اینترنت اشیا (IoT) به عنوان یکی از جدیدترین فناوری های ارتباطی برای کاربردهای مختلف در نظر گرفته می شود. از سوی دیگر با چالش های زیادی مواجه بوده است که یکی از مهم ترین آنها مربوط به امنیت است. به دلیل محدودیت های زیاد، اینترنت اشیاء در برابر حملات بسیار آسیب پذیر است و به دلیل کاربردهای حساس، به شدت در معرض حملات قرار می گیرد. مطالعات مختلفی برای بهبود امنیت اینترنت اشیا معرفی شده است. اکثر روش ها بر بهبود امنیت پروتکل RPL بر اساس توسعه مدل های اعتماد متمرکز شده اند. با این حال، بیشتر این تحقیقات رفتارهایی را برای محاسبه ارزش اعتماد در نظر گرفته اند. این روش ارزیابی اعتماد به دلیل حملات گسترده گره های مخرب کافی نیست. در این مقاله، یک روش بهبود یافته بر اساس توسعه RPL با استفاده از مدل های اعتماد با سیستم تشخیص نفوذ پیشنهاد شده است. روش پیشنهادی بر سه اصل مهم، از جمله ایجاد توپولوژی مسیریابی ایمن و قابل اعتماد، ارزیابی اعتماد، و شناسایی گره‌های مخرب تمرکز دارد. در مرحله اول توپولوژی مسیریابی شبکه بر اساس اعتماد و شرایط گره ها شکل می گیرد. در مرحله دوم، مطابق با تبادل داده ها، اعتماد گره ها ارزیابی شده و عوامل مخرب شناسایی می شوند. نتایج شبیه سازی با استفاده از Cooja نشان دهنده برتری روش پیشنهادی در بهبود قابلیت اطمینان مسیریابی و تبادل داده نسبت به عملیات قبلی است. Manuscript profile

In customs management, the main problem is balancing the needs of trade facilita-tion as a process of simplifying and accelerating foreign business on the one hand and countering illegal trade, reducing government revenue, capital sleep and the level of controls and interventions on the other. Also, due to the financial crisis in recent years, risk management has been reconsidered, although this attention is related to various financial branches. Since risk analysis and identification is the main component of risk management, developing a suitable model for data analysis is of particular importance. The purpose of this study was to use data data analysis techniques to develop an intelligent model to timely predict the risk of import declarations in customs and thus prevent irreparable losses. In this study, data analysis techniques have been used according to the statistical population which is data-driven. Statistical data were extracted from www.eplonline.ir with 575006 import declarations of all Iranian customs during 2019-2020. having pre-processed and prepared the data using PCA, LDA and FastICA methods, attribute reduction and effective attribute extraction were performed using 14 data analysis algorithms. Using Python software, algorithms were trained and modeled with 80% of the final data. Then, 14 obtained models were tested and validated with 20% of the data. Finally, the results of these models were compared with each other and the model obtained from the random forest algorithm was selected as a comprehensive model for predicting and determining the level of risk of import declarations at customs. Manuscript profile

Since the emergence of the coronavirus and the international community's attention to this virus, researchers and medical staff in hospitals have put approaches and methods for dealing with it and early diagnosis on their agenda. In this study, one of the method that called the Impact Dose program has been presented to calculate the absorbed and effective dose of people infected with the Covid-19 virus using CT scans of the respiratory tract. This is an analytical cross-sectional study that was conducted at Isfahan Shariati Hospital. The study is included of the 42 patients with Covid-19, who were selected based on doctor's opinion. The effective and absorbed dose were estimated by using the Impact Dose program, which is based on Monte Carlo. The data were analyzed using SPSS-24 software. The effective dose and DLP were reported as 1.67 mSv and 80.08 mGy.cm, respectively. The absorbed dose and effective dose were found to be higher in women than in men. The multiple CT scans of patients for the diagnosis of Covid-19 resulted in a very high absorbed dose, and lung sonography was recommended as an alternative method due to its high accuracy, similar to CT scans, in diagnosing pneumonia. Manuscript profile