In this paper, we propose a joint estimation method for the sampling rate offsets (SROs) of multiple recording devices. In wireless acoustic sensor networks, distributed microphones are connected to different analog-to-digital converters, and thus SROs occur on non-reference channels, which degrades the performance of various array signal processing techniques. To address this problem, we propose to jointly estimate and compensate SROs of all the non-reference channels. Since the proposed method is formulated as a multivariate non-convex optimization problem, we derive an efficient optimization algorithm on the basis of the majorization-minimization and majorizationequalization algorithms. We further propose to update SROs with only low-frequency components in the initial iterations to avoid undesired local optima. Our experimental results validate the effectiveness of the joint estimation of all SROs and demonstrate its advantage in subsequent array signal processing.

In this paper, we investigate the effectiveness of spatial features for acoustic scene classification (ASC) with distributed microphones. Assuming that multiple subarrays, each containing multiple micro-phones, are distributed and synchronized, we consider two types of generalized cross-correlation phase transform (GCC-PHAT) as spatial features: the intra- and inter-subarray GCC-PHATs. They are obtained from channels within the same subarray and between different subarrays, respectively. The log-Mel spectrogram as a spectral feature and the intra- or inter-subarray GCC-PHAT are processed in the neural network. The experimental results show that increasing the number of channels did not markedly improve the ASC performance when using the spectral features alone. However, using either of the GCC-PHATs as the spatial feature together with the spectral features successfully improved the ASC performance.

Nowadays, a lot of e-Learning systems are widely deployed in educational schools. Typical e-Learning systems are implemented as client-server model. In the client-server model, the number of clients affects on the load of the server. In order to reduce the load on the server, we developed a P2P-based distributed e-Learning system. The proposed system consists of a lot of mobile agents which manage study contents and some functions such as scoring, showing questions, and correct answers. When a learner requests content, a mobile agent who has its content comes to the learner's computer, and then he/she can start the study. Here, a mobile agent has to manage multimedia data, which may be a huge size of data. Thus, a mobile agent has to migrates to the learner's node with a huge size of data. Then, the learner cannot start the study until the mobile agent finished to migrate. In order to solve this problem, we divide multimedia data into fragments and prepare mobile agents which manages each fragments. Since each mobile agents become small, a learner can start the study soon without waiting for the migration of a mobile agent which has a huge size of multimedia data. We, however, have to search a mobile agent which manages their fragments. Therefore, a mobile agent which manages n-th fragments informs its location to the agent which manages (n+1)-th fragments, and vice versa. Since each agent knows the location of a mobile agent which manages next and previous fragment, and fragments are cached into learner's node, a learner can play multimedia data smoothly without finding the location of mobile agent which manages holding next fragment. Experiment results show the effectiveness of our method.

A mobile agent is an autonomous software module that can migrate between different computers. A mobile agent is designed and implemented like a human, and mobile agents work together by interactions among them like a human community. Thus, a mobile agent technology is helpful when we develop distributed systems with an easy-to-understand design and implementations for humans. Many researchers have proposed various applications through mobile agent technologies. However, mobile agent technologies are not used much as compared to other networking or programming technologies in the real world because migrations of mobile agents make it difficult to debug the system. This paper discusses problems of mobile agents for debugging, and proposes a remote debugger in order to solve these problems. Our proposed remote debugger supports functions of searching, single stepping execution, breaking, and viewing variables for a mobile agent who behaves with migrations on a running system.

This paper proposes an automatic vehicle locating (AVL) system which gains the geographic location data of a vehicle by passengers' sending the data via their smartphones in order to support public transportation users. While the existing AVL systems with specialized facilities costs a large amount of money and currently only large cities can introduce the system to their public transportation system, such user-generated approach will reduce the costs to collect the location data of a vehicle, and will enable small transportation system to adopt an AVL system. This paper developed a matching technique of location with a timetable, and demonstrated the system in the route buses in Tottori Prefecture with 70 smart phones, and evaluated its performance.

We previously proposed a meeting scheduling system based on mobile agent technology. The users of our system do not need to input all of their schedules unlike the existing groupwares. When a user intends to call a meeting, he only inputs information about the meeting. On behalf of the inviter, mobile agents move around each invited user's computer to ask whether he can join the meeting and negotiate with him if necessary. In this paper, we propose a NAT traversal method for a meeting scheduling system based on mobile agent technology. In our former system, the users inside a NAT network cannot use the system because mobile agent cannot migrate from outside the NAT network to inside NAT network. For this reason, our proposed system does not work with some network including a NAT router, such as a home network and an office network. To solve this problem, we implemented a new migration method for mobile agents.

In this paper, we present network address translator (NAT) traversal technique suitable for pure P2P applications, such as a distributed asynchronous Web-based training (WBT) system that we have previously proposed. P2P systems have the advantages of scalability and robustness, however, P2P systems face difficulties in constructing its network over the Internet, since end-to-end communication through NATs is sometimes restricted. In this study, as a solution of this issue, we ensure uniqueness of identification of nodes through the Internet and enable to establish a connection over the NAT by means of adding static routing entry to the NAT table with the help of universal plug and play (UPnP). Our NAT traversal technique is implemented in the proposed WBT system and is examined by experiments.

In this paper, we present mobile agent systems supporting persistency. In order to develop the mobile agent systems, a mobile agent framework has to have functions to support persistence of agents and persistence of an agent runtime environment. Our developed Java-based mobile agent framework named Maglog is taken up an example of a mobile agent framework. Maglog consists of three basic components, which are agents, agent servers and fields. Agent server and fields are corresponding to an agent runtime environment. The effectiveness of persistency is confirmed through descriptions of two mobile agent systems: a distributed e-learning system and a scheduling arrangement system.

Bus route system is the fundamental transportation device for aged people and students, and has an important role in every province. However, passengers decreases year by year, therefore the authors have developed the shortest path searching system called "BUS-NET" as a web application to sustain the public transport. Here, we take up generation of timetables. In general, timetable generation is considered as a very simple problem. However, there are some difficult problems. Firstly, bus stops of the same name are managed as one bus stop in the database of our system on grounds that error and cost decrease. Therefore, when timetables are generated, such bus stops must be distinguished. Secondly, there are so many bus stops in the timetables of bus route. However, all of them are not needed for an individual user in most cases. Thus, some bus stops may be excluded when a user intends to obtain a timetable of a bus route. In this paper, the authors report on the algorithms to solve those problems and the implementations, experiments of the functions.