Optimization method for joint interference elimination and resource allocation in D2D communication

Abstract

The invention relates to the technical field of remote sensing measurement and control, in particular to an optimization method for joint interference elimination and resource allocation in D2D communication, which comprises the following steps: constructing an SCDN network model, using StarBlock as a basic module for coding and decoding, and using WAM to increase the feature extraction capability of the model so as to improve the landslide extraction capability of the model; specifically, a WAM module is adopted, and the utilization efficiency of the model on the features is enhanced by aggregating the weight of the spatial dimension and the weight of the channel dimension, so that the identification capability of the model on the landslide is improved. And after the improved SCDN network model is obtained, training and verification are carried out by using a GVLM public data set to obtain an optimal model after iteration, and finally, the optimal model is used for prediction to obtain a landslide extraction result. According to the invention, by distinguishing similar spectrum and texture features on the image, the precision and accuracy of the extraction capability of the model are improved.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to an optimization method for joint interference elimination and resource allocation in D2D communication. Background Art

[0002] D2D communication is a new technology that allows neighboring terminals to communicate directly using cellular resources under the control of the base station. D2D users establish D2D communication by reusing resources within the authorized frequency band of cellular users. Because they use the same frequency resources as cellular devices, they will increase system throughput and spectrum resource utilization while also causing certain co-frequency interference to the existing cellular system. In order to reduce interference in the system and effectively improve system throughput, the main solutions are interference control and interference elimination.

[0003] In hybrid systems, the interference control, interference elimination and power control methods currently used are all aimed at improving the throughput of D2D communication and the total throughput of the system. However, the introduction of D2D communication in cellular systems will more or less bring co-channel interference to existing cellular users. These methods will reduce the communication quality of cellular users in order to improve the system throughput, which obviously damages the interests of cellular users. Summary of the invention

[0004] The object of the present invention is to provide an optimization method for joint interference elimination and resource allocation in D2D communication, so as to maximize the throughput of the system while ensuring that the communication quality of cellular users is not reduced.

[0005] To achieve the above object, the present invention provides an optimization method for joint interference elimination and resource allocation in D2D communication, comprising the following steps:

[0006] Step 1: Establish a system model that generates co-channel interference. In the system model, co-channel interference only occurs between D2D communication pairs and cellular users that reuse the same frequency, and the number of cellular users is greater than or equal to the number of D2D communication pairs.

[0007] Step 2: Establish a mathematical model, where the mathematical model is to match appropriate CUE frequency resources to N D2D communication pairs so as to maximize the total throughput of the system.

[0008] Optionally, each D2D communication pair includes a D2D transmitter and a D2D receiver. In step 1, the co-channel interference of the D2D transmitter to the cellular system is eliminated by zero-forcing precoding technology, and the interference of cellular users to the D2D receiver is minimized by channel selection.

[0009] Optionally, during the execution of step 2, let the set D N(n=1, 2, ..., N) represents a set of D2D communication receivers, and it is stipulated that each D2D receiver can only reuse the frequency resources of one CUE.

[0010] Optionally, in the process of matching appropriate CUE frequency resources for N D2D communication pairs in step 2, an optimal matching algorithm is used for matching. When the number of CUEs is greater than the number of D2D communication pairs, appropriate virtual D2D receivers are added to form a bipartite graph; when the number of CUEs is exactly equal to the number of D2D communication pairs, a bipartite graph is directly formed without the need to add additional virtual users.

[0011] Optionally, the execution process of the optimal matching algorithm includes the following steps:

[0012] Step 1: Construct graph G(V 1 ,V 2 ), V 1 ={x 1 ,x 2 ,…,x M} represents the set of cellular users, V 2 ={y 1 ,y 2 ,…,y N} represents the set of D2D receivers in the D2D communication pair; according to

[0013] l(x)+l(y)=w(x,y),(x,y)∈E(G)

[0014] Construct an equal subgraph G of graph G i ;

[0015] Step 2: Input weight matrix W and resource allocation matrix B t×t = 0, where t = max{N,M};

