5G for improving urban transport

Congestion, accidents and negative environmental impact are some consequences of increasing road traffic in cities around the world. We are convinced mobile technology can help alleviate many such problems, and are doing research on several transportation use cases together with Scania and the Royal Institute of Technology in Stockholm. 5G networks are essential to these use cases, as they comprise technologies used to reliably and timely transfer mission-critical data between vehicles, roadside infrastructure and software in the cloud.

Beyond MBB feature bus on the road
Leonid Mokrushin

Principal Researcher

Category

The number of vehicles on the roads is constantly growing. One study predicts the figure to surpass 2 billion by 2030. This is due partly to global urbanization, where the UN estimates that 21.1% of the world’s population will live in cities by 2050, up from 9.2% in 1990 and 11.7% in 2013.

To address contemporary and future transportation challenges, Ericsson Research earlier joined Scania and the Royal Institute of Technology (KTH) in the Integrated Transport Research Lab (ITRL). Under the ITRL umbrella, we have established the Connected Mobility Arena (CMA) project, where we will prototype a number of solutions for sustainable urban mobility. The initial phase of the project involves creation of a test site in the suburb of Kista, Stockholm, where prototypes of transport solutions will initially be built and demonstrated on a small scale. A major component of the test site is inclusion of early prototypes of 5G enabling technology. We expect, that the process of building and testing transport use cases in the test site will lead to a better understanding of the requirements these use cases will eventually put on 5G networks.

We envisage building use cases incrementally, and initially focus on public urban transport - in particular people and goods mobility. In general, the use cases fall in three categories.

  • Driver Assistance: buses and trucks with human drivers receive contextual information to improve efficiency of the respective transport service. The real-time recommendations to drivers (e.g. “wait 2 more minutes at the next bus-stop to pick up 10 more passengers”) and passengers (e.g. “take the bus at stop B in the next 5 minutes”) will offer a compelling service at a lower operational cost and reduced environmental impact.
  • Semi-Automation: vehicles such as buses or trucks are autonomous and supervised by off-board human operators; or are remotely driven by off-board human operators under certain circumstances.
  • Full Automation: vehicles are autonomous and supervised by software, such as an automated off-board operator.

Vehicle platooning is an example of semi- and fully automated vehicles. It creates a convoy in which vehicles travel in line very close to each other, coordinating breaking and acceleration. This increases road capacity, and, in the case of urban transportation, can help address commuting demands at peak hours while reducing marginal cost. Dynamically adding driverless buses into a route to meet additional passenger demand is an attractive situation for fleet operators, as the cost of a driver is typically the largest operational cost.

In addition to these use cases, cities around the world are now exploring intelligent transport systems (ITS) to improve commuting by using demand prediction, dynamic trip planning and integrated payment solutions. ITS continually analyzes commuting patterns and takes into account upcoming sports events, concerts or road maintenance projects. Traffic authorities can also benefit from using data provided by ITS, for example location, speed and intended route, to improve urban traffic flow by optimizing traffic signals. Various strategies, such as traffic signal preemption, can be applied to prioritize public transport and emergency vehicles.

Urban transport solutions, like the ones mentioned above, are expected to put specific Quality of Service (QoS) requirements on the network connection, for example maximum network delay and/or minimum guaranteed throughput. This is a departure from the best-effort mobile broadband solution, which is currently the core offering of 3G and 4G networks. Moreover, traditional mobile broadband services will co-exist with critical network services, for instance those used to optimize public transport. For example, emergency response or traffic flow services require priority over infotainment services. The ability to secure prioritized network resources even in high network traffic scenarios allows the 5G system to be used as a foundation on which to build urban transport optimization solutions.

Instead of building a completely new network from scratch, 5G networks will evolve from existing 4G networks technologies, of which one example is the Evolved Packet switched System (EPS) bearer. The EPS bearer is the basic way to separate data traffic from User Equipment (UE), through Evolved Universal Terrestrial Access Network (E-UTRAN) to Evolved Packet Core (EPC). A very important feature of EPS is the assigning of QoS parameters to each EPS bearer in order to classify its bandwidth, latency, packet loss and priority. Thus, the EPS bearer provides a logical, edge-to-edge transmission path with defined QoS between UE and packet data network. Furthermore, it is possible to define the mapping between QoS class in the EPS bearer and the Differentiated Service Code Point (DSCP) used in transport networks, so that the network between the cloud and the EPC also respects the QoS parameters defined in the radio network. This combination of EPS bearer and DSCP makes it possible to define an end-to-end QoS control from UE to the application in the cloud. Using this approach, a different QoS class is assigned to the network traffic flow for different purposes, for example, high priority class for mission critical vehicle control signaling, medium priority class for sensor data from road network, low priority (best-effort) class for infotainment network traffic.

There are two different ways to define EPS bearer on IP flows between UE and network: dynamic and static. Using the dynamic way, the network control function sets up an EPS bearer for specific IP flows on the fly. For example, if a remote surgery is taking place in an ambulance, one way to ensure the network QoS is to set up a dedicated EPS bearer for critical IP flow. This approach is flexible but not scalable. Alternatively, using the static way, EPS bearers are pre-provisioned in the user database. Several EPS bearers for specific IP flows can be provisioned on a group of mobile subscriptions, and network will initiate these EPS bearers when UEs with these subscriptions attach to network. This approach can be applied to massive amount of UEs, but on the other hand, the parameters for configuring such EPS bearers will be more static, for example, the IP flow between the sensors on the road network and traffic management command center.

Figure1


In addition to QoS requirements, urban transport solutions are inherently more dynamic than mobile broadband, as they can have much shorter lifespan and up/down scale many times during operation. Therefore, the lifecycle management process for these solutions (or "5G management") would need to automatically allocate free network resources to serve the needs of a particular solution. In previous work, we have presented concepts for orchestrating transport, network and cloud resources from a common resource pool to form isolated, logical networks known as “network slices”. These networks are isolated from each other and provide end-to-end QoS from UE to software hosted in cloud infrastructure. We published a paper: "Towards automated service-oriented lifecycle management for 5G networks", in the Proceedings of the IEEE 9th International Workshop on Service-Oriented Cyber-Physical Systems in Converging Networked Environments, in September of this year.

Leonid Mokrushin, Rafia Inam, Elena Fersman, Hongxin Liang, Keven Wang, Athanasios Karapantelakis, Ericsson Research.


ABOUT THE CONTRIBUTOR
Leonid Mokrushin
Leonid Mokrushin is a Principal Researcher in AI, in the area of cognitive technologies at Ericsson Research.
The Ericsson blog

In a world that is increasingly complex, we are on a quest for easy. At the Ericsson blog, we provide insight, news and opinion to help make complex ideas on technology, business and innovation simple. If you want to hear from us directly, please head over to our contact page.

Contact us