CN110958102B - Pilot pollution suppression method based on pilot distribution and power control joint optimization - Google Patents

Abstract

The invention discloses a pilot frequency pollution suppression method based on pilot frequency allocation and power control joint optimization, comprising the following steps: (1) receiving a pilot frequency signal, and performing channel estimation; (2) receiving an uplink data signal, and performing signal detection; (3) According to the derivation of the progressive signal-to-interference-noise ratio SINR expression, the optimization objective of maximizing the minimum frequency efficiency is established; (4) According to the optimization objective, the WGC-PD-UPC algorithm is used to optimize the pilot frequency allocation and uplink power control. The method of alternating iteration is mainly used. In the iterative process, the large-scale fading factor is first fixed for uplink power control optimization, and then the uplink transmit power is fixed for pilot frequency allocation optimization. The present invention provides a method for joint optimization of pilot frequency allocation and uplink power control, which effectively improves the minimum frequency efficiency in the system under the condition of low computational complexity, thereby greatly improving the communication quality of cell edge users.

Description

A pilot pollution suppression method based on pilot allocation and power control joint optimization

technical field

The invention belongs to the technical field of signal processing in wireless communication, and in particular relates to a pilot pollution suppression method based on pilot frequency allocation and power control joint optimization.

Background technique

Massive MIMO technology has become one of the key technologies of 5G because it can greatly improve the performance of frequency efficiency, energy efficiency and link reliability. In a massive MIMO system, a time division multiplexing TDD mode is generally used for data transmission. This is because in the TDD mode, the reciprocity of the uplink and downlink channels in the same coherence time can be easily utilized to perform channel estimation according to the uplink pilot signal sent by the terminal. However, since the number of orthogonal pilots is limited by the coherence duration of the channel, and in order to ensure that more valid data is sent within a certain coherence duration, massive MIMO systems generally adopt a full multiplexing pilot scheduling strategy, that is, the same set of orthogonal pilots Pilots are fully multiplexed by terminals in all cells. This strategy leads to mutual interference between signals generated by terminals multiplexing the same pilot between different cells, that is, the problem of pilot pollution. The problem of pilot pollution has seriously affected the accuracy of the channel estimation by the base station and the communication rate of the user, and has become a bottleneck for the further improvement of the performance of the massive MIMO system.

At present, the problem of pilot frequency pollution has been widely studied. The specific solutions are as follows: 1. Starting from the perspective of pilot frequency allocation, that is, according to the difference of the corresponding channel gain of the terminal and the degree of signal interference received by different pilot frequencies, the pilot frequency is intelligently scheduled to Maximizing the total system capacity is the optimization goal to find the best pairing relationship between the terminal and the pilot, thereby reducing the pollution of the pilot and improving the system performance, but this kind of method is often at the expense of the communication quality of the terminal with poor performance, and The algorithm has high complexity; 2. Starting from the perspective of time-shifted pilots, the cells are first classified, and then the terminals of different types of cells are asynchronously sent pilots in time slots, which effectively improves the performance of edge terminals with serious pilot pollution. However, the pilot frequency asynchronous sending will compress the pilot frequency transmission time, which reduces the number of terminals that the system can serve, and affects the number of terminals served by the system to a certain extent; however, the above two schemes only consider the design of the pilot frequency. , and does not consider the power allocation of pilot and uplink data to reduce pilot pollution; 3. Use user grouping to allocate pilot power to suppress pilot pollution. Although this method can obtain better performance, it does not perform power and pilot frequency allocation for uplink data, so it still has a large room for improvement.

SUMMARY OF THE INVENTION

Aiming at the problem of pilot frequency pollution in the prior art, the purpose of the present invention is to provide a pilot frequency pollution suppression method based on pilot frequency allocation and power control joint optimization. To improve the minimum communication rate of cell users.

