LAN PROTOCOL ARCHITECTURE

LAN PROTOCOL ARCHITECTURE

             The LAN protocol architecture consists of layering of protocols that contribute to the basic functions of a LAN. The standardized LAN protocol architecture encompasses 3 layers. They are Physical layer, Medium Access control layer (MAC), and Logical Link control layers. The physical layer deals with the topology and transmission medium.

IEEE 802 Reference Model
             The protocols of OSI Reference Model can be classified as Network Support Layers( Physical, Data link, Network) and User support layers(Session, Presentation, Application). LAN protocols are concerned with the network support layers i.e., the lower layers. Protocols defined specifically for LAN and MAN transmission address issues related to the transmission of blocks of data over the network. The higher layers of the OSI model are independent of network architecture and are applicable to LAN, MAN and WANs. Therefore LAN protocols are related to the lower layers of the OSI model.
          As the figure  shows, the lowest layer of the IEEE 802 reference model corresponds to the physical layer of the OSI model. This layer has the following functions.

1.     Encoding / Decoding of signals

2.     Preamble generation / removal

3.     Bit transmission reception.

It also includes a specification of the transmission medium and topology. The layers above the physical layers are

1.     Logical Link control layer

2.     Medium Access control layer

             
                                                                   

The logical link layer’s functions are

Provide an interface to the higher layers and perform flow control and error control.

1.     At the sender, assemble data into a frame with address and error detection fields.

2.     At the receiver, disassemble the received frame and perform address recognition and error detection.

3.     Govern access to the LAN transmission medium

Of these, the last three functions are treated as a separate layer called, Medium access control layer.

Relationship between the levels
As the figure shows data from the application layer is passed to the TCP layer along with its header. TCP layer appends its header to the data it received and passes it to the IP layer. It appends its header to the data it received and passes it to the Logical Link Control layer. LLC appends its header creating a Logical Link Control Protocol Data Unit( LLC PDU).This entire LLC protocol data unit is passed on to MAC layer. The MAC layer adds a header at the front of the data and adds a trailer at the back. This is called the MAC Frame.

                                       
      
Fig shows Relationship between the levels of the architecture

It is important to note here that the headers are added to the data units to include control information such as, sequence number, source address, destination address etc., In transmission, the control information is necessary for the operation of a protocol.

Logical Link Control (LLC)

LLC is concerned with the transmission of a link level PDU between two stations. There is no need for an intermediate switching node here. LLC has two unique characteristics.
1. There is no primary node involved here.
2. Half of the Link access details are taken care of by the MAC layer.

The LLC user addresses are called Service Access Points (SAPs). Generally LLC users are the higher layer protocols or some network management function.
          
LLC Services
The Logical Link Control specifies the following mechanisms

1. for addressing stations across the medium
2. for controlling the exchange of data between two users.

            There are three types of services provided by the Logical Link Layer.

They are

1. Connection mode service – A logical connection is set up between two users.
2. Acknowledged connectionless service- No connections are involved. But data grams are acknowledged.
3. Unacknowledged connectionless service- Operates in datagram style. No guarantee for the delivery of data. The service does not include flow control or error control mechanisms.
          
LLC Protocol

The LLC protocol is modelled after the HDLC protocol. So, the format and the functions of LLC and HDLC are similar. LLC offers the following types of operations. 

a). Type  1 Operations
b). Type 2 Operations
c). Type 3 Operations

            Type 1 Operation
The type 1 operation supports unacknowledged connectionless service. So, the Unnumbered Information (UI) PDU is used to transfer data. Since the service offered is connectionless, there is no flow control, error control, and acknowledgement.

            Type 2 Operation
The type 2 operation supports connection mode service. So, for data exchange to take place, a connection is established between two LLC Service Access Points. In this mode data transfer is possible only if a connection is established. Once this
is done, the data transfer process starts the using information PDUs. These information PDUs include the sequence number as the LLC Type 2 operation is concerned with sequence and flow control.

