Application quality monitoring and visualizing technology

①From network operation to service operation

With “telephone services ” using standard terminals, network quality nearly equals service quality. However, with networks for ICT infrastructures, network quality does not equal service quality. That is, we do not know whether customers can comfortably use services or not if we monitor network quality only. Accordingly, it is important to monitor the service quality itself and use the results to provide network quality operation.

Click on image for larger version.

We are using techniques based on quality assessment technology for speech and video to develop methods that quantify the QoE of recent applications such as cloud services. These methods enable us to estimate QoE from monitorable performance indicators of networks, terminals, and servers [Application quality estimation technology], and to monitor the quality of various services.

②From providing quality to co-creating quality

The number of users who have high ICT literacy is expected to increase in the future. Moreover, in line with the development of FMC (Fixed Mobile Convergence) and mobile offload, and the appearance of new types of carriers such as MVNOs (Mobile Virtual Network Operators), users are expected to have more choices in the networks they use. In view of these circumstances, carriers must provide high quality to users, and users must be able to improve QoE by devising and choosing networks. That is to say, users and carriers cooperate and “create” quality. To be able to devise and choose networks, users must have the necessary “information” for making a decision. For example, for train users, it is most important to know whether trains are running on a regular schedule (quality provided by carriers). However, for users who drive their own cars, the information necessary for making travel plans is most important, such as information on traffic conditions and weather forecasts.

In the transportation analogy above, along with communication quality, carriers should provide quality information to users. To do that, carriers need methods to grasp the quality of networks and services quickly and easily, and to express it using indicators that are easy for users to understand. Additionally, methods to predict network conditions may be necessary in order to do “a quality forecast” like a weather forecast. Moreover, inputting such information to applications, it may be possible to control quality suitable for networks. We are now working on the following technologies in order to realize these frameworks.

[Application quality estimation technology] grasps the actual state of provided services online, and map it to the condition of networks, terminals, and servers
[QoS tomography technology] analyzes the end-to-end quality structure by analyzing the service quality of various users
[Technology for expressing quality] extracts the necessary information for users to choose networks and services, and expresses it easily
[Technology for analyzing user behavior] predicts the behavior of users when providing the above information.

Click on image for larger version.