Ichiro Satoh

A Framework for Data Processing at the Edges of Networks

This paper proposes a distributed processing framework inspired by MapReduce processing. It is unique to other distributed processing approaches to large-scale data, i.e., so-called big data, because it can locally process data maintained in distributed nodes, including sensor or database nodes with non-powerful computing capabilities connected through low-bandwidth networks. It introduces mobile agent technology so that it distributes data processing tasks to distributed nodes as a map step and aggregates their results by returning them to specified servers as a reduce step. The paper describes the architecture of the framework, its basic performance, and its applications.

MapReduce-Based Data Processing on IoT

This paper presents a framework for data processing on the Internet of Things (IoT). It enables data on multiple nodes on IoT to be locally processed as much as possible by using the Map Reduce processing without transmitting the data to data-centers or clusters, and assembled into a few nodes. This paper describes the design and implementation of the framework, basic performance, and practical application.

Self-Adaptive Resource Allocation in Cloud Applications

Cloud computing has several differences from distributed computing in that the former is a special form of the latter. Whereas redundant functions and data are needed in distributed systems for reason of dependability, cloud applications should not support such functions and data, because cloud computing infrastructures support dependability on behalf of the applications. This paper addresses resource allocations for applications on cloud computing rather than applications on typical distributed computing. The approach proposed in this paper is to enable redundant architectures, e.g., master-slave architectures, to be dynamically transformed into less redundant architectures with the objective of minimizing the costs associated with them and meeting customer demands and application requirements in a self-adaptive manner.

RFID-enabled carbon offsetting and trading

This paper presents a novel approach to carbon credit trading with pervasive computing technologies, particularly RFID (or barcode) technology. It introduces RFID tags as certificates for the rights to claim carbon credits in carbon offsetting and trading. It enables buyers, including end-consumers, that buy products with carbon credits to hold and claim these credits unlike existing carbon offsetting schemes. It also supports the simple intuitive trading of carbon credits by trading RFID tags coupled to the credits. The approach was constructed and evaluated with real customers and real carbon credits in a real supply chain. It can also be used to encourage industries and homes to reduce greenhouse gas emissions.

Evolutionary Mechanism for Disaggregated Computing

This paper proposes a framework for configuring software agents on distributed system consisting of heterogenous computers. It is unique to other existing approaches for self-configuration because it introduces the notions of differentiation and dedifferentiation in cellular into real distributed systems. When an agent delegates a function to another agent coordinating with it, if the former has the function, this function becomes less-developed and the latters function becomes well-developed. Agents can escape to other computers whose capabilities can satisfy their requirements. The framework was constructed as a middleware system and allowed us to define agents as Java objects written in Java Bean. We present several evaluations of the framework in a distributed system instead of any simulation-based systems and describes practical applications.

Resilient Architecture for Complex Computing Systems

This paper proposes a self-organizing framework for adapting software agents on complex computer systems, in particular distributed systems. It is unique to other existing approaches for software adaptation because it introduces the notions of differentiation and dedifferentiation in cellular slime molds into complex computer systems. When an agent delegates a function to another agent. the formers function becomes less-developed and the latters compatible function becomes well-developed. The framework was constructed as a middleware system and allowed us to define agents as Java objects. We present several evaluations of the framework in a real distributed system.

Self-adaptive mechanism for distributed computing

This paper proposes a framework for adapting software components for a disaggregated system, which dynamically composes devices, e.g., displays, keyboard, and mice, which do not attached to the same computer into a virtual computer over a distributed system, including ubiquitous/pervasive computing systems. It introduces the notions of differentiation and dedifferentiation in cellular slime molds into real distributed systems, including disaggregated systems. When software components delegates a function to another component coordinating with it, if the former has the function, this function becomes less-developed and the latters function becomes well-developed. The framework was constructed as a middleware system and allowed us to define agents as Java objects written in JavaBean. We present several evaluations of the framework in a distributed system instead of any simulation-based systems and describes a practical application.

Agent-based MapReduce Processing in IoT

This paper presents an agent-based framework for processing data at nodes on the Internet of Things (IoT). The framework is based on MapReduce processing, where the MapReduce processing and its clones are popular but inherently have been designed for high-performance server clusters. It aims at enabling data to be processed at nodes on IoT. The key idea behind it is to deploy programs for data processing at the nodes that contain the target data as a map step by using the duplication and migration of agents and to execute the programs with the local data. It aggregates the results of the programs to certain nodes as a reduce step. We describe the architecture and implementation of the framework, its basic performance, and its application are also described here.

A Data Processing Framework for Distributed Embedded Systems

A MapReduce-based framework for processing data at nodes on the Internet of Things (IoT) is presented in this paper. Although MapReduce processing and its clones have been designed for high-performance server clusters, the processing itself is simple and generalized, so it should be used in non-high-performance computing environments, e.g., IoT and sensor networks. The proposed framework is unique among the other MapReduce-based processing approaches, because it can locally process the data maintained in nodes on the IoT rather than within high-performance server clusters and data centers. It deploys programs for data processing at the nodes that contain the target data as a map step and executes the programs with the local data. Finally, it aggregates the results of the programs to certain nodes as a reduce step. The architecture of the framework, its basic performance, and its application are also described here.

Specifying Distributed Adaptation through Software Component Relocation

This paper proposes a theoretical foundation for specifying and reasoning about adaptations based on our middleware system that introduces the relocation of software components to define functions between computers as a basic mechanism for adaptation on distributed systems. It provides a language for specifying adaptations policies. The language is useful to reason about adaptations and can be executed in the middleware system.