CN106788934B - A Joint Pilot Allocation Method for Multiple Cells in Massive MIMO System - Google Patents

Abstract

The present invention proposes a multi-cell joint pilot allocation method in a massive MIMO system, which is used to solve the technical problems of low system average channel capacity and high time complexity existing in the existing pilot allocation method, and the implementation steps are as follows: 1. Set communication scene parameters; 2. Obtain the location information of all base stations and all users through information interaction between base stations, users and base stations; 3. Calculate the large-scale fading information of all users based on the location information of base stations and users , to obtain the useful information set and the potential interference set; 4. Continuously according to the useful information set and the potential interference set, assign pilots to the user corresponding to the minimum value element in the useful information set and the user with the least interference to the user, and then update the useful Information collection and potential interference collection until the pilots of all users are allocated. The invention effectively improves the average channel capacity of the system and reduces the time complexity of pilot allocation.

Description

A Joint Pilot Allocation Method for Multiple Cells in Massive MIMO System

technical field

The invention belongs to the technical field of communications, and relates to a pilot allocation method, in particular to a multi-cell joint pilot allocation method in a massive MIMO system under a time division duplex mode.

Background technique

Multiple Input Multiple Output MIMO (Multiple Input Multiple Output) technology refers to the use of multiple transmitting antennas and receiving antennas at the transmitting end and receiving end, respectively, so that signals are transmitted and received through multiple antennas at the transmitting end and receiving end, thereby improving communication quality. Massive MIMO refers to a technology that configures dozens or even hundreds of antennas at the base station and serves multiple mobile terminals at the same time. As one of the key technologies of 5G, in multi-cell communication scenarios, it is generally necessary to obtain channel state information Using the traditional way of estimating the channel based on the pilot training sequence, in theory, with the increase of the number of antennas at the base station, the interference and non-correlated noise in the cell gradually tend to zero, and the channel capacity should continue to increase, but due to the limitation of the channel coherence time, There is not enough coherence time to assign orthogonal pilot sequences to users in different cells, and the same pilot group will be reused by users in adjacent cells, so that when performing channel estimation, the base station cannot distinguish The received pilot signals come from users in the own cell or other users in adjacent cells, so pilot pollution will occur, and the advantages of large-scale antenna arrays will be weakened, which also becomes the performance bottleneck of massive MIMO systems .

In order to overcome the performance bottleneck caused by pilot pollution, the current methods to solve pilot pollution mainly include the following three categories: (1) efficient channel estimation; (2) multi-cell cooperative precoding method considering pilot pollution; (3) pilot pollution Frequency scheduling and allocation methods. The pilot scheduling and allocation method refers to a method of reducing pilot pollution by changing the position of sending pilots, or by rationally allocating different pilots to different users. The traditional pilot scheduling and allocation method is typically a random pilot allocation method. By allocating the kth pilot to the kth user in each cell, this allocation method has low time complexity, but there is a system average channel The problem of poor capacity performance. In order to overcome the problem of poor average channel capacity in the random pilot allocation method, the author li Ku published an article "Low Complexity PilotAllocation in Massive MIMO System" at IEEE International Conference OnCommunication Software and Networks (ICCSN) in 2016, and proposed a method based on The greedy pilot allocation strategy first obtains all possible combinations of users who allocate the same pilot, and then continuously selects a group of users with the largest average channel capacity to allocate the same pilot by calculating the average channel capacity of the system in each case , this method can obtain relatively good system average channel capacity performance when the number of base station antennas is very large, but there is still a certain gap between its performance and the system average channel capacity performance under optimal allocation, and when the number of users is large , and the complexity is relatively high.

Contents of the invention

The purpose of the present invention is to address the shortcomings of the above-mentioned prior art, and propose a multi-cell joint pilot allocation method in a massive MIMO system, which is used to solve the problems of low system average channel capacity and Technical problems with high time complexity.

The technical idea of the present invention is: firstly, through the information interaction between base stations, obtain the location information of all base stations and users, then calculate the large-scale fading information from all users to the base station, construct a useful information set and a potential interference set, and finally according to User channel quality conditions, assign pilots to users with poor channel conditions first, and ensure that they suffer from less interference, so as to improve the average channel capacity performance of the system and reduce time complexity.

According to above-mentioned technical train of thought, the technical scheme that realizes the object of the present invention to take comprises the following steps:

(1) Let the number of cells be L, each cell includes a base station and K single-antenna users, where L≥2, K≥2, and each base station and user includes an orthogonal pilot group F={φ ₁ ,…,φ _p ,…,φ _K }, φ _p is the pth pilot frequency, and the value of p is 1～K;

(2) Information exchange is performed between each base station and the user first, the base station obtains the user location information of the cell, and then the base station location information and the user location information are exchanged between the base stations, and all user location information U and base station location information B are obtained;

(3) According to the user location information U and the base station location information B, calculate the large-scale fading β _i(i,k) of all users, and obtain the useful information set Where i represents the i-th cell, and k represents the k-th user;

