無線網路環境中，forwarding sets的概念被許多廣播協定廣泛得應用來減緩廣播風暴問題(broadcast storm problem)。在這些協定中，當一個節點收到廣播訊息後，會從自己的鄰居當中選擇部分鄰居(即forwarding sets)來協助重傳訊息。在本篇論文，我們發現在異質網路的環境中，存在一個隱藏接收點的問題(the hidden receiver problem)，因為此問題，透過局部拓樸資訊計算出每一個節點的forwarding set之任何廣播協定無法保證100%滲透率(penetration rate)。基於此考量，我們應用skyline sets來設計一個廣播演算法，並且對我們提出的廣播演算法進行優化，讓每一個收到廣播訊息的節點能自行決定是否需要進行重傳。模擬結果顯示，我們提出的廣播演算法不但可以保證100%滲透率，同時能有效降低重傳次數與延遲。除此之外，我們的演算法在計算過程僅需要1-HOP鄰居資訊，這將會有效減少計算forwarding set所造成延遲，同時也適用於車載網路這類型網路拓樸經常改變的環境中。 The concept of forwarding sets is widely adopted in many broadcast protocols for wireless networks to alleviate the broadcast storm problem. In these protocols, after receiving a broadcast message, each node that is requested to relay the message instructs a subset of its neighbors, a.k.a. the forwarding set, to further relay it. In this thesis, we first demonstrate that, in heterogeneous wireless networks where nodes have different transmission radius, of the broadcast protocols that select forwarding set based on local topology can guarantee the 100% penetration rate even under the ideal MAC assumption. The reason is because of the hidden receiver problem. To address this problem, we propose an enhanced version of skyline-based broadcast protocol for broadcast in heterogeneous wireless networks. Our protocol further optimizes the skyline-based broadcast protocol by having each receiving node computes whether or not it should relay the broadcast message distributively. The simulation results show that our proposed broadcast protocol is able to reach 100% penetration rate with relatively low number of retransmissions and low delay. This shows that our proposed broadcast protocol is more suitable for environments with a frequently changed network topology, such as vehicular ad hoc networks.
