User Experience Enhancement on Smartphones using Wireless Communication Technologies

Abstract

Recently, various sensors as well as wireless communication technologies such as Wi-Fi and Bluetooth Low Energy (BLE) have been equipped with smartphones. In addition, in many cases, users use a smartphone while on the move, so if a wireless communication technologies and various sensors are used for a mobile user, a better user experience can be provided. For example, when a user moves while using Wi-Fi, the user experience can be improved by providing a seamless Wi-Fi service. In addition, it is possible to provide a special service such as indoor positioning or navigation by estimating the users mobility in an indoor environment, and additional services such as location-based advertising and payment systems can also be provided. Therefore, improving the user experience by using wireless communication technology and smartphones sensors is considered to be an important research field in the future. In this dissertation, we propose three systems that can improve the user experience or convenience by usingWi-Fi, BLE, and smartphones sensors: (i) BLEND: BLE beacon-aided fast Wi-Fi handoff for smartphones, (ii) PYLON: Smartphone based Indoor Path Estimation and Localization without Human Intervention, (iii) FINISH: Fully-automated Indoor Navigation using Smartphones with Zero Human Assistance. First, we propose fast handoff scheme called BLEND exploiting BLE as secondary radio. We conduct detailed analysis of the sticky client problem on commercial smartphones with experiment and close examination of Android source code. We propose BLEND, which exploits BLE modules to provide smartphones with prior knowledge of the presence and information of APs operating at 2.4 and 5 GHz Wi-Fi channels. BLEND operating with only application requires no hardware and Android source code modification of smartphones.We prototype BLEND with commercial smartphones and evaluate the performance in real environment. Our measurement results demonstrate that BLEND significantly improves throughput and video bitrate by up to 61% and 111%, compared to a commercial Android application, respectively, with negligible energy overhead. Second, we design a path estimation and localization system, termed PYLON, which is plug-and-play on Android smartphones. PYLON includes a novel landmark correction scheme that leverages real doors of indoor environments consisting of floor plan mapping, door passing time detection and correction. It operates without any user intervention. PYLON relaxes some requirements for localization systems. It does not require any modifications to hardware or software of smartphones, and the initial location of WiFi APs, BLE beacons, and users. We implement PYLON on five Android smartphones and evaluate it on two office buildings with the help of three participants to prove applicability and scalability. PYLON achieves very high floor plan mapping accuracy with a low localization error. Finally, We design a fully-automated navigation system, termed FINISH, which addresses the problems of existing previous indoor navigation systems. FINISH generates the radio map of an indoor building based on the localization system to determine the initial location of the user. FINISH relaxes some requirements for current indoor navigation systems. It does not require any human assistance to provide navigation instructions. In addition, it is plug-and-play on Android smartphones. We implement FINISH on five Android smartphones and evaluate it on five floors of an office building with the help of multiple users to prove applicability and scalability. FINISH determines the location of the user with extremely high accuracy with in one step. In summary, we propose systems that enhance the users convenience and experience by utilizing wireless infrastructures such as Wi-Fi and BLE and various smartphones sensors such as accelerometer, gyroscope, and barometer equipped in smartphones. Systems are implemented on commercial smartphones to verify the performance through experiments. As a result, systems show the excellent performance that can enhance the users experience.