Дэлгэрэнгүй мэдээлэл

Манай улсын ууган их сургууль МОНГОЛ УЛСЫН ИХ СУРГУУЛЬ байгуулагдсаны 80 жилийн ой энэ жил тохиож байна. МУИС нь анхдагчийн хувьд үндэсний дээд боловсролтой судлаач эрдэмтэд, багш, эмч нар, техникийн дээд боловсролтой төрөл бүрийн инженер бэлтгэх өргөн хүрээтэй тогтолцоог бий болгон өөрөөсөө олон их, дээд сургууль салбарлуулан төрүүлсэн билээ. Энэхүү үйл ажиллагааны хүрээнд харилцаа холбооны үндэсний дээд боловсролтой мэргэжилтэн бэлтгэх ажлыг 55 жилийн өмнө эхлүүлэн 1967 онд МУИС-ийн физикийн ангиас радио-физик мэргэжлээр анхны 7 оюутан төгсгөсөн. Тус төгсөгчид нь Холбооны яаманд хуваарилагдаж, улс орныхоо бүтээн байгуулалтад гар бие оролцон ажилласан юм. Цаашид хөтөлбөрөө улам өргөжүүлэн МУИС нь радио, радио-физик, радио-электроникийн чиглэлээр инженер бэлтгэж байв. Улмаар техник технологийн хөгжлийн хурдцыг даган электроникийн мэргэжилтэн дагнан бэлтгэх онцгой шаардлага гарсныг үндэслэн 1990 онд МУИС-ийн электроникийн ангийг нээн 1994 онд анхны төгсөлтийг хийжээ. Өнөөдөр сургалтын хөтөлбөрүүд нь улам өргөжсөөр МУИС нь компьютерын сүлжээ, сүлжээний технологи, сэргээгдэх эрчим хүчний инженер мэргэжлээр мэргэжилтэн бэлтгэсээр байна.

Энэ ажлаар агаарын РМ2.5 тоосонцорд цаг уурын параметрүүд болон температур, чийгшил, даралт, салхины хурд хэрхэн нөлөөлж байгааг куантайль регресс арга ашиглан судлав. Куантайль регресс нь хувьсагчдын хамаарлыг том зургаар харах боломжийг олгодог. Цаг уурын параметр тус бүр нь агаарын бохирдолд хэрхэн нөлөөлж байгааг судлахаас гадна аль параметр агаарын бохирдолд илүү нөлөөлж байгааг куантайль регресс ашиглан судалсан. Судалгааны үр дүнгээс үзэхэд температур болон салхины хурд агаарын бохирдолд илүү нөлөөтэй байв. Мөн температур, салхины хурд бага үед РМ2.5 тоосонцрын түгэлтийн варианс их байсан бол температур, салхины хурд их үед варианс бага байв.

Outliers can affect misspecification of models, biased estimation of parameters, incorrect results and poor forecasts. Outliers in weather data arise due to human error and instrument error. Outliers can be classified into three categories: point outliers, collective outliers, and contextual outliers. In this paper, we propose a point outlier detection method based on a moving median filter. The method consists of three key steps. They are autocorrelation, moving median filter, and threshold determination. The autocorrelation step gives the information of similarity between immediate neighbors. The result of the autocorrelation step defines window width of the median filter. The output of the median filter gives candidate outliers and exact outliers detected based on the threshold values. Out analysis of the results demonstrate in this paper is successful and effective in detecting outliers in meteorological data.

