Cognitive Radio Technology Application To Solve The Problem Of Low Spectrum Utilization In Wireless Communication
With the increase in the number of wireless communication users, the limited resources available for signal transmission become more and more scarce, 70% of the radio frequency band is not effectively used according to the current fixed allocation spectrum transmission strategy. In the face of the growing shortage of spectrum resources and the low utilization rate, Dr. Mitola proposed a new technology for cognitive radio, a feasible solution is to achieve more efficient access to the free frequency band by allocating, the free spectrum of the specific time and space to the unauthorized user by sensing the spectrum environment and adaptively learning of the surrounding environment interaction information spectrum sharing.
Due to restrictions and reduce the occurrence of the collision, so it can be more reasonable and effective use. Cognitive radio is a crucial technology to solve the problem of low spectrum utilization in wireless communication. The wireless communication device having the cognitive function has an opportunity to access free spectrum without disturbing the communication of the authorized user. In each time slot, the secondary user must sense the channel state assigned to the authorized user in a short period of time, and access the time slot for data transmission when the perceptual result of the channel state is idle, to avoid interference with authorized users. The channel state prediction method mainly includes the spectrum prediction method based on the Hidden Markov model. Spectrum prediction method based on Auto Regressive model. Spectrum prediction method based on neural network and so on. Real-time analysis and prediction of band occupancy and idle state by using Markov model, realizing channel capacity estimation, and verifying the accuracy of experiment simulation. By adjusting the embedding strategy, the optimal embedding dimension and the delay variable of the model are judged according to the optimal prediction error. Compared with the system variables in the traditional auto regressive model, the improved method can improve the prediction accuracy.
The radial basis function neural network algorithm of K-means clustering is studied to predict the spectrum of the future spectrum based on the previous spectral state information, and the energy consumption of the collision is reduced and the energy consumption of the perceptual process is reduced. With the spectrum prediction, the SUs have high opportunity of selecting idle channels for sensing. The channels are sensed which are predicted to be idle based on spectrum prediction mechanisms, while the channels predicted to be busy are not sensed. So the probability of the SU idle channel sensing is significantly improved. SU's throughput is enhanced through spectrum prediction. With the spectrum prediction, the SUs can forecast the channel states and pre-known the time when PUs return. The SUs can remove before PUs return to the channels.
The performance of the whole cognitive radio network (CRN) can be improved by the probability increasing of selecting idle channel for sensing. In this paper, we propose a new cooperative spectrum prediction mechanism to improve the prediction accuracy with fusion the SU prediction results. In high traffic CRN, the true channel state for SUs are not visible, called hidden state. SUs acquire the channels states through sensing, called observation state. SUs may predict the state of a channel based on the past sensing results and select the channel predicted to be idle for sensing.
For cognitive radio spectrum prediction, the wavelet theory and fuzzy logic and neural network are combined to propose a spectrum forecasting method based on fuzzy wavelet neural network. The structure of fuzzy wavelet neural network model is introduced, and the complexity of the model are studied. Optimize the effectiveness of predictive results, improve the accuracy of spectrum access in real time, and decrease unnecessary consumption.
