I have been meaning to do this blog for some time .. It discusses the core ideas of net neutrality and how they play in the mobile telecoms arena and why making the network smart could be a dumb move!
Excuse the very basic starting points of this article, but it’s important to explain the perspective from first principles.
Lets start with the Internet
The Internet comprises of any device that uses the IP(Internet Protocol) . The most common user interface (service layer) to the IP protocol is the Web. The Web is built using the HTTP protocol (Hypertext Transfer Protocol) and mark-up languages like HTML (Hypertext Mark-up language).
The wikipedia references give more information about these three protocols.
It might appear frivolous to explain the entire Internet and the Web in a mere three sentences; but none the less – the Internet is nothing more than a set of devices connected by the IP protocol with HTTP/HTML (Web) as the most common user interface.
The primary goal of the Internet is to connect people(and don’t let anyone convince you otherwise!)
Hence, it is suited to person to person communication type applications. The Internet is also a packet switched network . In a packet switched network, the message is broken up into a set of packets (a unit of information) and these packets are routed between nodes over data links.
The philosophy of the Internet lies in the belief that ‘All packets are created equal’. This is also known as Net neutrality.
In other words, the network is dumb and intelligence shifts to the edge of the network(i.e. to the nodes or the applications). This philosophy of ‘dumb pipes and smart nodes’ has fulfilled the vision of ‘Let a thousand flowers bloom’ and has led to the unprecedented innovation we have witnessed in the last decade. We take this innovation for granted .. But its worthwhile spending a bit longer on some of the ideas of Net neutrality.
The effect of net neutrality is so prevalent around us now that the best way to understand it is to consider it’s converse i.e. what happens if ‘All packets were NOT created equal’.
In a scenario where all packets are not created equal, the network becomes ‘smart’ i.e. has some intelligence. In that case, the proposition is: fewer new services will arise leading to a drop in innovation.
We can best explain the ideas of net neutrality by considering a simple networked example where you want to play chess over the Internet. This example is adapted from the book : The 3G IP Multimedia Subsystem (IMS): Merging the Internet and the Cellular Worlds (Hardcover)
It is the best illustration I can find to explain the idea of Open systems and net neutrality.
Suppose you want to design a system where two players are playing Chess over a network. Since this is a connected(networked) application, to create this application, we need to know two things:
a) Knowledge of the network protocols to program the connected application and
b) Knowledge of the application itself (in this case Chess).
Now consider further, two scenarios:
a) The protocols of the network are free and open.
b) The protocols of the network are closed and proprietary
Scenario (a) is the Internet. With an Open system like the Internet, the knowledge of the network is openly available. In this case, the application developer can focus on creating the best Chess application. This is the case of ‘Smart nodes – dumb pipes’ and the ‘Intelligence shifting at the edge of the network’ i.e. to the chess application. It is what we see on the Internet today and it is so common that we take it for granted.
The converse happens if the network is not built on open protocols. If the network is not built on open protocols, then we have a ‘smart’ network. By extension, this leads to a ‘dumber’ application and less innovation.
Taking the example of Chess, if knowledge of the network was a necessity to create the application; then the person with the best knowledge of the network(and not necessarily the best knowledge of chess) would create the best application.
Obviously, the application created by the person who knows Chess best and operates in a scenario of Open systems (i.e. he does not need to know anything special about the network) will be the superior application.
This is the crux of the net neutrality argument i.e. All packets are created equal
The observation is: The Mobile Internet is also now mirroring the fixed line Internet and hence will display the same characteristics(for instance a blooming in applications) as the fixed line Internet. Admittedly, this is taking some time but the trends in this direction are very clear.
Historically, telecoms networks were not IP based (i.e. followed protocols which were different from the Internet protocols). With the global spread of the Internet, the telecoms networks are also increasingly adopting the IP protocol. This is facilitated by a technology called IMS (IP Multimedia Subsystem).
I have had the dubious distinction of comparing IMS to the Sith lords (Of Sith lords and the dark side of IMS). And some have an even more interesting take on that blog
This might sound extreme .. but it underpins a vital observation.
In a wireless network, we all accept that conditions are different(and harsher) than on the fixed network. Hence, some form of optimization(AKA intelligence) is necessary. In theory, that defeats the principle of ‘All packets are created equal’.
The real question is: What is the basis of the ‘inequality’?
If it is truly for technical reasons, then it is fine.
However, if commercial considerations creep in, then we have a big problem. With commercial considerations, the person paying the most will have their packets moved faster.
In any case, making the network smart is a dumb move .. because it penalises the ‘chess experts’ i.e. domain experts and ultimately is a reversal of the innovation mindset we have witnessed in the last ten years.
Let us not take this blossoming of innovation for granted. Competitive natures of whole countries could be decided on the basis of the decisions we make today(both for fixed and mobile Internet)
Cartoon source: slowpokecomics.com