東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Intrusion Detection: Embedded Softwa...

Abdulhammed, Razan.

FindBook

Google Book

Amazon

博客來

Intrusion Detection: Embedded Software Machine Learning and Hardware Rules Based Co-designs.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Intrusion Detection: Embedded Software Machine Learning and Hardware Rules Based Co-designs./
作者:	Abdulhammed, Razan.
出版者:	Ann Arbor : ProQuest Dissertations & Theses, : 2019,
面頁冊數:	177 p.
附註:	Source: Dissertations Abstracts International, Volume: 80-12, Section: B.
Contained By:	Dissertations Abstracts International80-12B.
標題:	Computer Engineering. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=13897424
ISBN:	9781392243084

Intrusion Detection: Embedded Software Machine Learning and Hardware Rules Based Co-designs.
Abdulhammed, Razan.

Intrusion Detection: Embedded Software Machine Learning and Hardware Rules Based Co-designs. - Ann Arbor : ProQuest Dissertations & Theses, 2019 - 177 p.

Source: Dissertations Abstracts International, Volume: 80-12, Section: B.

Thesis (Ph.D.)--University of Bridgeport, 2019.

This item must not be sold to any third party vendors.

Security of innovative technologies in future generation networks such as (Cyber Physical Systems (CPS) and Wi-Fi has become a critical universal issue for individuals, economy, enterprises, organizations and governments. The rate of cyber-attacks has increased dramatically, and the tactics used by the attackers are continuing to evolve and have become ingenious during the attacks. Intrusion Detection is one of the solutions against these attacks. One approach in designing an intrusion detection system (IDS) is software-based machine learning. Such approach can predict and detect threats before they result in major security incidents. Moreover, despite the considerable research in machine learning based designs, there is still a relatively small body of literature that is concerned with imbalanced class distributions from the intrusion detection system perspective. In addition, it is necessary to have an effective performance metric that can compare multiple multi-class as well as binary-class systems with respect to class distribution. Furthermore, the expectant detection techniques must have the ability to identify real attacks from random defects, ingrained defects in the design, misconfigurations of the system devices, system faults, human errors, and software implementation errors. Moreover, a lightweight IDS that is small, real-time, flexible and reconfigurable enough to be used as permanent elements of the system's security infrastructure is essential.The main goal of the current study is to design an effective and accurate intrusion detection framework with minimum features that are more discriminative and representative. Three publicly available datasets representing variant networking environments are adopted which also reflect realistic imbalanced class distributions as well as updated attack patterns. The presented intrusion detection framework is composed of three main modules: feature selection and dimensionality reduction, handling imbalanced class distributions, and classification. The feature selection mechanism utilizes searching algorithms and correlation based subset evaluation techniques, whereas the feature dimensionality reduction part utilizes principal component analysis and auto-encoder as an instance of deep learning. Various classifiers, including eight single-learning classifiers, four ensemble classifiers, one stacked classifier, and five imbalanced class handling approaches are evaluated to identify the most efficient and accurate one(s) for the proposed intrusion detection framework.A hardware-based approach to detect malicious behaviors of sensors and actuators embedded in medical devices, in which the safety of the patient is critical and of utmost importance, is additionally proposed. The idea is based on a methodology that transforms a device's behavior rules into a state machine to build a Behavior Specification Rules Monitoring (BSRM) tool for four medical devices. Simulation and synthesis results demonstrate that the BSRM tool can effectively identify the expected normal behavior of the device and detect any deviation from its normal behavior. The performance of the BSRM approach has also been compared with a machine learning based approach for the same problem. The BSRM tool for this application yielded more effective results. The FPGA module of the BSRM can be embedded in medical devices as an IDS and can be further integrated with the machine learning based approach. The reconfigurable nature of the FPGA chip adds an extra advantage to the designed model in which the behavior rules can be easily updated and tailored according to the requirements of the device, patient, treatment algorithm, and/or pervasive healthcare application.

