Point-to-Point Protocol

Point-to-Point Protocol (PPP) is a data link layer protocol, used to establish a direct connection between two nodes on a network. Connect two routers directly without any computer or other network devices between them. It is standardized in RFC 1661. It can provide authentication, transmission encryption, and compression.

PPP is used in various types of physical networks, including: serial cable, telephone line, trunk line, cellular telephone, specialized in radio link and fiber optic link such as SONET (Synchronous Optical Network). It is also used in Internet access connections (advertised as broadband). Internet service providers (ISPs) have used PPP to provide access to the Internet for users of a switched line, since IP packets cannot be transmitted via modem, without having a data link protocol.

Two derivatives of the PPP are:

• Point-to-Point Protocol over Ethernet (PPPoE),
• Point-to-Point Protocol over ATM (PPPoA).

They are commonly used by ISPs to establish digital subscriber line (DSL) Internet services for customers.

Therefore, it is a protocol associated with the TCP/IP stack used on the Internet.

Description

PPP allows communication to be established at the TCP/IP link layer level between two computers (point-to-point networks). Generally, it is used to establish the Internet connection of a private computer with its ISP through a telephone modem. Occasionally it is also used over broadband connections (such as PPPoE or PPPoA). In addition to simple data transport, PPP facilitates two important functions:

1. Authentication: usually via an access key.
2. Dynamic IP Allocation: Access providers have a limited number of IP addresses and have more customers than addresses. Naturally, not all customers connect at the same time. Thus, it is possible to assign an IP address to each customer at the time they connect to the provider. The IP address is retained until the PPP connection ends. It can then be assigned to another client.

PPP also has other uses, for example, it is used to establish communication between an ADSL modem and the ATM gateway of the telecom operator.

It has also been used to connect displaced workers (e.g. laptop) to their offices through a remote access center in their company. Although this application is being abandoned in favor of virtual private networks.

Automatic configuration

The Link Control Protocol (LCP) initiates and terminates connections, allowing users to negotiate connection options. It is an integrated part of the PPP, and is defined in the same specification standard. LCP provides automatic configuration of the interfaces of each end and selects optional authentication. LCP runs on top of PPP and uses the specific value in its protocol field of 0Xc021, and therefore a basic PPP connection must be established before LCP is configured.

Multiple network layer protocols

PPP allows multiple network layer protocols to operate on the same communication link. For each network layer protocol used, a separate Network Control Protocol (NCP) provides options for negotiating and encapsulating multiple network layer protocols. Negotiates network layer information such as network addresses or compression options after the connection has been established.

Cycle detection link

PPP detects a link cycle using a feature that involves magic numbers. When nodes send PPP LCP messages, these messages may include some magic numbers. If a line is cycling, the node receives an LCP message with its own magic number, instead of getting a message with the number.

DPI Settings

The previous section introduced the use of LCP options to find the specific requirements of a WAN connection. PPP may include the following LCP options:

• Authentication: the routers port exchange authentication messages. Two authentication options are:
  • Key Authentication Protocol (PAP) and
  • Mutual Challenge Authentic Protocol (CHAP).
• Compression: increases effective performance in PPP connections, reducing the amount of data in the plot to travel through the links.
• Error detection: identifies fault conditions. The quality and choice of Magic Numbers help ensure a reliable data link without repetitive cycles.
• Multilink: provides load balance of various interfaces using the PPP Multilink.25689

Operation

PPP consists of the following phases:

1. Connection setting: During this phase, a computer contacts the other and negotiates the link parameters using the LCP protocol. This protocol is a fundamental part of PPP and is therefore defined in the same RFC. Using LCP is negotiated the authentication method to be used, the size of the datagrams, magical numbers to be used during authentication, etc.
2. Authentication: not compulsory. There are two authentication protocols. The most basic and insecure is Password Authentication Protocol (PAP), although it is not recommended because it sends the user name and password in plain text format. A more advanced and preferred method by many ISPs is CHAP, in which the password is encrypted.
3. Network Settings: In this phase, parameters dependent on the network protocol are negotiated. PPP can carry many network protocols at the same time and it is necessary to individually configure each of these protocols. The corresponding NCP protocol is used to configure a network protocol. For example, if the network is IP, the IPCP protocol is used to assign the client's IP address and its DNS servers.
4. Transmission: During this phase, network information is sent and received. LCP is responsible for checking that the line is active during inactivity periods. Note that PPP does not provide data encryption.
5. Termination: the connection can be terminated at any time and for any reason.

