Computer network Learning Notes--networking layer Knowledge point finishing

Why divide the network layer?

No network has different protocols and standards, in order to be able to exchange visits between different networks, sharing the resources of both sides, and also to maintain the original independence of each computer network, so it was divided into a network layer, and developed a special protocol to responsible for inter-network communication.

The main functions of the network layer: 1. Shielding network differences, providing transparent transmission.

Services for the Transport layer: connection-oriented network services (virtual circuit services) and non-connected network services (datagram services).

2. Providing routing options for network traffic

　　　　What is route selection?　

According to a certain principle and routing algorithm, the process of selecting an optimal path to the destination node is selected in the communication subnet of multiple nodes. The policy that determines the routing selection becomes the routing algorithm.

3. Package and package of packets 4. Congestion control

In order to avoid the transmission delay or deadlock of the data in the network transmission path, we mainly adopt the reservation buffer , packet discard and so on.

Network layer data exchange related technologies: 1. Virtual circuit packet switching (similar to Circuit SwitchingTelephone communication):

Virtual circuit packet Switching features:

1. Establish a logical connection between the sender and receiver before each packet exchange

2. All the packets of a communication are transmitted through this virtual circuit sequence, and the Exchange path is fixed, so it is not necessary to encapsulate the destination address, source address and other information on each packet.

3. Each node in the communication subnet can establish multiple virtual circuits with any node. Each virtual circuit supports data transfer between a specific two nodes.

　　　　　　　　

Virtual circuit switching process, divided into three stages, 1: Establish virtual Circuit 2: Transmission Data 3: Removal of virtual circuit

2. Packet packet switching

What is a datagram? is to add the message segment of the source and destination address information fields to the front of the data .

Packet packet Switching features:

1. 1. 1. No need to establish a separate logical connection for each communication
      2. When the message is sent, it is first disassembled into several separate datagrams with serial number and address information, which is sent to the network node in turn.
      3. Each node of the network node provides independent routing to ensure that the datagram arrives at the destination node in order, nor does it guarantee that the datagram will not be lost in the middle.

3. Line switching (also known as circuit switching)

The most primitive way of data exchange, the most common example is the telephone Exchange system, in the network using a switchable physical communication line directly connected to the communication between the two parties for data exchange.

4. Storage-Forwarding

The working mechanism of the router is this kind of storage-forwarding data exchange, which has many advantages, such as when the data is sent to each node (such as router), error checking, error correction processing, congestion control, format forwarding and so on. Depending on whether the data unit being transmitted is a message or a packet, It can be said that the storage-forwarding switching method is subdivided into two kinds: message exchange and packet switching.

Network layer protocol and Message format: 1.IP protocol (one of the most important protocols in the network layer):

IP protocol is a kind of non-connected service, which belongs to the data exchange Mode of datagram packet Exchange . The main function of the IP protocol is to transmit datagrams between the original address and the destination address . Specifically, addressing , datagram encapsulation , fragmentation and reassembly . About the header format of the IP datagram, This blog post is more detailed, http://www.xuebuyuan.com/2038843.html, here is a schematic of the IP datagram header format:

2.ARP Protocol (Address Resolution Protocol):

The ARP protocol is the protocol that resolves an IP address to an Ethernet MAC address (Physical Address) . In a LAN, when a machine A to send data to machine B, it must know the other side of the network layer address, that is, the IP address, but the light knows the IP address is not enough, Because the data message is ultimately encapsulated as a frame in the physical network, sender A must also know the MAC address of B. Arr is the protocol that implements this functionality.

　　　　　Special attention: arp resolves the same LAN in the host (or router) on the map of the IP address and MAC address, if the destination host and the source host is not in the same LAN, then through the ARP to find a router located in the current LAN MAC address, the packet sent to this router, The router then forwards the packet to the next network, and the rest of the work is done by the next network.

3.ICMP Protocol (Control Message Protocol):

This protocol is used to transfer control messages between IP hosts, routers, such as network access, whether the host is accessible, and whether the routing is available, although these control information does not transmit user data, but also plays an important role in the transmission of user data.

