Multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing network

Abstract

The invention provides a multi-antenna channel transmission method for a large-scale unmanned aerial vehicle self-organizing network. The method comprises the following steps: acquiring a current channel state sensing range, channel state information feedback precision and a target communication parameter; determining an average achievable rate with feedback quantization loss according to the channel state information feedback precision, the channel state sensing range and the target communication parameters; determining the routing quality in the channel state sensing range according to the current channel state sensing range and the target communication parameter; judging whether the product of the routing quality and the average achievable rate with the feedback quantization loss in the current channel state sensing range is maximized or not under the constraint condition, and if not, taking the product of the routing quality and the average achievable rate with the feedback quantization loss as the target to be maximized under the constraint condition; and updating the channel state sensing range and the channel state information feedback precision. By implementing the invention, the average transmission delay can be reduced, and the transmission efficiency can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of self-organizing networks, and in particular relates to a multi-antenna channel transmission method for a large-scale unmanned aerial vehicle self-organizing network. Background Art

[0002] Large-scale intelligent unmanned clusters rely on ad hoc network technology to provide a reliable communication backbone for these unmanned clusters. Wireless ad hoc networks can be provided without the need for pre-deployment of infrastructure, making them flexible. Limited by the transmission capacity of a single antenna system, traditional ad hoc networks are difficult to meet the demand for broadband transmission between large-scale unmanned nodes in the future. In addition, due to severe interference between nodes, traditional Ad hoc networks need to consume more resources to implement access scheduling, thus limiting the expansion of the network scale.

[0003] Multiple-input multiple-output (MIMO) technology can achieve significant improvements in transmission performance. It can not only provide higher point-to-point transmission rates, but also reduce inter-node interference through beamforming. Therefore, MIMO is a key technology for building large-scale Ad Hoc networks. In related technologies, the asymptotic capability of multi-hop MIMO ad hoc networks has been demonstrated, and a transmission strategy selection method for MIMO ad hoc networks has been proposed, in which nodes can flexibly adjust the precoding model according to data transmission requirements. However, these studies assume that the transmitter can obtain accurate channel state information (CSI), which is difficult in practice. Some studies have obtained a closed-form expression for the precoding performance of MIMO systems with limited CSI feedback. On this basis, some studies further analyzed the impact of channel feedback loss on the transmission performance of MIMO ad hoc networks and proposed a flipping codebook design method with CSI feedback constraints to improve the precoding performance of MIMO systems.

[0004] However, multi-hop transmission is required in ad hoc networks, so it is not enough to analyze single-hop channel feedback alone. In order to obtain better routing performance, nodes in ad hoc networks must sense channel information within a multi-hop range. For multi-antenna ad hoc networks, in order to take advantage of the transmission advantages of MIMO systems, each node requires wide-range channel sensing and accurate feedback, which will lead to a lot of resource overhead when the network scale is large. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide a multi-antenna channel transmission method and device for large-scale unmanned aerial vehicle self-organizing networks to meet the needs of reducing resource overhead and improving transmission efficiency.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] The present invention provides a multi-antenna channel transmission method for a large-scale unmanned aerial vehicle self-organizing network, comprising: obtaining a current channel state perception range, channel state information feedback accuracy and target communication parameters; determining an average achievable rate with feedback quantization loss according to the channel state information feedback accuracy, the channel state perception range and the target communication parameters; determining a routing quality within the channel state perception range according to the current channel state perception range and the target communication parameters; judging whether the product of the routing quality within the current channel state perception range and the average achievable rate with feedback quantization loss is maximized under constraints, and if it is not maximized, updating the channel state perception range and the channel state information feedback accuracy with the goal of maximizing the product of the routing quality and the average achievable rate with feedback quantization loss under constraints.

[0008] Optionally, the target communication parameters include a first target communication parameter, and the routing quality within the channel state perception range is determined based on the current channel state perception range and the target communication parameter, including: determining the average number of sensed nodes based on the current channel state perception range; determining the average weight of the best path selection within the channel state perception range based on the average number of sensed nodes and the first target communication parameter; determining the routing quality within the channel state perception range based on the average weight of the best route selection.

