Part 1: Is There a "Next" for the Internet?
Reported by Tatsuya Kurosaka (Tatsuya Kurosaka Office)Is the Internet an infrastructure?
It has been about 10 years since the Internet started commonly being used. In the meanwhile, various private and public services using the Internet have become available, from services providing chat to services for using and purchasing content as well as filing tax returns. Nowadays, it is not even strange to consider that the Internet is a social infrastructure almost to the level of electricity, gas and water. Users also trust the Internet as a communication infrastructure and are starting to depend on it in many places in their lives and businesses.
In this way, the Internet has won people’s trust; however, few of the people who are building and are responsible for the Internet actually publicly acknowledge that the Internet is a social infrastructure. For example, the Government continues to take the position that "the Internet is something that the private sector has been developing" and they do not clearly state that it is a social infrastructure. Also in the private sector, you see few companies declare in a true sense that their business deals with the construction and operation of a social infrastructure no matter which layer they work on.
It is true that a structure based on many different people providing their "best effort" does not have any single entity clearly responsible for reliability and quality; to put it the other way around, this structure may make it hard for suppliers to see (or it may seem that there is no need for them to be aware) that "we are supporting a social infrastructure". Also, to take a different point of view, Internet use has become sophisticated and complicated particularly because the division of work among the layers has increased during the past few years; that is, it is no longer something that can easily be resolved and managed by any specific entity.
Either way, looking from the users’ viewpoint, when an Internet problem occurs, unless you can clearly identify where the cause of the problem is, it means that it is currently impossible to ask the relevant entity to fix the problem. You may simply say, "I can’t connect to the Internet"; however, there are a wide variety of possible causes for the problem, and it is quite a bit of work even to count how many possible causes there are. It is totally different from other services, where users can contact the power company when they lose electricity or they can contact the gas company if they have a problem with their gas service. This is probably the reason why a "person who is knowledgeable about the Internet" is appreciated in various places or why services for technical support are included in the ISP’s menu.
During the dawn of the Internet, users first recognized the issues, and while trying various solutions, they asked for help as the need arose. That is how the problems were supposed to be solved and there is no doubt that it became the breeding ground for the spread of the Internet. However, the Internet in Japan nowadays is moving from the growth phase to the mature phase, and users seek guaranteed quality and responsibility from suppliers. This trend will probably become stronger and stronger, and in order for the Internet to keep maturing in a real sense, there must be an "environment users can rely on".
The Internet is dilapidated
Then, is the Internet today an environment that users can rely on?
As a matter of fact, there are many worries. For example, in the lower layers, there are always concerns about the communication quality due to lack of bandwidth; the fact that the IPv4 Internet infrastructure is exploding or collapsing is already a common issue for those on the front lines. On the other hand, in the upper layers, as the convenience and diversity of services improve, concerns towards information management and security are increasing. So the number of situations requiring an integrated course of action that involves the lower layers are increasing.
At the same time, we are beginning to see disharmony generated between the lower and upper layers. For example, YouTube was hounded by speculations that "their communications provider billed them enormous amounts of money for the bandwidth use by their video distribution service so they will eventually run out of funds and their business will shut down." Then they were acquired by Google in the fall of 2006, which enabled them to continue service; however, this only resolved the immediate problem in the business sense. It was not the case that drastic measures were taken by the Internet industry as a whole in order to resolve problems faced by YouTube-like services.
Like this, two issues exist in the Internet: 1) bottlenecks and supply/demand gaps caused by its growing and increasingly sophisticated use, and 2) blame games among the layers. At the present point in time, both issues are avoiding a big collapse due to efforts at the operations site of the various layers; however, since this causes the issues to be hidden from users’ eyes, I feel that it is inviting users to further depend on and use the Internet even more, and hence the potential risk is increasing.
With such a situation in mind, Mr. Jun Murai points out that "the Internet is dilapidated." The bottom line of his words may be rooted in frustration, which is "the Internet is becoming unable to meet users’ expectations."
However, what we want to keep in mind here is that such comments and concerns are not a "denial of the Internet" at all. Rather, the Internet architecture and its idea still has meaning and an advantage towards the existing infrastructure in terms of "expandability, flexibility, and creativity"; of all things, the fact that users’ expectations are still growing is probably proof of this. In fact, the Internet has grown as a "major communication service" throughout the entire world due to these advantages. In order for users to continue using diverse (and unrestricted) communication methods, it is important for the Internet to continue growing and hence, the dilapidation needs to be fixed. These things are probably the needs for the next generation Internet.
