Abstract: For the high complexity of antenna selection algorithm based on optimal secrecy capacity,an improved antenna selection algorithm with low complexity based on the difference of column norm squared is proposed.First,the analytic formula of secrecy capacity is simplified by normalizing the fixed quantity with comparison of the difference between legitimate channel and eavesdropper channel,which is expanded and expressed by channel coefficient.Then the difference of channel coefficient square is traversed and sorted,and the antenna combination with the largest difference of channel norm square is selected.Meanwhile,combining the algorithm with the artificial noise,which is designed at the remaining legitimate channel null space,can obtain the optimal secrecy capacity.The simulation results show that compared with tranditional algorithm,the proposed algorithm can achieve optimal secrecy capacity with low complexity.In addition,the bit error rate of legitimate receiver is maintained and the bit error rate of eavesdropper is restricted maximumly.Meanwhile,the safety performance of the system is enhanced effectively.

Key words: Artificial noise vector forming, Low complexity, Norm difference antenna selection, Optimal secrecy capacity, Optimal sorting algorithm, Spatial modulation