An Adaptive Modulation Method for Multi-Cell Massive MIMO Systems Based on Incomplete Channel Information

Abstract

The present invention discloses an adaptive modulation method for multi-cell large-scale MIMO systems based on incomplete channel information. Based on the incomplete channel state information and zero-forcing detection technology, the continuous rate and discrete rate adaptive modulation technologies are utilized to maximize the spectral efficiency of multi-cell large-scale MIMO systems. The method proposed by the present invention aims at multi-cell large-scale MIMO systems under incomplete channel state information, and respectively gives the system adaptive modulation methods based on continuous rate and discrete rate, as well as the closed-form expression of system bit error rate, effectively improving the system spectral efficiency and providing an effective method for alleviating the shortage of frequency-domain resources in mobile communication systems.

Description

Technical Field:

[0001] The present invention relates to an adaptive resource allocation method for a mobile communication system, and particularly to an adaptive modulation method for a multi-cell massive MIMO system based on incomplete channel information, belonging to the field of mobile communications. Background Art:

[0002] As a key technology for the fifth-generation mobile communication, massive MIMO technology can provide a greater degree of freedom compared with traditional MIMO, and has the characteristics of low power consumption, high energy efficiency, less non-coherent interference, and high spatial resolution. This technology has been widely studied and applied. At the same time, adaptive modulation technology has been widely used in engineering practice because it can improve the performance of system capacity and spectral efficiency (SE) without power optimization, and has also received great attention in the field of theoretical research.

[0003] In Reference 1 (Goldsmith AJ, Chua S G. Variable-rate variable-power MQAM for fading channels[J]. IEEE Transactions on Communications, 1997, 45(10):1218-1230.), an adaptive modulation method with variable rate and variable power was proposed. It was pointed out that when using multi-level quadrature amplitude modulation and 5-6 constellation diagrams, a power gain of up to 20 dB can be provided compared with non-adaptive modulation schemes. In the research on adaptive modulation technology in Reference 2 (Xiaoxin Q, Chawla K. On the performance of adaptive modulation in cellular systems[J]. IEEE Transactions on Communications, 1999, 47(6):884-895.), it was found that the improvement of SE can also be obtained by only using adaptive modulation technology without power control, and the combination of power control and AM technology can further improve the spectral efficiency. In Reference 3 (Deng L, Rui Y, Cheng P. A unified energy efficiency and spectral efficiency tradeoff metric in wireless networks[J]. IEEE Communications Letters, 2013, 17(1):55-58), the tradeoff between system energy efficiency (EE) and SE was considered for the general point-to-point communication system, and an optimal power allocation scheme was designed to optimize the overall system performance under different SE and EE performance ratios.

[0004] As can be seen from the above, the above research on AM modulation technology has not yet involved multi-cell large-scale MIMO systems. Therefore, the impact of AM technology on the performance of multi-cell large-scale MIMO systems needs to be studied urgently. Summary of the Invention:

[0005] The present invention provides an adaptive modulation method for multi-cell large-scale MIMO systems based on incomplete channel information to solve the problems not solved by the prior art. This method considers incomplete channel state information and zero-forcing detection technology, and gives the spectral efficiency performance of this method in multi-cell large-scale MIMO systems under the constraint of instantaneous bit error rate.

[0006] The technical solution adopted by the present invention is as follows: An adaptive modulation method for a multi-cell large-scale MIMO system based on incomplete channel information, comprising the following steps:

[0007] S1. Establish a transmission model for a multi-cell uplink large-scale MIMO system. The number of cells is L. In each cell, there is a base station equipped with N antennas and K single-antenna users. Then, in the j-th cell, the received signal at the base station is

[0008]

[0009] where p i,k represents the transmission power of the k-th user in the i-th cell, x ik is the useful signal of the k-th user in the i-th cell, and satisfies represents the noise in the j-th cell; the channel from the k-th user in the l-th cell to the j-th base station obeys complex Gaussian distribution, where β jl,k represents the large-scale fading coefficient;

[0010] S2. Estimate the channel between the user and the base station. where is the estimated channel, is the estimation error, and there is where where τ is the pilot length, p u is the pilot power; After the zero-forcing detection technique, the detection matrix is The detected signal of the k-th user in the j-th cell is

[0011]

[0012] Then, the signal-to-interference-plus-noise ratio of the k-th user in the l-th cell is

[0013]

[0014] S3. According to the signal-to-interference-plus-noise ratio expression, give its corresponding PDF as

[0015]

[0016] where,

[0017] S4. Considering the instantaneous BER constraint, combine the PDF obtained in step S3 to give the calculation formulas for the system spectral efficiency and bit error rate.

