Various Transmission Control Protocol (TCP) congestion control strategies have recently been devised in response to the diversification of core and edge networks in Internet of Things (IoT) enabled environments. As the traffic on the Transmission Control Protocol/Internet Protocol (TCP/IP) network increases, the queues in the router increase, and subsequently the incoming packets are rejected. Because of this, congestion control studies have been conducted at various stages of the network to design methodologies that can better manage traffic congestion. As a result, it is critical to investigate, conventional algorithms for real-world edge network environments. The congestion control in the edge network is demanded to regulate traffic at the intermediate servers. Hence, this study aims to investigate and contrast different approaches to traffic congestion management, namely, Tahoe, old Tahoe, Reno, and modified New-Reno algorithms. The performance characteristics like average transmission latency, percentage of utilization, and packet received are analyzed for real-world Constant Bit Rate (CBR) and custom applications. The findings obtained indicate that the modified New-Reno congestion control technique improves the network by 3.57, 4.08, and 1.2% in terms of packet error, throughput, and delay as compared to the existing technique. Hence, the modified congestion control is suitable for IoT-enabled edge networks with high throughput and low error rates.