(4) According to the user location information U and the base station location information B, calculate the large-scale fading β _i(j,k) of the users in the other L-1 cells except the users in the i-th cell, and obtain the potential interference set Where j represents the jth cell;

(5) According to the useful information set α and the potential interference set α _ij , assign pilot information to all users, and the implementation steps are:

(5a) Select the user user _(i,k) corresponding to the minimum value of β _i(i,k ) from the useful information set α, and assign pilot frequency φ _p to the user;

(5b) Select the potential interference set α _ij corresponding to the user user _(i,k) from the potential interference set α _ij , and then select the minimum interference to the user user _(i,k) in each cell from the potential interference set α _ij users, and assign pilot φ _p to them;

(5c) Remove the user information that has been assigned pilots from the useful information set α and the potential interference information set α _ij , so as to update the useful information set α and the potential interference information set α _ij ;

(5d) Judging whether p is equal to K, if so, indicating that the pilots of all users have been allocated, otherwise, update p=p+1, and perform steps (5a)-(5d), until the pilots of all users are allocated .

Compared with the prior art, the present invention has the following advantages:

In the process of pilot allocation, the present invention uses the useful information set and potential interference set obtained through calculation, and preferentially allocates pilots to the user corresponding to the smallest element in the useful information set and the user with the least interference to the user, It not only considers the channel conditions of all users, but also makes users with poor channel conditions suffer less interference. Compared with the greedy pilot allocation method in the prior art, it effectively improves the average channel capacity of the system and reduces Time complexity of pilot allocation.

Description of drawings

Fig. 1 is the realization flow diagram of the present invention;

Fig. 2 is the relationship diagram between the system average channel capacity and the number of base station antennas of the present invention;

FIG. 3 is a relationship diagram between the system average channel capacity and the edge signal-to-noise ratio (SNR) of the present invention.

Detailed ways

The present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments.

With reference to Fig. 1, the realization steps of the present invention are as follows:

Step 1, set communication scene parameters.

Let the number of cells L=3, each cell includes a base station and K=3 single-antenna users, and each base station and user includes an orthogonal pilot group F={φ ₁ ,...,φ _p ,...,φ ₃ }, φ _p is the pth pilot frequency, and the value of p is 1~3.

Step 2: Information exchange is performed between each base station and users first, the base station obtains the user location information of the cell, and then the base station location information and user location information are exchanged between the base stations, and all user location information U and base station location information B are obtained.

Each base station first interacts with users in the cell where the base station is located. The base station obtains user location information, and then each base station interacts with other base stations with the location information of the base station and the location information of users in the cell, so that each base station All user location information U and base station location information B can be obtained.

Step 3: According to the user location information U and the base station location information B, calculate the large-scale fading β _i(i,k) of all users, and obtain the useful information set The calculation formula of β _i(i,k) is as follows:

Where z _i(j,k) is the shadow fading from the kth user of the jth cell to the ith cell base station, and obeys the logarithmic distribution, that is, 10log ₁₀ (z _i(j,k) ) is the mean value of 0, The standard deviation is the Gaussian distribution of σ _shadow , d _i(j,k) is the distance from user k in the j-th cell to the base station of the i-th cell, R represents the radius of the cell, and γ is the path loss.

Step 4: According to user location information U and base station location information B, calculate the large-scale fading β _i(j,k) of other L-1 users except the users in the i-th cell, and obtain the potential interference set The calculation formula of β _i(j,k) is as follows:

Step 5, according to the useful information set α and the potential interference set α _ij , assign pilot information to all users, the implementation steps are:

(5a) Select the user user _(i,k) corresponding to the minimum value of β _i(i,k ) from the useful information set α, assuming that the minimum value of β _i(i,k) is β _3(3,1) , corresponding to user Be user 1 of the third cell, and assign pilot frequency φ _p to this user;

(5b) Select the potential interference set α _3j corresponding to user 1 of the third cell from the potential interference set α _ij , and then select the item with the least interference from other cells to user 1 of the third cell from the potential interference set α _3j if it is β _{3 (1,1)} and β _3(2,2) , assign the same pilot φ _p to user 1 of the first cell and user 2 of the second cell;

(5c) From the useful information set α and the potential interference information set α _ij , the user information that has been assigned pilots is removed, and the useful information set α and the potential interference information set α _ij are updated. From the useful information set α and the potential interference information set β _i(i,k) and β _i(j,k) corresponding to users who have been assigned pilots are removed from the information set α _ij to ensure that users who have been assigned pilots will not participate in the next allocation.

(5d) judge whether p is equal to 3, if so, show that the pilot frequency of all users has been distributed, otherwise, update p=p+1, and perform steps (5a)-(5d), until the pilot frequency distribution of all users is complete.

Compared with the greedy pilot allocation method, the present invention effectively improves the average channel capacity of the system, and the time complexity of the present invention is O(KL ² logK), and the method of exhaustive search is used to obtain the optimal allocation scheme The time complexity is O((K!) ^L ), and the time complexity of the greedy pilot allocation method is O(K ^L ), which shows that the present invention effectively reduces the time complexity, especially when the number of users K is large The effect of reducing the time complexity is more obvious.

