, user data-rates, spectrum usage, and latency.

5G represents more than just an evolution of mobile broadband.

It will be a key enabler of the future digital world and the next generation of ubiquitous ultra-high broadband infrastructure that will support the transformation of processes in all economic sectors. It will also represent a step-change in being able to meet the growing scale and complexity of consumer market demands.

While 5G is still in its evolutionary stage, its development will clearly be influenced by a need to support three specific use-cases (all of which will have an impact on emerging fields like autonomous vehicles, telemedicine and the Internet of Things):

  • Extreme Mobile Broadband (EMB)
  • Massive Machine-Type Communication (mMTC)
  • Ultra-Reliable Machine-Type Communication (uMTC)

Bringing these to life will require adaptation on both the radio and network side. For example, services may be centralized and in some cases distributed. This will depend both on the service function itself (some service functions will be naturally centralized or distributed) and, from a use-case point of view, access to technology and the type of performance required.

Having a technology that is potentially able to apply functions independent from the underlying protocol gives service providers the flexibility to implement services almost everywhere in the network.

Mobile Edge Computing (MEC) – which enables the edge of the network to run in an isolated environment from the rest of the network and creates access to local resources and data – is likely to have an impact here. Indeed, Research and Markets has identified it as a $80 billion market opportunity by 2021.

The optimization and acceleration of transport protocols will become even more important for networks requiring low latency and capability to hit high performance in a short amount of time. In this case, it is recommended to have a TCP optimization function capability running in different points of the network and, in particular, as close as possible to the end-user in terms of RTT/Latency. This will enable faster reactions in case of changes of network conditions, as well as service/applications requests.

Delving further into the detail, TCP optimization could become hierarchical and distributed where different proxies talk each other, creating “reliable” point-to-point intermediate connections. The purpose here is to enable faster re-transmission in case of any network drop irrespective of cause (congestion, IP traffic rerouting, temporary loss of connection on radio or fixed connection etc.)

Another important element to consider is the deploy-ability of policy enforcement and traffic steering functions on 5G networks in different parts of the network. From an architectural prospective, the same concepts and capabilities for TCP optimization apply here. In other words, the capability to distribute the functions can happen at any point of the network and for any kind of traffic. This can include traffic steering, manipulating video, or working as a gateway function for IoT-based services can be orchestrated by F5 technologies, removing and re-adding existing tunneling protocols.

F5 is starting to stand out from the crowd in this space due to its capability to manage, analyze and manipulate the traffic from Layer 4 up to Layer 7, injecting, removing or changing content. This includes both application layer traffic (such as HTTP, SSL etc.), as well as network protocols (like GPRS Tunneling Protocol-encapsulated traffic for mobile network transport). By running a Virtual Network Function (VNF), it becomes possible to achieve high levels of distribution and, ultimately the ability to better monetize, secure and optimize service providers’ networks.