Authors
Lebedev I.
|
Degree
|
Dr of Technique, Professor, SPIIRAS |
|---|---|
|
E-mail
|
lebedev@iias.spb.su |
|
Location
|
Saint Petersburg |
|
Articles
|
Identification of Economic Information System Text Information SemanticsThe article covers some issues of the natural lan¬guages' formalization. The problems are extremely pressing as the global computer networks are developing and huge distributed data volumes are formed that are presented in the form of a text. The author examines the methodology used to build the examined structures and to formalize the structure of the questions asked.
Read more...
The approach to the analysis of the information security wireless network statusThe article сonside the issues of information security specific wireless network architecture. An assessment
of the state of information security systems based on indicators of the events intensity that
occur in the process of malicious impact in terms of queuing theory. The analysis of the potential opportunities
for the offender «soft» attacks on a wireless network. The analytic dependence by which
to measure the state of information security elements of a wireless network architecture. Model of destructive
information impact offender information security. The results showing the accuracy of the
assumptions about the exponential distribution law for the duration of service requests network nodes.
Read more...
Identification of the state of individual elements of cyber-physical systems based on external behavioral characteristics
The task of determining information security state of objects using the information of signals of
electromagnetic emissions of individual elements of devices of cyber-physical systems was investigated.
We consider the main side channels of information with which it is possible to monitor the state
of the system and analyze the software and hardware environment. Such «independent» methods of
monitoring allow analyzing the state of the system based on external behavioral characteristics within
the framework of conceptual models of autonomous agents. The statistical characteristics of signals
allowing to identify changes in the state of local devices of systems are considered. Was described an
experiment aimed at obtaining statistical information on the operation of individual elements of cyberphysical
systems. The efficiency of the neural networks approach for solving the described classification
problem, in particular, two-layer feed-forward neural networks with sigmoid hidden neurons
was investigated. The results of the experiments showed that the proposed approach is superior to the
quality of detection of anomalous states by classification based on internal indicators of the functioning
of the system. With minimal time of accumulation of statistical information using the proposed
approach based on neural networks, it becomes possible to identify the required state of the system
with a probability close to 0.85. The proposed approach of the analysis of the statistical data based
on neural networks can be used for definition of states of information safety of independent devices
of cyber-physical systems.
Read more...
Identification of abnormal functioning during the operation devices of cyber-physical systemsThe article explores the task of determining information security state of autonomous objects using
the information obtained through a side acoustic channel. The basic prerequisites for using of external
independent monitoring systems for monitoring condition of objects at the risk of the influence
of threats to information security are considered. An experiment aimed at studying the functioning
parameters of unmanned vehicles in various functioning situations was performed. The appearance
and statistical characteristics of the signals, with the help of which it becomes possible to identify abnormal
deviations during the operation of unmanned vehicles, are shown. An algorithm of two- and
three-class classification of the states of the studied objects is presented. Analysis based on the obtained
sample is very sensitive to any changes in the software and hardware configuration. At the same
time, with a minimum time of accumulation of statistical information using the proposed approach
based on a given threshold, it becomes possible to determine the point at which the attack was began.
The proposed approach model implies the possibility of using various mathematical apparatus, statistical
methods, and machine learning to achieve specified indicators for assessing the state of information
security of an object.
Read more...
Classifiers ensemble training on unbalanced samples in the analysis of the network segments stateThe relevance of the topic considered in the article lies in solving problematic issues of identifying rare events in imbalance conditions in training sets. The purpose of the study is to analyze the capabilities of a classifier’s ensemble trained on different imbalanced data subsets. The features of the heterogeneous segments state analysis of the Internet of Things network infrastructure based on machine learning methods are considered. The prerequisites for the unbalanced data emergence during the training samples formation are indicated. A solution based on the use of a classifier’s ensemble trained on various training samples with classified events imbalance is proposed. The possibility analysis of using unbalanced training sets for a classifier’s ensemble averaging of errors occurs due to the collective voting procedure, is given. An experiment was carried out using weak classifying algorithms. The estimation of features values distributions in test and training subsets is carried out. The classification results are obtained for the ensemble and each classifier separately. An imbalance is investigated consists in the events number ratios violation a certain type within one class in the training data subsets. The data absence in the training sample leads to an increase in the scatter effect responses is averaged by an increase in the model complexity including various classifying algorithms in its composition. The proposed approach can be applied in information security monitoring systems. A proposed solution feature is the ability to scale and combine it by adding new classifying algorithms. In the future, it is possible to make changes during operation to the classification algorithms composition, it makes possible to increase the indicators of the identifying accuracy of a potential destructive effect. Read more... |
