3. Computer network architecture
The computer network is a very complex system involving many components:
§ hosts (hosts) routers (routers) various links (links) application (applications) protocol (protocols) hardware, software ......
To From what angle to understand and analyze the computer network architecture?
How to organize and implement this complex network function?
Let's first look at the structure of an air travel to
help travelers from one place to another.
Each layer completes a (class) specific service/function.
Each layer relies on the services provided by the bottom layer and completes the corresponding functions through actions within the layer.
Key points: realize complex network functions hierarchically, and use the services provided by the lower layer and the information of the peer layer to achieve This layer functions and provides services for the upper layer.
Second example
1. Computer network overview
1.1 Overview of computer network architecture
- The network architecture is a functional description of the computer network structure
- Computer network architecture referred to as network architecture (network architecture) is a layered structure
- Each layer follows a certain network protocol to complete the function of this layer
- The computer network architecture is a collection of the various layers of the computer network and their protocols
- Architecture is the definition of the functional levels and relationships of a computer network
- The architecture is abstract
1.2 Adopt a hierarchical structure, why
- The structure is clear , which is conducive to identifying the components and relationships of the complex system
- Hierarchical reference model (reference model)
- Modular layering is easy to update and maintain the system
- The changes implemented by any layer of service are transparent to other layers of the system (what transparent?????
- For example, changes in the boarding process do not affect other parts (layers) of the aviation system
- Conducive to standardization
1.3 Basic concepts of hierarchical network architecture
The upper layer uses the lower layer services, or in other words, the lower layer provides services for the upper layer
- Entity refers to any hardware or software process that can send or receive information .
- protocol is controlled by two peer entities of the set of rules of communication , protocols are "horizontal" .
- Any entity at any level needs to use lower-level services , follow the protocol of this layer, realize the functions of this layer, and provide services to the upper layer. **Services are "vertical".
- The realization of the lower-layer protocol is transparent to the upper-layer service users .
- The adjacent layer entities of the same system interact through interfaces , and exchange primitives through the service access point SAP (Service Access Point) to specify the specific service requested.
2. OSI Reference Model
- The Open System Interconnection Reference Model (OSI), or OSI model for short , is a layered network architecture model proposed by the International Organization for Standardization (ISO) in 1984. The purpose is support for heterogeneous network system interconnection heterogeneous network system interconnection international standards best understood learning tool (theoretical model) network communication , (theory Victory, Defeat Market) . 7 layer (function ), each complete a particular Network function
2.1 The communication process explained by the OSI reference model
Look at the picture and talk:
- Host A and Host B want to communicate
- Comply with the same protocol and exchange data between peer levels (two-way dashed arrow
- The two-way solid arrow is the actual process of data flow. Data transmission is vertical.
- Adjacent layers, interact through interfaces
- The above four levels correspond to the four levels of the destination host. End-to-end layer.
2.2 OSI reference model data encapsulation and communication process
as the picture shows:
- A-PDU: Application-Protocol Data Unit (Application-Protocol Data Unit)
- …-PDU:…
- AH, SH...DH. For example, AH, the control information attached to this layer is called Application Header, which constitutes the data packet of this layer protocol
- DT adds a tail at the data link layer
- The physical layer is not added to the beginning and end, and converted into a bit stream transmission...
- Finally, go to the head and tail, go to the head and go to the head... and finally reach the target host
——Look at the pictures and talk Closing remarks: I don’t speak well with pictures, and I don’t understand the subtleties of these pictures.
2.3 Why do we need data encapsulation?
It's like why you should put the letter in an envelope when you send it.
- Increase control information
Construct protocol data unit (PDU) - control information includes:
§ address (the Address) : identifies the transmitting / receiving ends
§ error detection encoding (-Error Detecting code) : for error detection or correction
§ protocol control (Control Protocol) : Additional information implement protocol functions, Such as: priority, quality of service (QoS), and security control, etc.
2.4 Physical layer functions
Interface characteristics
- Mechanical characteristics, electrical characteristics, functional characteristics, regulatory characteristics
Bit encoding
data rate (data transmission rate, "100M gigabit network"
bit synchronization
- Clock synchronization
Transmission mode
- Simplex (simple communication, such as TV broadcasting, one-way communication
Half-duplex (two-way communication, alternately, such as walkie-talkie), you can’t listen when you talk, and you can’t talk when you listen
Full-duplex (two-way communication, simultaneous, such as video calls
2.5 Data link layer functions
Logical , data link layer.
