Daisuke Arai

A user authentication based on human reflexes using blind spot and saccade response

The use of biometrics for user authentication has recently attracted attention. Biometrics makes it possible to authenticate a person accurately, but biometric information can easily be leaked and/or copied. It is therefore desirable to devise biometric authentication that does not require biometric information to be secret. This paper proposes a user authentication method that utilises human reflex responses. The expectation is that, even if a persons reflex characteristics are publicly known, it would be difficult for someone else to impersonate that person, since human beings are basically unable to control their own reflexes. In this paper, we demonstrate user authentication using blind spot position and saccade responses as the prototype of a system of authentication based on reflex responses, and also study the feasibility of such a system.

Service Environment for Smart Wireless Devices: An M2M Gateway Selection Scheme

Wireless standardization activities, such as the Bluetooth V4.x (Bluetooth smart), have led to an explosive growth in innovative short-range wireless devices capable of providing various services to assist everyday life. Allowing machine-to-machine (M2M) communication between these M2M devices also presents new commercial service opportunities. M2M service providers, including telecom carriers, may need to establish new networks to accommodate M2M interactions while identifying new business opportunities. Typically, M2M devices need another capable device, such as a smartphone, to act as a gateway to connect to the Internet to operate at their full potential. We propose a novel scheme in which any Internet-enabled device in the vicinity can act as an M2M gateway for M2M devices. This eliminates the need for the M2M device user to also own a smartphone or an Internet-enabled device to obtain services such as online-to-offline. This scheme faces new challenges, such as how to identify the best M2M gateway for a required service and the QoS. This paper proposes a method to select an M2M gateway from a large number of possibilities in order to increase service availability while also obtaining better signal strength for higher QoS. The methods presented are evaluated in terms of their performance, including energy consumption, and a service deployment guideline is derived using real-world data collected at an exhibition, which gave encouraging results.

M2M gateway selection scheme for smart wireless devices

Wireless standardization activities such as the Bluetooth V4.x (Bluetooth smart) have led to an explosive growth in innovative short-range wireless devices capable of providing various services to assist everyday life. Allowing Machine-to-Machine (M2M) communication between these M2M devices also enable new commercial service opportunities. The M2M service providers including telecom carriers may need to establish new networks to accommodate M2M interactions while identifying new business opportunities. Typically M2M devices need another capable device such as a smartphone to act as a gateway to connect to the Internet to operate at its full potential. We propose a scheme where any Internet enabled device in the surrounding environment can act as an M2M gateway for M2M devices. This eliminates the need for the M2M device user to also own an Internet enabled device such as a smartphone to obtain services such as Online-to-Offline. This scheme faces new challenges such as, how to identify a best M2M gateway for a required service and QoS etc. This paper proposes a method to select an M2M gateway among many others, in order to increase service availability while obtaining better signal strength for higher QoS. The performance of the proposed method is evaluated using real world data collected at an exhibition with encouraging results.

Short paper: experimental study of long-term operation of BLE tags for realizing indoor location based service

Recently, indoor location-based services (LBS) have become attractive with the rapid development of indoor location estimation and with the spread of smartphones. An estimation technique using Bluetooth low energy (BLE) tags becomes most attractive and promising among other techniques for estimating location because a majority of smartphones supports BLE. The BLE tags transmit advertisement (ADV) packets periodically, which include a unique ID related to each tag. When a smartphone receives the ADV packet, the smartphone sends a query to the server to retrieve the locationbased information for the ID. To provide such a service continuously, operators have to monitor whether BLE tags work well and repair them in the event of a problem. These operations are difficult because they require physical labor that cannot be automated for a large number of BLE tags spread over a wide area. Therefore, we evaluated the operations of BLE tags quantitatively by the metric of operational costs. We established an environment composed of 54 BLE tags in a four-story building and measured the lifetime of the battery, the service availability ratio, and the operational costs. Through the experiments, we obtained various results regarding the battery management of the BLE tags and the effectiveness of redundant deployment.

UE-based network access timing control scheme for avoiding signaling spikes

The rapid growth of smartphone based User Equipment (UE) and the wide development of LTE networks have changed the traffic pattern of mobile applications. Today, many applications running on smartphones periodically generate background traffic by sending and receiving messages or retrieving updates (i.e. without user interaction). Since each activity requires the UE to first attach and then to detach from the LTE network, a large volume of signaling traffic is generated. When the timing of these activities synchronizes among many UEs, a storm of signaling traffic is caused. We refer to this phenomenon as a signaling spike. Such a spike is a serious risk for telecom operators as it may overload its gateways and eventually disrupt services. As a way to suppress these spikes, we propose a new approach called a user equipment based network access control scheme (UENAC). The aim of UENAC is to autonomously decentralize the access timing among of UEs by adding a variable delay before sending packets upon attachment to the LTE network. We implement the UENAC entity on a commercial Android smartphone and show that the scheme has no negative effect on existing applications with delays up to 8 seconds. Furthermore, we simulate the signaling traffic based on commercial traffic pattern and show that the proposed scheme is able to effectively decentralize signaling spikes.

Study on relationship between user awareness and QoE in communication delay on smartphones

Network delay may occur on the LTE (Long Term Evolution) network due to signaling spikes. This decreases quality of experience (QoE). In this paper, we propose a method for mitigating the effect of such a delay. Our method acts as “a concierge” to divert the attention of users away from delay. The biggest advantage of our method is that it can be implemented at lower cost (e.g. capital investment) compared to other methods. To evaluate our method, we conducted a comprehensive user experiment using crowdsourcing. The results show the effectiveness of our method.