A TCP socket is one endpoint of a two-way communication channel between two computers connected over a network. It provides a reliable, ordered, and error-checked way to exchange data using the Transmission Control Protocol (TCP).
Before any data is transmitted, TCP establishes a connection between the client and the server through a process known as the three-way handshake. Once the connection is established, both sides can send and receive data independently, allowing full-duplex communication.
A TCP socket is uniquely identified by a combination of an IP address and a port number. The IP address identifies the device on the network, while the port number identifies the specific application or service running on that device.
One of TCP’s main advantages is reliability. The protocol ensures that:
- Data arrives in the correct order.
- Lost packets are automatically retransmitted.
- Duplicate packets are discarded.
- Data corruption is detected and corrected through retransmission.
- Flow control and congestion control help maintain efficient communication.
TCP sockets are widely used in applications where reliable data delivery is essential, including:
- Web browsers communicating with web servers (HTTP/HTTPS)
- Email services (SMTP, IMAP, POP3)
- File transfers (FTP)
- Database connections
- Remote login protocols such as SSH
- Many client-server business applications
From a programming perspective, a TCP socket is typically represented as an object that allows an application to establish a connection, send data, receive data, and close the connection when communication is complete.
Because TCP guarantees reliable and ordered delivery, it is the preferred protocol for applications where every byte of data must arrive correctly, even if this introduces slightly higher latency than connectionless protocols such as UDP.