Modeling Of IEEE 802.15.4 Medium Layer

IEEE Std. 802.15.4 services the low rate, low power consumption, low complexity wireless personal networks with capability of self organization. It mainly specifies medium access control layer (MAC) and the physical layer (PHY). It has led to more and more public interesting because it is considered as one of the most suitable protocol stack for Wireless Sensor Networks (WSN). For the nodes of the WSN are always powered by the batteries, the research on the power consumption is the key point for the network. Thurs the analytical model which can describe the mechanisms of the MAC is desired. The accurate model is the base for the research, such as cross-layer optimization.Most of the literature which work over the model of IEEE 802.15.4 MAC considered the network only with uplink traffic. To the best of our knowledge, only Misic et al. has modelled both uplink and downlink. The existing model on the uplink is not satisfying, because these model either lack considering the algorithm of the MAC generally or lost the accuracy when compared to the simulation result, such as without considering the unsaturated situation, delay line and the retransmission after collision.Based on the literatures we completed the model with uplink traffic, then modelled the network with both uplink and downlink traffic. In the first model we consider the unsaturated situation, formulate the nodes no more independent after the collision and analyze the performance of the network with generally algorithm like delay line, the probability of collision and the retransmission after collision. Otherwise we study the probability of collision in detail and improve the derivate method for the probability of sensing the channel busy during the second CCA. The good agreement during the model and simulation results with different configuration validates the accuracy of our model.Misic's model on downlink was complex and lost the agreement with the simulation results. They adopted the CSMA-CA mechanism when the coordinator transmitted the date frame, and analyzed the impact of the number of nodes and the data arrival rate. In our work we analyze the effect of the collision on the system with the theory of dependent nodes. Due to the priority the stability of the network is investigated, which is mainly influenced by the downlink traffic. Otherwise we solve the problem that the downlink transmission concentrates during the front part of the superframe when calculating the parameters of the model. The good agreement during the model and simulation also validates the accuracy of our model.