NEW

4. IoT-dedicated Networking with NFV

4.1 Concept of IoT-dedicated Network

We propose an IoT-dedicated network as the solution to the issues discussed above.

The IoT-dedicated network is a network optimized for IoT services and separated from the traditional smart phone oriented networks. It builds the EPC (Evolved Packet Core) nodes, including the MME (Mobility Management Entity), P-GW (Packet data network Gateway) and S-GW (Serving Gateway) for each service. Fig. 3 shows the concept of the IoT-dedicated network and Fig. 4 shows an example of its configuration.

For example, when building a dedicated sensor network, it is possible to build a mobile network that is tuned optimally according to the C-Plane/U-Plane ratio proper to a sensor network in which a large number of devices are coected.

By building such a network on the same platform as an ordinary network by using the NFV function, it is possible to extend the equipment according to the properties of IoT traffic with reduced CAPEX/OPEX.

Additionally, optimization of network services and resources using the MANO (Management and Orchestration) and use of low-latency processing using MEC (Mobile Edge Computing) enable the flexible handling of the diverse traffic requirements of the IoT.

In the following sections 4.2 to 4.4, we describe the technologies necessary for building an IoT-dedicated network.

4.2 NFV for Flexible Architecture

NFV is the technology that turns the network functions previously implemented using dedicated hardware into software, and runs it on a general-purpose server. The improvement in performance of general-purpose servers and the progress of server virtualization as seen with the hypervisor have made it possible to secure performance, scalability and reliability that is applicable even to carrier networks. By building a virtualized IoT-dedicated network utilizing NFV technology and by optimally deploying resources according to service requirements, it becomes possible to implement network infrastructures that are flexible, economical and optimized for the IoT.

4.3 MANO for Dynamic Configuration Changes

MANO is the orchestration function that operates and manages services and resources in an integrated maer in the NFV environment. When the virtualized IoT-dedicated networks are diffused, it becomes necessary to optimize the resource and function deployment of the entire network dynamically, as well as to introduce services quickly.

The use of MANO enables the dynamic processing of resource allocation to the core network as well as achieving flexible and quick configuration changes to the IoT-dedicated network according to the real-time traffic situation.

4.4 MEC

The MEC is the technology that performs the IT processing, which is usually done in the cloud of the Internet, at the edge of the mobile network and closer to the users (base stations, etc.). The use of MEC makes it possible to handle the service requirements that necessitate real-time processing or analysis of huge amounts of data without imposing too much burden on the core network and data center. For example, in the implementation of a mission-critical service that needs data transmission/reception at an ultra-low latency, such as for collision avoidance of automatically operated vehicles. Building the network by using MEC makes it possible to eliminate the effects of the transmission delay according to distance.

5. Applications to Usage Cases

5.1 Traffic Control

Fig. 5 shows an outline of a traffic control usage case. The network collects and analyzes big data (sensor information and images) from multiple IoT devices on the roads (vehicles, signal systems, etc.) Thereby predicting traffic situations such as collisions and congestions in real time provides notification via road-to-vehicle and inter-vehicle communications. This procedure enables traffic control for congestion easing, collision avoidance

and coordinated driving.

This usage case is required to reduce the latency in road-to-vehicle and inter-vehicle communications to between 1 and 100 ms. Particularly, in the mission-critical services such as in collision avoidance, the required latency should be achieved with a reliability close to 100%. It is also required to provide the vehicles moving at ultra high speeds with a highly reliable means of communication for coecting them with millions of IoT devices on the road. Nevertheless, with the traditional mobile network, the long data transmission distance between the IoT devices and the cloud system provides service results that produce propagation latency according to the distance. This leads potentially to latency in the road-to-vehicle and inter-vehicle communications being unable to be reduced to the range between 1 and 100 ms. In addition, optimization is also required to provide highly reliable communications for vehicles moving at ultrahigh speeds.

It is only after the IoT-dedicated network using the NFV is tuned optimally to meet the low latency an high reliability requirements specific to the traffic control and is built using the MEC that the service requirements for the low latency and high reliability can be fulfilled perfectly.

5.2 Real-time Image Distribution for a Stadium Audience

Fig. 6 shows an example of real-time image distributions according to the wishes of viewers as a usage case of IoT-based entertainment. When the 4K movie streaming of ultra high definition became popular, a viewer of a soccer game, for example, could automatically track the performance of a favorite player or refer to the related statistical information using the image attribution information. In the stadium, it is also possible to collect image samples from cameras and wearable terminals at the site, generate images from viewpoints according to the demands of spectators (viewers) and distribute them selectively. The spectators (viewers) can also find facilities in the stadium using the navigation service or receive ads matching their demands.

This usage case requires accommodation of huge communication traffic from multiple viewers, particularly for a large-scale event. To provide the individual viewers with ads and navigation with AR (Augmented Reality), it is also required to analyze the viewers’ positions and demands and to distribute contents to them in real time (about 100 ms).

However, an ordinary mobile network may be unable to provide high-quality services due to congestion particularly when a large number of coections and a large amount of communication traffic is produced in a specific area. The IoT-dedicated network using NFV can tune the network functions and resources dynamically according to the traffic requirements for multiple coections and large capacities, and can meet the service requirements under such conditions.

6. Conclusion

In the above, we introduced the IoT-dedicated network as a network technology for supporting communications in the age in which the IoT is becoming more popular.

The technology introduced here can meet the diverse traffic requirements of the IoT and build a communication network that is safe, secure and economic.

At NEC, we believe that this solution can lead to improve the social value of the networks that support IoT and we are determined to pursue further social ideals based on applying the present technology.