Если у вас возникли сложности с курсовой, контрольной, дипломной, рефератом, отчетом по практике, научно-исследовательской и любой другой работой - мы готовы помочь.
Предоплата всего
Подписываем
Cloud computation protection by verifying software for revealing it destructive properties
Abstract
Cloud computations are becoming more and more important in business and social information exchange processes. With the growing part of cloud computations in our life, the need of securing data and calculations in cloud computations is also growing.
There are a lot of problems to solve in the aspect of cloud computation protection, but my research dedicated to:
My research offers the methods, which can describe the set of software destructive properties, and verify software for absence of such properties. The offered methods will allow for cloud computation distributor to select sets of software destructive properties, and check software for their absence before executing in cloud environment. Also offered methods are completely mathematically described, so the absence or existence of malicious properties in the analyzed software sample is rigorously proved.
This is my Ph.D results overview, so they are brief, I hope that is enough.
Keywords
Cloud computation protection, software verification, software destructive properties
There is no model of threat for a cloud computation process for a current moment. So, distributors of cloud computations creates new systems of protection for themselves, without using common model. There exist some common models, such as administrative and juristic, but they are useable to detect or try to prevent destructive action against cloud computation not to detect the source of threat.
Also there are a lot of static and proactive methods to detect destructive properties of software, but all of them have cons in case of using in cloud computations. Pros and cons of observed methods of destructive properties detection are described in table 1.
Table 1 – methods of destructive properties characteristic compare for cloud computations.
|
Methods requirements |
Observes methods |
||
|
Signature methods |
Proactive methods |
Verification methods |
|
|
The absence of opportunity of software, that has known destructive property |
+ |
– |
+ |
|
Opportunity to atomize methods |
+ |
+ |
– |
|
Detection of new and modified software destructive properties |
– |
+ |
+ |
|
Absence of need for continuous usage of cloud computation resources |
+ |
– |
+ |
|
The option of formal verdict about absence or existence of software destructive property availability about |
– |
– |
+ |
As we can see from the table all of observed methods have cons, if they are to use in cloud computations. My goal is to create methods without con for usage in case of cloud computations.
For reception of formal verdict about existence of destructive property in observed software, there were created mathematical model of software, represented as specially painted flow graph, model of software destructive property, designed in dominator/post-dominator painted graph theory, and algorithm of software classification.
Usage of supposed models allows to receive rigorously proved result.
In the main work there are described the architecture, modules and the possible paths to create system, which allows to detect software with destructive properties.
The offered model allows to create described system of protection for any model of cloud computations, whatever is in a distributor sight. Also it allows to any analyze software, which coders provide him, with absolute degree of paranoia. So this allows for a user of protection system to decide which level of detection he requires.
Туманов Ю.М. Обнаружение вредоносных сценариев JavaScript на основе поведенческих сигнатур. //Безопасность информационных технологий 2009. №4. С.63-65
Филоненко А. В., Исаев И. К., Сидоров Д. В., Туманов Ю. М.. Поиск уязвимостей по бинарному коду с помощью проверки выполнимости ограничений. //Безопасность информационных технологий 2010. №2. С.83-86
Гаврилюк С. В., Туманов Ю. М. Разработка метода защиты вычислительных Грид-сетей от намеренного искажения результата вычислений //Безопасность информационных технологий 2010. №1. С.53-54
Варфоломеев А. А., Коренева А. М., Краснопевцев А. А., Туманов Ю. М., Фомичев В. М. «О реализации метода полного опробования ключей криптосистем в условиях различных математических распределенных вычислений» //Безопасность информационных технологий 2011. №1. С.82-83
Станкевичус А. А., Туманов Ю. М. –Разработка системы обнаружения вредоносных сценариев JavaScript на основе поведенческих сигнатур //Бизнес и Безопасность в России 2009 июнь С.130-131
Гаврилюк С. В., Туманов Ю. М. Использование поведенческого анализа с применением поведенческих сигнатур для выявления вредоносного кода на примере JavaScript сценариев //Безопасность информационных технологий 2010. №1. С.115-117
Моисеев А.В. Станкевичус А. А., Туманов Ю.М. «Защита среды распределённых вычислений при помощи искусственных иммунных систем» //Безопасность информационных технологий 2011. №4. С.103-105
Туманов Ю. М. Выявление интерпретируемого вредоносного кода на основе поведенческих сигнатур. – Научная сессия НИЯУ МИФИ-2011. Аннотации докладов в 3 томах. Т. 3. М.:НИЯУ МИФИ, 2010. – с. 165.
Yury Tumanov «Behavioral detection of malicious interpreted code» [Электронный ресурс]: Microsoft Research Summer School 2011. Режим доступа к ресурсу: http://blogs.msdn.com/b/msr_er/archive/2011/07/12/top-students-descend-on-microsoft-research-cambridge.aspx
Туманов Ю.М. «Использование поведенческого анализа для выявления вредоносного кода, на примере JavaScript сценариев» – сборник аннотаций работ «7-я курчатовская молодежная школа» 2009г. С.157
Туманов Ю.М., «Разработка программы обнаружения вредоносных сценариев JavaScript на основе поведенческих сигнатур» [Электронный ресурс] «РусКрипто’2009». Режим доступа к ресурсу: http://www.ruscrypto.org/sources/conference/rc2009/