دانلود رایگان مقاله روش دانلود همیاری اختلال اجتناب تطبیقی

Abstract

Applications of heterogeneous wireless networks can help to achieve ubiquitous services. Cooperative downloading is a download technique used on cellular networks and wireless self-organized networks. It helps wireless users who are nearing the limits of their data plan to download data from the Internet. However, the existing studies on cooperative downloading techniques omit the problem of data disorder. Data disorder can decrease the quality of services experienced by users and increase memory usage. In this paper, we model cooperative downloading using queue theory and propose a calculation method for solving data disorder and decreasing download time. Based on the calculation method, an adaptive disorder-avoidance cooperative downloading method is proposed. The method consists of two parts: an adaptive task dissemination algorithm and a dynamic task delay prediction mechanism. The algorithm is implemented based on the calculation method that takes into account the dynamic features of wireless networks. We also propose a prediction model based on neural network learning and moving average, then use the model in the prediction mechanism to enhance the performance of the proposed method in scenarios with dynamic download rates. We used Network Simulation version 2 for the simulation, and simulation results show that the proposed method can solve the data disorder problem and be adapted to mobile scenarios. Furthermore, it can decrease the download time.

7. Conclusions

The data disorder problem in the cooperative downloading system increases the memory usage and decreases the quality of services. In this paper, we analyse the disorder problem and propose an adaptive disorder-avoidance cooperative downloading method that considers the dynamic features of wireless networks. We decrease the effect of unstable users on performance by applying recursion and propose a dynamic task delay prediction mechanism for enhancing the performance of our method in dynamic download rate scenarios. The simulation results show that our method can decrease the MUD by 85% and reduce the download time under various scenarios, including those with dynamic download rates and those with mobility. Furthermore, we only use direct neighbours as cooperative users in this work. Our future work will focus on solving the disorder problem when there are multiple hops between cooperative users and the primary user.