Unlike the ARP protocol, ICMP relies on the IP protocol to complete its task, so the IP header is encapsulated in the ICMP packet. For example, a diagnostic such as Ping, and tracert are content that belongs to ICMP.

Routing and Routing algorithms

　　The concept of routing : The behavior of transferring information from the source node to the target node.

　　The implementation of a route : Depends on the routing table in the router or layer three switch.

　　Classification of Routes :

Static routes: 1. Manual configuration 2. Routing path fixed 3. Unidirectional, no reverse route 4. Higher priority 5. For small networks.

Static routing is feasible for small and infrequently changing networks, such as LAN, but for larger WANs, dynamic routing is required because of the complexity of the topology and the frequent changes in the network structure.

Dynamic routing: 1. Auto-Generate 2. Auto-Adjust 3. Automatically generate bidirectional routes 4. You can only generate routing table diameters between networks, and you cannot generate routing table entries that reach specific nodes or hosts.

　　　　　

Routing Algorithm Basics:

The routing algorithm is to find the best path from the source node to the target node, given a set of routers and connecting router links.

Routing algorithms are divided into two main categories: non-adaptive routing algorithm and adaptive routing algorithm

The typical representative of non-adaptive routing algorithm is static routing, and the algorithm used in dynamic routing belongs to the adaptive routing algorithm.

Routing Table Basics:

A routing table is a spreadsheet or similar database that is stored on a router or networked computer. The routing table stores the path that points to the characteristic network address.

Each static route table entry contains at least the following information:

1. Network IP address of the destination address

2. Subnet mask: Used to determine the destination address belongs to the network

3. Next hop address/interface.

Some of the main routing algorithms:

Static routing algorithm:

1. Shortest path routing algorithm.
2. Diffusion algorithms

Dynamic Routing algorithm:

1. Distance Vector routing algorithm

2. Link state routing algorithm

Network layer main device router:

　　　　A router is a dedicated computer with multiple input ports and multiple output ports whose task is to forward groupings . That is, the packet received by a router input port, according to the packet to go to the destination (i.e. destination network), the packet from a suitable output port of the router forwarded to the next hop router

Hardware technology for routers:

ASIC technology: Used as packet forwarding, Anza, with a significant increase in speed

Distributed Processing Technology:

Switching structure and scheduling algorithm:

Routing Table Hardware Lookup technology: The traditional use of a number or hashing algorithm, the process is slow, and later the use of CIDR (Classless Inter-domain routing) greatly reduce the routing table entries.

High-density and multi-port

Software Technology for routers:

VPN (Virtual private network) technology: VPN is a new network technology that establishes virtual private network on public network.

Qos (Quality of Service) technology: Classification of data streams, such as packet classification technology

MPLS (Multiprotocol Tag switching) technology:

Multicast technology: The same piece of data is sent to multiple users at the same time.

Network Management System:

Three-layer switch:

Layer three switch is a two layer switch with routing function , the technology of three layer switch is two layer switch book + three layer forwarding technology

　　　　　　

is a Huawei three-layer switch. The three-layer switch works like this:

　　　　

　　　 What is the difference between layer three switches and routers?

Layer three switches can only be used to interconnect between VLANs, only for TCP/IP networks, so layer three switches are also called IP switches, and routers support interconnection between multiple types of networks.

Compared to the overall performance, the performance of layer three switches is much better than that of routers, and it is very suitable for local area networks where data exchange is frequent; the router has better routing capability than the three-layer switch.

Layer three switch three-layer forwarding function mainly relies on ASIC chip, is hardware, not easy to upgrade, router routing function is based on embedded network operating system, easy to upgrade, update.

In conclusion, three layer switches and routers have advantages, they are different applications, three layer switch main user enterprise IP LAN in different VLAN or different subnets between the three layer of communication, and routers are mainly used in WAN (WAN) network interconnection, who can not replace who.

(Note: The above is only a general over the network layer of the general knowledge point, I was reading while watching the tutorial summary organized out, and did not involve IP addresses and subnets, each of which is a very large knowledge point, need to learn to master, a lot of things I do not understand, can only first have a general understanding of the concept)

Computer network Learning Notes--networking layer Knowledge point finishing