Intercepting wireless signal has been of a great interest for law enforcement and military users as it provides precious information about illegal activity or the enemy intention. To intercept the signal, first, the signal existence must be detected. In case of using the spread spectrum, signal detection becomes more challenging since the spread spectrum system could bury the signal below the noise level to hide it. Therefore, a detection system that detects the signal existence even in a low signal to noise ratio (SNR) regime is needed to detect the hidden signal. In the unintended reception scenario, moreover, the receiver does not have any prior knowledge about the signal activity or the detection environment. Hence, the detection system should also operate without any prior knowledge about signal activity in the targeted frequency band or the signal statistics.
This thesis develops an algorithm that utilizes the cyclostationary detector to detect the signal in low SNR regime. Furthermore, the cyclic frequency mismatch concept is used to determine the existence of the signal without any prior assumption about the signal activity. The bootstrap technique is then used to determine the threshold of the algorithm regardless of the noise distribution. To prove the applicability in real time situation, the proposed algorithm is implemented on the SoC FPGA Xilinx board.
The research in this thesis offers the potentials of enhancing the detection performance while reducing the complexity of achieving this enhancement. The numerical results vali- date the efficacy of the proposed detection system in improving the detection performance compared to its counterparts. Finally, the FPGA implementation validates its applicability in real-world usage.