Abstract:

An anti-jamming algorithm based on single channel semi-blind source separation by exploiting the periodicity of the spread spectrum code for direct sequence spread spectrum (DSSS) communication is presented to overcome the technical bottleneck of traditional anti-jamming methods based on blind source separation (BSS) which need more receive antennas. Firstly, the single channel of DSSS communication’s multipath slow fading channel model is transformed into a multichannel instantaneous mixture model by symbol level oversampling based on the periodicity of the spread spectrum code. Then, training sequence is introduced as a priori information to the cost function of BSS in the constraints form to extract the expected user data from the mixed signal directly, which finally achieves the goal of antijamming. Simulation results show that in the condition of jamming such as the multiple access jamming and single tone/ multi-tone jamming, when the signal to jamming ratio fixes to -60 dB. with the increasing of the normalized signal to noise ratio, the bit error rate of the unseparated signal is almost constant and fluctuates in the vicinity of 10e-0.3. while the bit error rate of the proposed algorithm gradually decreases, and finally stays between 10e-4~10e-5.