Chenhao Qi

Optimized Pilot Placement for Sparse Channel Estimation in OFDM Systems

Compressed sensing (CS) has recently been applied for pilot-aided sparse channel estimation. However, the design of the pilot placement has not been considered. In this letter, we propose a scheme using the modified discrete stochastic approximation to optimize the pilot placement in OFDM systems. The channel data is employed to offline search the near-optimal pilot placement before the transmission. Meanwhile we also get a criterion to select CS algorithms based on the mean squared error (MSE) minimization. Simulations using a sparse wireless channel model have validated the effectiveness of the proposed scheme, which is demonstrated to be much faster convergent and more efficient than the exhaustive search. It has been shown that substantial performance improvement can be achieved for OMP and YALL1 based channel estimation, where YALL1 is preferred.

Pilot Design for Sparse Channel Estimation in OFDM-Based Cognitive Radio Systems

In this correspondence, sparse channel estimation is first introduced in orthogonal frequency-division multiplexing (OFDM)-based cognitive radio systems. Based on the results of spectrum sensing, the pilot design is studied by minimizing the coherence of the dictionary matrix used for sparse recovery. Then, it is formulated as an optimal column selection problem where a table is generated and the indexes of the selected columns of the table form a pilot pattern. A novel scheme using constrained cross-entropy optimization is proposed to obtain an optimized pilot pattern, where it is modeled as an independent Bernoulli random process. The updating rule for the probability of each active subcarrier selected as a pilot subcarrier is derived. A projection method is proposed so that the number of pilots during the optimization is fixed. Simulation results verify the effectiveness of the proposed scheme and show that it can achieve 11.5% improvement in spectrum efficiency with the same channel estimation performance compared with the least squares (LS) channel estimation.

A Study of Deterministic Pilot Allocation for Sparse Channel Estimation in OFDM Systems

In this letter, we investigate the deterministic pilot allocation for sparse channel estimation in OFDM systems. Based on the rule of minimizing the coherence of the DFT submatrix, we derive that the pilot design according to the cyclic different set (CDS) is optimal. However, the CDS only exists for some specific number of OFDM subcarriers. For those cases where the CDS is unavailable, we propose a scheme using discrete stochastic approximation to obtain a near-optimal pilot pattern. Simulation results demonstrate that our scheme is much faster convergent and more efficient than the exhaustive search; and it has been shown that substantial improvement for channel estimation can be achieved.

Pilot Design Schemes for Sparse Channel Estimation in OFDM Systems

In this paper, we consider the pilot design based on the mutual incoherence property (MIP) for sparse channel estimation in orthogonal frequency-division multiplexing (OFDM) systems. With respect to the length of channel impulse response (CIR), we first derive a sufficient condition for the optimal pilot pattern generated from the cyclic different set (CDS). Since the CDS does not exist for most practical OFDM systems, we propose three pilot design schemes to obtain a near-optimal pilot pattern. The first two schemes, including stochastic sequential search (SSS) and stochastic parallel search (SPS), are based on the stochastic search. The third scheme called iterative group shrinkage (IGS) employs a tree-based searching structure and removes rows in a group instead of removing a single row at each step. We later extend our work to multiple-input-multiple-output (MIMO) systems and propose two schemes, i.e., sequential design scheme and joint design scheme. We also combine them to design the multiple orthogonal pilot patterns, i.e., using the sequential scheme for the first several transmit antennas and using the joint scheme to design the pilot pattern for the remaining transmit antennas. Simulation results show that the proposed SSS, SPS, and IGS converge much faster than the cross-entropy optimization and the exhaustive search and are thus more efficient. Moreover, SSS and SPS outperform IGS in terms of channel estimation performance.

A hybrid compressed sensing algorithm for sparse channel estimation in MIMO OFDM systems

Due to multipath delay spread and relatively high sampling rate in OFDM systems, the channel estimation is formulated as a sparse recovery problem, where a hybrid compressed sensing algorithm as subspace orthogonal matching pursuit (SOMP) is proposed. SOMP first identifies the channel sparsity and then iteratively refines the sparse recovery result, which essentially combines the advantages of orthogonal matching pursuit (OMP) and subspace pursuit (SP). Since SOMP still belongs to greedy algorithms, its computational complexity is in the same order as OMP. With frequency orthogonal random pilot placement, the technique is also extend to MIMO OFDM systems. Simulation results based on 3GPP spatial channel model (SCM) demonstrate that SOMP performs better than OMP, SP and interpolated least square (LS) in terms of normalized mean square error (NMSE).

