The Research On Key Technologies Of Multi-carrier And Low-Cost Distributed Antenna Systems

It is well known that multicarrier technology and Multiple Input Multiple Output(MIMO)are two critical technology in the 5th Generation Mobile Communication Systems(5G),and rate and power consumption are two important performance index.For multicarrier technology,because of huge equipment connections,it calls for a multicarrier technology which is friendly in demand of synchronization and robustness to Carrier Frequency Offset(CFO).Benefit from the filtering by multiple filter banks,Universal Filtered Multi-Carrier(UFMC)system reduce the sidelobe power,so that it can better confront the CFO than Orthogonal Frequency Division Multiplexing(OFDM).Still,the CFO make a bad influence to the UFMC system,besides,nowadays,there are not enough efforts on the estimation of CFO in UFMC system.To the already exist technology OFDM,considering the accuracy of channel estimation concerns the performance of coherent reception performance.it is unavoidable that inference will be induced,while using exist method of add the pilots to the data.So,the way to eliminate the interference become a research emphasis,newer method Partial-Data Superimposed Training(PDST)can degrade the interference from data to pilot and get a better performance without complete removing the data added at the same time by introduced a new interference control factor.But through a deep research,we find out that the interference can be further degraded if the optimal positions of pilots added are take into consideration.MIMO technology has a huge superiority in improving the system capacity.When using the distributed deployment in MIMO system,the capacity can be further improved.However,with the continuously increasing of antennas,system energy consumption and cost become very high with rapidly increasing RF links.So,it takes a very important practical significance in reducing the energy consumption and cost using technology such as low resolution ADC.After introducing low resolution ADCs in distributed system,the performance of sum-rate and energy efficiency performance in new scenario should be derived again.Upon the analysis above,the research in the thesis focus on multicarrier technology and MIMO system and its main innovation points are as follows:Firstly,the thesis estimates the integer part of CFO in UFMC system and proposes a low complexity estimation method.Utilizing the property of Constant Amplitude Zero Auto Correlation(CAZAC)sequence in autocorrelation and the nature of correlation bandwidth,at the receiver,through correlation operation between the CAZAC sequence and the receiving signal and finding out the maximization correlation value,we get the value of CFO.In order to estimate the CFO,we assume that the training sequence is consist of three the same CAZAC sequences in UFMC system.At the receiver,after getting receiving signal considering CFO,we correlate the receiving signal in subcarriers corresponding to those occupied by the training sequence with CAZAC sequence locally generated.Then,by working out the correlation values corresponding to different shifting length and comparing each other,the maximization one,namely CFO value,is get,making use of the character of autocorrelation of CAZAC sequence and coherent bandwidth.As a further research,to lower the impact of corrupting autocorrelation property of CAZAC sequence coming from the slow channel change in reality,the thesis proposes a method of using more segments in training sequence to address this problem.Hereafter,in full consideration of general non-integer CFO estimation,towards fractional part CFO estimation,the already existing method is applied.After compensating the signal with the entire CFO estimation value,we use this processed signal into channel estimation.Analysis shows that the method proposed by our thesis is much lower to compared method in computation complexity.At the same time,it can reach almost the same performance in low Signal to Noise Ratio(SNR)regime,even better under channel with small attenuation parameter,which all confirmed by simulation results.Secondly,the thesis further studies interference reducing in OFDM system,which generates from superposing pilots with data.The proposed method improves the exist PDST method.One characteristic of PDST method is that a new factor is induced to control the interference level besides the data and pilots power controlled when adding the precoded data to pilots.Based on PDST method,considering equal insert of pilots in data,the method of the thesis utilizes the character of dramatic power change of precoding signal in frequency domain to find the series position with lowest power after calculating the power of precoding data in a series evenly spaced positions every time by shifting signal one by one.Then,by adding the pilots in these positions,and getting signal with pilots adding in new positions,the receiving signal is obtained after passing channel.With Least Squares(LS)channel estimation and corresponding signal detection and iteration,the final signal later after detection is acquired.The complexity analysis and simulation results demonstrate the proposed method reaches much higher Symbol Error Rate(SER)performance and improves the accurate of channel estimation in the same order of calculation complex compared to PDST method.Thirdly,the thesis studies the achievable rate and energy efficiency in uplink transmission of distributed MIMO in single cell using mixed Analog-to-Digital Converter(ADC)and Bussgang quantization model and derive close expression of them.The specific procedure is divided into two phases.In training phase,all Remote Antenna Unit(RAU)adopts the mixed structure,which means some of RAUs utilizing the high resolution ADCs while the others low resolution ADCs.By transmitting the pilots signal and passing channel,the receiving signal is quantized using Bussgang quantization model.Then,taking the Linear Minimum Mean-Squared Error(LMMSE)estimation with the help of arcsin criterion to quantized signal,the estimation channel is obtained.In data transmitting phase,the data signal is quantized under Bussgang model after transmitting and receiving.Following with Maximal Ratio Combining(MRC)or Zero Forcing(ZF)detection,we get the Signal to Interference and Noise(SINR)expression by definition.Then,solving the expectation in rate expression to achieve the rate analysis expression of current user,and summing the rate over all users,finally,we obtain the sum rate of systems.After further analyzing the power model,we also get the close expression of energy efficiency.The simulation results demonstrate the correctness of expressions mentioned above.Finally,the thesis studies the system achievable rate and energy efficiency in distributed Massive MIMO system of single cell using low resolution ADC and Digital-to-Analog Converter(DAC)in uplink and downlink respectively and co-frequency co-time full duplex(CCFD),and derives the asymptotically close expression of sum rate and energy efficiency utilizing the large dimensional random matrix theory.Towards downlink channel,original signal is transmitted after precoded under Regularized Zero Forcing(RZF)model and quantized by quantizer adopting Additive Quantization Noise Model(AQNM)model.Considering the interference and noise existing among users in downlink communication under full duplex,we get the receiving signal and preliminary expression of SINR in downlink.Solving the aforementioned expression by employing the large dimension random matrix theory to get the asymptotically expression of SINR and also based on the capacity formula,we acquire the asymptotically close expression of achievable sum rate in downlink of full duplex system.In uplink channel,transmitting signal is superposed by imperfect canceled interference from downlink BS antennas to uplink BS antennas in full duplex communication and noise.After quantized by AQNM for receiving signal,the corresponding preliminary expression of SINR is obtained.With the help of large dimensional random matrix theory,we get the asymptotically close expression of SINR,and further,asymptotically close achievable sum rate in uplink.After further analyzing the power model,we also get the asymptotically close expression of energy efficiency in uplink and downlink.Simulation results demonstrate the accuracy of asymptotically close expressions mentioned above.