Research On Secrecy Capacity Enhancement And Code Design For MIMO Wireless Communications

The Physical Layer Security(PLS) technology can ensure the perfect security of information transmission. The good characteristics of PLS include ubiquity, computability, and facile realization. For the above reasons, the research about PLS has become the focus in the field of secure communications. PLS technology maybe one of the key technologies in the next-generation mobile communiction systems, especially, in military communication system. In PLS technology, the core issue is how to calculate and improve the secrecy capacity, as well as the code design in Multiple Input Multiple Output(MIMO) technology. The research priorities of this paper are how to improve the secrecy capacity in MIMO wiretap channel and how to design the correlating codes near the secrecy capacity. The main innovations are as follows:1. If the quality of channel between the transmitter(Alice) and the eavesdropper(Eve) is superior to the quality of channel between the legitimate receiver(Bob) and Alice, or if the antenna number of receiver is more than the antenna number of transmitter, the integrated signal-to-artificial noise(ISAN) design can inprove the secrecy capacity in MIMO wiretap channel, meanwhile, ensure the positive secrecy capacity. The paper has researched the above issues from two angles: Multiple Input Single Output Single antenna Eavesdroppers(MISOSE) and Multiple Input Multiple Output Multiple antenna Eavesdroppers(MIMOME). The ISAN codes have been designed respectively in MISOSE and MIMOME. Alice selects the eigenvector with maximum eigenvalue of covariance matrix of the legitimate receive channel as the signal vector; the artificial noise vector is the linear combination of the other eigenvectors; the signal vector and the artificial noise vector are mutually orthogonal, and without interference.2. The random complex field(RCF) coding has been applied in MIMO wiretap channel. The RCF code has been designed from two perspectives: Eve knows the RCF code matrix or not. If Eve does not know the RCF code matrix, the information precoded by RCF is transmitted directly; if Eve knows the RCF code matrix, the ISAN-RCF is applied and used in MIMO wiretap channel in which the eavesdropper is passive. The results show that if Eve does not know the RCF code matrix, the receiver error rate can always be kept on 0.5, meanwhile, Eve cannot get any information about source; if Eve knows the RCF code matrix, ISAN can deteriorate the wiretap channel, and ensure the quality of channel between Bob and Alice being superior to the quality of channel between Eve and Alice, as a result, the secrecy capacity will always be positive. Because of the feature of full diversity and code rates, the RCF code can help Bob to counteract the channel fading effectively, at the same time, the safety performance of transmission system will be improved. The simulation results show that ISAN-RCF can ensure the reliability and security of data transmission. The Low Density Parity Check(LDPC) is proposed in MIMO wiretap channel. LDPC based on scrambling code can improve the security of information transmission effectively. Because of the characteristic of the error diffusion, the random scrambling code can diffuse the error rate of LDPC to 0.5 when the signal to noise ratio(SNR) of receiver is low. The security gap is selected as the security index in above technology. Compared with LDPC without scrambling code, LDPC with random scrambling coding can reduce the security gap; therefore, the BER of eavesdropper will be more likely to be near 0.5. Based on the theory above, luby transform(LT) rateless code is designed as the scrambling code. When the ordinary random scrambling code is compared with the LT rateless code, there is the obvious difference. If the coding length increases about 10%, the LT rateless code can decrease the encoding complexity greatly, and ensure the security of information transmission at the same time.