Look at the picture and talk:
Responsible for node-to-node data transmission (node2node, two directly adjacent computing devices, with the frame as the transmission unit
Framing (framing, plus header) Add tail
Physical addressing (Physical addressing )
- Add the physical address of the sender and/or receiver in the frame header to identify the sender and/or receiver of the data frame
- Not physically addressable at the physical layer
Let's look at the
picture again:
- Flow control
- Avoid overwhelming the receiving end (let the receiving end and the sending end match the receiving speed and sending speed.
- Error control
- Detect and retransmit damaged or lost frames, and avoid duplicate frames. (Using error coding
- Access (Access control)
- Decide which device has the link** (physical medium) control usage right** at any given moment
2.6 Network layer functions
Look at the picture and talk:
Responsible for the delivery of data packets from the source host to the destination host
- Possibility to traverse multiple networks
Logical addressing
- A globally unique logical address to ensure that data packets are delivered to the destination host, such as an IP address
The data is sent from END S to END D and then
look at the picture to speak :
Routing
- The router (or gateway) interconnects the network and routes the packet to the final destination host
- Path selection
Packet forwarding
2.7 Transport layer functions
Responsible for source-destination (end-end) (inter-process) complete message transmission
Look at the picture:
Segmentation and reorganization
SAP addressing (x and y in the figure above are SAP addresses)
- Ensure that the complete message is submitted to the correct process, such as port number
Connection control
Flow control
Error control
2.8 Session layer functions
Dialog controlling (dialog controlling)
- Establish and maintain
Synchronization
- Insert a "sync point" in the data stream
The thinnest layer (in practice, this layer does not exist alone)
2.9 Presentation layer functions
Processing between two systems exchange information syntax and semantics (syntax and semantics) issues
data representing conversion
- Converted to host-independent encoding
Encryption/decryptionCompression
/decompression
(this layer does not exist independently
2.10 Application layer functions
Support users to use network (service) through user agent (such as browser) or network interface
Typical application layer service:
- File transfer (FTP)
Email (SMTP)
Web (HTTP)
……
3 Other reference models
3.1 TCP/IP reference model
3.2 Five-layer reference model
Model Figure
5 data encapsulation layer model:
3.3 Discussion after class
Hierarchical network architecture
Most modern network architecture adopts hierarchical architecture, and the advantages of hierarchical architecture are also introduced in the course. So what are the disadvantages of the layered network architecture? Is it necessary in some cases not to adopt a hierarchical structure? Check the information to see if there is a network design with exaggerated layer architecture? What are the advantages?
The disadvantages of the hierarchical network architecture are
1. It reduces the performance of the system, which is originally a simple problem, and becomes complicated after a multi-layer process.
2. It causes cascading modification problems. In some problems, if some of the problems need to be changed or optimized, it may be Will cause all layers to need to change.
2. Multiple layers will reduce efficiency and increase many protocols.
When the network is small, simple and the process is clear, it is not necessary to adopt a hierarchical structure, such as a private network with a small user base.
Cross-layer architecture design is to realize the cross-layer interaction between layers in the network architecture.
Advantages: It reduces unnecessary information interaction and improves efficiency.
The
advantages of wireless communication cross-layer structure are: system performance is enhanced, and the comprehensive performance of the network is improved.
Fzu Count Wei Rong...2020-2-20
In addition to store and forward packet switching, pass-through and fragment isolation can also be used.
1. Straight-through. Advantages: No need to store, very small delay, very fast exchange, this is its advantage. Disadvantages: Because the content of the data packet is not saved by the Ethernet switch, it is impossible to check whether the transmitted data packet is wrong, and it cannot provide error detection capability. Because there is no buffer, input/output ports with different speeds cannot be directly connected, and packets are easily lost.
