An edge server receives a first request message for transmission to the host device. The edge server determines a first congestion control algorithm based on the first request message, including characteristics of the first request message. The edge server applies the first congestion control algorithm to the transport connection for application to the transmission of the first request message. Subsequently, the edge server receives a second request message for transmission to the host device over the transport connection. Based on the second request message, including characteristics of the second request message, the edge server determines and applies a second congestion control algorithm to the transport connection for application to the transmission of the second request message, wherein the second congestion control algorithm is different from the first congestion control algorithm