In order to achieve the above object, the present invention provides the following technical solutions: a pilot pollution suppression method based on pilot frequency allocation and power control joint optimization, comprising the following steps:

Step 1: Receive pilot signals and perform channel estimation;

Step 2: Receive the uplink data signal and perform signal detection;

Step 3: According to the derivation of the progressive signal-to-interference-to-noise ratio SINR expression, establish the optimization objective of maximizing the minimum frequency effect;

Step 4: According to the optimization objective, use the WGC-PD-UPC algorithm to optimize pilot frequency allocation and uplink power control. The WGC-PD-UPC algorithm is specifically:

Step 4.1: Perform equal-power pilot allocation optimization;

Step 4.2: This step is to judge whether the iteration termination condition is satisfied. When the iteration reaches the preset number of iterations or tends to a stable value, the iteration is terminated, otherwise, the following two sub-steps are continued to be alternately iteratively executed;

Step 4.3: Fix the large-scale fading factor according to the previous pilot allocation scheme, and then perform uplink power control optimization;

Step 4.4: In this step, the uplink transmit power is fixed first, and then the pilot power allocation optimization is performed. This step is performed in the same manner as step 4.1, the only difference is that it is not equal-power pilot allocation, but the influence of fixed uplink power is considered when pilot allocation is optimized;

Step 4.5: Compare the saved results of each iteration, and select the pilot frequency allocation scheme and uplink power corresponding to the optimal result.

In a specific embodiment, in step 4.3, the large-scale fading factor is fixed according to the result of the previous pilot frequency allocation, and then the optimization problem is transformed as follows:

表示第i个小区中第k个用户发射导频的功率，

表示为第i个小区中第k个用户传输上行数据的功率，β_iik表示为第i个小区中第k个用户到第i个小区中心基站的大尺度衰落因子，

表示为第j个小区中第k个用户发射导频的功率，

表示为第j个小区中第k个用户传输上行数据的功率，β_ijk表示为第j个小区中第k个用户到第i个小区中心基站的大尺度衰落因子，V_L表示为功率放大器的截止功率，V_H表示为功率放大器的饱和功率。In the formula, ξ is expressed as the minimum approximate signal-to-interference-noise ratio in L cells,

represents the power of the kth user transmitting the pilot in the ith cell,

is expressed as the power of the kth user in the ith cell to transmit uplink data, β _iik is expressed as the large-scale fading factor from the kth user in the ith cell to the center base station of the ith cell,

is expressed as the power of the kth user transmitting the pilot in the jth cell,

It is expressed as the power of the kth user in the jth cell transmitting uplink data, β _ijk is the large-scale fading factor from the kth user in the jth cell to the center base station of the ith cell, and _VL is the power amplifier The cut-off power, _VH , is expressed as the saturation power of the power amplifier.

Obviously at this time this power control optimization sub-problem is a geometric programming (GP) problem, which can be solved using the MOSEK toolkit using the interior point method.

In a specific embodiment, in step 4.4, the uplink transmit power is fixed, then the sub-problems of the problem decomposition can be expressed as:

In this way, this problem is just a pilot allocation optimization problem, but in a multi-cell Massive MIMO system, if performing an exhaustive search for pilot allocation, it will take a lot of computing time, which is difficult to achieve real-time in practical engineering. deal with;

In order to solve this sub-problem well, a weighted graph coloring algorithm is used for processing. The main idea of the algorithm is to first measure the metric of the mutual pilot contamination of two users potentially having the same pilot in different cells. The metric can be expressed as:

构建边缘权重干扰图(EWIG)；Then according to this metric weight

Build Edge Weight Interference Graph (EWIG);

Finally, greedy pilot allocation is performed according to EWIG.

In a specific embodiment, in step 4.5, this step is to judge whether the iteration termination condition is satisfied, and terminate the iteration when the iteration reaches a preset number of iterations or tends to a stable value, otherwise continue to alternate the following two sub-steps Iterative execution.

The present invention provides a pilot pollution suppression method based on pilot frequency allocation and power control joint optimization, so as to improve the communication rate of cell users. Since the problem is an NP problem, its computational complexity is unbearable. In order to reduce the computational complexity , the present invention proposes a joint optimization algorithm, which decomposes the problem into two sub-problems: pilot frequency allocation and power control.

The method to solve the sub-problem of pilot frequency allocation is to first fix the power of pilot frequency and uplink data transmission, and then carry out the pilot frequency allocation design. Since the computational complexity of this sub-problem is relatively large if the exhaustive search method is used, the present invention uses the graph coloring method to allocate pilot frequencies.

