Network technology is very popular nowbecause most offices and households use the internet forcommunication and carry out their day-to-daytransactions. With the present high demand for internetusers, they may experience congestion because of theincrease in the number of users and demand on thenetwork. Hence, controlling congestion forpacket-switched data on the wired and wireless networkusing an artificial intelligence technique is very importantto replace the current TCP protocols. This research usedthe neural network technique to realize this goal bytraining the model with some data set values which itsresult was later compared with the predicted result. Thisresearch works demonstrate great potential to controlcongestion for packet-switched on both wired andwireless networks. MATLAB was used for both thetraining of genetic programming and the ANN simulationfor the prediction. When the message sent is higher thanthe capacity of the router, it can cause congestion.Therefore, this application is designed to be installed onthe server to control congestion before its occurrence andcongestion prediction. If some of the packets are queuingup, the application controls it by passing a message to thesender that congestion is about to occur along with thelink. Therefore, the sender adjusts the speed oftransmitting packets by reducing the level of the messagethe sender is sending and the flow of packets is thenregulated.
展开▼
