在無線感測網路中以跳頻為主的通訊方式被廣泛的用來解決頻譜短缺的問題。然而，因為次要用戶在無線感測網路中沒有自己的專屬頻譜，所以很容易受到惡意的攻擊。大部份可以抵抗攻擊的分法都需要傳送端和接收端預先分享的資訊(如自己跳頻序列)，並利用此資訊建立一條互相連通的通訊，藉以閃躲這些惡意的攻擊。但是因為無線感測網路中的節點的鄰居資訊會動態的改變，所以這種預先分享的資訊的防禦方法在無線感測網路中是不實際的。因此，發展可以抵抗攻擊而且不用預先分享資訊的跳頻演算法變得越來越重要雖然有一些不用預先分享的資訊的防禦方法被提了出來，但是這些方法要不無法保證兩個節點在有限的時間相遇，要不然就是需要先預設節點所扮演的角色(傳送端或接收端)。但是預設角色這種做法在無線感測網路中也是不太可行的，因為一個節點時常會同時進行資料的傳送和接收。在這篇論文中，我們提出了一個可以抵抗攻擊而且不用預先分享資訊的跳頻演算法，SP-CH，同時我們的演算法也能在有限的時間內相遇，並且不需要先預設節點扮演的角色。;Cognitive radio network (CRN) is a promising technology to solve unlicensed spectrum shortage problem and enhance the utilization of licensed spectrum. Channel hopping scheme are widely used to develop communication links between users (SU) in CRNs. However, existing channel hopping for CRNs, which have fixed hopping patterns, are extremely vulnerable to malicious attacks in CRNs. Most of existing anti-jamming CH protocols need pre-shared informations (e.g. hopping sequences) to the communication link between the communication pairs. Due to the nature of opportunistic spectrum access, pre sharing informations is impractical in CRNs. Hence, anti-jamming channel hopping approaches without pre-shared secrets have gained more and more research interests. For anti-jamming CH protocols without pre-shared informations, they either have unbounded time to rendezvous or require role pre-assignment (i.e. every SU is pre-assigned as either a sender or a receiver). Role pre-assignment is inapplicable to scenarios that a SU may be a sender and receiver simultaneously. In this paper, we propose an anti-jamming channel hopping algorithm, SP CH, with bounded rendezvous time, but without pre-shared secrets.