Радио хугарлын илтгэгч нь радио долгионы тархалт болон радио системийн төлөвлөлтөд чухал параметр болдог. Энэ судалгааны ажлаар газрын гадаргуу орчмын радио хугарлын илтгэгчийн улирлын явцын загварыг Улаанбаатар хотын хувьд Гауссын функцийг ашиглан гаргасан. Бид радио хугарлын илтгэгчийн улирлын явцын загварыг сонгоход үндсэн хоёр шаардлага тавьсан. Нэгдүгээрт, загвар нь улирлын явцын ерөнхий шинж төрхтэй таарах ёстой. Хоёрдугаарт, загвар нь цөөн параметртэй байх ёстой. Ингэснээр загвар энгийн бөгөөд тооцоолол хялбар болно. Энэ үндсэн шаардлагуудаас гадна статистик хэмжүүрийг тохирох загвар сонгохдоо ашигласан. Үр дүнд нь есөн параметртэй Гауссын функцийг сонгож загварчилсан. Ингэж загварчилснаар тухайн жилийн аль ч өдрийн радио хугарлын илтгэгчийн утгыг таамаглах боломжтой болсон.

Abstract—This paper presents a hybrid model for surface radio refractivity. This model is a combination of deterministic and probabilistic models. The deterministic model represents regular changes in refractivity and the probabilistic model indicates random variations in the regular behavior. The both model parts determined in the seasonal and the diurnal cycles.

Abstract—This paper presents a hybrid model for surface radio refractivity. This model is a combination of deterministic and probabilistic models. The deterministic model represents regular changes in refractivity and the probabilistic model indicates random variations in the regular behavior. The both model parts determined in the seasonal and the diurnal cycles.

In this paper, we analyzed the diurnal cycle of the hybrid model of surface radio refractivity in Ulaanbaatar, Mongolia. It is seen that diurnal variations vary in different seasons. The original hybrid model did not consider these changes. In this paper, diurnal model has determined in four different seasons for both deterministic and probabilistic model parts. Also, the methodology of the hybrid modeling for seasonal and diurnal models is presented.

In this paper, we analyzed the diurnal cycle of the hybrid model of surface radio refractivity in Ulaanbaatar, Mongolia. It is seen that diurnal variations vary in different seasons. The original hybrid model did not consider these changes. In this paper, diurnal model has determined in four different seasons for both deterministic and probabilistic model parts. Also, the methodology of the hybrid modeling for seasonal and diurnal models is presented.

Intrusion Detection Systems has become crucial in computer network security. In the recent years, data mining and machine learning algorithms have been used by researchers in the intrusion detection with significance on improving the accuracy of the classifiers. In this paper, we present the relevance of the feature set in NUM15 and KDD99 datasets to classify packets into an attack and normal traffic. The result illustrates that the NUM15’s 10 features were more appropriate than the KDD 99 datasets features for detecting attacks with the trained module of Bayes Net classifier. The experiment is conducted using Bayes Net algorithm in the Weka data mining tool.

Intrusion Detection Systems has become crucial in computer network security. In the recent years, data mining and machine learning algorithms have been used by researchers in the intrusion detection with significance on improving the accuracy of the classifiers. In this paper, we present the relevance of the feature set in NUM15 and KDD99 datasets to classify packets into an attack and normal traffic. The result illustrates that the NUM15’s 10 features were more appropriate than the KDD 99 datasets features for detecting attacks with the trained module of Bayes Net classifier. The experiment is conducted using Bayes Net algorithm in the Weka data mining tool.

Intrusion Detection Systems have become a necessary in computer networking security of largest networks. In the recent years, the system needs to identify new intrusion in largest datasets in a timely manner because internet to instantly access information at anytime from anywhere. That is a massive increasing of data traffic and internet nodes. Therefore, to refine an IDS’s performance and computing time is a one of the important challenges in computer network security field. We are introducing by this paper studying the effects of TCP based attacks on AI algorithms computing time and detection ratio using KDDCUP dataset and our collected dataset. We are to gather network traffic; normal and abnormal containing attack are collected by SNORT. We extract features in TCP headers of the packets in the collected dataset such as sequence and acknowledge numbers, window size, control flags, and an event which is time between neighbor segments. First we normalize the feature set to reduce dimensionality of our input feature space and apply linear correlation to measure the dependability of the relationship. Finally, the selected subset of the features is given to learn the classifiers: J-48, Naïve Bayes and ANNs. By adopting the concepts of machine learning and data-mining, we could detect 98% of abnormal traffic containing attacks.

