The Point-to-Point Protocol (PPP) provides a standard method for transporting multi-protocol datagrams across serial, point-to-point links. It also have the mechanisms for network-protocol multiplexing, link configuration, link-quality testing, authentication, header compression, error detection and link-option negotiation. PPP is comprised of three main components:
The PPP Data-Link Layer is based on the concepts, terminology, and frame structure of HDLC. HDLC may be used in both the synchronous and asynchronous communications. A description of the specific PPP frame format follows.
The two PPP hosts use the LCP to establish, configure and terminate the data-link connection. LCP procedures have been defined to allow agreement on the encapsulation format options, handle varying limits on packet size, authenticate the identify of the hosts, determine when a link is functioning properly and when it has failed, detect a looped-back link, and identify other common configuration problem. LCP information is carried in the PPP/HDLC frame.
In essence, two parallel protocol stacks reside on the top of LCP. Network-Layer protocol packets, containing high-layer information, are used to exchange user information; NCP packets are used to control the Network-Layer protocol connections. Network-Layer protocol problems are handled by a family of NCPs that manage the specific needs of their particular Network-Layer protocol.
| Network-Control | Network-control |
| Link_Control Protocol | |
| HDLC | |
| Physical Layer | |
Figure 1. PPP protocol stack. For communication to be established, the Physical Layer
and HDLC Data-Layer must first be operational (go back)
| Flag(7E) | Addr.(FF) | Control(03) | Protocol | Information | FSC | Flag(7E) |
Figure 2. PPP frame fields include: Flag, which indicates the beginning and end of a frame; Address, which
specifies the HDLC All-Station address; and Control, which indicates frame type.
As Figure 2 shows, the fields of the PPP frame are:
All LCP packets have the same general format (see Figure 3). the one-octet Code field identifies the type of the packet; all LCP packet types are listed in Table1. The one-octet Identifier field is a number chosen at radom used to associate requests and replies. Any packet sent in response to a request must contain the same Identifier-field value as the request. The two-octet Length field indicates the length of the entire LCP packet, in octets. Any additional information, such as data or configuration options, follows the Length field.
| Code | Identifier | Length |
| Data/Options | ||
| Type Packet Name | Type Packet Name |
| 1 Configure-Request | 7 Code-Reject |
| 2 Configure-Ack | 8 Protocol-Reject |
| 3 Configure-Nak | 9 Echo-Request |
| 4 Configure-Reject | 10 Echo-Reply |
| 5 Terminate-Request | 11 Discard-Request |
| 6 Terminate-Ack | 12 RESERVED |
LCD packets are classified into one of three types according to function. the first type is Link-Configuration packets, used to establish and configure a PPP link. Following are the types of Link Configuration packets:
The second type of LCP packets are Link-Termination packets, used to terminate a PPP link. A PPP host wishing to close an LCP connection sends a Termination-Request packet and the other host will respond with a Teminate-Ack packet. Any desired data may be exchanged in the Data field of these packets.
The third type of LCP packets are Link-Maintenance packet, used to manage and debug a link. Types of this packets are:
In order to establish communication over a point-to-point link, connections are established one layer at a time, starting at the bottom of the protocol architecture, as Figure 1 shows. First , the Physical Layer and HDLC Data-Link must be operational. Next LCP packets are exchanged to configure and authenticate the link. NCP procedures will be used to establish Network-Layer connections for one or more Network-layer protocols. The link remains operational until LCP or NCP procedures are used to terminate the PPP link, or it otherwise fails.
Figure 4 shows the PPP phases of operation.
Each Network-layer protocol must have its own NCP and may establish and terminate connections in any appropriate fashion at any time during the Network-layer protocol phase. For each Network-layer protocol, a different control procedure exists; the first NCP that was specified, naturally , was for IP and is called the Internet Protocol Control Protocol(IPCP), although subsequent NCPs have been written for AppleTalk, DECnet, and the OSI Network-layer protocol.
As an example of one set of control procedures, consider IPCP( described in RFC1332). the Information field of PPP frames carries IPCP packets with a Protocol-field value of hex 8021. These packets have the same format as LCP packets, although IPCP only uses packet types one through seven(see LCP packet formats).
The originating PPP host initiates an IP connection by sending an IPCP Configure-Request packet , which may include such IP-specific options as the local IP address and an indication of the compression protocol to be employed. The destination host will either agree with the options, disagree with the options, or begin a negotiation process. As an aside, compression can provide a significant improvement in performance on slow serial lines, particularly for interactive traffic.
Once the IPCP connection has been successfully established, the two PPP can exchange IP datagrams. The Information field of PPP frames carries the IP datagrams, where the Protocol field has a value 0021. After IP data exchange is concluded, the hosts can close the IPCP connection by exchanging IPCP Terminate packets. (go back)
Today more and more sites, especially the personal computers, are accessing the Internet and other TCP/IP hosts and networks via the use of the PPP connections. And the PPP is becoming more popular than before since it became an official Internet standard.
Melanie McMullen ,1994 , "Network Remote Access"
David M.Piscitello, A.Lyman chapin, 1993 , "Open Systems Networking, TCP/IP and OSI "