The solution to the power control sub-problem is to first fix the large-scale fading factor based on the previous pilot allocation scheme, and then perform the power allocation. At this point this sub-problem becomes a geometric programming problem. This problem can be solved well using the MOSEK toolkit. By alternately and iteratively processing the above two sub-problems, the present invention can improve the minimum rate of cell users with lower computational complexity.

The WGC-PD-UPC algorithm of the present invention mainly adopts an alternate iterative method. In the iterative process, the large-scale fading factor is first fixed for uplink power control optimization, and then the uplink transmit power is fixed for pilot frequency allocation optimization.

The invention provides a method for joint optimization of pilot frequency allocation and uplink power control, which effectively improves the minimum frequency efficiency in the system under the condition of low computational complexity, which is close to the boundary in the ideal state, thereby greatly improving the Communication quality of cell edge users.

Description of drawings

FIG. 1 is a schematic block diagram of the work flow of a pilot contamination suppression method based on pilot frequency allocation and power control joint optimization of the present invention.

FIG. 2 is a schematic block diagram of the flow of the jointly optimized WGC-PD-UPC algorithm based on pilot frequency allocation and power control provided by the present invention.

FIG. 3 is a schematic diagram showing the comparison of the cumulative distribution curves of the minimum frequency efficiency performance of the WGC-PD-UPC algorithm based on pilot frequency allocation and power control joint optimization provided by the present invention and other algorithms.

FIG. 4 is a schematic diagram showing the comparison of the minimum frequency efficiency performance curves of the WGC-PD-UPC algorithm based on pilot frequency allocation and power control joint optimization provided by the present invention and other algorithms as the number of users in a cell increases.

FIG. 5 is a schematic diagram of the convergence situation of the joint optimization of the WGC-PD-UPC algorithm based on pilot frequency allocation and power control provided by the present invention.

Detailed ways

The solution of the present invention will be further elaborated below with reference to specific embodiments and accompanying drawings.

In this example, we consider a multi-cell multi-user TDD Massive MIMO system, which is composed of L hexagonal cells. In each cell, the base station located in the center of the cell is configured with M antennas, and serves randomly distributed single-antenna users (K<<M) in K cells at the same time. Without loss of generality, the channel gain h _ijk from the kth user in cell j to the base station in cell i can be expressed as:

where g _ijk is a small-scale fading factor, which obeys a circularly symmetric complex Gaussian distribution, namely CN(0, I _M );

β _ijk is the large-scale fading factor, which can be expressed as:

Here z _ijk represents shadow fading, and its logarithmic distribution (ie, 10log ₁₀ (z _ijk )) obeys the Gaussian distribution CN(0,σ _shadow ), and r _ijk represents the kth user in the jth cell to the The distance from the center base station of i cells, R represents the radius of the cell; β _ijk is constant in a coherence time, and changes slowly and is easy to track in several thousands of channel coherence time.

A method for suppressing pilot pollution based on the joint optimization of pilot allocation and power control in this embodiment, the specific working flowchart is shown in FIG. 1 , and includes the following steps:

Step 1. Receive pilot signals and perform channel estimation;

且满足ΨΨ^H＝I_K。每个小区中的每个用户使用Ψ中的一列作为导频，其序列长度为τ，且同一小区内的不同用户使用Ψ中不同的列向量。In order to overcome the interference in the cell, the system uses orthogonal pilots for the users in the cell. At the same time, in order to increase the frequency efficiency, the pilot frequency between cells adopts a full reuse strategy. Therefore, the pilot sequence used in this system can be expressed as

And satisfy ΨΨ ^H =I _K . Each user in each cell uses a column in Ψ as a pilot with a sequence length of τ, and different users in the same cell use different column vectors in Ψ.

可表示为：During the pilot transmission phase, the signal received by the base station in the i-th cell

can be expressed as:

表示的是第j个小区中的第k个用户导频发射的功率，ψ_jk表示的是在第j个小区第k个用户发射的导频序列。该用户是使用Ψ中第k列作为导频序列的，

代表的是加性高斯白噪声，其独立同分布服从CN(0,σ_p)。here

represents the power of the pilot frequency transmitted by the kth user in the jth cell, and ψ _jk represents the pilot frequency sequence transmitted by the kth user in the jth cell. The user uses the kth column in Ψ as the pilot sequence,

It represents the additive white Gaussian noise, and its independent and identical distribution obeys CN(0,σ _p ).