[0009] Optionally, the target communication parameters include a second target communication parameter and a third target communication parameter, and an average achievable rate with feedback quantization loss is determined according to the channel state information feedback accuracy, the channel state perception range and the target communication parameters, including: determining the time overhead of the sensing phase according to the channel state perception range and the second target communication parameter; determining the time overhead of the feedback phase according to the channel state information feedback accuracy and the third target communication parameter; determining the average achievable rate with feedback quantization loss according to the time overhead of the sensing phase and the time overhead of the feedback phase.

[0010] Optionally, the second communication parameters include the channel information quantization length, the number of antennas for diversity transmission, the bandwidth, and the average signal-to-noise ratio of the receiving antenna. The time overhead of the sensing phase is determined according to the channel state perception range and the second target communication parameters, including: determining the average broadcast overhead of the sensing phase according to the channel state perception range of each node and the channel information quantization length; determining the sensing transmission rate according to the number of antennas for diversity transmission, the bandwidth, and the average signal-to-noise ratio of the receiving antenna; determining the time overhead of the sensing phase according to the average broadcast overhead and the sensing transmission rate of the sensing phase.

[0011] Optionally, the third communication parameter includes codebook information and node computing performance in the feedback phase, and determines the time overhead of the feedback phase according to the channel state information feedback accuracy and the third target communication parameter, including: determining the time required to calculate the optimal precoding vector according to the codebook information and node computing performance in the feedback phase; determining the time required to transmit the index to the sender through the feedback channel according to the index of the optimal precoding vector and the channel state information feedback accuracy; determining the time overhead of the feedback phase according to the time required to calculate the optimal precoding vector and the time required to transmit the index to the sender through the feedback channel.

[0012] Optionally, the routing quality formula under the channel state perception range is as follows:

[0013]

[0014] in, E[τ(Q(n))] represents the average weight of Q(n), Q(n) represents the optimal path selection for the sensing range n, ω(n) represents the average number of sensed nodes, n represents the channel state sensing range, M represents the number of antennas, and s represents the number of path hops from the source to the destination. and p s denote the binomial distribution parameters of the cardinality of a set of s-hop paths.

[0015] Optionally, the time cost calculation formula of the sensing phase is:

[0016]

[0017] in, E[A n ] represents the average number of n hops, K represents the number of nodes in the ad hoc network, p represents the probability of connectivity between any two nodes, n represents the channel state perception range, and b 2 Represents the channel information quantization length, R 1 =Wlog 2 (1+Mρ), W represents the bandwidth, ρ represents the average signal-to-noise ratio (SNR) of each receiving antenna, and M represents the number of antennas.

[0018] Optionally, the time overhead of the feedback phase includes:

[0019] T 2 =T t +T c ;

[0020] Among them, T c represents the time required to calculate the optimal precoding vector, b 1represents the channel state information feedback accuracy, μ represents the node computing performance, that is, the time it takes for the node to calculate a complex operation, T t It represents the time required for the index to be transmitted to the sender through the feedback channel. R 2 =Wlog 2 (1+Mρ), W represents the bandwidth, ρ represents the average signal-to-noise ratio (SNR) of each receiving antenna, and M represents the number of antennas.

[0021] Optionally, determining an average achievable rate with feedback quantization loss according to a time overhead of a sensing phase and a time overhead of a feedback phase includes:

[0022]

[0023] in, T 0 Indicates the start time, T 1 Indicates the time consumed in the sensing phase, T 2 represents the total time overhead of the feedback phase, W represents the bandwidth, ρ represents the average signal-to-noise ratio (SNR) of each receiving antenna, and λ max Representation Matrix The maximum eigenvalue of 1 represents the channel state information feedback accuracy, and M represents the number of antennas.

