Hello, continuing with the course...
What does each layer of the OSI model determine?
Physical layer
The physical layer is the first layer of the OSI model. It has as PDU the bits that will define the digital signal format. It will be responsible for defining the electrical, optical and/or mechanical interfaces with the physical environment. It will also define the way in which these bits will use the medium, let's go to them:
• Simplex – Defines a one-way path between source and destination.
• Half duplex - In this case, the bit stream happens in both directions, but not simultaneously.
The 10Mbps Ethernet standard works on the Half Duplex system.
• Full duplex – In this case, the bit stream works both ways simultaneous. However, it is important to mention that you need to use 2 channels different, that is, 2 channels operating in simplex mode for each direction.
Evolutions of the Ethernet standard such as FastEthernet and GigabitEthernet use this form of channel allocation.
Link layer
The data link layer uses the frame as a PDU. It is responsible for providing a reliable medium between devices. For this, it uses error detection and correction techniques that can occur in the physical environment.
The physical medium will always be subject to interference, resulting in lost or corrupted frames, so we cannot say that the medium is free from errors. However, the link layer uses techniques that allow detection and correction of these errors, being transparent to the upper layer.
From the perspective of the upper layer, it simply assumes the condition that it operates in a reliable medium.
The link layer is also responsible for sequencing the frames to be transmitted. We must remember that frames are nothing more than a way to segment the sequence of bits that travel in the physical layer. So, a question arises. How to delimit each frame?
In this sense, there are four techniques that can be used, let's go to them:
1. Character Counting – Because it is extremely error sensitive, it is no longer used today.
2. Flag bytes, with insertion of bytes – Sensitive to a sequence of repeated data that arise from patterns accidentally. Thus, the method gets lost and generates flags that shouldn't exist.
3. Starting and ending flags, with insertion of bits – Less sensitive to the repetition problem due to insertion of “stuffed bits”. It is the method used by the Ethernet standard.
4. Physical Layer Encoding Violations – Method applicable only to networks where the decoding on the physical medium contains some type of redundancy.
Switching
An important concept for the Data Link layer is switching. By switching we understand the capacity of a device to receive information on a port or interface and pass it on to one or several ports of the same device.
There are 2 main switching techniques: circuit and packet.
• Circuit Switching:
In this way, the devices that mediate the communication between the source and destination must define the physical or logical path end-to-end for the communication to take place. When the source device wants to start sending information, the circuit is previously established. Upon termination of use, the circuit is closed. A classic example of circuit switching we have is the analog telephone circuit.
Its main advantages are:
- Resource Guarantee since the medium is not shared;
- Once access to the medium for establishing a connection is gained, there is no more competition to the medium;
- The ends have proper control of the circuit;
- It does not depend on processing at the intermediate nodes for defining and controlling the circuits;
Its disadvantages are:
- Errors can only be handled and recovered at the ends;
- In periods of idleness, that is, when there is no data traveling, there is a waste of bandwidth since the circuit remains established;
- If it is not possible to establish the circuit before communication, it will simply not occur, generating a block;
• Packet Switching:
Packet switching does not depend on the prior establishment of a dedicated path. It is characterized by the sharing of resources through the medium. The packet with the information is transmitted node-to-node so that they define the routes according to different criteria, which can be fixed or dynamic.
Its main advantages are:
- Greater efficiency in terms of using and taking advantage of the shared medium;
- Ability to handle errors in intermediate nodes;
Its main disadvantages are:
- Information usually takes longer to reach the destination;
- Does not guarantee transmission rate under native conditions;
Link layer subdivision
The data link layer of the OSI model is divided into two sublayers:
• Media Access Control (MAC): It is the lower sublayer of the link layer. It is responsible for providing the necessary resources for the device can access the network media. Media access technologies to avoid collisions are applied in this sublayer, such as CSMA/CA and CSMA/CD.
• Logical Link Control (LLC): Responsible for controlling the checking and correction of frame errors, as well as synchronizing the packets received from the network layer. As the name implies, it is responsible for maintaining the logical relationships between devices.
