PPP Internet Protocol Control Protocol (Chinese Version)

PPP Internet Protocol Control Protocol (ipcp)

1. Introduction

PPP has three main components:

1. Encrypts data packets on a serial line;
2. The Connection Control Protocol (LCP) used to establish, configure, and detect data link connections ).
3. The Network Control Protocol (NCP) protocol family is used to establish and configure different network layer protocols.

To establish communication over point-to-point connections, each PPP end must first send an LCP package to configure and detect data links. After the connection is established and all the Optional options have been set by LCP, PPP must send an NCP package to select and configure one or more network layer protocols. After the protocols at each layer are configured, you can communicate with each other on the blockchain. When LCP or NCP explicitly closes the connection, connection 1 remains open.

2. PPP network control protocol for IP addresses

The IP Control Protocol (ipcp) is responsible for establishing, enabling and suspending the IP module. The ipcp and LCP protocols use the same packet exchange mechanism. Ipcp packets cannot be exchanged before PPP reaches the network layer protocol phase. If an ipcp packet arrives before this phase, it is discarded. Ipcp and LCP are basically the same except the following:

Data Link Layer Protocol domain. An ipcp packet is included in the information domain of the PPP data link layer frame. The Protocol domain in the frame has a value of 8021 indicating the type. The encoding field only uses encoding 1 to 7. Other codes are considered unidentifiable and discarded. Timeout. Ipcp packets cannot be exchanged before PPP reaches the network layer protocol. In implementation, you should wait for the completion of the authentication and link quality decision before receiving the configuration confirmation or other responses. In implementation, the connection should not be interrupted unless the user interferes or exceeds the limit. Configuration option type. Ipcp has different set of configuration options, which will be discussed below.

2.1. Send IP Datagram

Before IP packet communication, PPP must reach the network layer protocol stage, and ipcp must be enabled. An IP packet is included in the information domain of the PPP data link layer frame. The Protocol domain in the frame indicates the type value of 0021. The maximum length of the IP packet transmitted over a PPP connection is the same as the maximum length of the information domain of the PPP data link layer frame. Too large IP datagram must be segmented. If the system does not want segment reorganization, the maximum TCP Segment Size Option and MTU must be used for discovery.

3. ipcp configuration options

You can select the desired IP parameter for the ipcp configuration option. Ipcp uses the same configuration option format as LCP. The value of the ipcp option type field is as follows:

1 ip address 1

2. IP compression Protocol

3 IP address 2

3.1. IP address 1

It is difficult to use the IP address of the configuration option, which is proved in implementation. The IP Address Configuration option can replace this domain, and the IP Address Configuration option should be used. If the received configuration request contains the IP address or IP address option, this option should not be included in the configuration request. This option is sent only when the configuration is rejected because of the IP address option, or when the received configuration does not confirm that the IP address option is included as an additional option.

3.2. Compression Protocol

This configuration option can use a specific compression protocol. By default, compression is not required. The format of the IP compression Protocol configuration option is as follows. The transfer order ranges from left to right.

Type 2

Length> = 4

The IP address compression protocol domain consists of two bytes, indicating the desired compression protocol. The value of this domain is consistent with the compression Protocol indicated by the PPP data link layer protocol domain. Currently, the following compression protocol values can be used: 002d represents van jakbson compressed TCP/IP. The data field is 0 or multiple bytes, which stores some parameters of specific compression protocols.

3.3. IP address 2

This domain provides a method to determine the IP address used for the local connection. It allows the request sender to request the desired IP address or the other party to provide the relevant information. The other party can provide this information by returning the naking option and the legal IP address.

If it is determined that the remote IP address is required, and the other party does not provide it in its configuration request options, this option should be added after the configuration rejection option. The IP address value is either a valid remote IP address or requires the other Party to provide information. By default, no IP address is specified. The following is the IP Address Configuration option format. The transfer sequence is from left to right.

Type 3

Length 6

The four-byte IP address is used to configure the local address that the request sender wants. If all four digits are zero, the other party is required to provide IP address information.

4. Van jacbson TCP/IP Header Compression

Van Jacob son TCP/IP packet header compression reduces the size of the TCP/IP packet header to nearly three bytes, which is a major improvement for slow serial lines. The IP compression Protocol configuration option is used to indicate the ability to receive compressed packages. If both directions are required for compression, both parties must apply for it independently. When an IP packet is transferred, the PPP protocol domain is set to the following values:

Type 0021 IP. The IP protocol is not TCP, or the package is a data segment, or has not been compressed.

TCP compressed by 002d. The TCP/IP header is replaced by the compressed header.

002f unzipped TCP. IP protocol domain replaced by slot mark.

4.1. Format of configuration options

The format of the IP compression Protocol configuration option is as follows:

Type 2

Length 6

IP compression Protocol

002d indicates van Jacob's compressed TCP/IP header. The max-Slot-Id field is a byte that indicates the maximum number of slots. It is one less than the actual number of slots, and the slot mark ranges from 0 to max-Slot-id. In practice, if Max-Slot-id = 0, the problem may occur. The comp-Slot-Id field is a byte that indicates whether the slot mark field is compressed.

The 0-slot mark cannot be compressed. All compressed TCP packets must be c-bit in each change mask (Change mask), and the slot mark must be included.

1 can compress the slot mark.