Machine Learning Spectrum Sensing in Cognitive Radio
Budget: $30 – $250 USD
Simulation based project to investigate machine learning techniques for spectrum sensing in Cognitive radio networks with moderate complexity and best performance metric.
The idea is to allow more license users to use the spectrum for a certain amount of time for optimum spectrum usage because 80% of the time the spectrum is not used effectively,most of it remains idle.So the sensing will allow for unlicensed users to utilize the radio frequency spectrum for transmission.The sensing will eliminate spectrum overcrowding which causes interference.
The secondary users and primary users are not allowed to co-exist.The secondary users must sense the spectrum to check idle spaces in the spectrum to occupy.Then they can use it until the primary/licensed user occupies the same space and they have to vacate the spectrum to give priority to PU.Hence the SU has to constantly sense the spectrum to detect presence of the PU.
The idea is to allow more license users to use the spectrum for a certain amount of time for optimum spectrum usage because 80% of the time the spectrum is not used effectively,most of it remains idle.So the sensing will allow for unlicensed users to utilize the radio frequency spectrum for transmission.The sensing will eliminate spectrum overcrowding which causes interference.
The secondary users and primary users are not allowed to co-exist.The secondary users must sense the spectrum to check idle spaces in the spectrum to occupy.Then they can use it until the primary/licensed user occupies the same space and they have to vacate the spectrum to give priority to PU.Hence the SU has to constantly sense the spectrum to detect presence of the PU.
Related categories:
Matlab and Mathematica
Machine Learning (ML)
Telecommunications Engineering
Deep Learning