[0016] Step 3: If any initial match M saturates the CUE user set V1, then M is the best match, and update B t×t , the algorithm terminates;

[0017] Step 4: If M is not a perfect match, let l(y j )=0,l(x i )=max(w(x i ,y j )), execute the Hungarian algorithm based on the initial matching to obtain the optimal matching M';

[0018] Step 5: If M' is saturated with V1, then M' is the best match and the corresponding B is generated. t×t , the algorithm terminates;

[0019] Step 6: If M' is not the best match, execute:

[0020] α 1 =min{l(x)+l(y)-w(x,y)|x∈S,y∈V 2 -T};

[0021] Step 7: According to the formula:

[0022] Determine a new feasible top label l' and construct a new equal subgraph G l' ;

[0023] Step 8: Execute the Hungarian algorithm and generate a new B if a perfect match is obtained. t×t , the algorithm terminates; otherwise, return to step 6 and repeat the operation.

[0024] The present invention provides an optimization method for joint interference elimination and resource allocation in D2D communication. By establishing a system model that generates co-frequency interference, a pair of D2D communication pairs reuses only the frequency resources of one cellular user, that is, the D2D transmitter will interfere with the base station, and the cellular user will interfere with the D2D receiver. Specifically, the maximum weight matching algorithm is used to reasonably allocate resources for the D2D communication pair, so as to minimize the co-frequency interference in the system and enable the D2D communication pair to obtain the maximum throughput. Then, the zero-forcing precoding method is used to eliminate the co-frequency interference of D2D communication to cellular users. The simulation results show that the present invention can not only ensure the quality of cellular communication, but also effectively improve the system throughput. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0026] Figure 1 It is a schematic diagram of intra-system communication of an optimization method for joint interference elimination and resource allocation in D2D communication of the present invention.

[0027] Figure 2 It is a terminal distribution diagram of a specific embodiment of the present invention.

[0028] Figure 3 It is a schematic diagram of the change of the network throughput with the number of D2D communication pairs according to a specific embodiment of the present invention.

[0029] Figure 4 It is a schematic diagram of the variation of the network throughput with the D2D communication distance according to a specific embodiment of the present invention.

[0030] Figure 5 It is a schematic diagram of the variation of the network throughput with the number of terminal antennas in a specific embodiment of the present invention. DETAILED DESCRIPTION

[0031] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0032] The following is an explanation of the English terms used in this invention:

[0033] D2D: Device-to-Device, also known as terminal direct communication;

[0034] CUE: Cell User Equipment, cellular user;

[0035] CSI: Channel State Information, instantaneous channel state information;

[0036] KM: Kuhn-Munkres, optimal matching algorithm.

[0037] See also Figure 1 The present invention provides an optimization method for joint interference elimination and resource allocation in D2D communication, comprising the following steps:

[0038] S1: Establish a system model that generates co-channel interference, in which the co-channel interference is generated only between D2D communication pairs and cellular users that reuse the same frequency, and the number of cellular users is greater than or equal to the number of D2D communication pairs;

[0039] S2: Establish a mathematical model, where the mathematical model is to match appropriate CUE frequency resources to N D2D communication pairs so as to maximize the total throughput of the system.

[0040] The following is a further explanation based on the specific steps:

[0041] In step S1, if Figure 1 As shown in the figure, a system model for generating co-channel interference is established for the case where a pair of D2D communication pairs reuses only the frequency resources of one cellular user. It can be stated that the D2D transmitter will interfere with the base station, and the cellular user will interfere with the D2D receiver. This co-channel interference only occurs between the D2D communication pair and the cellular user that reuse the same frequency.

[0042] There are M cellular users, and C is used to represent the cellular user set C = {C1 ,C 2 ,…,C k ,…,C M}, where C K represents cellular user k;

[0043] There are N pairs of D2D communication pairs, and D represents the set of D2D users, D = {D 1 ,D 2 ,…,D i ,…,D N}, where D i represents a D2D user pair i, and the number of M is greater than or equal to the number of N.

[0044] Use ρ i,k ∈{0,1} is the indicator variable for channel allocation. If D i Reuse C K channel, then ρ i,k =1, otherwise ρ i,k =0.

