Skip to main content

How LLC works

LLC

The primary responsibilities of LLC are:

Network Layer protocol Multiplexing/De-Multiplexing

Interfacing with the Network (Layer3) above by doing L3 protocol multiplexing/de-multiplexing. On receiving a frame from the physical layer below, the LLC is responsible for looking at the L3 Protocol type and handing over the datagram to the correct L3 protocol (de-multiplexing) at the network layer above. On the sending side, LLC takes packets from different L3 protocols like IP, IPX, ARP etc., and hands it over to the MAC layer after filling the L3 protocol type in the LLC header portion of the frame (multiplexing).

Logical Link Services

LLC can optionally provide reliable frame transmission by the sending node numbering each transmitted frame (sequence number), the receiving node acknowledging each received frame ( acknowledgment number) and the sending node retransmitting lost frames.  It can also optionally provide flow control by allowing the receivers to control the sender’s rate through control frames like RECEIVE READY and RECEIVE NOT READY etc.

Based on whether a logical connection is established betweeen the layer 2 peers and based on whether frames are acknowledged by the peer, LLC can be classified to provide the following types of service modes:

a) Connectionless  Unacknowledged Service : This is a best effort service like IP datagram service, with no connection establishment between L2 peers and also no acknowledgment for data frames from the peer. Whenever there is data to be transferred to the peer, it is sent directly, without any connection establishment. Flow control may be optionally provided in this service. In Internet, since reliability, flow control and error control are provided at the transport layer by TCP, a simple connectionless unacknowledged service is enough at the data link layer, provided the link is of good quality with low error rates. Hence, this service mode is the most widely used mode in the Internet, where TCP/IP is the basic protocol stack. This mode  is used on almost all high quality wired links like Ethernet and Optical.

b) Connectionless Acknowledged Service: In this mode, data is directly sent between Layer2 peers without any logical link establishment. But each frame is numbered using sequence numbers and the peer acknowledges each frame received using an Acknowledgment number field. This service mode is used in scenarios where the overhead of a connection establishment is costly due to the extra delay involved, but where data reliability is needed. The sender can track lost or damaged frames and retransmit such frames to achieve reliability. This type of service is used in wireless links, where the quality of link is not as good as wired links and so frequent link establishment and teardown are unnecessary overheads, as these control frames may themselves be corrupted or lost.

c) Connection Oriented Service: In this mode, procedures are laid out for logical link establishment and disconnection. Before data transfer, a logical connection is established between peers, before data transfer starts, through the exchange of control frames, known asSupervisory Frames. The logical connection is closed after the data exchange phase is over. Actual data transfer starts  after the connection establishment phase and frames carrying higher layer data are known as Information Frames. A third category of frames, known as Unumbered Acknowledgment frames are used to acknowledge received Supervisory frames.

In this mode too, there are two variants that are used, namely one without acknowledgement and another with acknowledgement.

Connection oriented service Without Acknowledgment

Here, though a logical link is established before actual data transfer happens, there is no concept of frames being numbered and acknowledged through Sequence number and acknowledgment number fields. It is just a best effort service, with reliability left to the higher layer protocol. Many WAN protocols like HDLC, PPP, LAPB, LAPD etc. use this mode of service.

Connection oriented service with Acknowledgment

Here, apart from a logical link being established before data transfer happens, reliability and flow control services are also provided by the LLC. Reliability is provided through the use of sequence number, acknowledgment number and retransmission of lost frames using strategies like Go Back N or Selective Repeat.Flow control is provided by using a sliding window mechanism. This service mode is rarely used in the Internet, because Internet uses TCP, that supports reliability and flow control at the transport layer. This service mode is used in properietary protocols like Microsoft’s NetBIOS.

Note that though connection establishment, reliability and flow control are optional services at the data link layer, error detection is still a basic service provided by the data link layer, through the use of CRC/checksums in the frame trailer, that is added by the MAC sub-layer framing functionality

Comments

Popular posts from this blog

Three-Way-Handshake

Definition - What does Three-Way-Handshake mean? A three-way-handshake is a method used in a TCP/IP network to create a connection between a local host/client and server. It is a three-step method that requires both the client and server to exchange SYN and ACK (acknowledgment) packets before actual data communication begins. A three-way-handshake is also known as a TCP handshake. Techopedia explains Three-Way-Handshake A three-way-handshake is primarily used to create a TCP socket connection. It works when:  Upon completion of this process, the connection is created and the host and server can communicate.

Difference Between TCP and UDP?

What is the Difference Between TCP and UDP? TCP and UDP are both protocols used to send data over the internet but they do this in a slightly different way. Transmission Control Protocol (TCP) is the most commonly use protocol on the internet. When information is sent as data packets through TCP, your browser will stitch these packets back together to present you, for example, with a web page that you requested. TCP will make sure that the packets are all received and in the right order before it servers you the page so it guarantees delivery. User Datagram Protocol (UDP) on the other hand, will not perform any of the error checking that TCP does making it significantly faster. This is why UDP is better for video streaming or gaming, for example, while TCP is the go-to protocol for your normal browsing. UDP TCP Reliability: UDP is less reliable as it is a connectionless protocol. When you send a message using UDP internet protocol, it is not guaranteed if it will reach the des...

Star Topology

Star A star topology is designed with each node (file server, workstations, and peripherals) connected directly to a central network hub, switch, or concentrator Data on a star network passes through the hub, switch, or concentrator before continuing to its destination. The hub, switch, or concentrator manages and controls all functions of the network. It also acts as a repeater for the data flow. This configuration is common with twisted pair cable; however, it can also be used with coaxial cable or fiber optic cable.    Star topology Advantages of a Star Topology Easy to install and wire. No disruptions to the network when connecting or removing devices. Easy to detect faults and to remove parts. Disadvantages of a Star Topology Requires more cable length than a linear topology. If the hub, switch, or concentrator fails, nodes attached are disabled. More expensive than linear bus topologies because of the cost of the hubs...