Below in conjunction with simulation experiment, technical effect of the present invention is described further:

Simulation conditions: Consider L cells, each cell contains a base station and K single-antenna users, each base station is equipped with M antennas, and there are no users within a radius of r meters near the base station, the parameter settings are as follows:

Number of districts L=3 Number of base station antennas M 50≤M≤800 The number of users in each cell K K=3,8 Cell radius R R = 500 Limit radius r r=30 path loss gamma γ=3.8 Logarithmic shadow fading coefficient σ_shadow 8dB

Simulation content: The present invention, the traditional random pilot allocation method, the greedy pilot allocation method, and the exhaustive search method under optimal allocation are simulated, and the results are shown in Figure 2 and Figure 3;

Refer to Figure 2: For the two cases where the edge signal-to-noise ratio SNR=20dB, the number of users K=3 and the number of users K=8, the average channel capacity is simulated with the change of the number of base station antennas, and the results are shown in Figure 2(a) and Figure 2 as shown in (b);

It can be seen from Fig. 2(a) that when the number of users K=3, as the number of base station antennas M increases from 50 to 800, the present invention, the traditional random pilot allocation method, the greedy-based pilot allocation method, and the optimal The average channel capacity under the exhaustive search method under allocation is also constantly increasing. Under the same number of antennas, the average channel capacity performance of the present invention is significantly better than the traditional random pilot allocation method and the greedy pilot allocation method, and can The average channel capacity performance of the system close to the optimal allocation method is obtained, which shows that the present invention can effectively improve the average channel capacity of the system;

Combining Figure 2(a) and Figure 2(b), when the number of users K=8, under the same number of base station antennas, compared with the case of the number of users K=3, the average channel capacity under the traditional random allocation method is almost unchanged , the increase of the average channel capacity is small under the greedy allocation method, and the average channel capacity of the present invention has a significant increase, and the average capacity of the system is increased by the present invention as the number of users increases.

Referring to Figure 3: In the case of the number of users K=3 and the number of base station antennas M=800, the average channel capacity is simulated with the change of the edge signal-to-noise ratio SNR, and the results are shown in Figure 3;

It can be seen from Fig. 3 that with the continuous increase of edge SNR in the range of 0-20dB, the average channel capacity of the present invention, the traditional random pilot allocation method and the greedy pilot allocation method all increase with the edge SNR SNR Increase and increase, after the edge SNR is greater than 20dB, the average channel capacity increase under the three methods tends to be gentle, and under the same SNR, the average channel capacity of the system of the present invention is obviously better than the other two methods.

Claims (1)

1. the pilot distribution method of multi-plot joint in a kind of extensive mimo system, it is characterised in that the following steps are included:

(1) number of cells is set as L, and each cell includes a base station and K single-antenna subscriber, wherein L >=2, K >=2, each Base station and user respectively include an orthogonal guide frequency group F={ φ₁,…,φ_p,…,φ_K, φ_pFor p-th of pilot tone, the value of p is 1 ~K；

(2) advanced row information interaction between each base station and user, base station obtain the customer position information of cell, then base station it Between carry out base station position information and customer position information interaction, obtain all customer position information U and base station position information B；

(3) according to customer position information U and base station position information B, the large-scale fading β of all users is calculated_i(i,k), had Use information aggregateβ_i(i,k)Calculation formula it is as follows:

Wherein z_i(j,k)For j-th of cell k-th of user to i-th of cell base station shadow fading, and obey log series model, That is 10log₁₀(z_i(j,k)) be mean value be 0, standard deviation σ_shadowGaussian Profile, d_i(j,k)It is the user k to i-th of j-th of cell The distance of a cell base station, R represent radius of society, and γ is path loss；

(4) it according to customer position information U and base station position information B, calculates except other L-1 with open air of i-th of cell are a small The large-scale fading β of the user in area_i(j,k), obtain potential interference setj ≠ i, β_i(j,k)Calculation formula it is as follows:

(5) according to useful information set α and potential interference set α_ij, pilot frequency information is distributed to all users, realizes step are as follows:

(5a) chooses β from useful information set α_i(i,k)The corresponding user user of minimum value_(i,k), and pilot tone is distributed to the user φ_p；

(5b) is from potential interference set α_ijMiddle selection and user user_(i,k)Corresponding potential interference set α_ij, then it is potential dry from this Disturb set α_ijIt chooses in each cell to user user_(i,k)The smallest user is interfered, and distributes it pilot tone φ_p；

(5c) is from useful information set α and potential interference information aggregate α_ijIt is middle to remove the allocated user information for crossing pilot tone, It realizes to useful information set α and potential interference information aggregate α_ijIt is updated；

Whether (5d) judges p equal with K, if so, showing that the pilot tone of all users has distributed, otherwise, updates p=p+1, and Step (5a)-(5d) is executed, until the pilot tone of all users is assigned.