Sparse channel estimation based on compressed sensing for massive MIMO systems

The sparse channel estimation which sufficiently exploits the inherent sparsity of wireless channels, is capable of improving the channel estimation performance with less pilot overhead. To reduce the pilot overhead in massive MIMO systems, sparse channel estimation exploring the joint channel sparsity is first proposed, where the channel estimation is modeled as a joint sparse recovery problem. Then the block coherence of MIMO channels is analyzed for the proposed model, which shows that as the number of antennas at the base station grows, the probability of joint recovery of the positions of nonzero channel entries will increase. Furthermore, an improved algorithm named block optimized orthogonal matching pursuit (BOOMP) is also proposed to obtain an accurate channel estimate for the model. Simulation results verify our analysis and show that the proposed scheme exploring joint channel sparsity substantially outperforms the existing methods using individual sparse channel estimation.

Joint Design of Pilot Power and Pilot Pattern for Sparse Cognitive Radio Systems

Existing works design the pilot pattern for sparse channel estimation, assuming that the power of all pilots is equal. However, equal power allocation is not optimal in cognitive radio (CR) systems. In this correspondence, we jointly design the pilot power and pilot pattern for sparse channel estimation in orthogonal-frequency-division-multiplexing-based CR systems, based on the rule of mutual incoherence property that minimizes the coherence of the measurement matrix used for the sparse recovery. Under the sum power constraint and peak power constraint, the pilot design is formulated as a joint optimization problem, which is then decoupled into tractable sequential formations. Given a pilot pattern, we formulate the design of pilot power as a second-order cone programming. Then, we propose a joint design algorithm, which includes discrete optimization for pilot pattern and continuous optimization for pilot power. Simulation results show that the proposed algorithm can achieve better channel estimation performance in terms of mean square error and bit error rate and can further improve the spectrum efficiency by 2.4%, compared with existing algorithms assuming equal pilot power.

Estimation of Extended Targets Based on Compressed Sensing in Cognitive Radar System

In this paper, the ranges and velocities of multiple extended targets are estimated by exploiting the target sparsity in the cognitive radar system. Different from the point targets in the traditional compressed sensing (CS) radar, the parameters of extended targets are expressed and estimated by using a novel CS-based model. Since the echo signals from extended targets are the convolutions between the transmitted waveform and target impulse responses (TIRs), the dictionary matrices in the proposed cognitive radar for all extended targets must be first established in the CS-based reconstruction algorithm. Then, the target parameters are estimated by reconstructing the nonzero entries of a sparse vector. To further improve the performance of CS reconst-ruction, a novel two-step method is proposed to minimize the mutual coherence of the dictionary matrix by optimizing the transmitted waveform. Simulation results demonstrate that the estimation performance of the extended targets is significantly improved by optimizing the transmitted waveform.

Near-Optimal Signal Detector Based on Structured Compressive Sensing for Massive SM-MIMO

Massive spatial-modulation multiple-input multiple-output (SM-MIMO) with high spectrum efficiency and energy efficiency has recently been proposed for future green communications. However, in massive SM-MIMO, the optimal maximum-likelihood detector has the high complexity, whereas state-of-the-art low-complexity detectors for small-scale SM-MIMO suffer from an obvious performance loss. In this paper, by exploiting the structured sparsity of multiple SM signals, we propose a low-complexity signal detector based on structured compressive sensing (SCS) to improve the signal detection performance. Specifically, we first propose the grouped transmission scheme at the transmitter, where multiple SM signals in several continuous time slots are grouped to carry the common spatial constellation symbol to introduce the desired structured sparsity. Accordingly, a structured subspace pursuit (SSP) algorithm is proposed at the receiver to jointly detect multiple SM signals by leveraging the structured sparsity. In addition, we also propose the SM signal interleaving to permute SM signals in the same transmission group, whereby the channel diversity can be exploited to further improve signal detection performance. Theoretical analysis quantifies the gain from SM signal interleaving, and simulation results verify the near-optimal performance of the proposed scheme.

Energy-Efficient Transceiver Design for Hybrid Sub-Array Architecture MIMO Systems

Millimeter-wave (mmWave) communication operated in frequency bands between 30 and 300 GHz has attracted extensive attention due to the potential ability of offering orders of magnitude greater bandwidths combined with further gains via beamforming and spatial multiplexing from multi-element antenna arrays. mmwave system may exploit the hybrid analog and digital precoding to achieve simultaneously the diversity, array and multiplexing gain with a lower cost of implementation. Motivated by this, in this paper, we investigate the design of hybrid precoder and combiner with sub-connected architecture, where each radio frequency chain is connected to only a subset of base station antennas from the perspective of energy efficient transmission. The problem of interest is a non-convex and NP-hard problem that is difficult to solve directly. In order to address it, we resort to design a two-layer optimization method to solve the problem of interest by exploiting jointly the interference alignment and fractional programming. First, the analog precoder and combiner are optimized via the alternating-direction optimization method where the phase shifter can be easily adjusted with an analytical structure. Then, we optimize the digital precoder and combiner based on an effective multiple-input multiple-output channel coefficient. The convergence of the proposed algorithms is proved using the monotonic boundary theorem and fractional programming theory. Extensive simulation results are given to validate the effectiveness of the presented method and to evaluate the energy efficiency performance under various system configurations.