無線感測網路中基於儲存限制模型下的金鑰分配系統

Abstract

這篇論文提出了在無線感測網路中，基於儲存限制模型下的金鑰分配系統。主要分為下列兩種架構：有Base Station跟無Base Station。在有Base Station的架構中，每一個Base Station所覆蓋範圍內擁有多個Sensor Node。由Base Station廣播一連串的隨機字串，而每個Sensor Node則隨機收取某幾個字元。最後再跟鄰居節點比對共有的字元，進而得到共有的金鑰。在無Base Station的架構中，每個Sensor Node產生一連串的Pseudorandom的字串，並且廣播給鄰居節點，而鄰居節點則隨機收取某幾個字元。最後再將所收取的字元位置回傳給原先的節點，原先的節點可以重複產生Pseudorandom的字串來拿取所得到的共同字元，進而得到共有的金鑰。我們的方法最特別之處在於不需預先讀取資料，較為符合sensor network的系統。此外，我們也透過數學及程式模擬來分析相關結果，以及討論驗證及部署的方法。跟之前的方法比較，在連結率及安全性上我們的方法有較好的結果。

In this paper, we propose a new key distribution system for wireless sensor networks in the bounded storage model. It contains two frameworks: with base station and without base station. In “with base station,” one base station covers several sensor nodes. The base station broadcasts random bits, and then each node randomly receives some bits and stores them. Finally, each node communicates with their neighbors and finds the common bits which are stored by the node. In “without base station,” each node generates pseudorandom bits and broadcasts them. Their neighbor randomly receives some bits and stores them. Finally, each node communicates with their original neighbor, and then their original neighbor can re-generate the pseudorandom bits and find the common bits which are stored by their neighbor. The common bits are the shared key. In our scheme, nodes don’t preload secrets, and it is easy to fit the sensor networks. Moreover, we analysis and simulate this scheme, and discuss the authentication and deployment. In the connectivity and security, the result of our scheme is better than preceding scheme.