ISBN: 9781392243084Subjects--Topical Terms:

1567821
Computer Engineering.

Intrusion Detection: Embedded Software Machine Learning and Hardware Rules Based Co-designs.
LDR:04850nmm a2200337 4500 001 2264561
005 20200504070446.5
008 220629s2019 ||||||||||||||||| ||eng d
020 $a 9781392243084
035 $a (MiAaPQ)AAI13897424
035 $a (MiAaPQ)bridgeport:10082
035 $a AAI13897424
040 $a MiAaPQ $c MiAaPQ
100 1 $a Abdulhammed, Razan. $3 3541686
245 1 0 $a Intrusion Detection: Embedded Software Machine Learning and Hardware Rules Based Co-designs.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2019
300 $a 177 p.
500 $a Source: Dissertations Abstracts International, Volume: 80-12, Section: B.
500 $a Publisher info.: Dissertation/Thesis.
500 $a Advisor: Faezipour, Miad.
502 $a Thesis (Ph.D.)--University of Bridgeport, 2019.
506 $a This item must not be sold to any third party vendors.
520 $a Security of innovative technologies in future generation networks such as (Cyber Physical Systems (CPS) and Wi-Fi has become a critical universal issue for individuals, economy, enterprises, organizations and governments. The rate of cyber-attacks has increased dramatically, and the tactics used by the attackers are continuing to evolve and have become ingenious during the attacks. Intrusion Detection is one of the solutions against these attacks. One approach in designing an intrusion detection system (IDS) is software-based machine learning. Such approach can predict and detect threats before they result in major security incidents. Moreover, despite the considerable research in machine learning based designs, there is still a relatively small body of literature that is concerned with imbalanced class distributions from the intrusion detection system perspective. In addition, it is necessary to have an effective performance metric that can compare multiple multi-class as well as binary-class systems with respect to class distribution. Furthermore, the expectant detection techniques must have the ability to identify real attacks from random defects, ingrained defects in the design, misconfigurations of the system devices, system faults, human errors, and software implementation errors. Moreover, a lightweight IDS that is small, real-time, flexible and reconfigurable enough to be used as permanent elements of the system's security infrastructure is essential.The main goal of the current study is to design an effective and accurate intrusion detection framework with minimum features that are more discriminative and representative. Three publicly available datasets representing variant networking environments are adopted which also reflect realistic imbalanced class distributions as well as updated attack patterns. The presented intrusion detection framework is composed of three main modules: feature selection and dimensionality reduction, handling imbalanced class distributions, and classification. The feature selection mechanism utilizes searching algorithms and correlation based subset evaluation techniques, whereas the feature dimensionality reduction part utilizes principal component analysis and auto-encoder as an instance of deep learning. Various classifiers, including eight single-learning classifiers, four ensemble classifiers, one stacked classifier, and five imbalanced class handling approaches are evaluated to identify the most efficient and accurate one(s) for the proposed intrusion detection framework.A hardware-based approach to detect malicious behaviors of sensors and actuators embedded in medical devices, in which the safety of the patient is critical and of utmost importance, is additionally proposed. The idea is based on a methodology that transforms a device's behavior rules into a state machine to build a Behavior Specification Rules Monitoring (BSRM) tool for four medical devices. Simulation and synthesis results demonstrate that the BSRM tool can effectively identify the expected normal behavior of the device and detect any deviation from its normal behavior. The performance of the BSRM approach has also been compared with a machine learning based approach for the same problem. The BSRM tool for this application yielded more effective results. The FPGA module of the BSRM can be embedded in medical devices as an IDS and can be further integrated with the machine learning based approach. The reconfigurable nature of the FPGA chip adds an extra advantage to the designed model in which the behavior rules can be easily updated and tailored according to the requirements of the device, patient, treatment algorithm, and/or pervasive healthcare application.
590 $a School code: 0749.
650 4 $a Computer Engineering. $3 1567821
650 4 $a Artificial intelligence. $3 516317
650 4 $a Computer science. $3 523869
690 $a 0464
690 $a 0800
690 $a 0984
710 2 $a University of Bridgeport. $b Computer Science and Engineering. $3 3181424
773 0 $t Dissertations Abstracts International $g 80-12B.
790 $a 0749
791 $a Ph.D.
792 $a 2019
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=13897424