Research On Cooperative Transmission And Channel Reciprocity In Wireless Communication Systems

In order to meet the increasing demands on higher rate and reliability of datatransmission，Multiple Input Multiple Output (MIMO) technique has been chosen asone of the indispensably basic techniques used in future wireless communicationsystems. The multiplexing gains and diversity gains obtained by MIMO techniqueprovide technical guarantee for higher data rate and reliability respectively. For thereason above, MIMO technique as well as some related techniques have been studiedand applied widely, including cooperative transmission and multi-antenna transmitpreprocessing.Cooperative transmission is the natural extend of MIMO technique in single-antennaenvironment. Usually multi-antennas are not available for a cheap and small-sized userterminal, whereas several single-antenna user terminals can share their antennas to forma virtual MIMO system to reap the various benefits of MIMO technique. However, it isdifferent from the conventional MIMO systems that the distributed antennas in acooperative transmission system can’t achieve perfect synchronization, so that manytechniques which are suitable for conventional MIMO systems can’t be directlyemployed. Therefore, it’s necessary to study the cooperative transmission inasynchronous environment so as to solve this problem.Another technique related to MIMO is multi-antennas transmit preprocessing. Withchannel state information (CSI) at transmitter, transmit preprocessing can significantlyimprove system performance. In FDD systems, the uplink and downlink use differentfrequency bands, which means no reciprocity between the two channels can be tookadvantage of, so that a mobile station (MS) can’t help sending CSI to base station (BS)via a feedback channel. It is not only a waste of system resource but also an increasingin system complexity. However, the uplink and downlink channels in a TDD systemshare the same frequency band, which means that there exists reciprocity between theproperties of the two channels. Thus, BS can perform transmit preprocessing accordingto the CSI obtained by means of uplink channel estimation rather than via a feedbackchannel, which is believed to be an inherent superiority of TDD systems. Unfortunately,various factors can damage the channel reciprocity; so that it’s a vital issue how tocompensate the influence resulted from such factors.This dissertation mainly researches on the problems of asynchronous cooperativetransmission and channel reciprocity mentioned above. The author’s main contributionsare as follows. 1. Considering non-feasibility of OSTBC in asynchronous environment, the dissertationproposes an asynchronous cooperative transmission scheme based on single-carrierblock transmission. In the scheme, the relay nodes segment the signal received from thesource node into several blocks and retransmit them based on single-carrier blocktransmission mode according to the predefined cooperation rules. For the multipleblocks of received signal, the destination node performs FFT respectively, separates theoverlapped ones among them by means of linear combination, and then performsfrequency domain equalization, IFFT and symbol decision to recover the original data.The proposed scheme can solves the problems of orthogonality damage and channeldispersion resulted from the asynchronization, and can achieve full diversity gains.Based on this idea, a generalized space-time cooperative transmission code withmodified OSTBC which is suitable for asynchronous environment is presented. It cansolve the problem of non-feasibility of OSTBC in asynchronous environment andachieve full diversity gains.2. In TDD systems with I/Q imbalance, the negative effect of I/Q imbalance will resultin the non-reciprocity between uplink and downlink channels, so as to degrade thesystem performance. Aiming at this problem, a compensation scheme is proposed,which is based on bidirectional channel estimations. Before starting formal datatransmission, BS and MS perform channel estimations to get the CSI of uplink anddownlink respectively, then MS send the downlink CSI to BS which is responsible forcalculating the calibration matrices used for BS and MS, and at last BS send MS’scalibration matrix to MS. In this way, BS and MS can use calibration matrices fortransmit preprocessing to maintain the channel reciprocity.3. Time-variance property of wireless channel will lead to differences between uplinkand downlink CSI of TDD systems to damage the channel reciprocity. Aiming at thisproblem, a compensation method based on transform-domain detection and prediction isproposed. Firstly the channel frequency response (CFR) sequence of pilot channelsobtained by least square estimation is transformed into transform-domain by DFT, andthen significant components in transform-domain is picked up with minimumdescription length (MDL) criterion, and at last prediction is made for each component tocompensate channel time-variances. It is through transform domain processing that theproposed method effectively reduces the estimation and prediction errors, and henceremedies the system capacity loses. Meanwhile the proposed method greatly lowers thesystem complexity compared with the frequency domain prediction method. 4. Channel estimation error is another causation of channel non-reciprocity. InOFDM systems with virtual subcarriers and non-strictly equidistant pilots, seriousenergy leakage of channel impulse response (CIR) emerges when the pilots’ CFRsequence obtained by LS criterion is transformed into time domain using IFFT, so thatthe algorithms based on time domain processing including FFT interpolation can’t beused to suppress estimation error. Aiming at this problem, a low complexity channelestimation method is proposed to compensate the channel non-reciprocity caused byestimation error. Firstly the CFRs of virtual subcarriers are added by linear interpolation,then strictly equidistant subcarriers are chosen as new pilot channels for CIR recoveryby IFFT, and at last a threshold-based denoising is performed to the recovered CIRsequence to suppress the estimation error and compensate the channel non-reciprocity.All the proposed methods mentioned above are validated for their correctness andfeasibility by means of theoretical analysis and simulation experiments.