Page 161 - DCAP207_NETWORKS_DCAP406_COMPUTER_NETWORKS
P. 161

Computer Networks/Networks




                    Notes

                                      Task  Differentiate between Class A, B, C, D and E type of IP addresses.

                                   Self Assessment

                                   Fill in the blanks:

                                   1.  ……………………………. is the collection of several independent networks, which  are
                                       interconnected with one another.
                                   2.  The ………………………. is basically network identifier or network address portion of an
                                       IP unicast address.
                                   3.  ………………………… is reserved for IP unicast addresses.
                                   4.  CIDR stands for …………………………….

                                   5.  Technically, IP addresses are expressed using binary notation with ……… bit long string.
                                   10.2 Congestion Control


                                   Congestion causes chocking of the communication channel. When too many packets are present
                                   in a part of the subnet, the performance of the subnet degrades. Hence, a communication channel
                                   of a network is called congested if packets traversing the path experience delays largely in excess
                                   of the paths propagation delay. It is called heavily congested when the packets never reach the
                                   destination indicating that the delay approaches infinity. The reasons for congestion are not one
                                   but many. When the input traffic rate exceeds the capacity of the output lines, the input part of
                                   the subnet gets chocked and creates congestion. Congestion is also happened when the routers
                                   are too slow to perform queuing buffers, updating tables, etc. Lack of capacity of the routers’
                                   buffer is also one of many factors for congestion. However, enhancing memory of the router
                                   may be helpful up to a certain point.  Beyond a certain point of time, congestion gets worse
                                   because of timeout retransmission will create more traffic load. Briefly, the apparent causes of
                                   congestion are jamming by several input lines, slow processors, low bandwidth, finite number
                                   of buffers, etc.



                                     Did u know?   Congestion  control  and  flow  control  are  two  different  phenomenons.
                                     Congestion is a global phenomenon involving all hosts, all routers, the store-and-forward
                                     processing within the routers, etc., whereas, flow control is concerned with point-to-point
                                     traffic between a given source host and a given destination host. The example of congestion
                                     control is a situation when a store-and-forward network with 1-Mbps lines and 1000 large
                                     minicomputers, half of which were trying to transfer files at 100 kbps to the other half. An
                                     example of flow control is when a fiber optic network with a capacity of 1000 gigabits/sec
                                     on which a supercomputer was trying to transfer a file to a personal computer at 1Gbps.

                                   10.2.1 General Principles of Congestion Control


                                   According to control theory, the computer network, which is also a system, is divided into two
                                   groups. They are open loop and closed loop solutions.
                                   The open loop solutions: provide good design to ensure that the problem does not occur in the
                                   first place. The designing tools include decision for accepting new traffic, discarding packets and
                                   scheduling of the packets at various points in the network. The open loop solution’s decisions
                                   are independent of the current state of the network.


          154                               LOVELY PROFESSIONAL UNIVERSITY
   156   157   158   159   160   161   162   163   164   165   166