Table of Contents
In the traditional OSI model, the data link layer plays a crucial role by providing services to the network layer while receiving services from the physical layer. Its primary responsibility is node-to-node communication. As a packet traverses the Internet, the data-link layer of each node, whether it be a host or router, takes on the responsibility of delivering the datagram to the next node in the communication path. Before delivery, it performs essential checks for flow control and error control on the communication channel. Additionally, each data packet is assigned a physical address, representing the current device's unique identifier.
This section will delve into a detailed discussion of the functions, services, and protocols employed by the data link layer.
Physical Addressing
When devices communicate on a network, they employ two distinct addressing concepts:
- Physical Address: Also known as the MAC address, the physical address pinpoints the precise physical location of a computer device. It serves as a unique identifier for a given device on the network, facilitating its distinctive recognition.
- Logical Address: Represented by the IP address, the logical address plays a vital role in identifying the network connection through which a device is linked to the network. Unlike the physical address, which focuses on the device itself, the logical address emphasizes the network connection used by the device.
Throughout this unit, we will explore the intricacies of the physical address within the network context.
Physical Address/The MAC Address
In computer networks, each device, such as a computer or printer, possesses a distinct and necessary identification number, akin to a Social Security Number for individuals in the US. This identifier, often known as the Physical or Hardware Address, is referred to as the MAC address. It is embedded in the Network Identification Card (NIC) during manufacturing and remains unique to that specific device.
|
|
Network Identification Card (NIC)
|
The MAC address, functioning as the physical hardware identifier, is crucial for the Data Link Layer to ensure accurate data delivery to the intended device. This address is assigned to every physical network connection on a device, resulting in one MAC address for a computer with a single interface and multiple MAC addresses for a router with several physical connections.
Distinguishing it from an IP address, a logical network address, a MAC address is a fixed 48-bit identifier represented by 12 hexadecimal digits. The first 6 digits constitute the Organizational Unique Identifier (OUI), identifying the manufacturer, while the subsequent 6 digits typically represent a vendor-assigned serial number.
Importantly, a MAC address is unalterable and is permanently programmed into the device's hardware, residing in ROM (Read-Only Memory) during initialization and transferred to RAM (Random Access Memory) thereafter. Unlike an IP address, the MAC address does not affect routing messages between machines.
Error Detection
Error detection is the straightforward process of determining whether any bit errors occurred while transmitting a frame. In most data links, a Frame Check Sequence or cyclic redundancy check field is included in the data link trailer to facilitate this. The field holds a value calculated and transmitted by the sender, and the goal is for this value to align with the calculation performed by the receiver.
It's crucial to note that error detection does not inherently involve error recovery. Many data links, such as Ethernet and token ring, do not offer error recovery. Nevertheless, within the 802.2 standard, an option known as LLC2 does provide error recovery capabilities.
Identifying Encapsulated Data
In the last segment of a data link, the focus is on determining the nature of the data enclosed within the frame's data field. Upon receiving the data, the PC directs it to the relevant software based on the content within the data field. For instance, if the data originates from a Novell server, the PC transfers the data to the Netware client code. Similarly, if the data is potentially from a Sun FTP server, the PC delegates the data to the TCP/IP code.
Framing
Primarily, the initial function offered by the data-link layer is framing. At every node, the data-link layer is responsible for packaging the datagram (a packet received from the network layer) into a frame before transmitting it to the subsequent node. Conversely, when receiving a frame on the logical channel, the node must extract or de-capsulate the datagram from the frame. While we've illustrated only a header for a frame here, it's worth noting that frames may, in later chapters, incorporate both a header and a trailer. Various data-link layers exhibit distinct formats for framing, and at the data-link layer level, a packet is commonly referred to as a frame.
Flow Control
In scenarios involving a producer and a consumer, the concept of flow control becomes crucial. When the producer generates items at a pace surpassing the consumption rate, item accumulation becomes inevitable. In the context of data transmission, the sending data-link layer, situated at one end of a link, acts as a producer of frames, while the receiving data-link layer at the opposite end functions as the consumer.
If the production rate of frames exceeds the consumption rate, frames at the receiving end must be buffered until they can be consumed or processed. However, maintaining an unlimited buffer size on the receiving side is impractical. Two options arise in this situation. The first option is to permit the receiving data-link layer to discard frames if its buffer reaches full capacity. The second option involves the receiving data-link layer sending feedback to the sending data-link layer, requesting it to either stop or slow down the frame transmission.
Various data-link-layer protocols employ diverse strategies for implementing flow control. It's worth noting that flow control also extends to the transport layer, where its significance is heightened.
Error Control
In the data-link layer, a frame undergoes a process at the sending node, where it is converted into bits, transformed into electromagnetic signals, and transmitted through the transmission media. At the receiving node, electromagnetic signals are received, translated into bits, and assembled to reconstruct the original frame. However, due to the susceptibility of electromagnetic signals to errors, the received frame may contain errors.
The first step in managing errors is detection. Once an error is detected, the next step involves deciding whether to correct it at the receiving node or to discard the frame and initiate a retransmission from the sending node. It's crucial to acknowledge that error detection and correction are challenges encountered at every layer, whether it's node-to-node or host-to-host communication.
Congestion Control
While a link may experience congestion due to an abundance of frames, leading to potential frame loss, it's notable that many data-link-layer protocols do not directly employ congestion control mechanisms to address this issue. Some wide-area networks may incorporate congestion control, but, as a general rule, congestion control is typically regarded as a concern more fitting for the network layer or the transport layer. This is primarily due to its end-to-end nature, where the overall network or transport layer is better positioned to manage and mitigate congestion-related challenges.