进行最小二乘法LS信道估计。因此第i个小区中的k用户与第i个小区的基站之间的信道估计值

可表示为：When the base station in the cell receives the pilot signal transmitted by the user, the base station can

Least squares LS channel estimation is performed. Therefore the channel estimate between k users in the ith cell and the base station of the ith cell

can be expressed as:

Step 2: Receive the uplink data signal and perform signal detection;

In the uplink data transmission stage, the data received by the base station can be expressed as:

表示的是第i个小区中心基站接受到的上行数据，

表示的是第j个小区中第k个用户发射上行数据时的发射功率，

表示的是第j个小区中第k个用户发射的上行数据，

表示的是在上行数据发射阶段所产生的加性高斯白噪声，其服从CN(0,σ_u)。here

Indicates the uplink data received by the i-th cell center base station,

represents the transmit power when the kth user in the jth cell transmits uplink data,

represents the uplink data transmitted by the kth user in the jth cell,

represents the additive white Gaussian noise generated in the uplink data transmission stage, which obeys CN(0,σ _u ).

和信道估计的

进行匹配滤波检测，检测后恢复第i个小区中第k个用户发射的数据

可表示为：According to the signal received by the base station in the uplink transmission phase

and channel estimation

Perform matched filter detection, and recover the data transmitted by the kth user in the ith cell after detection

can be expressed as:

表示的是小区内的干扰和其他非相关噪声之和，其表达式为：The second item represents the interference between cells,

It represents the sum of interference and other uncorrelated noise in the cell, and its expression is:

Step 3: Derive the progressive signal-to-interference-to-noise ratio SINR expression, and establish the optimization objective of maximizing the minimum frequency effect;

会随着基站的天线数目M的增多会大幅减少,当M→∞，

再根据表达式(6)可以推出第i个小区中第k个用户的上行信干噪比

可表示为：It can be seen from expression (7) that

With the increase of the number of antennas M of the base station, it will be greatly reduced. When M→∞,

Then according to expression (6), the uplink signal-to-interference and noise ratio of the kth user in the ith cell can be deduced

can be expressed as:

When M→∞,

式(9)可以进一步简化为：when

Equation (9) can be further simplified as:

可以表示为：Frequency efficiency of the kth user in the ith cell

It can be expressed as:

Here μ=τ/t represents the loss of uplink frequency efficiency, and t is the uplink coherence time interval.

The purpose of the present invention is to strive to improve the frequency efficiency of the user with the smallest frequency efficiency in the system, thereby improving the communication quality of the user located at the edge of the cell with serious pilot frequency pollution. So the problem can be expressed as:

According to equation (12) and the properties of the logarithmic function, the optimization problem can be transformed into:

And according to formula (9), the problem can be transformed into:

In addition, considering the linear range of the transmit power amplifier, both the pilot frequency transmit power and the uplink data transmit power should be limited to a certain range. That is to say, the maximum uplink transmit power should not exceed V _H , and the minimum should not be less than _VL . Among them, V _H is related to the amplification saturation power of the power amplifier, and _VL is related to the cut-off power of the power amplifier. Then, with this constraint, the optimization problem of the present invention can be re-expressed as:

Step 4: According to the optimization objective, use the WGC-PD-UPC algorithm proposed by the present invention to perform pilot frequency allocation and uplink power control optimization;

Obviously Equation (15) is not a convex optimization problem, it is a joint problem of discrete pilot allocation optimization and continuous uplink transmit power optimization. This problem is a hard NP problem to solve. In order to solve this problem well, the present invention decomposes the problem into two sub-problems and completes them in two steps, and alternately iterates the two sub-problems. When a certain number of iterations reaches stability, the optimal value in the iterative process is taken for pilot frequency allocation and uplink power control.

The specific algorithm flow is shown in Figure 2:

Step 4.1:

In order to optimize without falling into local optimization, the first step assumes that the uplink transmit powers are all equal, then the problem can be expressed as:

Thus the problem is just a pilot allocation optimization problem. However, in a multi-cell Massive MIMO system, it will take a lot of calculation time to perform pilot frequency allocation in an exhaustive search method, which is difficult to realize real-time processing in practical engineering. In order to solve this sub-problem well, the present invention adopts a weight map coloring algorithm for processing. The main idea of the algorithm is to first measure the metric of the mutual pilot contamination degree of two users potentially having the same pilot in different cells, and the metric can be expressed as:

构建边缘权重干扰图(EWIG)；最后再按照EWIG进行贪婪导频分配。Then according to this metric weight

Construct an edge weighted interference graph (EWIG); finally, perform greedy pilot allocation according to EWIG.