[0045] For a D2D receiver, its received interference signal-to-noise ratio is expressed as SINR D express,

[0046]

[0047] Where P D Indicates the signal transmission power of the D2D transmitter, N 0 represents additive white Gaussian noise, N C Indicates the co-channel interference of cellular users to the D2D receiver.

[0048] N C Expressed as

[0049] N C =υ+22log10(d k,i )

[0050] Where, d k,i represents the distance between the D2D receiver and the multiplexed cellular user, and υ represents the path-independent signal fading variable in the air.

[0051] υ indicates

[0052] υ=28.0+20log10(f c )

[0053] In the formula, f c Indicates the frequency used by user communication.

[0054] For the base station, the interference it suffers mainly comes from additive white Gaussian noise and co-channel interference generated by the D2D transmitter. Its signal-to-interference-to-noise ratio is expressed as SINR. C express.

[0055]

[0056] Where P C Indicates the uplink signal transmission power of the cellular user, N 0 represents additive white Gaussian noise, N D Indicates the co-channel interference of the D2D transmitter to the base station.

[0057] N D express

[0058] N D =υ+22log10(d i,k )

[0059] Where, d i,k represents the distance from the D2D transmitter to the base station, and υ represents the path-independent signal fading variable in the air.

[0060] The zero-forcing precoding technology is used to eliminate the co-channel interference of the D2D transmitter to the cellular system, and the interference of the cellular user to the D2D receiver is minimized through channel selection. The total throughput of the system is expressed as C T express.

[0061] C T =C C +C D

[0062] In the formula, C C represents the total throughput of cellular users, C D Indicates the total throughput of D2D users.

[0063] C C express

[0064]

[0065] Where P k represents the uplink signal transmission power of the kth cellular user, which is equal to P C .

[0066] C D express

[0067]

[0068] Where P i represents the signal transmission power of the transmitter in the i-th D2D communication pair, where the magnitude is P D .

[0069] Step S2 is the process of establishing a mathematical model, specifically:

[0070] Let the set D N (n=1, 2, ..., N) represents a set of D2D communication receivers, and it is stipulated that each D2D receiver can only reuse the frequency resources of one CUE.

[0071] The model is to match appropriate CUE frequency resources to N D2D communication pairs so that the total throughput of the system is maximized. The set U M (1,2,…,M) represents M cellular users, that is, M≥N.

[0072] When the number of CUEs is greater than the number of D2D communication pairs, appropriate virtual D2D receivers are added to form a bipartite graph. When the number of CUEs in the system is exactly equal to the number of D2D communication pairs, a bipartite graph is directly formed without adding additional virtual users.

[0073] Construct a bipartite graph G = (U, E, W), where D N (n=1,2,…,N) represents the number of D2D communication pairs, and the set E is a series of unordered edge sets of the elements in the vertex set U, which can be expressed by the formula E={e 1 ,e 2 ,…,e n} indicates that w n,m ∈W is the connected D2D receiver D N and CUE side E n,m The weight of represents the interference of CUE to D2D receiver when the D2D communication pair reuses the frequency resources of CUE. Taking the example that the number of CUE M is equal to the number of D2D communication pairs N, the weight matrix W is:

[0074]

[0075] Define an N×N resource allocation matrix B N×N =[b n,m ], b n,m ∈{0,1}, b n,m =1 indicates that D2D communication pair i and CUE user k share resources. The specific process is shown in Table 1.

[0076] Table 1 KM algorithm

[0077]

[0078] When analyzing a pair of D2D communication pairs and a cellular user to eliminate interference to the cellular user, the cellular communication channel is represented by H 1 The interference channel of D2D transmitter to CUE is represented by H21 The transmission channel of the D2D link is represented by H 2 express.

[0079] It is assumed that before the D2D connection is realized, the cellular base station has fully obtained the instantaneous channel state information (CSI) of all links, including the CSI of the D2D link, and the base station notifies the two DUEs to realize the D2D link and communicate using the frequency resources previously allocated to the multiplexed CUE.

