在車載隨意網路中，由於道路環境的多樣性，使得如何解決資料封包繞徑問題成為相關領域中的一項挑戰。雖然各種新的方法相繼被提出，然而對於在不同環境中如何選擇最佳方法的相關問題，卻缺乏較廣泛的研究及實證。有鑑於此，有別於單一改良型的繞徑方法，我們提出一個整合式的架構來解決面對不同環境時的繞徑問題，以整合多重繞徑方法於單一系統中。我們並以跨層式方法設計繞徑模組的選用模式，再配合廣播封包整合的機制等三種策略方法，讓車輛在不同的道路環境下，能在多個繞徑方法提供的路徑當中，選擇最佳的路徑來傳遞封包，並藉由取得其他網路階層的相關資訊，作為最佳化選擇的依據。同時，封包若在傳遞過程中，原來的路徑發生中斷時，系統能改以其他繞徑方法繼續傳遞封包至目的節點，減少來源端封包重傳的需求。最後為了減輕因數個繞徑方法同時進行所可能產生的廣播風暴問題，我們以封包列車的方式來整合廣播封包。從模擬實驗結果證明，本系統能在不同的道路車輛環境及車輛分布密度，提升資料的傳輸率，克服因節點位置快速變動而導致傳輸率下降的問題。 / Design of data routing between vehicle and vehicle (V2V) is a critical issue for VANET due to dynamics of traffic road environment. Therefore, there are some proposed solutions to cope with such problem in the past years. It is difficult, however, to choose a best approach from those solutions for a car node that could communicate by wireless under all kinds of rigid environment. In this paper, instead of designing new routing method, we proposed an architecture to cope with the problem. We use three building blocks to construct our methodology: combination of multiple routing protocols, path selection by cross-layer design and integration of broadcast routing packets. A car node may find more paths provided by the multiple routing modules for delivering packets to the destination. By using cross-layer design we demonstrate how to determine one from the routing modules that can provide the best path for packet according to information collected from the other network layers. Besides, a recovery of forwarding can be performed by this architecture while a broken path occurs during the packet delivering. Finally, we proposed a packet train idea to solve the broadcast storm which could be incurred by the multiple routing protocols operating at the same time. The experiment results show that our proposed architecture can raise the data throughput under different mobility environments of VANET.

在vehicular network中應用Peer to peer (p2p)是目前尚在研究的一項領域。不同於一般wired network的環境，vehicular network環境中變動性相當高，且車輛移動速度快，車輛間無法長時間維持連線，所以若要在vehicular network上分享檔案資源，會遭遇到許多問題，像是如何在網路上有效率地搜尋所需要的檔案、相距甚遠的車輛如何分享彼此的資源、在有限的頻寬中如何發揮最大的分享效益，皆是十分重要的課題。本研究提出Urban Multi Layered Chord (UML-Chord)架構，在市區環境的vehicular network中，引入infrastructure-based的架構，以解決ad-hoc環境中只能靠車輛機會相遇時交換訊息，僅能做範圍內搜尋的缺點。另外我們使用multi-layered chord來管理網路上分享的檔案資源，此種結構化的p2p架構可以有比較好的搜尋效率。在檔案傳輸上，我們加入交通號誌的狀況後，讓車輛間連線預估時間(connection lifetime)更加精確，得以選出最適合的節點下載檔案，增加檔案的傳輸效率，提昇整體p2p應用的效能。 / Peer-to-peer (p2p) applications on vehicular network is a significant research topic today .Vehicular network is different from traditional wired network environment in some characteristics. Topology is very dynamic and vehicles driving in high speed, the connection between vehicles can’t maintain in a long time in vehicular network. If we want to share file resources on vehicular network, there are some problems to confronted, like the efficiency of query operations, sharing files with distance vehicles and fully utilizing the limited bandwidth. Our paper proposes an Urban Multi-layered Chord (UML-Chord) method in urban Vehicular network. Introducing infrastructure based architecture to deal with the problem that cars can exchange their files only when they meet each other in ad-hoc mode. In addition, we use multi layered chord to manage file resources which should be shared on Internet. This kind of structured p2p architecture has better efficiency in searching files. For file transfer part, we take stop light into account. We modify connection lifetime, let it be used in choosing appropriated download peers, enhancing the efficiency of p2p application.

