Research On Generalized Frequency Division Multiplexing Modulation Technology For 5G

The 5th generation of mobile communications is moving in the direction of network diversification,broadbandization,integration,and intelligence,and has imposed higher requirements on the performance of latency,transmission rate,connection number,and frequency band utilization.Although orthogonal frequency division multiplexing(OFDM)has been widely used in 4G,OFDM may not be the best choice for 5G physical layer in the face of new scenarios and new demands in future communications.In order to effectively support these services and related scenarios,5G should use a more flexible waveform design as a breakthrough.This thesis focuses on the new multi-carrier technology,and studies the synchronization techniques,the estimation and compensation of RF non-ideal factors,and the channel estimation techniques of the generalized frequency division multiplexing(GFDM).The main work of this thesis is as follows:Firstly,the modulation and demodulation process of GFDM is studied in detail,and comparison of bit error rate,peak-to-average power ratio and power spectral density are conducted betwwen OFDM and GFDM.The basic principles,advantages and disadvantages of filter bank multicarrier(FBMC),universal filtered multicarrier(UFMC)and filtered orthogonal frequency division multiplexing(F-OFDM)are analyzed.Simulation results show that GFDM has a lower out-of-band power leakage,while spectrum efficiency has also been improved,and it can better meet the needs of future mobile communications.In addition,GFDM data block structure and pulse shaping provide more degrees of freedom,resulting in flexible allocation of timefrequency resources and support for diversified services.Secondly,the synchronization problem in GFDM is studied,and a low-complexity synchronization scheme is proposed.Because of the non orthogonality of GFDM,the correlation of synchronization sequences is impaired,so existing schemes require additional processing at the receiving end,resulting in high complexity.In this thesis,by analyzing the influence of GFDM modulation waveform,a special synchronization sequence is designed,and a corresponding low complexity synchronization algorithm is proposed.Simulation results show that the proposed scheme can effectively avoid data interference and ensure excellent synchronization performance.In addition,another advantage of this algorithm is that it greatly reduces the complexity of the synchronization estimation,so it can be applied to low-power scenarios such as IoT and MTC.Thirdly,the GFDM channel estimation problem is studied,and a precoding based pilotaided channel estimation scheme is proposed.Because GFDM exists self-induced interference,it can't get the pure pilot like OFDM at the receiving end,so the interference caused by the data needs to be eliminated in the scheme design.By precoding at the transmitter and placing the pilots in a special way,the interference-free equivalent pilot can be obtained at the receiver,and the channel estimation can be realized.By combining GFDM with multi-antenna technology,the proposed channel estimation scheme is extended to MIMO-GFDM.The simulation results show that the proposed scheme can achieve the same channel estimation performance as OFDM,and the performance is not affected by the filter roll-off factor.In addition,the precoding at the transmitter can effectively reduce the PAPR of the transmitted signal.Finally,the phase noise and IQ imbalance estimation and compensation schemes are studied.Two phase noise suppression schemes for GFDM and IQ imbalance and channel joint estimation algorithms are proposed.Due to the special data block structure and the influence of pulse shaping,GFDM cannot use the traditional ICI mapping.For this reason,a new mapping scheme is proposed to implement GFDM phase noise suppression.The precoding based pilotassisted phase noise suppression scheme uses precoding at the transmitter,and utilize the block cycle characteristics of GFDM modulation matrix to reconstruct the equivalent pilot without data interference at the receiver,then the phase noise estimation is realized.According to a special training sequence,the proposed IQ imbalance and channel joint estimation algorithm can separate the channel and IQ imbalance parameters on odd and even carriers,and obtain independent parameter estimation.The simulation results show that the proposed scheme has good performance.In addition,the proposed scheme requires only one training symbol and simple linear operation,which reduces system overhead and complexity and has more implementation value.