[0080] The signal received by the base station in cellular communication is y 1 :

[0081] y 1 =H 1 W 1 X 1 +H 21 W 2 X 2 +Z 1

[0082] Where W 1 is the precoding matrix for cellular communication, X 1 is the signal sent by the base station, W 2 is the precoding matrix for D2D communication, X 2 is the signal sent by the D2D transmitter, Z 1 is the noise interference corresponding to the cellular.

[0083] The trace of the matrix is ​​expressed as:

[0084] tr(E[X 1 X 1 H ])=P 1

[0085] Where tr() is the trace of the matrix, E[] is the expectation of the matrix, and X 1 H is the matrix X 1 The conjugate transpose of 1 It is the signal transmission power of cellular users during uplink.

[0086] Orthogonal column vectors are represented as:

[0087] W 1 =[w 1 (1)…w 1 (S 1 )]

[0088] W 1 is a cellular communication complex precoding matrix (W 1H W 1 =I), signal 1 to S 1 is the signal to be sent. The size of the matrix depends on the number of transmitting antennas. By performing singular value decomposition on the channel matrix and precoding the transmitted signal, the cellular signal can be obtained.

[0089] After the linear receiver is used to receive the signal, the received signal is processed to obtain the transmitted signal. The received signal is expressed as:

[0090] Q 1 =[q 1 (1)…q 1 (S 1 )]

[0091] The sum of the transmission power of all transmitting antennas is fixed, and the value is P 1 By constraining the D2D signal to 21 It is sent in the null space to form zero interference transmission.

[0092] Interference channel H 21 The SVD decomposition of is:

[0093] H 21 =U 21 ∑ 21 V 21 H

[0094] V 21 It is expressed as:

[0095] V 21 =[V 21 (1)…V 21 (A 2 )V 21 (A 2 +1)V 21 (A 2 +A 1 )]

[0096] Define V 21 part of express:

[0097] V 21 ⊥ =[V 21 (A 2 +1)V 21 (A 2 +A 1 )]

[0098] As the basis of null space, it has the following properties:

[0099]

[0100]

[0101] It is orthogonal to the interference channel and is used to eliminate interference.

[0102] Now form a projection matrix

[0103]

[0104] The precoding matrix can be expressed as:

[0105]

[0106] It is used to ensure that the D2D precoding vector and the interference channel are orthogonal, and the D2D link will always be transmitted in the null space, which can avoid interference with cellular communications. 1 is rewritten as:

[0107] y 1 =H 1 W 1 X 1 +Z 1

[0108] Since the D2D link communicates by reusing the frequency resources of the CUE, when the null space of the interference channel is found, the co-frequency interference caused by D2D communication to cellular communication can be resolved, effectively ensuring the communication quality of the cellular link.

[0109] Furthermore, in order to verify the effectiveness of the method of the present invention, a specific embodiment is proposed and the following simulation experiments are performed:

[0110] Reasonable and random allocation of users within the coverage area of ​​the base station. The main simulation parameters are shown in Table 2.

[0111] Table 2 Simulation model

[0112]

[0113] 10 cellular terminals and 5 D2D communication pairs are used as an example, and the communication radius of the D2D communication pair is indicated by a small circle. Figure 2 .

[0114] Three experiments were set up. The first experiment was to show the change of network throughput with the number of D2D communication pairs. The number of CUEs was 30, the number of D2D communication pairs was 5-30, and the distance between D2D communication pairs was set to 20m. The results are shown in the figure below. Figure 3As shown, ZF represents the interference elimination method for cellular users, and joint represents the method of the present invention, which combines the throughput of CUE under the ZF method and the throughput of DUE under the KM method. It can be seen from the figure that the system throughput under the joint method is 50% higher than that of the random method.

[0115] The second experiment is the variation of network throughput with the distance of D2D communication pairs; the number of CUEs is 30, the number of D2D communication pairs is 5, 15, and 30, and the distance of D2D communication pairs is set to 20-100m. The results are as follows Figure 4 As shown, using the ZF method to eliminate interference to cellular communications can avoid losing about 25% of the channel capacity compared to random conditions, effectively protecting the rate of cellular communications.

