ITS Conference in San Lorenzo del Escorial offered an excellent opportunity to integrate and discuss different visions about the development of the telecommunications market. Interesting ideas came out when the Net Neutrality issue was on the floor. In this post I will try to summarize some of the main point in discussion:
Quality of service (QoS) parameters and mechanisms are important to enable network operators to design, build and manage their networks, but they are not directly visible to end-users. Crucial for the end-users, however is the quality that they personally experience when they use a service. QoS involves tracking of jitter, latency and other measurable parameters. If the QoS score is not good enough, operators can identify the problem and fix it. With QoE the solution is less straight-forward. QoE is a subjective measure of how the viewer is judging the content delivered by the network. This means the same type of content might be evaluated in a different way depending on the user profile and expectations. In that sense, meeting user expectations would require from the content provider and the network operator a deeper understanding of the user interests, awareness of the content that is traversing the network, and new ways to manage/prioritize the traffic. In these two cases, one of the key tools to provide awareness and activate traffic management policies is deep packet inspection (DPI). However, a challenge arises when data encryption makes difficult the analysis of the information in the network . A combination of sensors in the field, combined with machine intelligence and big data analytics will produce a tremendous amount of data that will help to improve analysis of problems, aid in the planning of system upgrades and even support sales efforts, generating positive effect on users’ QoE . However, it remains to be seen how deeply it can be leveraged by operators. The uncertainty about Net Neutrality is causing some operators to move a bit more slowly on QoE in general and DPI and sensing technologies in particular.
Traffic management is a collection of technologies and policies which lead to different types of traffic being treated differently, which in principle goes beyond the best effort principles that support the original Internet idea. Without traffic management, different data packets are treated more or less equally, which means that under congested conditions traffic management would cause some data to have a greater chance of being delivered than others. Traffic management can be implemented in different ways, which include:
- guaranteeing delivery of data or reserving bandwidth for that data;
- prioritizing certain types of data in the event of queuing;
- de-prioritizing certain types of data;
- restricting certain types of data or the bandwidth allocated;
- blocking certain types of data.
Such discrimination between data types would probably affect users’ QoE; in the extreme some applications would not be able to function. Of course, congestion could also cause applications to fail, but the distinguishing feature of traffic management is that it involves purposeful discrimination. In one hand, the traffic management could guarantee or prioritize data for sensitive applications and reduce the congestion to manageable levels, allowing fair use for all the users, increasing their satisfaction levels. On the other hand, the traffic management can restrict or block certain applications and make other people’s traffic take priority, which can generate a negative impact on the user’s perception.
On the other hand, full transparency would involve providing data that describe the effects of policies over time and therefore the resulting quality of experience for users. This implies the need for diagnostic tools to help users understand whether and in what way traffic management is affecting them.
Finally, we could see that Net Neutrality may influence QoE in two ways: how deeply operators are allowed to examine the packets flowing through those networks in order to use the extracted information to feed mechanisms to improve users’ QoE, and the transparency about the prioritization policies implemented to fulfill user’ expectations and requirements.