According to the National transport strategy of Ukraine for the period up to 2030, one of the tasks is to ensure consistent and coherent development of regional airports of Ukraine and consequent development of Ukrainian air transportation network. One of the ways to achieve this is to deploy a remote tower concept in small regional airports of Ukraine. The main change intro-duced by the remote tower concept compared to the conventional air traffic service (ATS) provided from a local tower is that the aerodrome control tower will be provided from a remote location and air traffic control and aerodrome flight information service officers (ATCO/AFISO) do not have to be physically present at the airport. Provision of remote ATS requires a continuous exchange of information between the infrastructure at airport side and a Remote Tower Module (RTM) to ensure that the data received by ATCO/AFISO can be used to provide safe and orderly control of traffic. Due to the amount of video and audio data that has to be exchanged between an RTM and the local airport site in addition to other required data such as data from surveillance sensors, meteorological sensors, etc., the requirements for the network connection are very high. One of the factors that have a high impact on network performance is latency. The aim of this article is to investigate an application of a method of gradient descent to reduce the network latency or data transmission delay by minimizing the distance between airport sites and remote tower centres to facilitate selection of an optimal location for them.