Lowcomplexity beamdomain channel estimation and power allocation in hybrid architecture massive MIMO systems
 74 Downloads
Abstract
In this journal, we investigate the beamdomain channel estimation and power allocation in hybrid architecture massive multipleinput and multipleoutput (MIMO) communication systems. First, we propose a lowcomplexity channel estimation method, which utilizes the beam steering vectors achieved from the directionofarrival (DOA) estimation and beam gains estimated by lowoverhead pilots. Based on the estimated beam information, a purely analog precoding strategy is also designed. Then, the optimal power allocation among multiple beams is derived to maximize spectral efficiency. Finally, simulation results show that the proposed schemes can achieve high channel estimation accuracy and spectral efficiency.
Keywords
Beam domain Channel estimation Massive MIMO Hybrid architecture Spectral efficiencyAbbreviations
 AWGN
Additive white gaussian noise
 BS
Base station
 CC
Channel capacity
 CS
Compressive sensing
 DFT
Discrete fourier transform
 DOA
Directionofarrival
 DOD
Directionofdeparture
 EPA
Equal power allocation
 ESPRIT
Estimation of signal parameters via rotational invariance techniques
 FDD
Frequency division duplex
 MIMO
Multipleinput and multipleoutput
 mmWave
Millimeter wave
 MSE
Mean square error
 MUSIC
Multiple signal classification
 OPA
Optimal power allocation
 RF
Radio frequency
 RMSE
Root mean square error
 RPA
Random power allocation
 SE
Spectral efficiency
 SVD
Singular value decomposition
 UE
User equipment
 URA
Uniform rectangular array
1 Introduction
Massive multipleinput and multipleoutput (MIMO) has been envisioned as a crucial technology for the nextgeneration cellular systems, due to its high energy efficiency and spectral efficiency [1, 2, 3]. To exploit the benefits of massive antennas, a large number of radio frequency (RF) chains are needed so that each antenna can own one dedicated RF chain. Combating the expensive cost, hybrid architecture has been a promising scheme which combines costeffective phase shifters and limited RF chains. For hybrid architecture, there are two most common precoding techniques: hybrid precoding and analog precoding [4, 5].
1.1 Related work
Hybrid analog/digital precoding has been investigated extensively including precoding design and channel estimation strategy [4, 5, 6, 7], discrete Fourier transform (DFT)based hybrid architecture [8, 9, 10], angle domain hybrid precoding and channel tracking [11, 12], and performance analysis on hybrid architecture [13, 14, 15]. Most of the existing hybrid precoding designs heavily rely on the perfect knowledge of the channel, which brings a great challenge because only limited RF chains are available at base station (BS) and user equipment (UE). An adaptive channel estimation algorithm was proposed in [6] for hybrid architecture millimeter wave (mmWave) systems, which applies the compressive sensing (CS) technology based on the poor scattering nature of mmWave. With the CS method, the channel estimation algorithm in [6] yields a limited accuracy and a high computational complexity. For analog precoding, an RF beam search method was proposed in [16, 17], where the channel estimation was obtained by sweeping all beam directions over the spatial region. However, the beam search algorithm in [16] is timeconsuming due to the exhaustive beam searching. Besides, orthogonal training sequences have also been widely used to obtain the angle information for channel estimation in the DFTbased hybrid architecture [8, 9], which caused high computational complexity and feedback cost. Therefore, our objective is to design a simple and effective signal processing scheme for hybrid architecture massive MIMO systems.
1.2 Summary of contributions
In this journal, we investigate lowcomplexity beamdomain channel estimation, precoding, and power allocation in hybrid architecture massive MIMO systems. The major contributions of this journal are summarized as follows:
(1) A lowcomplexity channel estimation algorithm is proposed, which avoids high pilot overhead by utilizing estimated beams and beam gains.
(2) The RF precoder and combiner are designed based on estimated beam information.
(3) The power allocation among multiple beams is optimized to maximize spectral efficiency.
(4) Simulation results demonstrate high channel estimation accuracy and comparable spectral efficiency achieved by our proposed channel estimation and power allocation schemes.
2 System model
where ⊗ denotes the Kronecker product, a_{x}(θ_{l},ϕ_{l}) is an M×1 vector with the mth element being \(e^{j\frac {2\pi }{\lambda }d(m1)\cos \theta _{l}\sin \phi _{l}}\), and a_{y}(θ_{l},ϕ_{l}) is a N×1 vector with the nth element being \(e^{j\frac {2\pi }{\lambda }d(n1)\sin \theta _{l}\sin \phi _{l}}, d=\frac {\lambda }{2}\) is the distance between two adjacent antennas, and λ is the wavelength. M and N are the numbers of antennas in the horizontal direction and vertical direction of the URA, respectively.