PPP has all the properties of a link layer protocol:

• Reception guarantee.
• Orderly reception.
• Use of port 53 for bidirectional connection sockets.
• Used in load balancers (Load Balancer, LB) as a distribution protocol.

PPP phases and line activation

The phases of the point-to-point protocol, according to RFC 1661, are as follows:

1. Dead link: this phase occurs when the connection fails, or on one side it has been said to disconnect (e.g., a user has terminated their phone access connection).
2. Linking phase: this phase is where you try to negotiate with the Link Control Protocol. If it is successful, either control goes to the authentication phase or the network-layer protocol phase, depending on whether authentication is desired.
3. Authentication phase: this phase is optional. The sides are allowed to authenticate each other before a connection is established. If successful, control passes to the network layer protocol phase.
4. Link layer protocol phase: this phase is where each desired protocol of network control protocols is invoked. For example, the internet protocol control protocol (IPCP) is used in the establishment of IP services through the line. Data transport for all protocols that are successfully initiated with their network control protocols is also produced at this stage. Close network protocols are also produced at this stage.
5. Link termination phase: at this stage the connection is closed. This can happen if there is a authentication error, if there are so many check-in errors that the two parties decide to remove the link automatically, if the link fails suddenly, or if the user decides to hang his connection.

PPP vs. SLIP

The SLIP (Serial Line Internet Protocol) protocol fulfills the same function as PPP, but it is a much more outdated protocol. The advantages of PPP over SLIP are:

• It allows the connection through both synchronous and asynchronous lines.
• Allows dynamic allocation of IP addresses at both ends of the line.
• It allows the transport of several network protocols over it (SLIP only allows IP).
• Implement an NCP network control mechanism.

PPP can also be used to create virtual private networks (VPNs), both encrypted and unencrypted, but if encryption is desired it must be implemented below PPP.

PPP over multiple links

Multilink PPP

Multilink PPP or Multilink PPP (also known as MLPPP, MP, MPPP, or MLP) provides a method for broadcasting traffic over multiple connections of different PPPs. It can be used, for example, to connect a personal computer to an Internet service provider via two traditional 56Kb modems, or to connect a business via two leased lines. On a single line PPP frames cannot arrive out of order, but this is possible when frames are split between multiple PPP connections. Therefore, multilink PPP must enumerate the fragments so that they can be put in the correct order again when they arrive. Multilink PPP is an example of a link aggregation technology. Cisco IOS version 11.1 and later supports multilink PPP.

Multiclass PPP

With PPP, several different PPP connections cannot be established simultaneously over a single link. This is not possible with multilink PPP, since it uses contiguous numbers for all the fragments of a packet, and as a consequence it is not possible to suspend the sending of a sequence of fragments of a packet in order to send another packet. This prevents Multilink PPP from being run multiple times on the same links.

Multiclass PPP is a kind of Multilink PPP where each "class" traffic uses a separate space, sequence number and buffer reassembly.

IOS commands for DPI configuration

PPP is configured as follows on a Cisco router:

Router# configure terminal
Router(config)# serial interface 0/0/0
Router(config-if)# ip address 192.168.1.255.255.255.252
Router(config-if)# encapsulation ppp
Router(config-if)# no shutdown

Verification commands:

Router# show interface
Router# debug ppp authentication
Router# undebug ppp authentication (eliminar depuración de PPP)