Research

Uninformed Search using Lévy Walk in Graphs

Lévy Walk, a type of random walk, is a model for the foraging behavior of marine predators such as albatrosses, bees and killer whales, and is known to provide a statistical approximation to human behavioral patterns. Very little research has been done on Lévy Walk.

In this study, we perform a blind search based on Lévy Walk on unit disk graphs, which are models of road networks, to determine the relationship between the search performance of Lévy Walk and the characteristics of the graph (e.g., diameter) and the distance to the search target. We believe that Lévy Walk-based mobile sensor sampling can efficiently detect critical incidents, such as those in sensor networks that do not assume a network infrastructure (e.g., on a DTN).

Cuckoo Search based Replication Protocol for Mobile Ad-Hoc Networks

Cuckoos do not incubate their own eggs but beg for eggs in the nests of other birds. In this case, which nest to go to is very important to increase the success rate of hatching. Cuckoos use an original mendicant algorithm to select their nest for mendacity very skillfully, so that their offspring can flourish.

Cuckoo Search is a meta-heuristic algorithm which is based on the brood parasitic behavior of Cuckoos.

In this research, we aim to develop a data replication mechanism using an algorithm that mimics the cuckoo’s mendicance to prevent data loss on ad hoc networks while minimizing the consumption of resources on the network for low-demand, but very important information (e.g., personal safety information in a disaster).

Secure IoT Environment using Trusted Execution Environment

While the IoT environment is becoming the infrastructure for providing various services, the data collected by IoT devices, including personal information, requires a high level of security and availability to continue the services. However, IoT devices such as surveillance cameras are more vulnerable in terms of security than ordinary computer systems, and there is a possibility that an external attack could steal the collected data and leak personal information. The same thing can happen in a normal computer system if there is a vulnerability in the operating system. In recent years, the Trusted Execution Environment (TEE), which is a security mechanism that cooperates with hardware, has been attracting attention for these problems.

In this research, we realize advanced security and system availability using TEE in the IoT environment consisting of various devices such as IoT devices and computer systems.