Main Content

Troubleshooting TCP/IP Communication

Transmission Control Protocol (TCP) is a transport protocol layered on top of the Internet Protocol (IP) and is one of the most highly used networking protocols. You can use network socket communication to connect to remote hosts to read and write data.


If you are having trouble connecting to or communicating with your remote host, try these troubleshooting tips.

Possible Solutions

Supported Platforms

TCP/IP is supported on these platforms:

  • Linux®

  • macOS

  • Windows® 10

The TCP/IP interface is supported on the same platforms as MATLAB®. For updates to the list of currently supported platforms, see System Requirements.

Configuration and Connection

  1. Make sure you can create your TCP/IP object. You create a TCP/IP object with the tcpclient function, which requires the name of the remote host as an input argument. You also need to specify the remote port value.

    Each TCP/IP object is associated with one instrument. For example, to create a TCP/IP object for a Sony/Tektronix® AWG520 Arbitrary Waveform Generator, you use the remote host name of the instrument and the port number, which can be found in the instrument documentation.

    t = tcpclient("",4000);
  2. After you connect to the device, you can communicate with it. If sending and receiving does not work, you can check the following:

    • Make sure the data is being sent in the format expected by the server.

    • If you connect to a web server, you might need to send HTTP get or post commands. You can also use the urlread or webread functions to communicate with web servers.

    • Many TCP/IP servers expect header information inside the TCP/IP packet.

    See Write and Read Data over TCP/IP Interface for an example of communication over TCP/IP.

VISA TCP/IP Communication

You can also use the visa interface with a VISA TCP/IP resource name instead of the tcpclient interface for TCP/IP communication with instruments.

Incorrect Data Type

Make sure the correct data type—for example int16, uint16, double—is being used with read and write. Use the same data type as the instrument. If reading and writing data types other than uint8 or int8, make sure the ByteOrder is correct.