...from the user's view

Today we live in a world that is full of technical equipment.
By controlling our air conditioning, lighting, serving media etc. the little electronic helpers around support us to have an easier and more comfortable life.

Today these systems coexist non cooperatively in most cases: even though connecting the functionality isles is possible (some sort of network interface exists) they do not interact.
This is more by accident then by purpose: the different components come from different manufacturers and are therefor incompatible.

As result of this coexistence the user has to be the bridge the gap between the different specialized functionalities:
For instance if he starts watching a movie the lights might automaticaly get dimmed, the shutters closed, the air conditioning set to a less noisy mode etc. all depending on information like it is day, it is stormy etc. available through again other systems.

In the future the functionality isles will be connected. This is necessary in order to provide more comfort to the user. Since the user will gently pay for comfort this will happen.

I am currently developping a smart network operating system that integrates existing solutions and allows us to get information of and control all the available devices.
The whole system will contain more information than its parts together since it will be able to generate additional knowledge. This allows more automation and this again can lead to more comfort.

With the system I want to find out what will be necessary for such an integrating approach. I am curious to explore what might be possible in the future.
I am not limiting myself to the technical aspects here.
I am also concerned with usability aspects: such a "allmighty" system will have some complexity that must be shadowed from the user by not limitating his possibilities at the same time.
Also ethical aspects are very relevant in this context because the evolving information and especially its diffusion must be carefully judged and controlled. Automating things always brings a risk of taking control from the user. This must not happen.
Probably additional fields will open as soon as we have a running system.

I am always looking for encouraged students that want to be part of the project. So if this sounds interesting to you, I am happy if you contact me!

 

...from the network's view

Looking at todays computer networks we see that the knowledge is in the end nodes: for instance congestion and flow control are things controlled by the TCP-instances of the end nodes, routing is controlled by more or less isolated routers only sharing a minimum knowledge.

So today's networks don't know much about their components.

This situation is not surprising taking into regard the fact that the Internet of today is technically compatible to the ARPANET of the late 1960s.
Only the fact that the components worked independent from each other allowed such an immense growth of the network without running into much trouble. Other approaches would have needed much more bandwidth and CPU-power that was not available.

Today we have the bandwidth as well as the CPU-power in most cases.
Providing the knowledge that is in only in the end nodes today to the core network would allow much better performance: we would be able to react to problems encountered at the traffic ends as well as to heavy traffic in some areas of the Internet using deviations that are not known to todays systems.
This might be similar to the street traffic where we installed a measurement and routing infrastructure over the years allowing traffic management and therefor better performance.
In a network we already have the physical connectivity between as well as the needed entities themself.

What is missing is some sort of interconnecting middleware allowing the implementation of more abstract concepts with wider impact.

It will be possible to evaluate the interaction possibilities of existing components with the framework I am currently developping. So as stated in the paragraph above I am curious to figure out what is possible and which impact it has.

I am always looking for encouraged students that want to be part of the project. So if this sounds interesting to you, I am happy if you contact me!

 

Briefly: I am a researcher in the field of autonomic networking.

...for further information to the respective course refer to the teaching menu on the left.

ss2009

* Grundlagen Rechnernetze und Verteilte Systeme
* Internetpraktikum

ws2008/ 2009

* Internetpraktikum

ss2008

* Grundlagen Rechnernetze und Verteilte Systeme