Step 4.2:

This step is to judge whether the iteration termination condition is satisfied, and terminate the iteration when the iteration reaches the preset number of iterations or the region is stable, otherwise, continue to alternately execute the following two sub-steps.

Step 4.3:

In this step, all β _iik and β _ijk in Equation (15) are fixed according to the results of the previous pilot allocation, and then Equation (17) is converted as follows:

Obviously at this time this power control optimization sub-problem is a geometric programming (GP) problem, which can be solved using the MOSEK toolkit using the interior point method.

Step 4.4:

In this step, the uplink transmit power is fixed first, then the sub-problem of the problem can be expressed as:

This step is performed in the same way as step 4.1, the only difference is that the measurement of the mutual interference degree considers the uplink power, and its measurement expression is:

Step 4.5:

Compare the saved results of each iteration, and select the pilot frequency allocation scheme and uplink power corresponding to the optimal result.

For the convenience of numerical analysis, we briefly describe some algorithms compared with the algorithm of the present invention (WGC-PD-UPC):

(a) RPA: This algorithm randomly allocates pilots with equal power;

(b) WGC-PD: This algorithm performs equal-power pilot allocation according to the weighted graph coloring method;

(c) ESPA: This algorithm adopts an exhaustive pilot allocation scheme, and then searches for the optimal pilot allocation scheme according to the optimization objective of equation (16);

(d) RPA-PPC: This algorithm only considers the pilot frequency control optimization, the pilot frequency is allocated in a random manner, and the pilot frequency optimization is performed according to step 4.3 of the algorithm provided by the present invention. At this time, all uplink data transmission power is considered to be equal;

(e) WGC-PD-PPC: This algorithm is similar to the algorithm of WGC-PD-UPC provided by the present invention, the only difference is that when step 4.3 is executed, the WGC-PD-PPC algorithm considers that all uplink data transmission powers are equal;

(f) ESPA-PPC: The algorithm is completed in two steps. The first step uses the ESPA algorithm for pilot allocation, and the second step uses the RPA-PPC algorithm for pilot power control optimization;

(g) RPA-UPC: This algorithm considers pilot frequency and uplink data power control optimization, pilot frequency allocation adopts a random manner, and power optimization is performed according to algorithm step 4.3 provided by the present invention;

(h) ESPA-UPC: This algorithm is completed in two steps. The first step uses the ESPA algorithm for pilot frequency allocation, and the second step uses the RPA-UPC algorithm for pilot frequency control optimization;

(i) idea optimal solution: This method adopts a combination of exhaustive search and geometric programming to solve the optimization problem of Equation (15). This method is an exhaustive pilot allocation scheme. The power control optimization is performed in step 4.3 provided by the invention, and then the pilot frequency allocation scheme and its corresponding power are selected according to the objective function of formula (15).

Figure 3 shows the cumulative distribution function of the minimum frequency efficiency in the system under different algorithm processing, where the number of cells in the system is set to L=3, and the number of users in each cell is set to K=4. It can be seen from the figure that the random pilot allocation (RPA) without pilot allocation optimization and power control has the worst performance; after using the WGC-PD algorithm to optimize the pilot allocation, the performance has been greatly improved; although Using the ESPA algorithm to optimize the pilot frequency allocation has a small improvement in performance compared with the WGC-PD algorithm, but it is difficult to use in practical applications due to its high computational complexity. The three algorithms, RPA-PPC, WGC-PD-PPC, and ESPA-PPC, are based on RPA, WGC-PD, and ESPA to perform pilot power control optimization. It can be found that compared with the corresponding algorithm without pilot control optimization, its performance is greatly improved. The performance of the RPA-PPC algorithm is close to that of the ESPA algorithm. This shows that in order to improve the minimum frequency efficiency, not only pilot frequency allocation optimization but also pilot frequency control optimization is required.

