Study of Queue Management Algorithms for TCP Traffic in 5G mmWave Networks

In this paper, the fifth generation (5G) wireless networks that use the mmWave to provide high data rates applying the Transmission Control Protocol (TCP) were studied. This protocol has been originally developed for wired networks, which are characterized by an approximately constant data rate over time. However, in 5G mmWave networks, due to strong fluctuations in the capacity of the received signal, the data rate over the wireless channel can change significantly over time, which leads to the following problems: (a) underutilization of the available wireless bandwidth when transmitting data using TCP and (b) queue overflow on intermediate network nodes, which leads to an increase in delay in packet transmission. To solve these problems, various approaches are explored in the thesis, such as TCP congestion control algorithms, performance-enhancing proxy and active queue management algorithms. A comparative analysis of the effectiveness of various approaches using the simulation environment of the Network Simulator 3 (NS-3) was performed. Based on the results obtained, recommendations are given for the usage of the proposed approaches to improve the performance of 5G mmWave networks in transmitting TCP data.