Сүлжээний хөнөөлт урсгалыг өгөгдөл өлбөрлөх болон сигнатурт суурилсан системийн хамтарсан шалгалтаар илрүүлэх

Энэхүү судалгааны ажилд Фэйсбүүк болон Твиттерээр дамжиж байгаа байгаа хөнөөлт пакет дээр анализ хийж хөнөөл учруулах халдлага дайралтыг илрүүлэхийн тулд ...

The diurnal radio refractivity over Mongolia was studied. The values of radio refractivity have been determined 59 different locations. The diurnal refractivity was calculated first day of January, April, July and October. A total of more than twenty thousand refractivity measurements was considered in this analysis. The result showed that the refractivity values were lower in the morning and the night, and higher in the afternoon. This is the result of variations in meteorological parameters such as humidity, temperature and atmospheric pressure. The highest values observed in winter and the lowest values were in spring. The diurnal maximum radio refractifiy, 342 was in Khuvsgul aimag on January 1 at 12.00 and a minimum one, 292 was in Dundgovi aimag on April 1 at 21.00.

The purpose of this work was to evaluate the possibility of using digital processing of signals in the single channel system, registering signals from the output of preamplifier and amplifier, and comparing the results and see the advantages of digital processing of signals. The novelty of this study is first time signals from the single channel spectrometry system was digitally recorded and analyzed. The advantage of the digital system was evaluated and suggested for future applications. Measurements of signals were done using high precision ADC for Single Channel Nuclear Spectrometry system using Cs137 gamma sources, from two points of the system, output of preamplifier and the main amplifier. Results show that for the digital measurement, it can be done from the output of preamplifier, not using main amplifier.

The purpose of this work was to evaluate the possibility of using digital processing of signals in the single channel system, registering signals from the output of preamplifier and amplifier, and comparing the results and see the advantages of digital processing of signals. The novelty of this study is first time signals from the single channel spectrometry system was digitally recorded and analyzed. The advantage of the digital system was evaluated and suggested for future applications. Measurements of signals were done using high precision ADC for Single Channel Nuclear Spectrometry system using Cs137 gamma sources, from two points of the system, output of preamplifier and the main amplifier. Results show that for the digital measurement, it can be done from the output of preamplifier, not using main amplifier.

The diurnal radio refractivity over Mongolia was studied. The values of radio refractivity have been determined 59 different locations. The diurnal refractivity was calculated first day of January, April, July and October. A total of more than twenty thousand refractivity measurements was considered in this analysis. The result showed that the refractivity values were lower in the morning and the night, and higher in the afternoon. This is the result of variations in meteorological parameters such as humidity, temperature and atmospheric pressure. The highest values observed in winter and the lowest values were in spring. The diurnal maximum radio refractifiy, 342 was in Khuvsgul aimag on January 1 at 12.00 and a minimum one, 292 was in Dundgovi aimag on April 1 at 21.00.

...

...

Intrusion Detection Systems have become a necessary in computer networking security of largest networks. In the recent years, the system needs to identify new intrusion in largest datasets in a timely manner because internet to instantly access information at anytime from anywhere. That is a massive increasing of data traffic and internet nodes. Therefore, to refine an IDS’s performance and computing time is a one of the important challenges in computer network security field. We are introducing by this paper studying the effects of TCP based attacks on AI algorithms computing time and detection ratio using KDDCUP dataset and our collected dataset. We are to gather network traffic; normal and abnormal containing attack are collected by SNORT. We extract features in TCP headers of the packets in the collected dataset such as sequence and acknowledge numbers, window size, control flags, and an event which is time between neighbor segments. First we normalize the feature set to reduce dimensionality of our input feature space and apply linear correlation to measure the dependability of the relationship. Finally, the selected subset of the features is given to learn the classifiers: J-48, Naïve Bayes and ANNs. By adopting the concepts of machine learning and data-mining, we could detect 98% of abnormal traffic containing attacks.