In addition, we can also see from Figure 3 that WGC-PD-UPC, ESPA-UPC and idea Optimal Solution algorithms are superior to other algorithms in terms of minimum frequency efficiency performance. This shows that the joint optimization of pilot frequency allocation, pilot frequency and uplink data power control can further improve the performance. The WGC-PD-UPC algorithm provided by the present invention is higher than the ESPA-UPC algorithm in the average minimum frequency efficiency performance of 0.219b/s/Hz, and the ESPA-UPC algorithm takes a lot of time due to the exhaustive search pilot scheme. . The performance of the WGC-PD-UPC algorithm provided by the present invention is relatively close to that of the idea OptimalSolution algorithm, and the difference between them is only 0.243b/s/Hz. The idea OptimalSolution algorithm is actually an ideal optimization solution in terms of performance, but its computational complexity is very high. This shows that the WGC-PD-UPC algorithm is efficient and feasible.

中求和子项数目也相应增加了，在采用随机导频分配时，由于各子项中的大尺度衰落因子β_ijk都比较大，所以分母数值会显著增大而导致最小频效逐渐减小。而当进行导频分配优化后，虽然

中求和子项数目仍然会随着用户数目进行增加，但是各子项的值会大幅减少，因此造成最小频效反而逐渐增大。Figure 4 shows the minimum frequency efficiency change with the increase of users in the cell under different algorithms. It can be seen from Figure 4 that after the number of users increases to 4, the ranking of their performance from good to bad is WGC-PD-UPC>WGC-PD-PPC>WGC-PD>RPA-UPC>RPA-PPC>RPA. When the number of each user is 8, compared with the RPA algorithm, WGC-PD-UPC, WGC-PD-PPC, WGC-PD, RPA-UPC, RPAPPC increase by 4.551b/s/Hz and 3.757b/Hz respectively. s/Hz, 2.153b/s/Hz, 1.154b/s/Hz, 0.494b/s/Hz. This result shows that with the increase of the number of users, the change trend is still the same, and WGC-PD-UPC is still the best solution. As the number of users increases, the performance of the three algorithms WGC-PD-UPA, WGC-PD-PPC, and WGC-PD is getting better and better, and the performance of the three algorithms RPA-UPC, RP A-PPC, and RPA is getting better and better. Difference. This has also led to an increasing performance gap. When the number of users increases from 2 to 8, the performance gap of the WGC-PD-UPC algorithm increases from 3.126b/s/Hz to 4.551b/s/Hz compared to the RPA without any optimization. This is because when the number of users increases, the denominator of the objective function in Eq. (15)

The number of summation sub-items also increases accordingly. When random pilot frequency allocation is used, since the large-scale fading factor β _ijk in each sub-item is relatively large, the denominator value will increase significantly, resulting in a gradual decrease in the minimum frequency effect. And when the pilot allocation optimization is carried out, although

The number of summation sub-items will still increase with the number of users, but the value of each sub-item will be greatly reduced, so the minimum frequency effect will gradually increase instead.

Figure 5 shows the convergence of the WGC-PD-UPC algorithm provided by the present invention. It can be clearly seen from Figure 5 that after a certain iteration, the algorithm provided by the present invention will converge to a stable state, and its maximum value Close to the ideal idea Optimal Solution algorithm results. Since the algorithm provided by the present invention performs alternate iterations on the two sub-problems decomposed, the algorithm may converge to two fixed values, and this result is shown in channel 2 in FIG. 5 .

The above-mentioned embodiments only represent concentrated embodiments of the present invention, and the descriptions thereof are specific and detailed, but should not be construed as limiting the scope of the invention patent. It should be noted that, for those skilled in the art, without departing from the concept of the present invention, several modifications and improvements can be made, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention should be subject to the appended claims.

