Smart Reliable Flooding for Internet of Things (IoT)

碩士 === 國立臺灣大學 === 電信工程學研究所 === 107 === Internet of things (IoT) is the network of physical objects which are embedded with electronic hardware such as sensors, software and network connectivity. IoT allows objects to sense and to be controlled remotely across networks. As result, they can interact w...

Full description

Bibliographic Details
Main Authors: Nai-Syuan Li, 黎乃瑄
Other Authors: Chun-Ting Chou
Format: Others
Language:zh-TW
Published: 2019
Online Access:http://ndltd.ncl.edu.tw/handle/3pau7m
Description
Summary:碩士 === 國立臺灣大學 === 電信工程學研究所 === 107 === Internet of things (IoT) is the network of physical objects which are embedded with electronic hardware such as sensors, software and network connectivity. IoT allows objects to sense and to be controlled remotely across networks. As result, they can interact with each other and cooperate with their neighbors to reach common goals. Among variant objectives of the IoT applications, high reliability is required in many applications such as IoT-based healthcare system, smart factories, etc. In this thesis, our objective is to focus on reliable multi-hop communications in IoT-based applications. In a typical IoT network, different applications generate a wide range of traffic patterns. The traffic patterns of IoT can be either single hop or multi-hop. The multi-hop traffic can be further divided, depending on the number of source and destination nodes into three types of patterns including: one-to-one, one-to-many, and many-to-one. Although many existing routing schemes have been tailored for different traffic patterns, given the limitation of computing and storage, we cannot implement multiple schemes in a single device. Therefore, we adopt a simple flooding with implicit acknowledgment mechanism as baseline protocol. In this baseline protocol, reliability can be guaranteed but a redundant transmission becomes a real challenge. Our goal is to support all three patterns and to minimize redundant transmissions while guaranteeing reliability. We proposed three solutions to deal with redundant problems in the baseline. First, we introduce Neighbor Information Bitmap (NIB) for devices to share the information regarding which neighbors have relayed a given message, which resolves the redundancy caused by simple flooding implicit acknowledgement. Second, we propose Directional Flooding (DF) for devices to know the direction of the message by the hop count between the source and the destination nodes, which resolves the redundancy caused by Omni-directional flooding. Last, we propose Cross-layer Consideration (CLC) for devices to know the order of the received messages, which resolves the redundancy caused by higher-layer retransmission. In order to evaluate the performance of our solutions in real environments, a testbed composed of a gateway and 39 end devices is developed in BL building, National Taiwan University. According to our experiment results in high and low channel loads cases, NIB reduces 49% and 58% of redundant transmissions respectively on average, NIB with DF reduces 86% and 81% of redundant transmissions respectively on average, NIB along with DF and CLC reduces 98% and 85% redundant transmissions respectively on average. More importantly, in all of the cases, at least 97% reliability is guaranteed.