Amorphous flattening self-organization wireless network is made up of a group of wireless mobile nodes. Amorphous network means that the network structure is not fixed, and there are no fixed center base stations.Flattening network means that the three levels wireless network which contains core network, base station, terminals will be flattening into one level with network node only. Self-organization network means the network can adaptively adjust the network parameters to adapt to the environment. Amorphous flattening self-organization network is widely used because of its flexible structure, convenient, powerful functions and many other excellent features. The innovation of this thesis mainly includes three parts, shown as following:(1) in order to make amorphous flattening self-organization networks get the ability to adjust to changes in the structure of networks because of high-speed mobile node moving and make the network easily to be conducted, this thesis put forward "a Perceptron-based adaptive on-demand weighted clustering algorithm",designed to minimize the cluster computing cost, maximum load balancing factor. Ultimately achieve the network load balancing. The simulation results show that the PerAOW algorithm can improve the network load balancing factor 5.2%.The improvement of network load balancing can be a positive cycle to improve the energy limitation of the network nodes.(2) due to the limitation of bandwidth in wireless communication system, the bit rate of wireless communication system is limited. In the other hand, the high mobility of nodes causes the Doppler shift. In order to alleviate the contradiction of the validity and reliability, this thesis presents a method of realizing rate adaptation and anti-Doppler shift. at the sending side, each sub-channel use adaptive modulation technology to get adaptive rate and make the most use of wireless resource, then the system reached maximum bit rate under the condition of low bit error rate.(3) in order to make the amorphous flattening self-organization wireless network support the scene of nodes with relatively high-speed moving, this thesis designed a preprocess Doppler diversity receiver, the Doppler diversity preprocess is before the analog-to-digital converter, to compensate the shift and multipath fading from wireless channel. The result of experiment show that the proposed method can improve the data rate and reduce bit error rate. As the signal to noise ratio of the transmitter changing from 2dB to 30dB, the bit error rate improvement factor of the receiver is changed from 0.22% to 13.69%. |