Type 3 Operation
            The type 3 operation supports acknowledged connectionless service.
So, all the transmitted PDUs are acknowledged. For this, a new PDU is defined. It is called Acknowledged Connectionless (AC) Information PDU. A 1-bit sequence number is used to guard against the PDU loss.
Note: In all the three types of operations, there are two other protocols used to support management functions. They are XID and TEST. The XID PDU is used to exchange the following information.
1. Types of operations supported
2. Window Size.
            The TEST PDU is used to
1. conduct a loopback test of the transmission path between two LLC entities.
 

        

                          LLC PDU in a generic MAC Format

When the TEST PDU conducts a loopback test, it sends a TEST command PDU to the receiver of the transmission. Then the receiver issues a TEST response PDU immediately. The main uniqueness of LLC Layer is that it supports a multi access shared medium for a link.

Medium Access Protocol
                                                The devices in a LAN or a MAN share the transmission capacity of a network. So, there is a need for a mechanism to maintain order and to maximize the utilization of the transmission capacity. This mechanism must also control access to the transmission medium. Hence, Medium Access Protocol was designed to provide the above mentioned mechanism.

There are two easy ways of exercising control. They are
           1. Centralized control
           2. Distributed control
            Centralized Control:
Here a controller grants authority to access the network to the users. When a station wishes to transmit, it must request for permission to access the network from the controller.

            Advantages of Centralized Control:

            1. Greater Control
            2. Simple Access Logic 
            3. Avoids problems of distributed co-ordination among peer entities.                            

Disadvantages of Centralized Control:
1. Creates a single point of failure

The ‘How’ parameter is influenced by the following factors such as topology, cost, performance and complexity. Generally the medium access control technique can be categorized into two as Synchronous and Asynchronous.

Distributed Control:Here no controller is involved. All the stations collectively perform medium access control dynamically.

  In this technique, a specific capacity is dedicated to a connection. This is very similar to circuit switching and Frequency Division Multiplexing. (FDM)

Asynchronous MAC 
       This technique can be categorized into three. They are, 

1. round robin, 
2. reservation, and 
3. contention.

                                 

Round Robin:

1.  As the name suggests, each station takes turn to access the network and transmit.
2. It is up to the station to decide, whether to transmit or not during the time allotted. The right to access the network flows in sequential order.
3. When system1 has finished its transmission process, the control is passed on to system2, and so on.
4. This process of transfer of control may be either centralized or distributed. This technique is of immense help when the number of stations in a network is very high.
5. Eg., Polling. There are two types of traffic involved in a network. They are, Stream traffic and bursty traffic.
6. Instream traffic, the transmission process is continuous. Eg., Voice communication, and bulk file transfer. In bursty traffic, the transmission process is short.

Reservation:

1.  This technique finds application in networks that involve bursty traffic.
2.  Here the medium is assigned to stations based on time slots.
3. The time slots are equal.
4. The uniqueness of this technique is that the stations that wish to transmit can reserve time slots for their requirement.
5.  Reservation can be made in a centralized or distributed fashion.

Contention:

1. Here, all the stations in the network can claim their right to access the network irrespective of their position and duration of access.
2. This technique is an open method.
3. The logic is very simple and easy to implement.
4. Any station in the network can claim to access the network at any time.

MAC Frame Format:
                                    The protocol data unit (PDU) generated at the medium access layer is called the MAC Frame. 

The fields in the 

1.  MAC Frame are MAC control,
2.  Destination MAC address,
3.  Source MAC address, 
4.  LLC and CRC.

           1. MAC control: this field contains protocol control information.  Eg., Priority level

            2. Destination MAC address: The physical attachment point in the destination system

            3. Source MAC address:  The physical attachment point in the source system

            4. LLC: The LLC data from the next higher layers.

o    CRC: Cycle Redudancy Check    

     It is other wise called Frame Check Sequence. 

    This field is used to detect errors in received frames.