[0018] Furthermore, S4 includes the following sub-steps:

[0019] (a) The average spectral efficiency and average bit error rate of the entire system under the discrete rate adaptive modulation method are

[0020]

[0021] where π ( ψn ) is the system bit error rate parameter, is the incomplete gamma function, is the discrete rate modulation switching threshold corresponding to M modulation methods, denoted as ρ q are the numerical integration nodes, Q is the number of nodes, and

[0022] (b) The average spectral efficiency of the entire system under the continuous rate adaptive modulation method is:

[0023]

[0024] where A = -0.625ln(5BER0), is the exponential integral function.

[0025] The present invention has the following beneficial effects: When modeling the channel, the present invention considers the incomplete channel state information and estimates the channel, so that the analysis results are more in line with the actual situation; The calculation formula of the spectral efficiency of the multi-cell large-scale MIMO system under zero-forcing detection is given, and the spectral efficiency and bit error rate of the multi-cell large-scale MIMO system with discrete and continuous adaptive modulation based on the instantaneous bit error rate constraint provided by the present invention are compared through MATLAB platform simulation. Description of the drawings:

[0026] Figure 1 is the flow chart of the adaptive modulation method for the multi-cell large-scale MIMO system based on incomplete channel information of the present invention.

[0027] Figure 2 is the effect comparison diagram of the discrete rate adaptive modulation scheme and the continuous rate adaptive modulation scheme of the multi-cell large-scale MIMO system using different numbers of modulation methods under the target BER constraint between the theoretical value and the MATLAB simulation result of the present invention.

[0028] Figure 3 is the average bit error rate comparison diagram of the adaptive modulation scheme in the multi-cell large-scale MIMO system using different numbers of modulation methods between the theoretical value and the MATLAB simulation result of the present invention under the target BER constraint. Specific implementation manners:

[0029] The present invention will be further described below with reference to the accompanying drawings.

[0030] The adaptive modulation method for multi-cell large-scale MIMO systems based on incomplete channel information of the present invention comprises the following steps:

[0031] S1. Establish a transmission model for the multi-cell uplink large-scale MIMO system. The number of cells is L. There is a base station equipped with N antennas and K single-antenna users in each cell. Then, in the j-th cell, the received signal at the base station is

[0032]

[0033] where p i,k represents the transmission power of the k-th user in the i-th cell, x ik is the useful signal of the k-th user in the i-th cell, and satisfies represents the noise in the j-th cell; the channel from the k-th user in the l-th cell to the j-th base station obeys complex Gaussian distribution, where β jl,k represents the large-scale fading coefficient;

[0034] S2. Estimate the channel between the user and the base station, where is the estimated channel, is the estimation error, and there is where where τ is the pilot length, p u is the pilot power; through the zero-forcing detection technology, the detection matrix is The detected signal of the k-th user in the j-th cell is

[0035]

[0036] Then, the signal-to-interference-plus-noise ratio of the k-th user in the l-th cell is

[0037]

[0038] S3. According to the signal-to-interference-plus-noise ratio expression, give its corresponding PDF as

[0039]

[0040] where,

[0041] S4. Considering the instantaneous BER constraint, combine the PDF obtained in step S3 to give the calculation formulas for the system spectral efficiency and bit error rate;

[0042] Furthermore, S4 includes the following sub-steps:

[0043] (a) The average spectral efficiency and average bit error rate of the entire system under the discrete rate adaptive modulation method are

[0044]

[0045]

[0046] where π ( ψn ) is the system bit error rate parameter, is the incomplete gamma function, is the discrete rate modulation switching threshold corresponding to M modulation methods, denoted as ρ q are the numerical integration nodes, Q is the number of nodes, and It should be noted that due to the asymptotic upper limit of the signal-to-interference-plus-noise ratio in the multi-cell massive MIMO system caused by pilot contamination, the PDF in (11) has an upper limit of 1 / η l,k , so when the threshold value is greater than 1 / η l,k , let this threshold and all subsequent larger thresholds be equal to 1 / η l,k , that is, the corresponding modulation method and larger modulation methods are not adopted;

[0047] (b) The average spectral efficiency of the entire system under the continuous rate adaptive modulation method is:

[0048]

[0049] where A = -0.625ln(5BER0), is the exponential integral function; since the modulation order of the continuous rate adaptive modulation can vary continuously, it can make the instantaneous BER always equal to the target BER. Therefore, the BER of the multi-cell massive MIMO system based on the continuous rate adaptive modulation is always equal to the target BER.

[0050] Next, the performance evaluation of the adaptive modulation scheme in the multi-cell massive MIMO system proposed in the present invention is verified through simulation on the MATLAB platform, and the correctness of the derived formula is proved. Set the number of receiving antennas to 64, the number of users to 5, the number of cells to 4, and the parameter τ u related to the pilot is set to 0.3, and the target bit error rate is BER0 = 10 -3 . Refer to the 3GPP LTE specification to model the system parameters and large-scale fading. The large-scale fading coefficient is where is a Gaussian distribution with zero mean and standard derivative of 8. Denote the distance from the \(k\)-th user in the \(l\)-th cell to the \(i\)-th base station. Each cell is a square with a side length of 1 kilometer, and the base station is located at its center. The reference distance is 35 meters. Consider 6 candidate QAM modulation modes, namely 2QAM, 4QAM, 8QAM, 16QAM, 32QAM, and 64QAM.