Lemma 1: The steering vectors of different paths are asymptotically orthogonal for sufficient large MN. That is, \(\mathop {\lim }\limits _{M, N\rightarrow \infty }\mathbf a^{H}(\theta _{l_{1}}, \phi _{l_{1}})\mathbf a(\theta _{l_{2}}, \phi _{l_{2}})=0\) (l_{1}≠l_{2}).
Proof 1.
: See Appendix 6.
where \(\mathbf {s}\in \mathbb {C}^{L\times 1}\) is the transmitted signal with \(\mathbb E\{\mathbf {s}\mathbf {s}^{H}\}=\sigma _{s}^{2}\mathbf I_{L}, \mathbb E\{\cdot \}\) is the expectation operator, \(\mathbf P\in \mathbb {R}^{L\times L}\) is the power allocation matrix, \(\mathbf F_{t}\in \mathbb C^{M_{t}N_{t}\times L}\) denotes the RF precoder, \(\mathbf F_{r}\in \mathbb C^{M_{r}N_{r}\times L}\) denotes the RF combiner, and \(\mathbf {n}\sim \mathcal {CN}(\mathbf {0},\sigma _{n}^{2}\mathbf {I}_{M_{r}N_{r}})\) denotes additive white Gaussian noise (AWGN) vector. Let [F]_{i,j} denote the (i,j)th element of a matrix F. The RF precoder and RF combiner control phase changes by employing phase shifters with the magnitude \([\mathbf F_{t}]_{i,j}=\frac {1}{\sqrt {M_{t}N_{t}}}\) and \([\mathbf F_{r}]_{i,j}=\frac {1}{\sqrt {M_{r}N_{r}}}\), respectively.
3 Methods of beamDomain channel estimation and power allocation
In this section, we present the beamdomain signal processing in hybrid architecture massive MIMO systems. First, we discuss the DOA estimation based on the hybrid architecture with limited RF chains. Then, lowcomplexity channel estimation is proposed based on the estimated beam information. Finally, we study the power allocation to improve the system performance.
3.1 Beam information estimation
Due to the angle reciprocity, the directionofdeparture (DOD) of the forwardlink channel is the same as the DOA of the reverselink channel, even for the frequency division duplex (FDD) systems [8].
The DOA information \(\theta _{l}^{t}\) and \(\phi _{l}^{t}\) can be estimated using the limited RF chains in two time slots, respectively. For the reverse link, K_{t} contiguous antennas in a row at the receiver (which is the transmitter in the forward link) are first chosen in the horizontal direction as in Fig. 1, and then linked with K_{t} RF chains to estimate the azimuth angle \(\theta _{l}^{t}\) in the first time slot. Similarly, the elevation angle \(\phi _{l}^{t}\) can be estimated by employing K_{t} antennas in a column in the second time slot. Here, the estimated angles are defined as (\(\theta _{l}^{t}, \phi _{l}^{t}\)).
By using only a small number of antennas, this estimation strategy has a low complexity and a high accuracy, which have been analyzed in our previous work [18]. For example, while K_{t}=16, the root mean square error (RMSE) of the estimated DOA has been less than 1^{∘} by costing only 32 snapshots [18]. Many DOA estimation algorithms can be adopted in this proposed estimation strategy, such as the wellknown MUltiple SIgnal Classification (MUSIC) and Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT) [19, 20, 21]. In each time slot, there are a number of snapshots to obtain approximate covariance information for DOA estimation with MUSIC or ESPRIT algorithm.
Practically, considering the lowcost, finiteresolution phase shifters are implemented at the transmitter and receiver for RF processing. For the M_{t}×N_{t} URA, there are M_{t} and N_{t} phase shifters with finiteresolution in the horizontal and vertical directions, respectively. When all the phase shifters have the same resolution \(\frac {1}{P}\pi \), for the azimuth angle θ∈[0,π) in the horizontal direction, the beam codebook consists of P basic beams by sampling the angle space [0,π) as \(G_{M}=\left \{0,\frac {1}{P}\pi,\cdots,\frac {P1}{P}\pi \right \}\); for the elevation angle \(\phi \in [0,\frac {\pi }{2})\) in the vertical direction, the beam codebook consists of \(\left (\left \lfloor \frac {P}{2}\right \rfloor +1\right)\) basic beams by sampling the angle space \([0,\frac {\pi }{2})\) as \(G_{N}=\left \{0,\frac {1}{P}\pi,\cdots,\frac {\left \lfloor \frac {P}{2}\right \rfloor }{P}\pi \right \}\). Thus, for an M_{t}×N_{t} URA, the beam codebook consists of \(P\left (\left \lfloor \frac {P}{2}\right \rfloor +1\right)\) basic 2D beams (θ,ϕ).
