3.1 PPP Architecture
3.1.1 Point-to-point links
To make remote connections, remote-node users should have the application software (for example, File Transfer Protocol [FTP], Telnet), protocol stacks (for example, TCP/IP), and link-layer drivers (for example, PPP, Serial Line Internet Protocol [SLIP]) installed on their own remote devices, as shown in the Figure. The higher-layer protocols are encapsulated in the link-layer protocols (such as SLIP and PPP) when they are transmitted across the dialup line.

Point-to-point links between LANs, hosts, terminals, and routers can provide sufficient physical connectivity in many application environments. Many regional and commercial network services provide access to the Internet, and point-to-point links provide an efficient way to access the service provider locally.

The Internet community adopted two schemes for the transmission of IP datagrams over serial point-to-point lines:

  • SLIP -- SLIP is a standard protocol for point-to-point serial connections, using a variation of TCP/IP. SLIP was a predecessor of PPP.
  • PPP -- PPP provides router-to-router and host-to-network connections over synchronous and asynchronous circuits.

PPP and SLIP are data link layer protocols (Layer 2 of the Open System Interconnection [OSI] model). Although both SLIP and PPP were designed with IP in mind, SLIP is pretty much limited for use with IP, whereas PPP can be used for other network-layer protocols. Unlike SLIP, PPP supports multilink, which allows hosts to load balance over multiple channels. Multilink is especially useful with ISDN when both B channels can be used to achieve 128-kilobit-per-second (Kbps) throughput.

Note: ARAP and SLIP are not used very frequently in current network configurations, so they are not covered in the course. For additional configuration information, refer to Cisco Connection Online (CCO) at www.cisco.com.