Yuichi Miyaji
Toyohashi University of Technology
Network
Latest external collaboration on country level. Dive into details by clicking on the dots.
Publication
Featured researches published by Yuichi Miyaji.
international conference on ubiquitous and future networks | 2010
Yuichi Miyaji; Masahiro Kawai; Hideyuki Uehara; Takashi Ohira
Recently, many studies in wireless ad hoc networks assume the use of smart antennas which can electronically control their beam pattern. Communications using smart antennas can be expected to improve spatial utilization efficiency and to extend communication distance. However, there are some unique problems caused by the use of the smart antennas. In this paper, we focus on the directional hidden terminal and the deafness problems. We propose a directional monitoring scheme and a switching backoff scheme for overcoming these problems. The directional monitoring scheme can solve the directional hidden terminal problem by monitoring the direction toward the destination after sending a data packet. The switching backoff scheme can reduce deafness loss time by keeping backoff-interval as long as the terminal is deaf. Comparing the proposed scheme with conventional schemes by computer simulations, the results show that the proposed scheme can mitigate the effect of these problems and it achieves higher throughput.
vehicular networking conference | 2014
Hiroki Matoba; Yuichi Miyaji; Hideyuki Uehara
Routing protocols using directional antennas can utilize knowledge of a direction. For example, these protocols are possible to reduce redundant traffic by controlling a propagation direction of a packet based on the positional relation among nodes. To implement direction-oriented communication protocols, we must make a mechanism to detect the direction from which a packet comes. We construct a communication system equipped with a DOA (Direction Of Arrival) detection mechanism by using GNU Radio and USRP, which decides whether a signal is received from the front or rear. In our system, a node measures RSSIs (Received Signal Strength Indicator) at two directional antennas and detects a DOA by comparing each RSSI. In order to use two directional antennas, one node consists of two USRPs, and each USRP has a directional antenna. To show a usefulness of our system, we implement DORP (Direction Oriented Routing Protocol) which is a routing protocol using a direction of a neighbor and destination node. We show an experiment of a multi-hop wireless communication by a video. In our video, each node can flexibly select a relay node without additional control packets like a route recovery packet by utilizing the direction of neighborhood and destination nodes.
wireless communications and networking conference | 2017
Kazuki Komatsu; Yuichi Miyaji; Hideyuki Uehara
In-band full-duplex communications have been spotlighted because they can double the spectral efficiency of the current wireless communication systems. However, it is necessary to mitigate the self-interference (SI). Currently, several time-domain and frequency-domain SI cancellers have been proposed. Time-domain SI cancellers are based on the parallel Hammerstein (PH) model, and they have good flexibility with high computational cost. In contrast, frequency-domain SI cancellers can achieve high cancellation performance with low computational cost but they have less flexibility than time-domain PH based SI cancellers. In this paper, we propose a frequency-domain SI canceller based on the PH model. The proposed scheme estimates the frequency response of the SI channel and regenerates SI signals by the overlap- save method. Therefore, the computational complexity of the proposed scheme is less than time-domain PH based SI canceller. The performance of the proposed scheme is assessed by equivalent baseband signal simulations of a full-duplex transceiver. As a result, the proposed scheme achieves high SI cancellation as the time-domain PH based SI canceller with low computational cost. In addition, the convergence performance of the proposed scheme is faster than the time-domain scheme.
international symposium on wearable computers | 2017
Kikuya Miyamura; Yuichi Miyaji; Ren Ohmura
In cases where a user wears multiple devices, power supply and management for these devices can be a significant problem. In this research, we propose utilizing a wireless power transfer technique between items of clothing, with the aim of distributing power from a single (or a few) source(s) to multiple devices. Considering the power transfer between a pair of trousers and a shirt as a first step, we investigated three models of resonators attached to fabric on the surface of the clothing. Power transmission efficiencies were measured as a function of the misalignment of the resonator coils. The results showed that the helical model achieved stable wireless power transmission, while the single and array models could achieve higher efficiencies if the transmitter and receiver coils were kept in opportune alignment.
2017 International Conference on Computing, Networking and Communications (ICNC) | 2017
Yuichi Miyaji; Hideyuki Uehara
In wireless ad hoc networks, neighbor discovery is essential for medium access control and routing. In this paper, we propose a new neighbor discovery algorithm utilizing in-band full duplex. This algorithm employs a reservation scheme with full duplex into an existing algorithm based on half-duplex. We analyze neighbor discovery time of both algorithms with absorbing Markov chain models and simulations. These results show our proposed algorithm shortens neighbor discovery. Furthermore, we discuss an affinity between the proposed algorithm and an existing one with full duplex.
international symposium on computing and networking | 2016
Sho Sasaki; Yuichi Miyaji; Hideyuki Uehara
A large number of battery-driven sensor nodes are deployed to operate a wireless sensor network, and thus power saving of nodes is a problem of vital significance. Many routing methods have been proposed to reduce energy consumption. In light of this fact, if it can be known how much energy consumption will be required for sensor nodes to operate a network before deploying the nodes, it will help the insight into the energy-efficient network topology. In this paper, utilizing the stochastic-geometric analysis on hop-count distributions for random planar networks, the formulation of energy consumption in wireless sensor networks which employ the progress-based nearest forwarding is offered. The formulation enables us to show how much energy is required for all nodes in the network to forward sensed data in a pre-deployment phase. The evaluation results demonstrate that the derived analytical value can approximate the simulated cost with a deviation of approximately 10%. The discussion on the deviation is then made with geographical metrics.
international electronics symposium | 2016
M. Zen Samsono Hadi; Yuichi Miyaji; Hideyuki Uehara
Nowadays, mobile wireless sensor networks have many applications especially in a group movement such as evacuation system, search and rescue activity. To support the group movement and to achieve an energy-efficient protocol, we propose a Group mobility based Clustering (GC) scheme. This protocol reduces the control overhead in the setup phase of a clustering system by introducing a concept of group leader and group member. In this scheme, the communication with cluster-head is only done by the group leader to save the energy consumption. Based on the simulation results, GC Single increases the lifetime of the networks and the number of packets received at a base station, compared with LEACH, MN-LEACH.
IEICE Electronics Express | 2014
Akiro Shimada; Shinsuke Ohtake; Yuichi Miyaji; Hideyuki Uehara; Takashi Ohira
We derive theoretically the power division ratio (PDR) on multi-hop wireless power transfer (WPT) to two receivers coupling with two couplers. The derived expression suggests that the PDR can be controlled by adjusting the load resistance. However, the control range of PDR by adjusting the load resistance of one receiver is limited when a hop count to that receiver is odd number. Therefore we show that the PDR can be controlled by another two methods; control of load resistances of each receiver and control of the load impedance of one receiver. These two methods can control the PDR within coaxial misalignment around the radius of the coupler.
Journal of Signal Processing | 2014
Shinsuke Ohtake; Akiro Shimada; Yuichi Miyaji; Hideyuki Uehara; Takashi Ohira
IEICE Communications Express | 2017
Sho Sasaki; Yuichi Miyaji; Hideyuki Uehara