Finally, the estimated angles (\(\theta _{l}^{t}, \phi _{l}^{t}\)) should be mapped to the beam codebooks for practical transmission. The DOA information can be easily estimated as \((\hat {\theta }_{l}^{t},\hat {\phi }_{l}^{t})=\text {arg} \mathop {\text {min}}\limits _{\theta \in G_{M},\\ \phi \in G_{N}}\left \{\theta \,\,\theta _{l}^{t}\,+\,\phi \,\,\phi _{l}^{t}\right \}\) for practical RF processing. Notably, the DOA estimation strategy can also be directly applied at the receiver in forward link to obtain \((\hat {\theta }_{l}^{r}, \hat {\phi }_{l}^{r})\).
3.2 Channel estimation
Channel estimation is crucial but intractable in massive MIMO systems [22, 23]. In this subsection, we propose a lowcomplexity channel estimation method in the beam domain. The proposed method includes the following three steps.
where \(\mathbf A_{t}^{\dag }=\left (\mathbf A_{t}^{H}\mathbf A_{t}\right)^{1}\mathbf A_{t}^{H}\) is the pseudoinverse of A_{t}, and \(\left (\mathbf A_{t}^{H}\right)^{\dag }=\mathbf A_{t}\left (\mathbf A_{t}^{H}\mathbf A_{t}\right)^{1}\) is the pseudoinverse of \(\mathbf A_{t}^{H}\). By using the statistic characteristic, the estimated magnitude of the beam gain will have higher accuracy than the pilotbased estimation method.
where \(\mathbf X_{2}^{\dag }=\mathbf X_{2}^{H}\left (\mathbf X_{2}\mathbf X_{2}^{H}\right)^{1}\) is the pseudoinverse of X_{2}. With L beams, there only needs n_{s}= log2L bits pilots to estimate the phase of the beam gain. Due to the pseudoinverse in (10), there is a significant complexity reduction by using n_{s} bits pilots.
In summary, the proposed channel estimation strategy has a low complexity for three reasons: (1) It directly uses the results of DOA estimation to obtain the magnitude of the channel gain as an additional product; (2) the DOA estimation strategy has a low complexity; (3) the L beam gains are estimated with short training pilots. Furthermore, the proposed channel estimation strategy has a high accuracy. The reason is that the magnitude of channel gain is estimated from the statistic characteristic as shown in (8), which has a higher accuracy than that of the conventional pilotbased channel estimation.
The proposed channel estimation method can also be adopted for multiuser systems, resulting in a great degradation of pilot overhead. When multiple users have similar DOAs and bring in multiuser interference, the proposed channel estimation method is still available. The reason is that the channel estimation of all users is obtained from the DOA estimation which needs to be implemented in a timedivision method. As a result, the interference caused by the similar DOAs does not affect the timedivision channel estimation of multiple users.
3.3 Power allocation
where P_{t} is the total transmit power.
Theorem
where u^{+}=max{u,0}, and λ is the Lagrange multiplier satisfying
□
Proof 2.
where (a) and (c) follow from the Lemma 1 that the basis beams are asymptotically orthogonal for sufficiently large M_{r}N_{r},(b) and (d) are due to \(\mathbf a_{r}^{H}(\hat {\theta }_{l}^{r}, \hat {\phi }_{l}^{r})\mathbf a_{r}(\hat {\theta }_{l}^{r}, \hat {\phi }_{l}^{r})=M_{r}N_{r}\), and all results are proved in Appendix 6.
where (e) and (f) follow the Lemma 1 and \(\mathbf a_{t}^{H}\left (\hat {\theta }_{l}^{t}, \hat {\phi }_{l}^{t}\right)\mathbf a_{t}\left (\hat {\theta }_{l}^{t}, \hat {\phi }_{l}^{t}\right)=M_{t}N_{t}\), respectively, and all proofs are shown in Appendix 6.
Problem (12) with the simplified objective function (20) is a convex problem, and admits a waterfilling solution as Theorem 1.
To sum up, the main steps of the lowcomplexity beamdomain signal processing are summarized as follows:

In forward link, the receiver estimates beams \((\hat {\theta }_{l}^{r},\hat {\phi }_{l}^{r})\).

Estimate the beams \((\hat {\theta }_{l}^{t},\hat {\phi }_{l}^{t})\) and the magnitude of beam gain \(\hat {\gamma }_{l}\) by (8) in reverse link.

