[Network] computer network hierarchy principle & architecture

[Network] The hierarchy principle of computer networks & a series of agreements on the hierarchy and implementation of computer networks in terms of communication between computer networks A set of protocols at different layers. OSI hierarchy principle according to different levels of abstraction hierarchy each layer should implement a clearly defined function each layer selection should help to develop international standards for network protocols each layer boundary selection because this should be minimized the traffic layers of multinational interfaces are sufficient, to avoid mixing of different functions on the same layer, but not too much, otherwise the architecture is too complex the five-layer computer protocol architecture (ISO Open System Interconnection Reference Model) application layer) transport layer network layer data link layer) the physical layer TCP/IP model 1.TCP/ IP's IP layer corresponds to the OSI network layer, effectively solving the problem of heterogeneous network interconnection, shielding the heterogeneous network, and unified network identification (IP address ), the design concept is efficient and concise, providing unreliable No connection service, "best effort to transfer", assuming that the transmission quality of the physical channel can be guaranteed, the transmission layer error correction typical protocol: IP 2.TCP/ IP transport layer corresponds to the OSI transport layer, enables end-to-end connections between the source host and the target host peer object. Typical Protocol: Transmission Control Protocol TCP (connection-oriented protocol) Error Control: Traffic Control User Datagram Protocol UDP (connectionless protocol) 3. the TCP/IP network interface layer corresponds to the bottom two layers of OSI. The host uses some protocol to connect to the network, so that IP groups can be transmitted through the network. The similarities and differences between TCP/IP and OSI the similarities Between OSI and TCP/IP are that both adopt hierarchies and are hierarchical by function. Differences Between OSI and TCP/IP: OSI is divided into seven layers: physical layer, data link layer, network layer, transport layer, Session Layer, presentation layer, and application layer TCP/IP are divided into three layers: network Interface Layer, network layer (IP), and transmission layer (TCP ). Physical Layer features: Provides transparent bit stream transmission between two network physical devices. Physical Layer task: Determine some features of the interface with the transmission media DTE: data terminal equipment, data generation and processing DCT: data communication equipment, data circuit final device. Converts the input data to a signal output based on the user's network interface, and converts the received signal to a data output. Channel: logical line. A line can be divided into several channels. Bandwidth: The frequency band width modulation between the highest and lowest frequencies allowed-the process of converting a digital signal to a simulated signal-the process of converting a analog signal to a digital signal. Data transmission classification analog data uses analog signal transmission-carrier, digital data uses digital signal transmission-encoding, analog data uses digital signal transmission-sampling, digital data is sent by analog signal-modulation information interaction: Single-order communication, half-duplex: bidirectional alternate communication, full-duplex: the transmission rate of two-way simultaneous communication code element is limited by the nes' criterion. The information transmission rate is limited by the Shannon formula. Porter is the transmission rate unit of the code element (How many code elements are transmitted per second ). Code element transfer rate, also known as modulation rate bit, is the unit Frequency Division Multiplexing of information (OFDM, Frequency Division Multiplexing) that is, the total bandwidth used for the transmission channel is divided into several sub-bands (or sub-channels), and each sub-channel transmits one signal. Frequency Division Multiplexing requires that the total frequency width be greater than the sum of the frequencies of each sub-channel. To ensure that the signals transmitted in each sub-channel do not interfere with each other, an isolation band should be set up between the sub-channels, this ensures that each signal does not interfere with each other (one of the conditions ). The feature of frequency division multiplexing technology is that signals transmitted by all sub-channels work in parallel. transmission latency is not considered during each channel of signal transmission, therefore, the frequency division multiplexing technology has been widely used. TDM (Time Division Multiplexing) is to divide the Time that is provided to the whole channel for transmission information into several Time slices (Time slots), and assign these Time slots to each signal source for use, each signal is transmitted through an exclusive Channel in its own time slot. The time-division multiplexing technology is characterized by the time-division allocated in advance and fixed, so it is sometimes also called synchronous time-division multiplexing. The technical principle of CDMA is based on the Spread Spectrum Technology. Information data that requires certain signal bandwidth will be modulated using a high-speed Pseudo-Random Code (PN) with a bandwidth far greater than the signal bandwidth, the bandwidth of the original data signal is extended and then modulated and sent by the carrier. The receiving end uses identical pseudo-random codes for processing, the bandwidth signal is replaced with the narrowband signal of the original information data, that is, the signal