Network Performance Simulation Over WLAN Based On Monte-Carlo Means

Traditional techniques for simulation of multi-user wireless communications systems simulate the transmission and reception of individual bits or packets through the system. Such techniques are considerably detailed and well suited for applications such as verification of low-level communications protocols and simulation of small systems with a relatively small number of users. For applications consisting of a rather large number of network nodes and a large number of users, such techniques are extremely computationally intensive and often infeasible.This paper presents an alternative approach where simulation is executed at the session level rather than at the bit or packet level, which implodes for IEEE 802.11 WLAN. This technique uses a combination of analytical results and event-driven Monte-Carlo simulation to estimate bit or packet level statistics without actually simulating the transmission and reception of individual bits or packets through the system. Taking account of the characteristic of the wireless communication network, the program has been written to perform an event based Monte-Carlo simulation of cellular network. Protocol flow can be recurred from packet's state probability by solving equations, which get from combination of analytical results. Network traffic models and propagation models are used to build the network. Then, CELLSIM can convert bottom information to top-lever index, which users really want to know. This paper also presents the progress of CELLSIM contrast to NS-2.This technique is generic in nature and thus applicable to virtually any wireless communications system. This method significantly reduces the computational complexity we compared to traditional simulation methods, and thus allows for simulation of large wireless networks.