Channel Estimation For Massive MIMO And OFDM Wireless Communication Systems

Recently,all kinds of wireless applications make the wireless data demand rapidly growing.Future mobile communication systems are expected to further improve the spectral efficiency and support much more mobile terminals for the ambition of connected everything.In order to accomplish the above goals,the massive multiple-input multiple-output(MIMO)and the full-duplex operation are currently the promising candidates for the physical-layer techniques of the fifth-generation(5G)mobile communication system,while the orthogonal frequency division multiplexing(OFDM)technique of the current fourth-generation(4G)mobile communication system may well be inherited by the 5G mobile communi-cation system.In order to perform data transmission on a complex and dynamic wireless channel,future communication systems employing the above physical-layer techniques need to acquire wireless channel information,upon which we will investigate the acquisition of channel information for the massive MIMO and OFDM wireless communications.Firstly,we investigate channel estimation for OFDM systems and focus on transform domain techniques that include discrete cosine transform(DCT)and discrete Fourier transform(DFT).Transform domain channel estimators suffer from channel leakage,which significantly degrades their estimation performance;in order to solve this problem,we use partial minimum mean square error(MMSE)filtering to estimate channel leakage and then recover the channel leakage compo-nent.More precisely,for DCT-based estimators,we derive the channel impulse response in the DCT domain,based on which we suggest a threshold to identi-fy the energy-concentrated region.Further,we propose the leakage estimation outside the energy-concentrated region,and reduce its implementation complex-ity.For DFT-based estimators,we also derive the channel impulse response in the DFT domain,based on which we propose a low-complexity leakage estimator that approximates the optimal MMSE estimator.Numerical results validate the superior estimation performance of the proposed channel estimators.Then,we investigate a pilot scheduling algorithm for the massive MIMO and OFDM systems,which makes that the served users can efficiently acquire channel information.We propose to assign pilot resource by utilizing channel character-istics in the two dimensions that include the delay and spatial angle domains,where the channel characteristic in the delay domain is utilized by properly de-signing circularly shifted pilots.In the presence of contamination caused by the pilot reuse,we analyze the asymptotic behavior and design an enhanced match filter(E-MF)precoder.For both the simple match filter(MF)precoder and the sophisticated E-MF precoder,we derive achievable rates with both the limited and unlimited numbers of channel paths in the presence of pilot contamination,which are metrics for quality of service(QoS).Further,according to the analysis under the assumption of homogeneous users,the number of QoS guaranteed users proves to be maximized when an equispaced property in the angle domain is satisfied.The theoretical result facilitates a low-complexity user scheduling algorithm that can select out the maximum number of served users with guaranteed QoS.Nu-merical results confirm our analysis and validate the performance of the proposed scheduling algorithm.Furthermore,we investigate channel estimation for a two-layer wireless com-munication system consisting of a massive MIMO base station and a full-duplex relay.Under the full-duplex operation,the communication system suffers from self-interference and direct-link interference,which render channel estimation very challenging.In the presence of the above two kinds of interference,we address two channel estimation problems for such systems:individual estimation,where both the base station and the relay estimate their respective channels,and cas-caded estimation,where only the base station estimates the cascaded two-hop channel.For the base station,we propose an estimator that exploits the sparsity and slowly-varying nature of the channel in the beam domain.Then,we analyze the probability of correctly distinguishing the desired and interfering direct-link channels.For the relay,we present an estimator that simultaneously estimates both the source-to-relay and self-interference channels based on the expectation-maximization(EM)algorithm.The performance of this estimator is also analyzed.Numerical results demonstrate the excellent performance of the proposed channel estimators and corroborate the analysis results.At last,we investigate beam-domain channel estimation for frequency-division duplex(FDD)massive MIMO systems.In the macrocellular setup,the massive MIMO channel exhibits the sparsity in the beam domain due to the limited s-cattering.Therefore,for beam-domain channel estimation,we focus on utilizing a threshold-based method to extract effective channel taps of the sparse channel.Firstly,for threshold-based estimation,we derive its closed-form mean-squared er-ror(MSE)expression,based on which we propose an optimal threshold.Further,we propose a simplified threshold that is only related to noise variance.Fur-thermore,we present a threshold to identify the common support,with which an algorithm is designed to improve the estimation accuracy.As for channel feedback,we suggest to feed back only significant elements in the beam domain.Numer-ical results validate our derived MSE expression and demonstrate the superior performance of proposed threshold-based estimators.