雙層式長鏈狀車載網路之高效率資料收集與散佈機制

Abstract

本篇論文探討在雙層式網路架構下所形成之長鏈狀車載網路的資料收集與散佈效率議題，典型的應用為具有共同騎乘路線的自行車車隊網路，其上層網路介面為3G/3.5G ，下層為IEEE 802.11 Ad-hoc網路。在此自行車車隊網路架構中，我們著手研究通訊效能最佳化的問題，若所有使用者都透過3G/3.5G網路更新與下載資訊時，將對3G/3.5G網路造成頻寬嚴重不足與傳輸延遲過長的問題，因此，我們設計出自行車車隊成員分群與群組維護的機制，群組成員使用Ad-hoc通訊來交換彼此的資料，並由一個最佳的成員負責回報群組資訊以便Server分享給其他群組，如此一來，藉由在上下層網路通訊之間的最佳化，可針對3G/3.5G網路在頻寬方面作更有效的利用及減少不必要的花費，模擬實驗與數學分析結果顯示我們所設計的高效率資料收集與散佈機制可大幅度地降低3G/3.5G傳輸需求與建立連線數。除此之外，我們設計並實作出一套車隊通訊與管理系統，利用Android智慧型手機作為使用者的手持裝置，可以有效率地達成車隊成員更新位置與狀態資訊的目的。

The bicycling recently has gained a lot of popularity in many applications, such as transporta-tion, recreation, and exercise. In a cycling fleet, bikers may ride in different speeds due to their body strengthes and bike performance. Fleet members using smart phones form two-tier vehicular long-thin networks (VLTNs) along the common cycling path, which consist of a 3G/3.5G high-tier interface and an IEEE 802.11-based low-tier interface. This paper investi-gates the optimization of data collection and distribution in two-tier VLTNs. As all bikers up-load their data and download the fleet information via 3G/3.5G communications, it causes serious problems of insufficient bandwidth and uncertain delay. To reduce the bandwidth usage and transmission delay of 3G/3.5G communications, we propose a framework consisted of a dynamic grouping mechanism and a group maintenance scheme. For data collection, bikers belonging to the same group locally exchange data with each other via ad-hoc communications, and only the gateway needs to report the group data and obtain the fleet information to/from the backend server via 3G/3.5G communications. For data distribution, the gateway uses multi-hop forwarding to broadcast the fleet information to all group members via ad-hoc communications. Through the cooperation between 3G/3.5G and ad-hoc communications, the high-tier bandwidth usage and transmission delay can be minimized. Analytical and simulation results show our framework outperforms existing works, which can significantly reduce the amount of 3G/3.5G data and the number of 3G/3.5G connections. In addition, we implement a Google Android and Maps-based prototype on smart phones for cycling fleet communications.