[0116] The third experiment is the change of network throughput with the number of terminal antennas; the number of CUEs is 20, the number of D2D communication pairs is 20, the distance between D2D communication pairs is set to 50-450m, and the number of antennas is 2x2, 4x4, and 8x8. The results are as follows: Figure 5 As shown, three different antenna configurations are considered: (1) ignoring co-channel interference; (2) random conditions; and (3) the joint optimization method of the present invention. When the number of antennas is small, as the distance increases, the gain of D2D communication is not enough to compensate for the interference, and the method of the present invention has obvious advantages.

[0117] In summary, compared with the prior art, the present invention has the following characteristics:

[0118] (1) Frequency resources are fully utilized.

[0119] (2) Ensure the quality of cellular communications and effectively improve system throughput.

[0120] What is disclosed above is only a preferred embodiment of the present invention, and it certainly cannot be used to limit the scope of rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiment and equivalent changes made according to the claims of the present invention still fall within the scope of the invention.

Claims

1. A method for optimizing joint interference cancellation and resource allocation in D2D communication, characterized in that: The following steps are involved: Step 1: Establish a system model that generates co-channel interference. In the system model, co-channel interference only occurs between D2D communication pairs and cellular users that reuse the same frequency, and the number of cellular users is greater than or equal to the number of D2D communication pairs. Step 2: Establish a mathematical model, where the mathematical model is to match appropriate CUE frequency resources to N D2D communication pairs so as to maximize the total throughput of the system.

2. The method for optimizing joint interference cancellation and resource allocation in D2D communication according to claim 1, characterized in that: Each D2D communication pair includes a D2D transmitter and a D2D receiver. In step 1, the co-channel interference of the D2D transmitter to the cellular system is eliminated by zero-forcing precoding technology, and the interference of cellular users to the D2D receiver is minimized by channel selection.

3. The optimization method for joint interference cancellation and resource allocation in D2D communication according to claim 2, characterized in that: During the execution of step 2, let the set D N (n=1, 2, ..., N) represents a set of D2D communication receivers, and it is stipulated that each D2D receiver can only reuse the frequency resources of one CUE.

4. The optimization method for joint interference cancellation and resource allocation in D2D communication according to claim 3, characterized in that: In the process of matching appropriate CUE frequency resources to N D2D communication pairs in step 2, the optimal matching algorithm is used for matching. When the number of CUEs is greater than the number of D2D communication pairs, appropriate virtual D2D receivers are added to form a bipartite graph. When the number of CUEs is exactly equal to the number of D2D communication pairs, a bipartite graph is directly formed without the need to add additional virtual users.

5. The optimization method for joint interference cancellation and resource allocation in D2D communication according to claim 4, characterized in that: The execution process of the optimal matching algorithm includes the following steps: Step 1: Construct a graph G(V1,V2), V1={x1,x2,…,x M } represents the set of cellular users, V2 = {y1,y2,…,y N } represents the set of D2D receivers in the D2D communication pair; according to l(x)+l(y)=w(x,y),(x,y)∈E(G) Construct an equal subgraph G of graph G i ; Step 2: Input weight matrix W and resource allocation matrix B t×t = 0, where t = max{N,M}; Step 3: If any initial match M saturates the CUE user set V1, then M is the best match, and update B t×t , the algorithm terminates; Step 4: If M is not a perfect match, let l(y j )=0,l(x i )=max(w(x i ,y j )), execute the Hungarian algorithm based on the initial matching to obtain the optimal matching M'; Step 5: If M' is saturated with V1, then M' is the best match and the corresponding B is generated. t×t , the algorithm terminates; Step 6: If M' is not the best match, execute: α1=min{l(x)+l(y)-w(x,y)|x∈S,y∈V2-T}; Step 7: According to the formula: Determine a new feasible top label l' and construct a new equal subgraph G l' ; Step 8: Execute the Hungarian algorithm and generate a new B if a perfect match is obtained. t×t , the algorithm terminates; otherwise, return to step 6 and repeat the operation.

Images