Lower Layers of OSI Model

Understanding Lower Layers of OSI Model, Physical Layer and Data Link Layer

The lowest layers of the OSI model are: the Physical and Data Link layers. These layers specify standards for the computer’s interaction with the physical transmission media. As the network administrator, you should understand what each layers does.

Physical layer

Physical layer is the bottom layer of the OSI model. It contains standards for connecting the computer to the actual network transmission medium. The main purposes of the Physical layer are to:

• Specify standards for interacting with network media
• Specify media requirements for networks
• Format electrical signals for transmission over the network medium
• Synchronize signal transmission
• Detect errors during transmissions

At the physical layer, the computer sends a stream of bits across the transmission media. Since the computers use electrical signal to represents binary 0s and 1s, physical layer standards dealing with electrical signals include:

• Signal type (analog or digital)
• Voltage level
• Bit identification
• Bit synchronization

Transmission media standards

Protocols at the physical layer describe the characteristics of the transmission media and the electrical signal which include the specifications for the following:

• Physical connectors
• Connection devices, such as switches, multiplexers
• Data transfer rates
• Maximum transmission distances

Physical Topology

The term topology describes how all the devices on the network are physically connected together, such as:

• Bus Topology
• Ring topology
• Star topology
• Mesh topology
• Cellular topology

Hybrid topologies are combinations of different topologies used on the same network which include the following:

• Tree topology
• Star Bus topology
• Hybrid Mesh Topology

For more detail – click lan topology – the basic

Data link layer

The Data Link layer is the second-lowest layer of the OSI model. While Physical layer specifies standards for the actual physical media, the Data Link layer contains the standards for formatting data for transmission, and for transmitting data across physical media. The main purposes of Data Link layer are to:

• Format data into frames for transmission
• Provide error notifications
• Provide flow control
• Specify the logical network topology and media access methods

The Data Link layer is divided into the following two sub-layers.

• Media Access Control (MAC) Sublayer
• Logical Link Control (LLC) Sublayer

Media Access Control (MAC) Sublayer

Media Access Control Sublayer is the first or bottom sublayer in the Data Link layer. The MAC sublayer frames data before it is transmitted, and handles physical addresses for network devices. Devices such as switches and bridges use Data Link addresses to route user data through a network to specific hosts. The MAC sublayer handles the following three tasks.

1. Physical Device addressing, identify specific hardware devices. All devices on a network must have a unique physical address. For LAN networks, devices addresses are “burned in” to the network interface card (NIC). MAC addresses are 48-bit hardware addresses that appear as 12-digit hexadecimal numbers.

2. Media Access, media access method governs how network devices determine when to send signals through the network, what to do when two devices want to transmit at the same time. There are three common media access methods used on the computer networks.

• Contention (all devices have equal access)
• Token-passing (devices with the Token have access)
• Polling (devices are queried in a predetermined order)

3. Logical Topology, describes how messages travel from device to device. A specific physical topology can transmit messages in more than one way, so you can actually use a logical topology that is different than the physical topology of your network. There are three possible topologies:

a. Physical Bus, Logical Bus

b. Physical Ring, Logical Ring

c. Physical Star, Logical Bus

d. Physical Star, Logical Ring

e. Physical Star, Logical Star

Logical Link Control (LLC) Sublayer

The Logical Link Control (LLC) Sublayer is the second Data Link sublayer. It includes rules that control how multiple devices and protocols share a single link in a network. The LLC sublayer performs the following tasks:

1. Error detection, as frames and bits are transmitted through the network, errors can occur. Communication errors fall into one of two categories:

a. An expected packet is not received

b. A packet is received but contains corrupt data

Lost packets are identified through sequence numbers, and correction is performed in conjunction with flow-control features. Corrupt data within a packet is detected using one of the following methods: Parity bits and Cyclic Redundancy Check (CRC).

Parity Bits are used with simple asynchronous transmissions. Errors are detected by adding an extra bit, called a parity bit, to the end of each frame. This added bits ensures that an even or odd numbers of 1 bits are sent for each transmission. Error checking is performed by adding the number of 1 bits in the frame. If the number is not odd (or even, if even parity is used), an error has occurred.

Cyclic Redundancy Check (CRC) is a mathematical computation used to detect errors in synchronous communications.

The sending device applies the calculation to the data that is transmitted. The result is appended to the packet. As data is received the receiving device applies the same equation to the data. If the CRCs are different, it is assumed that an error in the transmission has occurred. This figure is just for an example to ease the understanding, the actual process is very complicated.

2. Flow Control, to prevent data transmissions from congesting the network or overwhelming the receiver, the LLC sublayer provides flow control which moderates the speed at which devices send data. There 3 methods:

a. Acknowledgment

b. Buffering

c. windowing

3. Multi-protocol Support, acts as a buffer or intermediary between lower, media-dependent protocols and upper, Network layer protocols. The LLC sublayer let you:

a. Run multiple upper-layer protocols on the same device at the same time.

b. Run the same upper-layer protocols on different transmission media

Connection-oriented and connectionless Services

“Connection services” is a term used to describe network functions that control and verify the sending and receiving of network messages. Connection services include items such as error detection, error correction, and flow control. Depending on the protocol implementation, connection services are implemented at various OSI layers, not just at the Data Link layer.

A protocol is often described in relation to the connection services it expects or provides. The following two classifications are typically used.

1. Connection oriented protocols

Connection oriented protocols assume that data will be lost during transmission, and therefore verify that each packet has reached its destination is a need. These protocols are slower because of the verifications also ensure reliable delivery between devices.

Connection oriented protocols require the devices to established communication session to transfer data. The three phases of connection-oriented communication are:

a. Session initialization (connection establishment)

b. Session maintenance (data transfer)

c. Session termination (connection release)

2. Connectionless protocols

Connectionless protocols assume that a reliable transmission path exists between communicating devices, and that all data will get through. Transmitting devices continue to send data without waiting for acknowledgement. Connectionless protocols are able to send data rapidly, since they do not add extra control information to each packet. Acknowledgments, if required, are provided by other upper-layer protocols.

The term reliable and unreliable are often used to describe protocols. A reliable protocol ensures packet delivery (connection-oriented), while un-reliable protocol does not (connectionless). However, unreliable protocols often have very predictable and satisfactory message delivery because they use error-free transmission media, or rely on other protocols to provide delivery reliability.

In WAN cloud connections, frame relay network and Point-to-Point-Protocol operate both at Physical and Data link layer. But ISDN network operates at Physical and Data link layer as well as network layer of the OSI model.

Understanding the OSI models is sometimes confusing for some IT students, just like understanding the IP addressing and subnetting. However, understanding both concepts will ease the network administrator in designing the network and troubleshooting any network problems.

See also:

• Spanning tree protocol
• WAN technologies
• Internet security download – free download