With the estimated beams, RF precoder at the transmitter and RF combiner at the receiver are derived from (4) and (5), respectively.

Perform the power allocation only with \(\hat {\gamma }_{l}\) according to Theorem 1.
4 Numerical results and discussion
In this section, we present the numerical results to evaluate the performance of the proposed signal processing methods in hybrid architecture massive MIMO systems. In the simulations, both the transmitter and receiver are equipped with 16 ×16 URA and 16 RF chains. The channel fading coefficients are generated from the urban micro model in the 3GPP standard [24]. The path loss is given by PL(dB)=36.7log_{10}(d)+22.7+26log_{10}(f_{c}), where d=50 m is the distance between the transmitter and receiver, and f_{c}=1 GHz. All simulation results are obtained by averaging over 1000 channel realizations.
5 Conclusion
In this journal, we have investigated the beamdomain channel estimation and precoding for hybrid architecture massive MIMO systems. We have proposed a lowcomplexity channel estimation algorithm based on the estimated beam information. The RF processing has been derived from the beam steering as well. Furthermore, we have optimized the power allocation to maximize the spectral efficiency. Simulation results have shown that the proposed signal processing methods can achieve high channel estimation accuracy and spectral efficiency.
6 Appendix
6.1 Proof of Lemma 1
where \(D=\frac {2\pi }{\lambda }d, \alpha _{1,2}=\sin \theta _{l_{2}}\sin \phi _{l_{2}}\sin \theta _{l_{1}}\sin \phi _{l_{1}}\), and \(\beta _{1,2}=\cos \theta _{l_{2}}\sin \phi _{l_{2}}\cos \theta _{l_{1}}\sin \phi _{l_{1}}\).
Thus, it is proved that the basis beams are asymptotically orthogonal for sufficiently large MN.
Footnotes
 1.
For example, the communication between two BSs, the communication between BS and UE, the system of mmWave communications, etc.
Notes
Acknowledgements
We would like to thank the anonymous reviewers for their insightful comments on the paper, as these comments led us to an improvement of the work.
Authors’ contributions
XC and LW designed the proposed channel estimation and power allocation algorithms. XC also implemented the simulation environment for experiments and completed the analyses of experimental results. ZZ and JD gave some advice on this manuscript and proofread it. All authors read and approved this submission.
Funding
This work was supported by the National Natural Science Foundation of China projects (61501109, 61571105, and 61601119), National Key Research and Development Plan (2016YFB0502202), Scientific Research Foundation of Graduate School of Southeast University (YBJJ1816), and the Scholarship from China Scholarship Council (201806090072).
Competing interests
The authors declare that they have no competing interests.
References
 1.F. Rusek, D. Persson, B. K. Lau, E. G. Larsson, T. L. Marzetta, O. Edfors, F. Tufvesson, Scaling up MIMO: Opportunities and challenges with very large arrays. IEEE Signal Process. Mag.30(1), 40–60 (2013).CrossRefGoogle Scholar
 2.E. G. Larsson, O. Edfors, F. Tufvesson, T. L. Marzetta, Massive MIMO for next generation wireless systems. IEEE Commun. Mag.52(2), 186–195 (2014).CrossRefGoogle Scholar
 3.L. Lu, G. Y. Li, A. L. Swindlehurst, A. Ashikhmin, R. Zhang, An overview of massive mimo: Benefits and challenges. IEEE J. Sel. Topics Signal Process.8(5), 742–758 (2014).CrossRefGoogle Scholar
 4.F. Sohrabi, W. Yu, Hybrid digital and analog beamforming design for largescale antenna arrays. IEEE J. Sel. Topics Signal Process.10(3), 501–513 (2016).CrossRefGoogle Scholar
 5.V. Raghavan, S. Subramanian, J. Cezanne, A. Sampath, O. H. Koymen, J. Li, Singleuser versus multiuser precoding for millimeter wave MIMO systems. IEEE J. Sel. Areas Commun.35(6), 1387–1401 (2017).CrossRefGoogle Scholar
