Performance of a LAN under different ethernet wiring standards

Abstract

In modern time people are not using 10BaseT Ethernet wiring standard in the implementation of a local area network (LAN), category 3 cables which is used for the implementation of that standard in a network is become obsolete now, it is only used in telephone lines. Our aim is to find out some parameters under which 10BaseT can work well as compare to 100BaseT standard. We want people to motivate to use 10BaseT standard because under some assumptions 10BaseT standard can work well as compare to 100BaseT. To support our idea we simulate an office LAN using IT Guru Opnet 9.1, then we calculate the performance of this network under 10BaseT and 100BaseT Ethernet wiring standards with increasing time and finally we compare the results i. e. which standard performs better. We find that numbers of collision counts at hubs are always more under 10BaseT, because of the nature of this standard. The hubs are more utilized under 10BaseT, because of more retransmission attempts due high number of collisions in 10BaseT. Hubs perform smartly in case of 100BaseT standard because they filter little bit traffic in this case but they did not filter any traffic in case of 10BaseT. Initially when time is less the switch performs better under 10BaseT because it filtered more traffic, but when we start increasing the time the performance of 100BaseT is going to improve and at one instance of time their performances are become equivalent then after that the performance of 100BaseT is become good. So in our simulation we find that if a LAN is used for a short duration of time, then we have to use 10BaseT standard in the implementation of a LAN because in this case switch filter more traffic so the performance of LAN is good here. When this LAN is used for the long time duration we have to use 100BaseT standard because in this case we got good results as compare to 10BaseT.