The protocols in the IEEE 802.3 Ethernet standard enable the computers in a local network to exchange messages with each other. In practice, most Ethernet networks also use Internet protocols such as TCP or UDP and IP. These provide defined and well-supported methods for accomplishing common tasks such as flow control and flexible addressing and routing of messages. Messages that travel on the Internet must use IP. And because TCP and UDP are designed to work along with IP, local communications that use TCP or UDP also use IP, even if they wouldn’t otherwise require it. This networking tutorial begins with a guide to connecting embedded systems to the Internet. Following this is an introduction to the Internet Protocol, including when and how embedded systems can use it in local and Internet communications.

Connecting to the Internet

To communicate over the Internet, a computer must have three things: an IP address that identifies the computer on the Internet, the ability to send and receive IP datagrams, and a connection to a router that can access the Internet. An Internet Service Provider (ISP) can provide one or more IP addresses and a connection to a router that can communicate over the Internet. Customers use a variety of ways to connect to ISPs. A high-volume user, including the networks at some large businesses, government offices, and schools, may have a dedicated, high-speed connection to an ISP. If your network is located at a facility that has this type of access, your network administrator can tell you if your system can use the connection. Connections that support low to moderate traffic typically connect to the ISP via a modem or other device that interfaces to a phone line or a cable from a cable-TV provider. Considerations in Obtaining Internet Service The type of Internet connection to use depends in part on its intended use.

A computer that hosts a Web page that other computers can request has different requirements than a computer used only to request Web pages but not serve them. In many Internet communications, one computer functions as a client, and the other as a server. A client requests resources from a server. A resource may be a Web page, file, or other data. In response to a request, a server sends the client the requested resource or a response such as an error message. Microsoft’s Internet Explorer and other Web browsers are clients. The text that you type or copy into the browser’s Address text box (such as http://www.Lvr.com or http://192.168.111.1) identifies the resource you’re requesting and the server you’re requesting it from. The computers that host the resources are functioning as servers, which detect, interpret, and respond to requests from computers on the Internet or in a local network. Many servers are huge systems that store thousands of files, but a server can also be a small embedded system that serves a few basic Web pages or other information on request. As Chapter 3 showed, many Ethernet-capable modules for embedded systems include software that enables the modules to function as Web servers. If you want users on the Internet to be able to request Web pages, download or upload information, or access other resources on your system, you’ll need three things: a computer that functions as a server, an Internet account that permits hosting a server, and network-security settings that enable the server to receive and respond to requests from other computers in the network without putting other local resources at risk. When selecting a method of connecting, you need to consider the speed in both the upstream (towards the Internet) and downstream (from the Internet) directions.

For many inexpensive accounts, the upstream speed is slower than the downstream speed. This arrangement is generally fine for home users, who tend to use Internet connections for activities such as Web surfing, where most of the traffic is downloads. Typical uploading activities for home users, such as sending moderate amounts of e-mail, aren’t time-critical, so a slower upload speed is fine. In contrast, a server sends most of its data upstream. Still, an embedded system that serves very basic Web pages or transfers moderate amounts of data may function fine with a slower connection. To host a server, it’s likely that you’ll need a business, or commercial, account with your ISP. In addition to limited speed for upstream communications, accounts offered to home users typically forbid hosting servers because a server is likely to draw more traffic than the ISP can support at home-user prices. For home accounts, some ISPs block unsolicited requests to port 80, which is the default port where Web servers receive requests. One option that uses a different approach is worth a mention for applications where an embedded system only needs to provide information periodically to a server on the Internet. Many ISPs and other companies offer Web hosting services that enable you to host Web pages on one of the company’s servers. You upload the files, typically via FTP, to the server, and the server responds to requests to view the pages. For some applications, you can program a device to send files to the server as needed and let the server handle the work of serving requests on the Internet. With this arrangement, the device doesn’t have to function as a server; it just needs to be able to transfer files as needed to a remote server.