A study at the college of engineering addresses the problem of wireless network congestion
At the department of electronic engineering and communications in the college of engineering, university of Baghdad, a public discussion of M.A student “Zainab Ghazi Faisal Najm” for her thesis entitled “password management in wireless sensor networks based on smart control algorithms”. The thesis discussed the nature of wireless sensor networks that lead to sudden password movement in the network and sending nodes when monitoring data related to only in case that certain cases occur. This study comes to develop solutions to the huge quantum of data that can easily lead to network congestion due to overflow that exceeds the available capacity of the temporary storage in the main node, leading to a decline in the overall performance of the network. So this study came to develop an approach that would steer and control that congestion control by sending all data packets to the main node. To achieve this goal, the researcher has proposed two approaches to minimize the effects of congestion and prevent it from spreading through the main node, the first approach is called “controller of nervous traffic with modified and developed waves (NNMWCC)” which can access to the main node in the grid (central control unit) and the second approach is called “controller of Gaussian traffic (FLCC)” which access the nodes close to the main node (distributed control unit). The first proposed approach (NNMWCC) has three main functions: the first is to detect indicators of the level of traffic, the second is to guess traffic rate to avoid traffic next time and the last function is to distribute the average fairly among the active nodes. As a result, the researcher adjusted the rate of password management by sending data packages while traffic is detected in the proposed second approach (FLCC) depending on the status of the temporary storage related to the node used which is close to the main node and modifying the transmission speed of active nodes. The researcher developed the simulation and tested it depending on the approach proposed by using (MATLAB R2012a) on a computer that has Windows 7 (64-bit), i7 processor, RAM, hard disk 8G .1T. The study analyzed the results of simulation to evaluate the performance of the proposed approaches and they showed that the proposed approaches were able to avoid network traffic and improving the quality of network services.The researcher compared between these proposed approaches and found that the nervous networks in the main node is more convenient than the nodes close to it because it an control and avoid the network traffic, optimizing the quality of the services in terms of percentage loss of data packages. The thesis was under the supervision of Dr. Nadia Adnan shiltagh, while the discussion committee consisted of professor Dr. Saleh Mehdi al-Qaraawi as chairman and the membership of assistant professor Dr. Najm Abdullah and assistant professor Omar Walid Abdul Wahab.