is unexpanded to achieve information communication. CDMA features that all sub-channels can use the entire channel for data transmission at the same time. It shares the channel and time resources. Therefore, the channel efficiency is high and the system capacity is large. The Data Link Layer establishes links between adjacent nodes through the bit stream service provided by the physical Layer to check and correct possible errors during transmission, provides error-free transparent transmission to the network layer. function: it is the basic function of implementing reliable data transmission data link layer protocol on unreliable physical links. The sender encapsulates Network Layer groups into frames and sends them to the physical layer; the receiver extracts frames from the physical layer and transmits them to the network layer. The traffic control function allows the sender to send the next frame only after receiving the confirmation frame; the receiver can send the confirmation Frame back after receiving the data frame. The sender defines the frame serial number, starts the timer, and times out the retransmission receiver defines the expected frame serial number, and submits the matching frames to the data link layer of the network layer using the Connection Methods: point-to-point (WAN) and broadcast (LAN ); the basic unit of transmission: Frame MAC: The 48-bit Sliding Window Protocol can slide forward only when the receiving window slides forward (and confirmation is also sent. The windows at both ends of the receiving and receiving devices continuously slide forward according to the above rules. Therefore, this protocol is also called the sliding window protocol. When the size of both the sending and receiving windows is equal to 1, the waiting protocol is stopped. The key point of the sliding window method is that each frame to be sent is assigned a serial number. The serial number ranges from 0 ~ 2n-1 (n-Bit Field) receives groups from the network layer in sequence, and stores up to n frames to be confirmed. When the window reaches the maximum value of n, the network layer is forced to close and submit the frame sequence to the network layer for entering the window, confirmation is generated. frames that fall outside the window are discarded. The retransmission mechanism of the wrong frame is error control. back n frames (go back n). The sender continuously sends them to the sending window when the receiving window is full. The receiving window is 1, if the discarded frame is not confirmed and the sender times out and re-transmits, the selective repeat option is selected from the unconfirmed frame. The receiving window stores all the correct frames after the error frame, and the sender only retransmits the error frame; the receiver receives retransmission frames and submits the packets to the network layer Traffic Control Technology in the correct order (1) stop-when the traffic control sending node sends a frame of data, it must wait for the recipient to send back to confirm the response information and then send the next frame. the receiving node checks the frame verification sequence. If there is no error, a confirmation frame is sent. Otherwise, a denial frame is sent and re-transmission is required. there is a problem: the two sides wait endlessly (the data frame or the confirmation frame is lost), the solution is to use the timeout timer after sending; the frame duplication phenomenon (the same two frames are received ), solution: number frames (2) Sliding Window traffic control refers to any time point Allows the sender/receiver to send/receive multiple frames at a time. The number of frames is called the size of the sending/Receiving Window, the size of the sending window is WT. The size of the receiving window is WR. The relationship between the size of the sliding window and the Protocol is WT> 1, WR = 1, and the Protocol is returned to the ARQ (continuous ARQ protocol) in N steps) WT> 1, WR> 1. The protocol is the ARQWT = 1, WR = 1 for selection of retransmission, and the protocol is the stop-equation ARQ bit-oriented Link Control Procedure HDLC, HDLC uses the zero-bit Filling Method to transmit any combination of bit streams, or, the Point-to-Point Protocol (PPP) of the transparent transmission over the data link layer uses a special character filling method. A reliable transport Bridge (Bridge) that uses serial numbers and validation is also called a Bridge. It is a storage and forwarding device that connects two local networks, it can be used to connect a network system with identical or similar architecture. The bridge works on the data link layer, connects two LANs, and forwards frames according to the MAC address, it can be considered as a "low-layer Router" (the router works at the network layer and is forwarded based on network addresses, such as IP addresses ). A vswitch is a network device that can encapsulate and forward data packets based on MAC address recognition. The switch can "Learn" the MAC address and store it in the internal address table. By creating a temporary exchange path between the initiator of the data frame and the Target receiver, the data frame can directly reach the destination address from the source address. The vswitch has a high-bandwidth back bus and an internal switching matrix. All the ports of the vswitch are attached to the back bus. After the control circuit receives the data packet, the processing port searches for the address table in the memory to determine the target MAC address (the hardware address of the NIC) NIC (Network Card) connected to which port, through the internal switching matrix to quickly send data packets to the destination port, if the destination MAC does not exist to broadcast to all ports, after receiving the port