基於Bloom Filter之雲端環境下使用者匿名機制查詢研究;Bloom Filter based Research on Anonymous Protection Mechanism in Cloud Environment

NCUIR > School of Management at National Central University > Graduate Institute of Information Management > Electronic Thesis & Dissertation > Item 987654321/74813

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/74813

Title:	基於Bloom Filter之雲端環境下使用者匿名機制查詢研究;Bloom Filter based Research on Anonymous Protection Mechanism in Cloud Environment
Authors:	張瑋倫;Chang, Wei-Lun
Contributors:	資訊管理學系
Keywords:	隱私保護;布隆過濾器;增量變數;布穀鳥過濾器;誤報;Privacy-Preserving;Bloom Filter;Variable Increment;Cuckoo Filter;False Positive
Date:	2017-08-08
Issue Date:	2017-10-27 14:40:37 (UTC+8)
Publisher:	國立中央大學
Abstract:	現今雲端服務的普及，使用者享受雲端服務所帶來便利下，並不曉得雲端服務商是否會要求過多個人及Android裝置上的辨識資訊（如IMEI碼），而後在未經使用者同意下利用其隱私資料進行使用者輪廓（User Profile）偏好分析或將資訊變賣給廣告商謀利，此行為大大侵犯到使用者的隱私與權利。本研究提供一機制以保護使用者隱私，使得使用者要求服務時並不需要傳送任何可辨識之個人資訊至雲端，即在完全匿名狀況下，雲端服務商依然能正確地提供服務給使用者。但在大量且不同的使用者同時提出請求狀況下，雲端服務商要如何能夠在使用者匿名狀況下快速辨別不同的使用者以提供適當服務?因此本研究首先使用一次性匿名化代號取代雲端上使用者傳輸之可識別資訊，接著使用布隆過濾器實作這項保護機制，讓使用者得知其服務是否得到滿足。我們創新性地使用雜湊表與布隆過濾器的配合，其優點在於時間複雜度為O(1)且布隆過濾器具有無法反查(Reverse)的特性，因此使用者在得到匿名化的保護下，不會消耗太多時間於此機制上，且雲端方面則是面對著陣列向量而無法獲知使用者為何者，我們的預期實驗結果顯示，我們結合了VI-CBF，在誤判率上較標準的布隆過濾器低上約四倍，如何可使搜尋過濾器有更好的效能，又本匿名機制在搜尋上採用Cuckoo Filter作為雜湊表優化，讓整體機制在使用者與雲端之間不會消耗太多的時間，且不會影響到使者者其他操作的運行。 ;The popularity of cloud services and the ease with which users enjoy cloud services do not know whether cloud service providers will ask for more information on personal and Android devices (such as IMEI) and then use it without the user′s consent Its privacy information for the user profile preferences analysis or information sold to advertisers for profit, this behavior greatly violated the privacy and rights of users. This study provides a mechanism to protect user privacy, so that users do not need to send any identifiable personal information to the cloud when the service is requested, and the cloud service provider is still able to provide the service to the user correctly. But how can cloud service providers quickly identify different users to provide appropriate services in the user′s anonymity situation with a large number of different users at the same time? Therefore, this study uses a one-time anonymization code to replace the user-friendly data transmission in the cloud, and then use the Bloom filter to implement this protection mechanism, so that users know whether their services are met. Our innovative use of the hash table with the Bloom filter is advantageous in that the time complexity is O (1) and the Bloom filter has the characteristic of being able to reverse, so that the user is protected by anonymity , It will not consume too much time on this mechanism, and the cloud side is facing the array vector and can not know why the user.
Appears in Collections:	[Graduate Institute of Information Management] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	264	View/Open

社群 sharing

Loading...