5G Transport demo - joint orchestration of radio and transport

Fresh from our Kista 5G Transport Lab: A demo of programmable transport showing how transport and radio resources can be jointly optimized in response to dynamic traffic patterns.

5G Transport feature

Principal Researcher

Principal Researcher

Category & Hashtags

The transport network is the part of the network that carries data traffic between the radio base stations and the core of the network and datacenters. 5G will bring a 1000-fold increase in traffic, a wide range of new services including both massive and critical machine type communications, and new business models. All that needs to be supported by the 5G transport network. The transport will also have to handle radio networks with an order of magnitude more cell sites and new features such as multi-site coordination and centralized RAN architectures.

In our Kista 5G Transport Lab, we at Ericsson Research work together with the Royal Institute of Technology (KTH) in Stockholm and Acreo Swedish ICT on solutions and concepts that can meet these tough requirements. Our latest achievement is a demo of programmable transport. A video about the demo is on display in the Ericsson booth at the NGMN Industry Conference taking place March 24-25.

We see flexibility and programmability as key features of 5G transport. Flexibility facilitates service deployment and efficient use of the network resources. Then, by exposing the flexibility of the transport network through an API, transport resources can be controlled in a programmable manner. This enables joint optimization of the transport resources together with radio and IT resources.

Our demo shows joint orchestration of radio and transport resources based on SDN, to improve capacity and resource utilization. It is shown for a centralized RAN deployment using dense wavelength-division multiplexing (DWDM) optical fronthaul. For the orchestration, we use live feedback from the actual traffic situation to optimize the network utilization: radio capacity is increased at locations where it is needed most.

We show how transport resources can be jointly optimized together with radio resources in response to dynamic traffic patterns. This enables better use of both transport and radio resources.

Such dynamicity is something we expect will play an important role in 5G. For example, when crowds of people gather for special events, like concerts and big sporting events, network traffic increases dramatically. Sharing of photos and videos from the event further increases the requirements on the network. What if we could increase the capacity locally during the event? Another example of dynamicity is the daily traffic variation in large cities; traffic in business areas peak during working hours whereas traffic in residential areas peak after office hours. Could we make the network adapt dynamically to regional traffic dynamicity? A third example is mission critical applications, like remote control of heavy equipment where a high performance connection is needed at all times. How do we dynamically make sure that this client always has a reliable connection to its control center?

In our 5G transport demo, we consider a potential deployment scenario, where baseband units are centralized as shown in the figure below.

At the antenna site, remote radio units (RRUs) – in the left part of the figure – are connected to baseband processing units in the more central part of the network, using CPRI fronthaul over an optical network. This optical network is based on DWDM where different channels are transmitted on different wavelengths in the same fiber. Although the demo considers optical fronthaul, the concept of programmable transport is not limited to this, but can also be applied to many real networks deployments which include e.g. microwave links, and converged transport of IP services for residential and enterprise customers.

We have designed a multi-layer control architecture based on Software Defined Networking (SDN) concepts outlined in the figure below.

The architecture provides unified control over multi-domain infrastructure with heterogeneous resources, with a particular focus on radio and transport. This is in contrast to today’s networks, where different domains are controlled and managed independently from each other. For example, the transport is controlled independently from the radio domain, resulting in an overprovisioned transport network and inefficient use of resources.

In order to show the benefits of the multi-domain orchestration we have developed a proof of concept demo for an “Elastic Mobile Broadband Service”. By this we mean that broadband services are provided to mobile users and the service capacity is dynamically and automatically scaled up and down—when and where needed. An LTE network with two macro cells and two small cells are used to show the features.

To realize the multi-domain orchestration according to the selected architecture, we need each domain be programmable, preferably through a single programming interface. To realize the programmable transport network we have developed an SDN transport controller based on the OpenDayLight controller. We have extended OpenDayLight with several additional features to enable control of transport networks including optical DWDM technologies.

A “Radio/Transport Optimizer” application is part of the demo. This application runs on top of the orchestrator and decides how many active small cells are needed and at which locations they are needed. To make this optimization, the application continuously monitors the traffic situation and the demand. It then configures the network across the different domains to adapt to the situation. It uses the interface to the orchestrator for activating or deactivating the affected small cells, and to setup the connection in the underlying fronthaul network. The snapshot below shows the monitored traffic situation and what happens when the traffic reaches a certain threshold: A small cell 2 is activated and the corresponding fronthaul connection is set up between the RRU and the baseband unit.

In a larger network, macro cells would provide coverage and a persistent connection for each end user or device, and always be turned on. Small cells would provide extra capacity at hot spots, complementing the macro network. There could several small cells for each macro cells in such deployments.

In summary, 5G transport will enable a unified orchestration of transport resources with radio and cloud to support the wide range of services, radio deployments and business models in the Networked Society.

Peter Öhlén, Ahmad Rostami, Björn Skubic, Zere Ghebretensaé, Ericsson Research.

The Ericsson Blog

Like what you’re reading? Please sign up for email updates on your favorite topics.

Subscribe now

At the Ericsson Blog, we provide insight to make complex ideas on technology, innovation and business simple.