Shigeki Shiokawa

Call admission scheme based on estimation of call dropping probability in wireless networks

An effective call admission control scheme is proposed in order to solve the problem due to handoff. The proposed call admission control is based on estimation of call dropping probability. When a new call is generated, the proposed scheme estimates the call dropping probability of the cell where the new call exists and accepts the call only when the estimated probability does not exceed a certain value of threshold set in advance. By computer simulation, we evaluate the call blocking probability and the call dropping probability of the proposed scheme and show the effectiveness of the proposed scheme.

Multi-path routing protocol considering inter-path interference in MANET environment

In an ad hoc network, an original routing protocol, which is different from those in wired networks, is used in general. Recently, a multi-path routing has been focused on, and many protocols have been proposed for the purpose of improvement in a packet delivery ratio or load balancing. As one of them, there is a multi-path routing protocol considering interference between constructed paths. This protocol can achieve higher throughput, however, it works under the assumption that every node grasps the position information of all the nodes in advance. Therefore, it cannot be applied to some dynamic network environments such as MANET. In this research, we propose a new multi-path routing protocol considering interference between constructed paths without assumption of grasping the position information in advance. Also we make efforts to improve the throughput by restricting the amount of control packets.

Location based clustering scheme considering node mobility in wireless sensor networks

In wireless sensor networks, one of key techniques to save power consumption is a clustering. There are several location based clustering schemes. In these methods, a network area is divided into cells, and one node is selected as a cluster head in each cell. Making all nodes except cluster heads sleep, these methods can save power consumption. However, they can be applied to the static network topology. Therefore, we propose a new location based clustering scheme. Our proposed scheme selects a cluster head on the basis of not only residual power but also estimated staying time in a cell. By computer simulations, we show the effectiveness of the proposed scheme with respect to the packet delivery ratio.

Multi-path routing considering inter-path interference using a mobile agent in the ad-hoc networks

In ad-hoc networks where different routing protocols from the wired network are used, a multi-path routing has been paid attention for the purpose of load balancing. The multi-path routing considering inter-path interference proposed in the past can reduce the amount of control packets. However, this routing protocol has a problem that an obtained path is not always suitable one. In this paper, we propose an effective multi-path routing considering inter-path interference using a mobile agent.

Dynamic Sleep Control Using Location Based Clustering Scheme in Mobile Sensor Networks

In wireless sensor networks, saving energy consumption is an important issue and one of key techniques to save energy is a clustering. There are several location based clustering schemes. In these schemes, a network area is divided into plural subareas, and one node is selected as a cluster head in each subarea. Making all nodes except cluster heads sleep, these schemes can save energy consumption of sensor nodes. However, they don’t consider the mobility of nodes and also set a constant sleep interval. Therefore, in this paper, we propose a new sleep control using location based clustering scheme in mobile sensor networks. Our proposed scheme selects a cluster head on the basis of not only residual energy but also estimated staying time in a subarea. Moreover, the sleep interval is determined dynamically according to the above staying time. By computer simulations, we show the effectiveness of the proposed scheme.

Performance analysis for use of mobile agent in wireless multihop networks

In wireless multihop networks, each of terminals exchanges control message in a distributed manner for network processes such as routing. However, such a distributed control is not efficient under the situation where the terminals move freely and network topology often changes dynamically. In this situation it degrades network performance. Therefore, the authors group has proposed an efficient use of mobile agent with the aim of performance improvement in the wireless multihop networks. In this paper, the performance about efficiency of use of the mobile agent is theoretically analyzed.

Sleep control considering relay node information in mobile Ad-hoc networks

In mobile ad-hoc networks (MANET), the suppression of energy consumption is very important, because the networks may be used in situations with disaster or lacking of infrastructures. In our study, we propose a sleep control method exploiting a routing protocol using mobile agent, where nodes not included in a derived route go to sleep. In the proposed method, a node sends a sleep control packet before a packet transmission. The node which received the packet decides to sleep or not autonomously. From performance evaluation by computer simulation, we show that the proposed method can decrease the energy consumption while maintaining the packet delivery rate.

Retransmission control method using location information in MANET

In mobile ad hoc networks, each node can communicate with disconnected nodes by multi-hop transmission. In wireless communication, flame transmission failures occasionally occur due to obstacle, radio wave interference, moving of node, and so on. In IEEE802.11 MAC, a retransmission process is executed when such a failure occurs. However, there is a possibility that retransmissions continuously fail because link status is not considered. In this paper, we propose a retransmission control method in order to solve the above problem. In this method when a packet transmission fails, an overhearing node retransmits the packet to more suitable node than the original destination. From computer simulation, we show effectiveness of the proposed method in terms of packet loss rate.

Cell based clustering scheme considering dynamic network topology in wireless sensor networks

A routing protocol is one of key technical issues for wireless sensor networks (WSN), and various protocols have been proposed. Some protocols use cell based clustering; however, few of them can be applied to the dynamic network topology. In this paper, we refine HGAF (Hierarchical Geographical Adaptive Fidelity), which has been proposed for the static network topology, and apply it to the dynamic one. When HGAF is applied to the sensor network which consists of mobile nodes, there is the problem that some cluster heads go out of their staying cell and no cluster head exists in a cell. Our proposed scheme selects the node with the longest staying period in a cell as the cluster head in the cell. Furthermore, this scheme dynamically changes the cell in which the cluster head node is selected. By computer simulations, we show the effectiveness of the proposed scheme with respect to the packet delivery ratio.

A migration algorithm for moblile agent based on predicted lifetime in MANET

In recent years, applications using mobile agents (MAs) in mobile ad-hoc networks (MANETs) are attracted. An information dissemination service is one of these applications. In this service, users can get local area information from an MA, which migrates from node to node within the local area. The system need not provide any server. Instead of servers, MAs collect local information and disseminate it. One of very important problems in such a service is to make MAs survive as long as possible in the service area. Several methods to manage mobile agents have been proposed. However, the migration algorithm for MA used in these studies is a simple one. In this paper, a new migration algorithm based on the predicted lifetime of MA is proposed. By computer simulation, the performance of the proposed scheme is evaluated.