An Efficient Key Management Scheme for Data-Centric Storage Wireless Sensor Networks

• Main Author: 楊舜博
• Other Authors: 黃志銘
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/82732033584316845677

碩士 === 逢甲大學 === 資訊工程學系 === 101 === In Data-Centric Storage wireless sensor Networks (DCSNs), the sensing data are usually sent to store on a specific node’s location in a network, based on their attributes, such as event type, happen place, and time of occurrence, and so on. Afterwards, the users, if necessary, can send requests to a specific node to retrieve the interesting data. The DCSN has been proven to be a promising data storage architecture. However, most of previous DCSN-related researches always focused on the discussions of data retrieval and maintenance, instead ignored the compromised attacks of storing data in the network. For instance, if attackers located and compromised a node, the data stored in its memory would be wholly explored to them. Even the compromised node has no data they are interested, they can also analyze the mapping rule to trace where the coming data will be stored on, and then obtain the valuable information. To the best of author’s knowledge, pDCS is the first study on providing security mechanism for DCS networks. In pDCS, the sensing area is first divided into a number of grid cells, the nodes belonging to a certain cell are geographically formed to a group. For different security requirements, a variety of cryptography keys are allocated to each group. In such situation, even an attacker captured a node and cracked its security information, the network administrator can still quickly initiate a rekeying task to avoid a deeper damage. However, the rekeying process in pDCS causes a very large amount of update messages needed, this would lead to a severe depletion of network energy, and thus shorten the network lifetime. In this thesis, based on the Exclusion Basis Systems (EBS), we propose an efficient distributed key management scheme, termed as ERP-DCS, to improve the deficiencies identified in the pDCS scheme. ERP-DCS attempts to distribute the key management tasks, including key distribution, rekeying, and key revocation, to each cluster (i.e. grid cell) to reduce the number of rekeying messages. The main objectives in this thesis are targeted at the analyses of key storage and rekeying messages needed in key management task. We conduct analyses and simulations, in C programs, to investigate the effects of many factors, such as transmission range and grid size, on the performance of our proposed scheme. The results show that, comparing to the pDCS scheme, the ERP-DCS is superior, in terms of update messages needed in the rekeying process, while at a little cost in key storage.