Two directions for the Internet to proceed
If the next generation Internet exists beyond the current Internet being repaired, then its form will probably be something that is based on the current Internet. The next generation Internet is not something that will appear out of nowhere; it will be developed by the players who currently already take part in the Internet.
However, various players already take part in the Internet and the details of their approach to solving issues differ from each other. This is because the path they take differs greatly depending on the factors that determine their services and their strategy, e.g. the market condition, assets, developing technologies, future perspectives, etc.
As stated above, currently the work is divided among the layers. Therefore, it is probably easy to understand the difference between approaches towards the next generation Internet if we divide the players according to the layer they work on. For example, if it is a company whose core business lies in the upper layers, their business is focused around that layer. They want the lower layers (network) to be convenient for the upper layers (especially the application layer) and seek to build a network based on that (see Figure 1). On the contrary, if it is a company whose business is focused around the lower layers, since taking upper layers’ trends into account reduces risks, they aim first to design a network based on their own expectation of the business and technological structure, and then they integrate the upper layers (applications) there (see Figure 2). After all, both upper and lower layers focus on technologies that are convenient for their own layer and hope the others will follow them.
Up until now, the balance between these two, as to which one is principal, has changed depending on the status of the spread of Internet services. For example, when there are changes in the infrastructure, just like during the dawn of broadband services, it leads to implementations based around the lower layers. Once the infrastructure spreads to a certain extent, how it should be used then becomes an issue, which leads to implementations based around the upper layers. Of course, since it is universally the case that the lower layers support the whole infrastructure, the situation changes if later innovations occur in the lower layers; however, currently it has not reached that point yet, hence work on the upper layers is very active. At the same time, it also looks as if that activeness is affecting the innovation in the lower layers.
The common flaw to both approaches is that they focus only on the layers that make up their core business. For example, as shown in Figures 1 and 2, in implementations based around the lower layers, it is not always the case that the upper layers (application and service vendors) would behave as the lower layers expect. On the other hand, in implementations based around the upper layers, a situation where the infrastructure cannot support the upper layers’ activities can occur. In fact, Google, the acquirer of YouTube, is in opposition with communication carriers in the Net Neutrality debate, and you can take it as carriers’ message that they cannot technically and business-wise support the enormous traffic that services like Google have in the future.
By the way, NGN is something that probably tries to use a different approach from the Internet at the architectural level (see Figure 3). In that sense, it can be said that NGN is the most challenging alternative to the Internet; however, as long as NGN’s identity is the "architecture that is not the Internet", horizontal interoperability with the Internet is probably difficult (although interoperability using an overlay called Internet over NGN is possible). Also, because of that, the economic framework of the entire network that NGN itself tries to support would be different from the Internet. Therefore, recognizing NGN as the direct answer to the next generation Internet seems odd.
The Internet lacks long-term vision
In order to fix the dilapidated Internet, no matter which layer’s approach you take, it is essential that each layer cooperates with the other layers so long as those layers constitute the network. However, the reality is that disharmony among the layers is starting to appear. Not only the above-mentioned Net Neutrality debate, but also some other matters are already being recognized as issues, including security related control points (which layer would take care of it), the relationship between applications and traffic (how can a situation where specific applications consume most of the traffic be controlled), etc.
If I am to raise the issue provocatively, I suspect that each player in the current Internet may not realize that they are "stake holders who support the Internet". Because of that, I feel that they fall into the thought of partial optimization that focuses on the service they are in charge of and the balance as a whole is falling apart.
In addition, if the Internet is intended to be an infrastructure, each player in the Internet is required to take "responsibility as an infrastructure company", which means that they can no longer take the lenient attitude that they are giving their "best efforts". That is precisely the attitude they should get rid of. I think they should learn from the engineers who support the telephone and broadcasting networks and take pride in their attitude of "never let it go down, even for a second".
If there is any merit to my concern, this issue comes before the technical solution, which means that it is necessary to organize things related to the vision. That vision is one that all the Internet players can share no matter which layer they belong to, a policy for each layer to be in agreement and for the Internet to upgrade as a whole.
Luckily, there have been already some visions like that and elemental technologies that reflect those visions developed by many predecessors. Since specific problems have actually started occurring and there is not much time left until those problems need to be resolved, we are very fortunate that we have these visions when you consider the trouble you have to go through to create a vision from scratch. In addition, there is no doubt that inheriting its history and assets leads to increasing the depth of the Internet.
Since these are still fragmented, it is necessary to organize what can be used, and what we want to use, for the next generation Internet, and to sketch anew that which is missing. In this series, I will identify such missing pieces for the "next generation Internet" through the interviews with people who take part in various layers of the Internet.
