Recent advance in large amount of data processing in parallel, and large amount of IoT data communicated over the network left us one issue, i.e., "how many computational resources are required to optimize the throughput? The more, the better?" It is required to establish a theoretical model to find a solution.

Therefore, my research objective is to establish a methodology to collaboratively process jobs among computers over the network. In particular, my current research themes are: (i) how to minimize the response time with a small number of computers, (ii) how to allocate each VNF (Virtual Network Function) to a virtual machine in order to avoid depletion of computational resources, and (iii) how to optimize the number of computers and data chunks for data analysis in a big data processing infrastructure.

Thus, algorithms for scheduling tasks and resource provisioning are required to achieve above themes. I am working on those themes from both theoretical aspect and practical aspect.

1. Scheduling algorithm for scientific workflow jobs
2. Assignment of VNFs (Virtual Network Functions) for service function chaining
3. Resource provisioning in big data processing

https://sites.google.com/edu.teu.ac.jp/netcomplab/