車載網路受到各方廣泛討論，激發出許多新的議題，由於車載網路的通訊品質不穩定，速度快、節點多，封包傳送不易，因此許多人都採用分群式架構增進效能，以集中式管理群組，避免封包被重複傳送，降低封包碰撞的機會。然而，現有的分群機制只能用在即時方面的應用，在檔案傳輸方面效能不足。本篇論文擬改善C. Shea等人[1]所提出的分群機制File-based Affinity Propagation Cluster, FAPC，建立兼具動態性和檔案相關性的叢集架構，並且提出改善失去叢集管理員的重建機制，以提升分群的穩定性及吞吐量（throughput）。最後，我們以模擬證明所提出的方法優於C. Shea [1]的方法，以query hit ratio、retrieve file ratio、average number of clusters及average cluster head duration為效能指標，觀察在不同時間、車輛數目及車輛速度時效能表現。 / Vehicular Ad-hoc Network (VANET) has been widely discussed and many issues have been proposed. Due to VANET’s unstable quality, varying speed, lots of mobility nodes, it’s not easy to deliver packets. Thus many researchers suggested using cluster architecture to enhance performance. Because of the central management, we can avoid duplication of packets in the same cluster and decrease the probability of packet collision. However, we find most of the cluster architectures are suitable for real-time applications, but not for file transfer. In this research, we improve C. Shea’s [1] method by adding file-similarity by classifying into groups and reselecting cluster head, when the group of nodes have not cluster head. This cluster architecture can enhance stability and throughput. Finally, we use simulation to prove that our method outperforms Chen’s [1] cluster method in terms of query hit ratio, retrieve file ratio, average number of clusters and average cluster head duration.

車載網路在未來社會將會越來越普及，車載網路資料傳遞技術，不僅能連接上網提供多元化的網路應用，甚至提供行車安全性服務。然而車載網路的網路品質不是十分理想，乃因存在著許多不確定因素，因為網路品質低，所以需要更完善的機制來提升車載網路品質。如車輛間速度不一，車輛間的距離變化相當頻繁，若因此而需要重新建立連線路徑是相當浪費傳輸時間並影響到傳輸的頻寛。本篇論文提出Cookie-Cooperative Automatic Repeat request (CCARQ) 的機制是針對這類型的間歇性斷線做修補，其方法是透過MAC (Media Access Control) 層協助網路層 (Network Layer) 的轉送。因為車輛是行駛在道路上，所建立的車載網路拓樸是沿著道路佈建，所以當連線路徑中斷時，周圍鄰近的車輛將是最好的替代選擇。本模擬將比較CCARQ, AODV與Bypass-AODV，透過 packet delivery ratio、end to end delay、control overhead及goodput為效能指標，調整在不同的hop數、車輛數目及車輛速度來觀察效能表現。模擬結果顯示packet delivery ratio增加14.6%、end to end delay減少41.3%、control overhead減少38%。 / VANET (Vehicular Ad-hoc Network) is expected to become more popular in the near future. VANET transmission technology will offer not only bountiful network applications but also safety driving information. Network quality of VANET is usually bad due to speed variation, high mobility, driver behavior variation, traffic density, traffic signal, environment complexity and swarm effect, et al. In this research, we propose a Cookie-Cooperative Automatic Repeat reQues (CCARQ) to enhance network performance by repairing intermittent broken connection. As mentioned above, speed variation will dynamically change the distances of car-to-car frequently and result in broken connections. Each re-establishment of broken connections will waste transmission time and reduce transmission bandwidth. The proposed CCARQ mechanism will help network layer transmission through MAC layer. Since VANET’s topology deployed along with road, the surrounded cars will be the best candidate to be a substitute car when a broken connection occurs. The performance indices are set to varied hops, varied speed and number of cars. Simulations show that CCARQ outperforms AODV and Bypass-AODV in packet delivery ratio by 14.6% increase, end to end delay by 41.3% decrease and control overhead by 38% decrease.

