Searching for Autonomous Distributed Algorithms

Free and logical thinking: that is where innovation is born.

Masuzawa, Toshimitsu

Graduate School of Information Science and Technology
Algorithm Engineering Lab

Masuzawa, Toshimitsu

Professor Masuzawa's research theme is "Autonomous Distributed Algorithms". This research relates to distributed algorithms that facilitate communication among networked computers configured in a distributed system to direct their behavior.

Developing distributed systems that can adapt autonomously

Within distributed systems, unforeseen problems can arise, such as hardware failures, user input errors, and changes to connectivity due to the movement of users. We are working towards developing systems that can adapt themselves autonomously when problems like these arise.
One example of what we are working on is a system where wirelessly-networked devices can automatically establish connections and then operate without any conflicts as a whole. Such systems could be used for devices brought to the site of an earthquake or other disaster. “Operate without any conflicts as a whole” means that the system detects drops in performance or periods of unstable behavior but is still able to guarantee an eventual return to ordinary usability.


Applying the knowledge of our natural world

There is growing agreement among those involved in interdisciplinary research that in the near future, computer communication networks will function with the same environmental adaptability as we see in the natural world. Our research lab has considered using the relationship between population and food supply in the living world as a model for the redundancy management of files and data. The application of the mechanisms of the nature can be thought of as one way to approach our research.
However, even when applying knowledge of biology into our research, theoretical research is still essential. In other words, rigorous mathematical proofs must be made as well. This is an important point to keep in mind for our work.

There are no rules for logical methods

masuzawa3In research, it is important to think logically. However, I don’t think there are rules that say which kind of thinking is the most logical. Instead, I feel that ways of thinking should be unfixed and free to arise from various perspectives. I think the key point of this process is to develop methods logically through a process of properly identifying valid and invalid arguments and evaluating their applicability.
One excellent aspect of the Humanware Innovation Program is that it provides opportunities for students from different fields to hold discussions together on a regular basis. The program is designed to allow students to develop their ideas freely without interference, and to grow as individuals capable of demonstrating leadership in fields of interdisciplinary research. However, as this program merely provides the opportunities for interdisciplinary research, just how these opportunities will be taken advantage of is up to the students themselves. I want all participants in this program to make the most of what this program offers.