The LLC service can be divided into 3 types:
- Not connection oriented and no delivery confirmation;
- Makes logical connection with delivery confirmation;
- Non-connection-oriented service with delivery confirmation.
Network layer
The network layer PDU, which is the packet, also called the DATAGRAM. Its characteristic is the ability to forward datagrams between two devices that are in different networks, unlike the link layer that interconnects points in the same network.
It is capable of applying various routing and packet forwarding techniques, which may or may not guarantee a minimum level of quality of service.
Its characteristic is the ability to fragment packets and reassemble them between nodes in order to adapt them to the link's capacities, which we call MTU (Max Transfer Unit). This unit defines the number of bytes the link supports, disregarding the link layer header only.
It has the ability to exchange control messages between nodes. It is the routers that act in this layer that allow the interconnection between the various networks through the use of logical addresses (IP), a premise for the routers to be able to identify points in the network and carry out the routing.
And here we can define the concept of the Internet. The Internet can be objectively described as the network of networks, a global system. It uses different standards, protocols, services and devices to form a large interconnected network for sharing resources.
We can also define the concept of Intranet and Extranet:
Intranet – Internal and private network of a certain organization with its own and local services;
Extranet – Extension of the internal network or intranet. The communication of partners external to the network as if they were located within the network itself, that is, as if they were part of the Intranet. This service is secure and fully independent of other external users without access permission. such resource can be implemented via VPN protocol.
There are several routing techniques, with different characteristics that can be explored according to the network profiles. However, all of them have in common the use of routing tables, which is nothing more than a table of information that maps devices and networks.
When a router needs to forward a packet to some destination, it consults its table to obtain information on which interface should be used to forward the packet.
The assembly of route tables can use several parameters such as link quality, congestion criteria, factors related to distance (geographic or number of links), and others.
For this reason, we say that the network layer provides point-to-point communication, that is, node-to-node. These points are the routers spread across the network or Internet that route packets.
Transport layer
The PDU is the segment. These are defined through the segmentation of the data received from the session layer, that is, the respective processes opened in a terminal. In this context, there is the concept of multiplexing the demultiplexing of processes in relation to the ports used for communication.
This layer allows for flow control features, packet ordering, error detection and correction, and packet loss detection. To do so, it uses features that are connection-oriented, such as the TCP protocol.
On several occasions, the services only want the information to be delivered as quickly as possible and for this reason they may opt for a functionality that is not connection-oriented, such as the UDP protocol. In these cases, there is no control of content error and confirmation of receipt service by the recipient.
Just as the network layer works with the MTU parameter, as we saw earlier, the transport layer has a similar unit, called MSS (Max Segment Size). This parameter defines the maximum segment size based on information obtained from the lower layer MTU.
The main function of the MSS is to avoid the fragmentation of packets by the network layer so that the segments are already properly adjusted to the link's capacities, thus optimizing the communication and sending of packets through the network. In this case, there is no need to consume resources in the intermediate routers for packet fragmentation and reassembly.
Session Layer
As PDU we have the data. It provides resources for the establishment and support of service sessions required by the upper layers. This layer is said to manage communication between applications on devices after session establishment.
According to Tanenbaum:
"A session offers several services, including dialog control (keeping control of who should transmit at which point), token or token management (preventing two parties from trying to perform the same critical operation at the same time) and synchronization ( performing periodic verification of long transmissions to allow them to continue where they were when a failure occurred).”
Presentation Layer
It has the data as PDU. It is responsible for formatting the data received from the application layer into a common format that is understandable by the protocols used by both source and destination. It has compression and encryption features.
Application Layer
It has the data as PDU. This layer acts closer to the user, allowing the communication of application and user processes with network services. All features of this layer act in the software realm. It has several features that are defined by its various protocols such as: Remote access, virtual terminals, electronic messages (e-mail), directory services, resource sharing, among others.
Website: https://www.intelligencesoftware.com.br
Ads:
https://ads.intelligencesoftware.com.br
Comentários
Postar um comentário