在耐延遲網路(Delay Tolerant Network)中，因節點之間的高移動性、連接的不確定性環境嚴苛限制，採用Store-And-Forward 訊息傳輸的模式提供一個可接受的網路表現。常見的路由協定可分為機會路由、基於預測的路由以及調度路由，然而這些路由協定使用在市區環境中，有些許不足的地方，因此我們提出適用在市區封包轉發策略。 我們提出的以公車亭為基礎之耐延遲車載網路封包轉發策略，是在市區環境中建立一個以公車亭為基礎的資料傳送架構，包含汽車、公車、公車站、公車轉運站四種節點。我們建立節點與節點相遇時資料傳送規則，例如汽車與汽車相遇、汽車與公車相遇、公車與公車站相遇、公車與公車轉運站相遇、公車轉運站與公車相遇、公車站與汽車相遇、汽車與目的地相遇時各自有不同的資料傳送判斷與限制。 實驗結果也證明所提出的演算法，除了可以有效地減少延遲傳送時間並提高訊息成功傳送率，及在各節點有限的緩衝區大小下，我們的封包轉發策略有著最突出的效能。 / In Delay Tolerant Networks (DTNs), there does not exist an end-to-end path due to intermittent connectivity and high node mobility. Messages are stored for a period of time at network nodes and are conveyed hop-by-hop to the destination. The current DTN routing protocols can be summarized into three categories: opportunistic, prediction-based and scheduling protocols. However, these routing protocols have some deficiencies and are not specifically focused on the urban areas. Based on the characteristics of urban areas, we proposed a kiosk based packet forwarding strategy for vehicular delay tolerant networks. We established the rules of data transmission when one node contacts other nodes. More specifically, Car-to-Car, Car-to-Bus, Bus-to-Bus stop, Bus-to-Bus transfer station, Bus transfer station-to-Bus, Bus stop-to-Car ,Car-to-Destination contacts, have different judgments and restrictions for data forwarding. The simulation results demonstrate that we proposed packet forwarding strategy would reduce the delivery delay, and improve the successful delivery rate. Especially with limited buffer and little overhead, our proposed strategy has the most prominent performance.

許多國家正面臨因嚴重的交通壅塞所帶來龐大的經濟損失，同樣在台灣也面臨相同問題，根本解決之道就是找出壅塞路段，並即時地告知車輛駕駛以迴避此路段。因此如何有效率地偵測出壅塞路段，是當前重要的議題。以往大多透過長期所累積的統計數據，針對各路段收集行駛於該路段所有車輛的平均速度，來獲得該路段的交通狀況，但卻無法立即反應即時路況。 因此我們提出以耐延遲車載網路方式，收集路況資訊與彙整旅途時程，透過具有全球定位系統(GPS)以及無線網路的車輛來進行路況交通資訊的收整與交換。再透過本研究所提出路況交通覺知路由協定(RTARP)，以One-Hop Controlled Flooding的傳輸方式在節點與節點相遇時進行路況資訊的傳送，進而交換節點彼此間各自所存放各路段的路況資訊。 模擬實驗結果證明我們所提出的路由協定在路況交通資訊交換中，有效地減少網路傳輸負載、提高訊息送達成功率以及路況交通資訊(RTI)正確率。 / Many countries are concerning about the huge economic losses caused by the critical traffic congestion. We have the same problem in Taiwan. The solutions lie in finding the road section with traffic congestion first and informing drivers to avoid that section. But how to detect the road section with traffic congestion effectively is the most important issue of this research. Conventionally, the real-time road traffic is mostly predicted by long-term accumulated statistics calculated by the collection of the average speed of cars on the same road section. But this way can’t provide the real-time road traffic immediately. Based on vehicles with GPS and the capability of WiFi, we proposed gathering road traffic information and consolidating travel time estimation using vehicular delay tolerant networks. We also proposed the Road Traffic Awareness Routing Protocol (RTARP) to exchange the road traffic information preserved individually by One-Hop Controlled Flooding during nodes’ encounter. The results of simulations prove the routing protocol we proposed can effectively reduce the transmission overhead, improve the delivery ratio and the accuracy of the road traffic information.