Research On Multiple-mode MIMO Systems Based On Index Modulation Technology

In recent years,the field of wireless communication has developed rapidly,and various new technologies have emerged.With the continuous deepening of research,how to improve the efficiency of spectrum resource utilization and how to improve the accuracy of signal detection has become one of the key issues in the communication field.Index modulation(IM)technology has received widespread attention from academia and industry because of its ability to use additional information bits to transmit information without consuming energy and increasing communication links.In addition,multiple input multiple output(MIMO)channels can greatly increase the channel capacity and have high application value.In MIMO systems,using the singular value decomposition(SVD)theory,several equivalent virtual parallel channels can be obtained.Recently,a novel modulation technology named generalized virtual spatial modulation(GVSM)was proposed.By carrying additional information bits with the indices of the virtual parallel channels,the bit error rate performance of GVSM system is better than that of traditional MIMO systems.However,since only a few virtual channels are activated in each time slot,the spatial efficiency of GVSM system is relatively low.In order to solve the above problems,conducts three main tasks which are listed as follows.First of all,this thesis proposes a novel technology termed multiple-mode MIMO with index modulation(MM-MIMO-IM).By activating all virtual parallel MIMO channels,this scheme can transmit modulation systems from different signed constellations,and by utilizing the permutation of constellations,the additional information bits can be carried,which improves the spatial efficiency in a large scale.Simultaneously,in order to improve the diversity gain of the proposed MM-MIMO-IM scheme,coordinate interleaved multiple-mode MIMO with index modulation(CI-MM-MIMO-IM)scheme is proposed.As proven in this thesis,the bit error rate(BER)performance of MM-MIMO-IM and CI-MM-MIMO-IM are better than that of traditional MIMO schemes and GVSM.Secondly,in order to further improve the spectral efficiency of the MM-MIMO-IM system,in-phase and quadrature extension of MM-MIMO-IM(MM-MIMO-IM-IQ)scheme is proposed.The basic idea of this scheme is to implement MM-MIMO-IM independently and in parallel on the in-phase and quadrature branches of the channels,which can further improve the spectral efficiency in double scale.Simultaneously,the BERs of MM-MIMO-IM and MM-MIMO-IM-IQ are strictly deduced,and the computer simulation verifications are carried out.Finally,in order to further improve the detection accuracy of the received signal,a deep learning(DL)based detector termed as Deep IM is proposed.The BER performance for both traditional maximum likelihood(ML)detector and Deep IM detector are compared.Simulation results show that Deep IM detector outperforms ML detector for the case of imperfect channel estimation and has nearly the same performance for the case of perfect channel estimation.