To search, Click below search items.


All Published Papers Search Service


Deep Learning Enabled Spectrum Sensing Radio for Opportunistic Usage


Muneeb Aalam Khan, Aamir Zeb Shaikh, Shabbar Naqv, Saima Khadim, Talat Altaf


Vol. 19  No. 11  pp. 179-183


Radio spectrum is becoming overcrowded especially after the roll out of social media applications, live broadcasting, video calling and related applications. Additionally, under the umbrella of 5G and 5G+ wireless standards, these applications will require higher speed and higher bandwidth. Under the static allocation of frequencies regime, it is really difficult to accommodate the next generation users using the classic wireless technology. Thus, the current conditions in RF usage suggests the use of Cognitive Radio technology for using the RF spectrum in opportunistic fashion. This technology advocates the use of spectrum fashion such that the primary users are not faced with harmful interference. This paper presents the implementation of Deep Learning algorithm i.e. ADAM and Levenberg Merquardt algorithm (LMA) for prediction of spectral holes into Karachi city. Root Mean Square Error (RMSE) values computed for 1.9 GHz comes out to be 0.00063683 and for 3.4GHz the RMSE becomes 0.0010649, whereas RMSE for LMA is 2.91183e-03 at 1.9 GHz, appearing at epoch 5 and 6.0607e-3 at epoch 5 for 3.4GHz RMSE


Deep Learning (DL), Machine learning (ML), Cognitive Radio (CR), Levenberg Merquardt Algorithm (LMA), ADAM