Research On Pilot Contamination Elimination And PAPR Reduction In Wireless Communication Systems

The precipitously increasing cost and scarcity of the spectrum have forced all the current and the future wireless communication systems to be optimized for spectral efficiency to meet the expeditious increase in the wireless data traffic.Irn the recent years,several spectral efficient techniques are developed,including massive multiple-input multiple-output(M-MIMO)systems and multi-carrier transmission techniques.The key distinguishing feature of M-MIMO from conventional MIMO is the deployment of very large number of antennas at base stations(BSs),which offers favorable propagation conditions,high data rates,enhanced link reliability,large coverage,and good energy efficiency.However,the performance of M-MIMO systems depends on the accuracy of channel estimation,which is often contaminated due to the use of non-orthogonal pilot sequences in the adjacent cells.Multi-earrier transmission technique has several advantages over the single-carrier scheme,including high rate transmission,frequency diversity,and small inter-symbol interference in the multipath fading channel.Therefore,it is widely used in several telecommunication standards(e.g.,ADSL,DAB,and WLAN).However,a major disadvantage of multi-carrier transmission schemes is the high peak-to-average power ratio(PAPR),which results in in-band distortion that degrades the bit error rate,and out-of-band radiation that interferes with adjacent frequency bands.To address these performance limiting factors of M-MIMO and multi-carrier transmission,the pilot contamination elimination and PAPR reduction are studied in this dissertation,and the main contributions are as follows:(1)Channel state information in time division duplex(TDD)M-MIMO systems is degraded due to pilot contamination that results in inter-cell interference.A pilot contamination elimination scheme for multi-cell TDD M-MIMO systems is proposed.The proposed scheme employs Zadoff-Chu sequences(ZCs)as uplink pilot sequences and implements distinct orthogonal variable spreading factor(OVSF)code rows at each BS.The set of ZCs at each BS is multiplied element-wise with the BS-specific OVSF code row to generate the orthogonality among the sets of ZCs across the neighboring cells.Then,a mobile station selects randomly one of these multiplied ZCs from a given set and transmits it on the randorm access channel at the commencement of coherence interval.The proposed scheme can promise uncontaminated channel state information and interference free downlink transmission,thus achieves significant downlink sum-rate gains without requiring the knowledge of the channels' second-order statistics and without increasing the training overhead.(2)Irn multi-cell TDD M-MIMO systems,the pilot sequences are reused by mobile stations in different cells due to limited coherence time,which causes pilot contamination and then degrades reverse-link and forward-link channel estimations.To design a TDD M-MIMO system properly under a practical multi-cell environment,it is necessary to assess the achievable rate of such systems.A performance assessment of multi-cell TDD M-MIMO forward-link system with linear precoders is proposed.The reverse-link and forward-link channels are first estimated by the BSs using least square channel estimator.Then,the closed-form approximate achievable forward-link sum-rate expressions of M-MIMO system with matched-filter(MF)and zero-forcing(ZF)precoders are derived.Next,a number of important implications for multi-cell TDD M-MIMO systems are obtained from the closed-form expressions.Finally,the performances of the achievable forward-link sum-rate for MF and ZF precoders are compared with respect to forward-link transmission signal-to-noise ratio(SNR),the number of BS antennas,cell size,frame length,reverse-link pilot SNR and the pilot length.The performance comparison results corroborate that MF precoding has better sum-rate performance than the ZF precoding.(3)A multi-carrier transmission,including multi-carrier code division multiple access(MC-CDMA),has a high PAPR,which results in nonlinear distortion and deteriorative system performance.To reduce the PAPR,an error control code selective mapping(SLM)method for MC-CDMA systems is proposed.The proposed method is based on SLM method and employs interleaver and error control coding.The 2n different candidate data sequences of an original data sequence,all representing the same information as the original data sequence,are first generated by adding,n-tuple PAPR control bits to it,then processed by the inter-leaver and error control coding.The error control codes,including convolutional,cyclic,Golay,Hamming and low-density parity-check codes,provide error detection and correction of coded data,which makes them a desirable choice for PAPR reduction.The proposed method can achieve significant PAPR reduction as well as better power efficiency,and also avoid the need for the side-information transmission.