[0051] Figure 2 The figure shows the comparison of the effects of the discrete rate adaptive modulation scheme and the continuous rate adaptive modulation scheme of the multi-cell large-scale MIMO system using different numbers of modulation methods under the target BER constraint. The abscissa average signal-to-noise ratio is defined as the ratio of the user transmission power to the noise power. It can be seen from the figure that the theoretical curve and the simulation curve are in agreement, which proves the accuracy of the theoretical formula. As the transmission power increases, the spectral efficiency of the discrete rate adaptive curve gradually improves and reaches a steady state when the power is large enough. When the power is large enough, the signal-to-noise ratio of the user tends to the upper limit \(1 / \eta\) l,k , and at this time, the highest modulation method with a corresponding threshold lower than the upper limit is adaptively selected, and it is not necessarily possible to select the maximum modulation method in the alternative scheme. For example, when the number of candidate modulation method types \(M = 6\), the upper limit \(1 / \eta\) l,k in some channels is less than the threshold value corresponding to 64QAM, resulting in the user being unable to select the highest modulation method 64QAM at this time. Therefore, the overall spectral efficiency performance is lower than 6 bit / s / Hz. And the continuous rate adaptive curve is always higher than the discrete rate adaptive curve because the continuous rate adaptive scheme has no discrete rate limit and can find the optimal method that meets the target BER constraint, while the discrete rate AM is rate-limited due to its discrete modulation order, which also verifies that the continuous rate adaptive is the performance upper bound of the adaptive scheme.

[0052] Figure 3 The figure shows the comparison of the average bit error rate of the adaptive modulation scheme in the multi-cell large-scale MIMO system using different numbers of modulation methods under the target BER constraint. The theoretical BER curves and the simulation BER curves of the discrete AM scheme under different parameters are basically in agreement, which indicates that the derived formula can effectively evaluate the system BER performance. In the discrete rate AM, the larger the number of selectable modulation methods, the worse its BER performance because a larger number of selectable modulation methods corresponds to higher selectable modulation methods, and higher-order modulation methods are more likely to cause bit errors. However, all three BER curves are below the target BER (\(BER_0 = 10\) -3 ), indicating that the discrete AM modulation scheme has met the target BER constraint, reflecting the effectiveness of the scheme.

[0053] In summary, the adaptive modulation method for multi-cell massive MIMO systems based on incomplete channel information proposed by the present invention can effectively improve the spectral efficiency of multi-cell massive MIMO systems while meeting quality requirements, providing an effective method for improving system performance.

[0054] The above are only the preferred embodiments of the present invention. It should be noted that for those of ordinary skill in the art, several improvements can be made without departing from the principle of the present invention, and these improvements should also be regarded as the protection scope of the present invention.

Claims

1. An adaptive modulation method for multi-cell large-scale MIMO systems based on incomplete channel information, characterized in that: It includes the following steps: S1. Establish a transmission model for a multi-cell uplink massive MIMO system. The number of cells is L. In each cell, there is a base station equipped with N antennas and K single-antenna users. Then, in the j-th cell, the received signal at the base station is where p i,k represents the transmit power of the k-th user in the i-th cell, and x ik is the useful signal of the k-th user in the i-th cell, and satisfies denotes the noise in the j-th cell; the channel from the k-th user in the l-th cell to the j-th base station obeys a complex Gaussian distribution, where β jl,k represents the large-scale fading coefficient; S2. Estimate the channel between the user and the base station, where is the estimated channel, is the estimation error, and there is where where τ is the pilot length, p u is the pilot power; After the zero-forcing detection technique, the detection matrix is The detection signal of the kth user in the jth cell is Then, the signal-to-interference-plus-noise ratio (SINR) of the k-th user in the l-th cell is S3. According to the SINR expression, give its corresponding probability density function (PDF). Among them, S4. Considering the instantaneous bit error rate (BER) constraint, combine the PDF obtained in step S3 to give the calculation formulas for the system spectral efficiency and bit error rate. S4 includes the following sub-steps: (a) The average spectral efficiency and average bit error rate of the entire system under the discrete rate adaptive modulation method are Among them π ( ψn ) is the system bit error rate parameter, is the incomplete gamma function, is the discrete rate modulation switching threshold corresponding to M modulation methods, denoted as ρ q is the numerical integration node, Q is the number of nodes, and (b) The average spectral efficiency of the entire system under the continuous rate adaptive modulation method is: where A = -0.625ln(5BER0), is the exponential integral function.

Images