Accurate Spectrum Prediction Based on Joint LSTM with CNN toward Spectrum Sharing
Liang Zhang, Min Jia
Abstract
In cognitive radio context, spectrum sensing is the vital technique for the cognitive users to acquire the spectrum of frequency band. It requires cognitive users to determine the usage state of the spectrum through spectrum perception and access the spectrum for communication when the spectrum is idle. However, the accuracy of single radio spectrum is low. Thus, the proposed algorithm adopts the joint Long and Short Term Memory (LSTM) and Convolutional Neural Network (CNN) as the prediction model for a combined design, which is used for spectrum prediction under multi-channel toward spectrum sharing. The simulation results show that the sensing accuracy of the radio spectrum is significantly improved.
Topics & Concepts
Cognitive radioComputer scienceSpectrum (functional analysis)Joint (building)Context (archaeology)Radio spectrumChannel (broadcasting)Convolutional neural networkArtificial intelligenceSpeech recognitionWirelessComputer networkTelecommunicationsEngineeringPaleontologyBiologyQuantum mechanicsArchitectural engineeringPhysicsCognitive Radio Networks and Spectrum SensingAdvanced Adaptive Filtering TechniquesBlind Source Separation Techniques