论文部分内容阅读
Ultra wideband impulse radio(UWB-IR) has important applications in high data rate communications,high precision positioning and penetrating target detection.Combined with cognitive radio(CR),UWB can also greatly alleviate the current spectrum scarcity.However,the existing UWB waveforms are faced with two challenges.The spectrum utilization of UWB signal is rather low,which reduces the communication reliability.Meanwhile,it appears difficult for UWB-IR to avoid spectrum collision with other incumbent services.So UWB will inevitably interfere other legal wireless systems.Motivated by these considerations,we model the UWB waveform design into an interpolation problem.Based on radial basis function neural network,a novel UWB pulse is presented in this paper.This UWB pulse is then realized though reasonable simplification on the already proposed theoretical network.The designed UWB pulses can adaptively track the external spectral environment where UWB devices work and reconfigure its emission spectrum according to the state of primer users without any modification on hardware.The frequency utilization is maximized when there is no active primary user.The adequate spectrum avoidance has also been achieved without much spectrum utilization degradation within the non-primary bands.Simulation and analysis show that our UWB-IR pulse is much superior to the other existing UWB signals in the frequency utilization,spectrum avoidance and in the transmission performance.
Ultra wideband impulse radio (UWB-IR) has important applications in high data rate communications, high precision positioning and penetrating target detection. Cined with cognitive radio (CR), UWB can also greatly reduce the current spectrum scarcity. Despite, the existing UWB waveforms are faced with two challenges. The spectrum utilization of UWB signal is rather low, which reduces the communication reliability. Meanwhile, it appears difficult for UWB-IR to avoid spectrum collision with other incumbent services. So UWB will inevitably interfere other legal wireless systems. Motivated by these considerations, we model the UWB waveform design into an interpolation problem. Based on radial basis function neural network, a novel UWB pulse is presented in this paper. This UWB pulse is then realized though. The designed UWB pulses can adaptively track the external spectral environment where UWB devices work and reconfigure its emiss ion spectrum according to the state of primer users without any modification on hardware. The frequency utilization is maximized when there is no active primary user. The adequate spectrum avoidance has also been achieved without much spectrum utilization degradation within the non-primary band. Simulation and analysis show that our UWB-IR pulse is much superior to the other existing existing UWB signals in the frequency utilization, spectrum avoidance and in the transmission performance.