於叢集架構之無線感測器網路分析並選擇最佳的叢集代表節點之個數

Abstract

機器對機器間通訊是目前第五代通訊技術的主要發展方針之一，而這裡我們討論的無線感測器網路是機器對機器間通訊的其中一種類型，該種網路的特性為感測器裝置皆為低移動率，我們利用此特性，可以將散佈在系統之中的所有裝置節點分成多個叢集，並在叢集中挑出叢集代表節點負責整併所有叢集內成員節點的資料及訊息後，再轉送至匯聚點(如基地台)，在此，我們分析並討論整個系統所有節點中叢集代表節點所佔的數量，如此一來，我們可以按照分析之結果，在不同的網路狀況下挑選出更節能的傳輸方式。

Machine to machine(M2M) communication is a key technology for 5G radio access. Wireless Sensor Networks we focus on here is a special use case in the M2M communication. Its characteristic is that the sensing devices or sensor nodes in the system are low mobility. With this characteristic, we can divide the field spread with sensor nodes into clusters and select a node in the cluster to be cluster head. The duty of cluster head is to aggregate the data or message from Non-head nodes(member nodes) and then transmit the data to the sink node (maybe an eNB). In this thesis, we analyze the cluster head selection number among all Sensor nodes, and then we can obtain an optimal number of cluster head in different WSN scenarios.