Claims (3)

1. A pilot pollution suppression method based on pilot allocation and power control joint optimization is characterized by comprising the following steps:

step 1: receiving a pilot signal and carrying out channel estimation;

step 2: receiving an uplink data signal and carrying out signal detection;

and step 3: establishing a maximized minimum frequency effect optimization target according to a deduced progressive signal to interference plus noise ratio (SINR) expression;

and 4, step 4: according to an optimization target, utilizing a WGC-PD-UPC algorithm to perform pilot frequency allocation and uplink power control optimization, wherein the WGC-PD-UPC algorithm specifically comprises the following steps:

step 4.1: performing equal-power pilot frequency distribution optimization;

step 4.2: judging whether an iteration termination condition is met, terminating iteration after the iteration reaches a preset iteration number or tends to a stable value, and otherwise, continuously performing alternate iteration execution on the following two substeps;

step 4.3: step 3) is not a convex optimization problem in the formula (15) obtained, and is a combined problem of scattered pilot frequency allocation optimization and continuous uplink transmission power optimization; therefore, all β in formula (15) are first treated_iikAnd beta_ijkFixing according to the result of the previous pilot frequency distribution; then, uplink power control optimization is carried out;

step 4.4: in the step, the uplink transmitting power is firstly fixed, and then pilot frequency power allocation optimization is carried out. The step is the same as the step 4.1, the only difference is that the pilot frequency distribution with equal power is not, but the influence of fixed uplink power is considered when the pilot frequency distribution is optimized;

step 4.5: comparing the saved iteration results each time, and selecting a pilot frequency distribution scheme and uplink power corresponding to the optimal result;

in the step 3), the specific process is as follows:

uplink signal-to-interference-and-noise ratio of kth user in ith cell

Can be expressed as:

when the rate of M → ∞ is,

frequency efficiency of kth user in ith cell

Can be expressed as:

where μ τ/t represents the uplink frequency loss, and t is the uplink coherence interval.

The communication quality of users with serious pilot frequency pollution at the edge of a cell is improved by improving the frequency efficiency of the users with the minimum frequency efficiency in the system; the problem can therefore be expressed as:

according to equation (12) and the nature of the logarithmic function, the optimization problem can be converted into:

again, this problem can be translated into:

in addition, the linear range of the transmitting power amplifier is considered, and the linear range is limited in a certain range regardless of the pilot frequency transmitting power or the uplink data transmitting power; that is, the uplink transmission power should not exceed V at maximum_HMinimum should be not less than V_L(ii) a Wherein V_HAmplified saturation power, V, of associated power amplifier_LA cutoff power of the associated power amplifier; with this constraint on, then, the optimization problem can be re-expressed as:

wherein:

represents the power at which the kth user in the ith cell transmits a pilot,

expressed as the power, beta, of the transmission of uplink data for the kth user in the ith cell_iikExpressed as the large-scale fading factor from the kth user in the ith cell to the central base station of the ith cell,

expressed as the power at which the kth user in the jth cell transmits the pilot,

expressed as the power, beta, of the transmission of uplink data for the kth user in the jth cell_ijkExpressed as the large-scale fading factor, V, from the kth user in the jth cell to the ith cell center base station_LExpressed as the cut-off power, V, of the power amplifier_HExpressed as powerThe saturation power of the amplifier;

as indicated by the channel estimate values, are,

indicating intra-cell interference and other non-facies.

2. The method for pilot pollution suppression based on pilot allocation and power control joint optimization according to claim 1, wherein in step 4.3, the large-scale fading factor is fixed according to the result of the previous pilot allocation, and then the optimization problem is transformed as follows:

max ξ

where ξ is the minimum approximate signal-to-interference-and-noise ratio in L cells,

represents the power at which the kth user in the ith cell transmits a pilot,

expressed as the power, beta, of the transmission of uplink data for the kth user in the ith cell_iikExpressed as the large-scale fading factor from the kth user in the ith cell to the central base station of the ith cell,

expressed as the power at which the kth user in the jth cell transmits the pilot,

expressed as the power, beta, of the transmission of uplink data for the kth user in the jth cell_ijkExpressed as the large-scale fading factor, V, from the kth user in the jth cell to the ith cell center base station_LExpressed as the cut-off power, V, of the power amplifier_HExpressed as the saturation power of the power amplifier.

3. The method for pilot pollution suppression based on pilot allocation and power control joint optimization according to claim 1, wherein in step 4.4, if the uplink transmission power is fixed, the sub-problem of the problem decomposition can be expressed as:

adopting a weight map coloring algorithm for processing, wherein the process of the algorithm is as follows: first, a measure of the mutual pilot pollution level of two users potentially having the same pilot in different cells is measured, which can be expressed as:

and then based on the metric weight

Constructing an edge weight interference graph;

and finally, carrying out greedy pilot frequency distribution according to the edge weight interference graph.

Images