Optimization of Round-Trip Time (RTT) on New Reno TCP Wireless Network Performance (Case Study: Banda Aceh City Coffee Shop)
The TCP protocol is one of the transport agents on the communication line, with the Quality of Service (QoS) parameter as a fixed indicator of network performance measurement including throughput, time delay, packet loss, and fairness. The use of wireless network systems often occurs disruption. The existence of wifi in a coffee shop in the city of Banda Aceh has become a modern lifestyle as the main facility, the better the internet connection, the more communities and people who come. thus motivating the author to conduct further research on the methods that have been applied before. The research results are expected to contribute to the increase in wireless network throughput with a strategy to reduce the Round-Trip Time (RTT) so that traffic will be better and this model can be applied to coffee shops in Banda Aceh City. The research objective is to simulate and analyze existing algorithms developed previously using the NS-3 application, which affects the measurement of RTT, cwnd, and throughput as well as to evaluate the performance of TCP New Reno based on topology and research parameters so that it is obtained specifically how to optimize RTT and throughput on wireless networks. Banda Aceh City coffee shop. The research method utilizes the literature and simulates the development of the TCP New Reno algorithm on wireless network coding so that the results can be applied and studied directly. The approach taken is to describe the various stages or techniques in increasing throughput, controlling the congestion window (Cwnd), and reducing RTO. The results of the study concluded that the performance of TCP New Reno is better than other TCP models when the Round-Trip Time (RTT) is smaller and the probability of Bursty Loss is lower. Then the simulation based on NS-3 produces an average value of RTO less in the point-to-point multi-hop adhoc topology than single-hop, meaning that the interference on the network that occurs is smaller. Meanwhile, if the point-to-point network is compared to the grid, there will be less disruption to the grid network.