response, the switch "learns" the new address and adds it to the internal address table. The difference between a vswitch and a HUB is that a vswitch can transmit data between multiple port pairs at the same time. Each port can be considered as an independent network segment, network devices connected to them enjoy all their own bandwidth and do not need to compete with other devices. At the same time, the HUB can only transmit data between one port. The network component of the NIC works on the data link layer. It is the interface connecting computers to the transmission media in the LAN, not only can physical connection and Electrical Signal Matching be achieved between the local area network transmission media, it also involves frame sending and receiving, frame encapsulation and unblocking, media access control, data encoding and decoding, and data cache functions. Functions of the network layer: Provides the transmission layer with the means to establish, maintain, and release network connections (point-to-point transmission) on top of the data link to solve the routing selection problem. When dividing subnets, the router forwarding group algorithm extracts the destination IP address D from the header of the received group. Use the subnet mask of each network and D-bit-by-bit to check whether it matches the corresponding network address. If yes, the group is delivered directly. Otherwise, the group is delivered indirectly and executed (3 ). If the route table contains a specific host route whose destination address is D, the group is sent to the specified next hop router. Otherwise, execute (4 ). The subnet mask of each row in the route table and D are "matched" by bit. If the result matches the destination network address of the row, the group is sent to the next hop router specified by the row; otherwise, execute (5 ). If there is a default route in the route table, the group is sent to the default router specified in the route table. Otherwise, run (6 ). An error occurred while reporting the forwarding group. Classless addressing uses network-prefix of various lengths to replace the network and subnet numbers in the classless addressing. Principles and Methods: route aggregation longest prefix match (longest-prefix matching) Internet Control Packet protocol ICMP in order to improve the chance of successful IP datagram delivery, internet Control Message Protocol (ICMP) is used at the network layer ). There are two types of ICMP packets: ICMP Error Report and ICMP inquiry message. PING is used to test the connectivity between two hosts. PING uses ICMP send-back request and send-back response packets. PING is an example of using network layer ICMP directly at the application layer. It does not use TCP or UDP at the transport layer. An autonomous system is a router and network group under the control of a management organization. It can be a vro that is directly connected to a LAN and connected to the Internet. It can be a LAN that is interconnected by an enterprise backbone network. All routers in an autonomous system must connect to each other, run the same routing protocol, and assign the same autonomous system number. The role of the transport layer is to provide a reliable data exchange mechanism between processes in two different systems. The Protocol has an end-to-end significance. The transport layer provides logical communication for communication application processes, while the network layer provides Flow Control for communication between hosts) it refers to the ability to control the number and speed of data sent by the sending end on a channel so that it cannot exceed the capacity of the receiving end, this capability mainly refers to the speed at which the receiver receives data and the size of the buffer for receiving data. Generally, the traffic is controlled by the stop method or sliding window method. Congestion Control refers to a situation where the network transmission performance decreases due to the large number of packets transmitted in the packet switching network and the limited resources of the storage and forwarding nodes. An extreme condition of congestion is the Deadlock (Deadlock), which usually requires network reset. Traffic control is set for limited resources in the end system, and congestion control is set for limited resources on intermediate nodes. The traffic control at the transport layer defines the traffic control between end users. The data link protocol defines the traffic control between two adjacent links. The congestion control algorithm is slow-start ), congestion avoidance, fast retransmit, and fast recovery ). The Error Detection data link layer provides reliable link transmission. However, when a router reformats frames containing groups, errors that affect the group content may occur. The frame checksum is calculated after the new frame is created and contains the error data. The error detection at the transport layer is used to detect this error. The TCP protocol is byte-oriented. TCP regards the packet to be transmitted as a data stream consisting of bytes, and makes each byte correspond to a sequence number. The transport layer must have two different transport protocols: connection-oriented TCP and connectionless UDP. Connection Management defines a rule that allows two users to talk like a direct connection. Many protocols at the application layer of TCP connections are based on the client server. FTP is based on the client server. Two connections: Data Connection (20) control connection (21) Simple Mail Transfer Protocol (SMTP) One Mail Server can be either a customer or a server. The unified Resource Locator url http is a transaction-oriented Client Server protocol. Dynamic Host Configuration Protocol DHCP Simple Network Management Protocol SNMP