 Professor, Electronics and Telecommunication Engineering, Valchand Institute of Technology, Solapur. Today, we will discuss frame relay network, learning outcome. At the end of this session, students will be able to describe key features of frame relay network, contents. In the following slides, we will consider different features of frame relay, virtual circuit, used in the frame relay, and congestion control. Let us see the frame relay. This is what figure 1 indicates, x dot 25 network, and figure 2 indicates frame relay network. As we see, there are different nodes are there, and these nodes are connected with each other for transferring data between any given node to any other node. But in frame relay, these nodes are connected to the frame relay network, and the transferring of data from a particular DTE to another DTE is done by the frame relay network. Here, if you look at x dot 25 is a point-to-point network communication. So if there are five nodes, as we see here, which wants to communicate with each other, it will require 10 different links. These are the 10 different links between nodes. This is avoided in a frame relay network. As we see here, these nodes are not connected in a frame relay network itself does the task of switching from a particular DTE to another DTE. So as we see, each DTE is connected to a DC, and the frame relay interval does the remaining thing for us. Here, figure 3 and figure 4 shows what type of data a particular network supports. For example, this is quoted by x dot 25, and this type of data is supported by frame relay. So here, if you look at this on x-axis, we have time, and on y-axis, we have data rate, and this nature of data is constant here, and this type of data is supported by x dot 25, whereas in this data is not constant. At some point in time, there is data. In another time, there is no data. And third time, a huge amount of data. That means this type of data, data rate, or this pattern of data rate is called busty nature of data. And this x dot 25 does not support this, because it is designed to communicate data rate of, that is, constant data rate, whereas frame relay supports this busty nature of data by using the buffering technique. Another feature is in x dot 25, in addition to data packets, many packets have to transfer, and that increases the load on the network. But in frame relay, there is no necessity of transferring other packets, because the links are more reliable, and that's why flow and error control is not required in frame relay. But because of higher speed in frame relay, a congestion may occur, and that's why it is necessary to perform congestion control in frame relay. Here, if you look at this is figure number 5 of x dot 25. These are the different layers present in the x dot 25. And this is figure number 6, frame relay. These are the two layers present in the frame relay. So if you look at this, to perform data transfer in frame relay requires only two layers, whereas in x dot 25, it requires three layers. Now let us have a question. Why congestion control is needed in frame relay? You pause the video and answer the question. You might have thought over the question. Transition control is needed in frame relay, because the data is transferred at a very high rate. Now let us study this virtual circuit. Virtual circuits are of two types. One is a permanent virtual circuit, and another is a switched virtual circuit. If you look at this frame format of a frame relay network, we have flags in the beginning and at the end, then address information, control information, information to be transmitted. So we have this address information is shown here in detail. There are different bits are used for congestion control. For example, dlci, this first bit, it is called data link connection identifier, which is used for congestion control. Another bit, c oblique r, that is command oblique response. Then ea extended address. Then fecn, becn, fecn stands for forward explicit congestion notification. And becn stands for backward explicit congestion notification. These two bits are also used for congestion control. And this de is used for discard eligibility. So this is what the frame format of frame network, which is used for congestion control. Now let us see dlci. In PVC, dlci are permanent and assigned by the network provider, whereas in SVC, switched virtual networks, dlci are temporary and assigned by the frame relay during the connection phase. Let us have a look at figure eight, how congestion control is done in a frame relay. Here, this is a node DTA, this is a DTB. And these two DTs want to communicate with each other. And there are different nodes. Communication network nodes are present. There are five nodes total. And at each node, we can have a table of dlci's. So this is an input dlci's number. And this is an output dlci number. For example, here we have expanded this one. Node two of the network, where one is an input node. This one, this is an input node. And other two are output node numbered as two and three. If you look at this table, this is input one. This is an input one. And when the output is two, that means if there is a connection between one and two, then dlci assigned to input is 48, this one. So this dlci assigned to this input is 48. And the dlci assigned to output is 65, as we see here. But if the connection is between one and three, dlci changes. So the dlci is 62 assigned for the input. And 98 assigned for the output, that means here, three. So this way, virtual circuits are created. And dlci are assigned. Congestion control. Because of higher data rate and no use of flow control, frame-related network is prone to congestion. It uses two bits for congestion control, that is BECN and FECN. And discard utility bit is used for packet discarding whenever other things do not work. Here we can have the table. The two bits are 0, 0. That means there is no congestion. When there is a 0, 1, there is a congestion in one direction. When it is 1, 0, the congestion is in other direction. And when both are 1, the congestion is in both the directions. So we can find out what type of congestion is occurred between using these two bits. Now, how to find out that congestion has taken place? For that, what is necessary is we should perform traffic control measurements. And that is done in the next slide. Here these are the different parameters for congestion control. One is the access rate. Then there is a committed bus size BC, committed information rate, and excess bus size. So when access rate is beyond a particular value, then we can say that there is a congestion. Here also the bus size is also if it is greater than BC, that is predefined maximum bits per second. Then we can say the congestion is occurred. Here this is the average information rates in bits per second. And this is what the excess bus size means. If the maximum number of bits per second exceeds this BC, then congestion may occur. These are the references, data communication by Professor Ajit Pal, Department of Computer Science and Engineering, Indian Institute of Technology, Kharapur. And I have taken the contents from NPTEL lecture number 24 on x.25 and frame reading. You can go through this link for additional information. Thank you very much.