 6.A. Alkhateeb, O. E. Ayach, G. Leus, R. W. Heath, Channel estimation and hybrid precoding for millimeter wave cellular systems. IEEE J. Sel. Topics Signal Process.8(5), 831–846 (2014).CrossRefGoogle Scholar
 7.R. W. Heath, N. Gonz álezPrelcic, S. Rangan, W. Roh, A. M. Sayeed, An overview of signal processing techniques for millimeter wave MIMO systems. IEEE J. Sel. Topics Signal Process.10(3), 436–453 (2016).CrossRefGoogle Scholar
 8.H. Lin, F. Gao, S. Jin, G. Y. Li, A new view of multiuser hybrid massive MIMO: Nonorthogonal angle division multiple access. IEEE J. Sel. Areas Commun.35(10), 2268–2280 (2017).CrossRefGoogle Scholar
 9.W. Tan, M. Matthaiou, S. Jin, X. Li, Spectral efficiency of DFTbased processing hybrid architectures in massive MIMO. IEEE Wirel. Commun. Lett.6(5), 586–589 (2017).CrossRefGoogle Scholar
 10.W. Tan, G. Xu, E. Carvalho, M. Zhou, C. Li, Low cost and high efficiency hybrid architecture massive MIMO systems based on DFT processing. Wirel. Commun. Mobile Comput.2018(10), 1–11 (2018).Google Scholar
 11.D. Fan, F. Gao, G. Wang, Z. Zhong, A. Nallanathan, Angle domain signal processingaided channel estimation for indoor 60ghz TDD/FDD massive MIMO systems. IEEE J. Sel. Areas Commun.35(9), 1948–1961 (2017).CrossRefGoogle Scholar
 12.J. Zhao, F. Gao, W. Jia, S. Zhang, S. Jin, H. Lin, Angle domain hybrid precoding and channel tracking for millimeter wave massive MIMO systems. IEEE Trans. Wireless Commun.16(10), 6868–6880 (2017).CrossRefGoogle Scholar
 13.W. Tan, D. Xie, J. Xia, W. Tan, L. Fan, S. Jin, Spectral and energy efficiency of massive MIMO for hybrid architectures based on phase shifters. IEEE Access. 6:, 11751–11759 (2018).CrossRefGoogle Scholar
 14.W. Tan, S. Jin, C. Wen, T. Jiang, Spectral efficiency of multiuser millimeter wave systems under single path with uniform rectangular arrays. EURASIP J. Wirel. Commun. Netw.181:, 458–472 (2017).Google Scholar
 15.W. Tan, W. Huang, X. Yang, Z. Shi, W. Liu, L. Fan, Multiuser precoding scheme and achievable rate analysis for massive MIMO system. EURASIP J. Wirel. Commun. Netw. 1:, 1–15 (2018).Google Scholar
 16.J. Wang, Beam codebook based beamforming protocol for multigbps millimeterwave WPAN systems. IEEE J. Sel. Areas Commun.27(8), 1390–1399 (2009).CrossRefGoogle Scholar
 17.Z. Xiao, T. He, P. Xia, X. G. Xia, Hierarchical codebook design for beamforming training in millimeterwave communication. IEEE Trans. Wireless Commun.15(5), 3380–3392 (2016).CrossRefGoogle Scholar
 18.X. Chen, Z. Zhang, L. Wu, J. Dang, P. S. Lu, C. Sun, in 9th IEEE Proc. Int. Conf. Wireless Commun. Signal Process. (WCSP). Robust beam management scheme based on simple 2D DOA estimation (Nanjing, China, 2017), pp. 1–6.Google Scholar
 19.R. O. Schmidt, A SIGNAL SUBSPACE APPROACH TO MULTIPLE EMITTER LOCATION AND SPECTRAL ESTIMATION (Stanford Univ. Press, Stanford, CA, USA, 1982).Google Scholar
 20.R. Schmidt, Multiple emitter location and signal parameter estimation. IEEE Trans. Antennas Propag.34(3), 276–280 (1986).CrossRefGoogle Scholar
 21.R. Roy, T. Kailath, ESPRITestimation of signal parameters via rotational invariance techniques. IEEE Trans. Acoust. Speech, Signal Process. 37(7), 984–995 (1989).CrossRefGoogle Scholar
 22.Y. Ding, S. E. Chiu, B. D. Rao, Bayesian channel estimation algorithms for massive MIMO systems with hybrid analogdigital processing and low resolution ADCs. IEEE J. Sel. Topics Signal Process.99:, 1–1 (2018).Google Scholar
 23.L. Wu, Z. Zhang, J. Dang, J. Wang, H. Liu, Y. Wu, Channel estimation for multicell multiuser massive MIMO uplink over rician fading channels. IEEE Trans. Veh. Technol.66(10), 8872–8882 (2017).CrossRefGoogle Scholar
 24.ETSI TR, Universal Mobile Telecommunications System (UMTS); Spacial Channel Model for Multiple Input Multiple Output (MIMO) simulations (3GPP TR 25.996 version 13.0.0 Release 13) (2016).Google Scholar
Copyright information
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.