2. Fragmentation isolation. It checks whether the length of the data packet is enough for 64 bytes, if it is less than 64 bytes, it means it is a fake packet, then discard the packet; if it is greater than 64 bytes, then send the packet. Disadvantages: This method also does not provide data verification. Advantages: Its data processing speed is faster than store-and-forward, but slower than straight-through. The disadvantages of hierarchical network architecture are 1. It reduces the performance of the system, which is originally a simple problem, and becomes complicated after multiple layers. 2. Causes Cascading modification issues. In some issues, if some of the issues need to be changed or optimized, all layers may need to be changed. 2. Multiple layers will reduce efficiency. Adding many protocols can eliminate the need for hierarchical structure when the network is small, simple and the process is clear, such as a private network with a small user base. Cross-layer architecture design is to realize the cross-layer interaction between layers in the network architecture. Advantages: It reduces unnecessary information interaction and improves efficiency. The advantages of the wireless communication cross-layer structure are: system performance is enhanced, and the comprehensive performance of the network is improved
mooc97533698...2020-5-10
(Supplement) 4. Service model
- Service: The ability of lower-level entities to provide communication between them to upper-level entities
service user
service provider - Primitive: The form in which the upper layer uses the lower layer service, the upper layer uses the service provided by the lower layer, and the lower layer provides services to the upper layer are all interacted through service access primitives—form
- Service Access Point SAP (Services Access Point): The upper layer uses the services provided by the lower layer through the interface between the layers—location;
Example: mailbox
Address: An entity in the lower layer supports multiple entities in the upper layer, and SAP has a logo The role of different upper-layer entities There can be different implementations, queues Example: SAP of the transport layer: port (port)
4.1 Types of services
Connection-oriented service and connectionless service-method
-
Connection-oriented Service
- Connection:
a combination established by two communicating entities for communication - The process of connection-oriented service communication:
connection establishment, communication, and disconnection - An example of connection-oriented services:
the connection at the network layer is called a virtual circuit - Scope of application:
for large data blocks to be transmitted; not suitable for small sporadic messages - Features: Preservation
- Type of service:
Reliable information flow delivery page (reliable acquisition, confirmed by the recipient)
Reliable byte stream remote login
Unreliable connection to digital voice
- Connection:
-
Connectionless Service
- Connectionless service:
Two peer-to-peer entities do not need to establish a connection before communicating, and do not reserve resources; do not need to communicate with both parties to be active; (for example: sending a letter) - Features: unreliable, may be repeated, may be out of sequence, IP packet, data packet;
- Scope of application: suitable for transmitting sporadic data;
- Type of service:
Unreliable datagram electronic correspondence
Confirmed datagram registered letter
Request to answer information inquiries
- Connectionless service:
4.2 Service and Agreement
- The difference between service and agreement
- Service : The ability of lower-level entities to provide communication between them to upper-level entities is operated through primitives, vertical
- protocol (Protocol) : between peer entity (peer entity) in the process of communicating with each other, a set of rules to be followed, the level of
- Service and agreement connection
- The implementation of this layer protocol depends on the services provided by the lower layer.
This layer entity provides higher-level services to the upper layer through the protocol.
- The implementation of this layer protocol depends on the services provided by the lower layer.
4.3 Data Unit
Describe in saliva words:
Layer n represents the current layer (the nth layer), then n-1 represents the previous layer, and n+1 represents the next layer.
The n+1 layer SDU comes to the n layer through the interface interface between the layers. The n-layer entities exchange n-pdu in their n-layer protocol, and then transmit it to become the SDU of the next layer
(it may be said that it is still very vague , Wait for me to start thinking about it again)
4.4 Hierarchical processing and the benefits of implementing complex systems
Deal with complex systems
- Conceptualization: clear structure, easy to label network components and describe their relationships
hierarchical reference model - structured: Modular easier to maintain and upgrade
change to achieve a certain level of service does not affect the other levels in the system
For other levels is transparent
eg, change in gate procedure does not affect other parts of the system
Changing the communication method used by the two secretaries does not affect the work of the two translators
Changing the language used by the two translators does not affect the work of the upper and lower levels - Where is the thought of stratification considered harmful? (The efficiency is relatively low, and the harm is taken away)
5. Protocol data units at various levels
Application layer: message
Transport layer: segment: TCP segment, UDP datagram
Network layer: packet (if no connection mode: datagram datagram)
Data link layer: frame ( frame)
Physical layer: bit
6. Development history (hang up a few pictures)
End of this article.