Software for Cloud Computing

Recently, in the rapidly emerging cloud computing technology, dynamically and automatically constructing as well as allocating computing environment by virtualizing computing resource based on software technology has been gathering attentions. In our research, we aim at manipulating computing resource by utilizing software technique. Specifically, Software Defined Networking (SDN) which is a software-based networking-control technique is in our main focus. Based on SDN, we hope to extend and advance software techniques that can support cloud computing-related technologies such as Cloud Gaming, Big Data Analysis, Machine Learning, IoT and etc.

Our Current Research topics

Software Defined Networking (SDN)

By using software techniques to virtualize computation resource, virtual machine technologies that can dynamically and automatically construct or allocate computer environment have been established. With the rising popularity of open source softwares such as KVN and Xen, or commercial product such as VMware, everyone can construct their own cloud environment by using the above infrastructure softwares. On the other hand, the virtualization of network is gaining attention. If network can be also configured and allocated automatically and dynamically based on software, we can provide users with a network that corresponds to their requirements. Furthermore, by integrating such network with cloud computer, we can construct more optimized computer environment. As a concept of technology to implement such environment, the idea of Software Defined Networking (SDN) is being proposed in recent years. As we can acknowledge from the name, SDN refers to defining network by using software. In SDN, the concept of Programmable Networking is being proposed and it aims at building an environment that makes network programmable.

In this lab, we are leading PRAGMA-ENT (Experimental Network Testbed), which is an international project for building and applying demonstration-purposed network that evaluates SDN research ideas in a distributed worldwide large-scale network environment. Therefore, students can directly apply their ideas in such international network environments to perform analysis and evaluations. For details please visit the project website.

Research for Application-aware Routing

In recent years, thanks to the versatility of network, we have thousands of applications that provide wide-range of services such as video streaming, web-browsing, chatting, SNS and etc. However, current network routing design is not optimized for coping with applications because it is only determined by the sender and the receiver. Therefore, the routing cannot be considered as efficient. Application-aware routing is to develop techniques that utilize SDN technology to determine the optimal path for applications. Please click here for details.

Research for Multipath Routing

In a network, multiple paths, including the alternative ones, usually exist between a sender and a receiver. However, current network routing is not able to exploit multiple paths simultaneously, which means an application can only choose one single path for the transmission. The technology of Multipath Routing aims at maximizing the performance of network by distributing multiple TCP transmissions generated by an application into multiple different routes. As the approaches to achieve multiple TCP transmissions, we apply MPTCP technique which is at the level of OS’s TCP transmission library. Furthermore, we also perform multiple TCP transmission at application level by implementing our own application. For more details, please click here and here

Research for Monitoring and Analyzing Network based on SDN Technology

By applying SDN technology, it becomes possible to monitor network in greater details. In order to realize Application-aware Routing mentioned in previous session, it is necessary to understand more details about the property of transmission carried out by an application and how it conducts such transmission. In this research, by applying machine learning technique such as deep learning to analyze and classify large-scale packet capture database, we aim at developing SDN-based approaches that can efficiently classify transmissions of applications in network.

Software Engineering for SDN

Based on the concept of Programmable Networking, software plays an important role in the network that can be controlled by program. Even though the techniques that are fostered in the field of software engineering have been applied in the program which implements SDN, the corresponding environment is not yet sufficiently organized. Therefore, from the perspective of software engineering, this research aims at developing approaches that make the development of program for SDN more efficient.

Research for Cloud Gaming

Besides the increasing popularity of cloud-based storage and computation service, the utilization of cloud in practical service also benefits entertainment-related business to move their products to cloud-based. Among them, cloud gaming is a new entertainment service that executes games on cloud server. Therefore, by merely connecting to network, cloud gaming allows users to play games with high-quality graphics at anytime in anywhere. In this research, by improving the performance of cloud gaming, we aim at developing a framework that can improve graphics quality. Furthermore, based on the same technology, we try to facilitate cloud-based VR gaming experience. Please click here for details.

Kohei Ichikawa
Kohei Ichikawa
Associate Professor