Ibrahim Ahmed Saleh

Recently, software cost estimation has become a more important issue in the software project development cycle, software quality, and decision-making management. In view of the common problem of inaccurate and difficult cost estimation in the software industry, in this article, the proposed bagging method is one of the ensemble learning methods to estimate the cost of software development. Five algorithms were used as the basic models: Random Forest, Decision Tree, AdaBoost, K-Nearest Neighbor, and Gradient Boosting, compiled using the bagging method. The proposed method was applied to a data set (ISBSG). The contribution of the paper suggests a more accurate method compared to previous studies, and applying it to high-quality data, which was prepared to obtain more accurate results when applying the proposed model, which showed its superiority over individual models in estimating the cost of software development. The results showed high accuracy in R2 prediction in ratio (97%) and gave a lower error rate (MMRE: Mean Magnitude of Relative Error) compared to previous studies in ratio (0.1). This indicates its accuracy in prediction is closer to the real cost, where the RF model was the basic estimator model in this method, because it surpasses the main models that were used in the proposed method. The KNN model gave the lowest accuracy ratio (73%) among the main models when trained on the ISBSG dataset.

This paper proposes intelligence methods for the prediction of wireless network traffic and the evaluation of its development. This solution analyses the changing patterns of actual business traffic data in cellular networks in the time dimension and uses ensemble learning (EL) to predict the tendency of traffic, providing guidance for operators' network optimization deployment from a short-term perspective. The paper depends on a dataset taken from the Eurecom ElasticMon 5G monitoring platform. The approach applied a multi-dimensional mapping relationship between other operational data and traffic in the network using four ensembles of learning: Extreme Gradient Boosting (XGBoost), ADA boosting, gradient boosting algorithm, and a voting regression algorithm. They applied to find the upper limit of traffic that the cellular cell can carry, optimizing the deployment of the network from a long-term perspective. The results showed that XGBoost outperforms other methods, whose best performance ensures the privacy of traffic data.

An Ad-hoc network is collection of mobile nodes without the necessity of existing any centralized access points. In this paper introduce novel protocol that used to improveAdHoc protocol with salp swarm algorithm to routed flying Ad Hoc Network (FANET) by method of simulation. FANET predicated unmanned aerial vehicles (UAV), it designed wireless network has nodes withwith high mobility, actively changing topology and movement in 3D space. The main problem for FANET manner of routing packets among managed nodes. The new protocol based on salp swarm algorithm called “SalpAdHoc” protocol to solve routing problem and less the conjunction, thesimulation results of an experimental study confirming the feasibility of using salp swarm algorithms for routing in FANET are presented.

Particle swarm optimization (PSO) has become universal due to its simplicity and effectiveness in solving many problems in various applications with low computational cost. This algorithm consumes time as dealing with large tasks programs. The main goal of this paper is to introduce a parallel particle swarm optimization (PPSO) on multi-core processing kernel to decrease the determination. In order to ease transfer information among particles of shared area and exchange information by switching randomly. Most of serial PSO algorithms allow updating information among particles which takes a long time during the implementation period. The algorithm was applied to the standard optimization test set CEC (Congress on Evolutionary Computation) 2014 and gave good results compared to the previous algorithm. The empirical results show the execution time of Shared-PSO is more efficient than the serial PSO's. The proposed algorithm using a multicore CPU technique to improve it via parallelization and enhanced the efficiency of an algorithm by increase the range of PSO application.