karrar khaleel A.Aljeleil Aljawaheri

Abstract The demand for the array antenna that consists of multiple ports has increased in recent years, because of its main importance in reducing noise and interference between users. In this paper, we propose a new method for designing an 8×8 (16-ports) multi-input and multi-output (MIMO) antenna. This method relied on the Micro Strip mechanism so that we presented a small antenna that operates at wide and multi-bands of millimeter waves (Mm-Waves). According to the information curves generated by the CST experimental software, it was observed that the proposed antenna operates well from 36 to 60 GHz. Therefore, the antenna achieved the best results in terms of many most important parameters, the reflection coefficient is <-10 dB, return loss is <-25 dB, and voltage standing wave ratio (VSWR) is < 2. In addition, the efficiency of the antenna for all frequencies from 70% to 97%, the envelope correlation coefficient (ECC) is <0.001, and the diversity gain (DG) is 10 dB for all frequencies, while the maximum gain achieved by the antenna is 9 dB at 46 GHz. All these good results achieved by the antenna make it the prominent and potential element in most of the future 6G wireless communication systems.

Recent advancements in technology, PC facial animation is utilized in a critically large number of fields that carried human and social to contemplate PC games, films, and intuitive sight and sound reality development. this study aims to pro-vide a survey of efficient computer facial animation techniques. The proposed approach consists of two basic techniques for the generation of extreme 3D facial animation expression; facial expressions were created and described using FACS. This system disintegrates facial expressions into minor units referred to as Action Units (AUs). The generation of the emotional facial expressions relied on the combination of the independent AUs. Then, crying teardrops were created following the rules of physics by using the SPH method. The SPH rule was used to simulate and control the particles in the tears drop. Expressions such as happiness and sadness were created with FACS, followed by the addition of the tears based on the SPH rules. This provides the avatar tears that portrayed the desired expression. The study succeeded in combining the FACS & SPH approaches in creating a system that can express and implement extreme facial animation expressions. there are still areas that need to be furthered, such as modeling of muscle movement in certain expressions like crying.

In our contemporary days, relying on machine learning has become a matter that we see permeates all life sciences and engineering, and there is no doubt that the exploitation of machine learning in various medical matters, the most important of which is the early detection of types of cancerous diseases, which has become an imperative due to its relevance to human life, in this research paper we will address Incorporating both machine learning in early detection of cervical cancer and then applying the recommendation system to the final results and comparing them with traditional results, and here it should be noted that after 80 years of the invention of the Pap test, this test is still the most used method. Cell pathologists search for microscopic abnormalities inside and beside the cells, and here in this paper, we introduce deep learning techniques for cell fragmentation, as these techniques are able to detect abnormal cell clusters with the high overlap rate of digital photos for conventional Pap smears. Our method of research relies on ignoring low-load images by applying a system of advice to the results received. In the database, there are images from real cervical swabs from 113 fields of view that contain abnormal cells, and there are 96 swabs containing normal cells out of millions of cells. The results of our search on cells show an accuracy ratio (MAP = 0.924) and it works faster than the currently tiring methods.

Current days’ attacks detection represents a vital security tool because of monitoring activity for single device data follow or networks of devices. Moreover, traditional intrusion detection system still facing several limitations such as high rate of false alarm and low detection rate. Last decade machine learning represents one of the important development based on intrusion detection. This work divided various intrusion detection system based on machine learning as single models and hybrid models. Which gives the researchers new opportunities in this field. Furthermore, evaluation and measurement for complexity, a new type of dataset and time for running because of all these measurements necessary for more accurate results in comparison with relating works.

Nowadays, image annotation has attracted extensive attention due to the explosive growth of image data. Large amount of researches on AIA have been proposed, mainly including classification-based methods and probabilistic modeling methods. In this paper, a detailed study on state-of-the-art of image annotation was presented devoted to a detailed study of image annotation methods. Differences between manual, semi-automatic and automatic annotation were completely distinguished. The criteria for evaluating annotation systems are also presented in this study. In conclusion, a synthesis of methods of automatic image annotation were shown by presenting the pros and cons of each. This synthesis allowed us to examine our choice for automatic image annotation and the importance of integrating user feedback and a semantic. Finally, we participated in our perspective on the issues and challenges in AIA as well as research tendency in the future.

In current days, accuracy of Intrusion detection system still represents one of main limitation of the system. Moreover, hybrid model based on intrusion detection system achieved better results in compared with model based single algorithm. This work presents a new intrusion detection system based on hybrid model based on Fast Learning Network (FLN) and Genetic Algorithm (GA) which called (GA-FLN), NSL-KDD dataset used to evaluate the proposed model. The results of new model represent with different numbers of hidden