چكيده به لاتين
Massive multiple-input multiple-output is one of the prominent technologies for the fifth generation and beyond wireless networks due to its amelioration of spectral and energy efficiencies. However, achieving a good performance from massive MIMO systems highly depends on the accuracy of channel state information which is usually acquired through pilot sequences. Any significant decrease in channel estimation accuracy can lead to major performance loss. Specifically, since wireless systems are prone to security attacks due to their broadcasting nature, adversarial devices can substantially deteriorate massive MIMO systems' spectral efficiency by transmitting jamming signals in both channel training and data transmission phases. Thus, detecting and suppressing such attacks is of great importance and should be addressed in designing massive MIMO networks.
To make the jamming attacks futile, the base station should first detect that there is a jammer in the system then take the proper counter-measures. In this thesis, we utilize directional information of received signals to propose a new jamming detection and suppression technique. The directional information is considered because massive MIMO systems have a great angular resolution due to exploiting a large number of antennas in the base station. At first, we use received signals in the channel training phase to estimate directions from which each pilot is received. Since a proportion of pilots are also received along with jammer's directions, we suggest a method to detect jammer and its directions. Furthermore, using the information obtained in the previous stages, we propose a technique for improving channel estimation precision and designing combining vectors which can suppress jammer's attack on uplink spectral efficiency. Simulation results illustrate that the proposed schemes are able to effectively detect jammer and reject its impact on the system's performance.