[0024] Optionally, the channel state perception range and the channel state information feedback accuracy are updated with the goal of maximizing the product of the routing quality and the average achievable rate with feedback quantization loss under constraints, including: maximizing the product of the routing quality and the average achievable rate with feedback quantization loss under constraints includes:

[0025] P 1 :

[0026] stn∈N + ;

[0027] n≤K-1,

[0028] b 1 ∈N + ,

[0029] T 1 +T 2 <T 0 ,

[0030] Among them, R real represents the average achievable rate with feedback quantization loss, R routing represents the routing quality, n represents the channel state perception range, b 1represents the channel state information feedback accuracy, N + represents a positive integer, K represents the number of nodes in the ad hoc network, T 0 Indicates the start time, T 1 Indicates the time consumed in the sensing phase, T 2 Represents the total time cost of the feedback phase;

[0031] The product of maximizing the routing quality and the average achievable rate with feedback quantization loss is converted into the first optimization objective and the second optimization objective, including:

[0032] P 2 :

[0033] P 3 :

[0034] Among them, E[τ(Q(n))] represents the average weight of Q(n), and Q(n) represents the optimal path selection within the sensing range n.

[0035] This embodiment provides a multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing networks, analyzes the impact of the sensing range and channel state information feedback accuracy on the transmission performance, and constructs an optimization problem based on the impact of the sensing range and channel state information feedback accuracy on the transmission performance, determines the optimal sensing range and feedback accuracy for multi-hop data transmission, and adjusts the self-organizing network with the optimal sensing range and feedback accuracy, which can obtain a lower average transmission delay, reduce resource overhead, improve transmission efficiency, and ensure the transmission advantages of the multi-antenna system. Simulation results show that compared with the traditional sensing and feedback method, the average transmission delay of the method proposed in this embodiment has been significantly improved.

[0036] Other advantages, objectives and features of the present invention will be described in the following description and will be apparent to those skilled in the art to some extent, or those skilled in the art may be taught from the practice of the present invention. The objectives and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to make the purpose, technical solution and beneficial effects of the present invention clearer, the present invention provides the following drawings for illustration:

[0038] Figure 1 A specific example flow chart of a multi-antenna channel transmission method for a large-scale unmanned aerial vehicle self-organizing network according to the present invention;

[0039] Figure 2 A schematic diagram of time resource allocation for a MIMO ad hoc network in the present invention;

[0040] Figure 3 is a graph showing the relationship between the broadcast quantity and the sensing distance in the present invention;

[0041] Figure 4 A graph showing the number of nodes and average transmission delay changes under different numbers of antennas for the method proposed in the embodiment of the present invention and the traditional method;

[0042] Figure 5 It is a principle block diagram of a specific example of an electronic device in an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, it can also be the internal connection of two components, it can be a wireless connection, or it can be a wired connection. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0045] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0046] The embodiment of the present invention provides a multi-antenna channel transmission method for a large-scale unmanned aerial vehicle self-organizing network. Figure 1 As shown, including:

[0047] S101, obtaining a current channel state perception range, channel state information feedback accuracy, and target communication parameters;

[0048] S102, determining an average achievable rate with feedback quantization loss according to a channel state information feedback accuracy, a channel state perception range, and a target communication parameter;

[0049] S103, determining a routing quality within the channel state sensing range according to a current channel state sensing range and a target communication parameter;

[0050] S104, judging whether the product of the routing quality under the current channel state perception range and the average achievable rate with feedback quantization loss is maximized under the constraint conditions, if not, jumping to S105; if reached, ending.

[0051] S105, with the goal of maximizing the product of the routing quality and the average achievable rate with feedback quantization loss under constraints, update the channel state perception range and the channel state information feedback accuracy, and re-enter step S101 until the maximization is achieved.

[0052] For example, in this embodiment, a self-organizing network including K nodes is considered, and each node is equipped with M antennas. Assume that the probability of connection between any two nodes i and j is p. When they are connected, the channel between i and j is represented by an M-dimensional square matrix H ij Without loss of generality, assume that every element h in H ij All follow a circularly symmetric complex Gaussian distribution, that is, h ij ~CN(0,1). When node i sends message s to node j i When , the signal received by node j can be expressed as:

[0053] y j =H ij cs i +w; (1)

[0054] Where c is the precoding vector and w is the additive noise at the receiver. Ideally, the transmitter can obtain perfect CSI H ij , then the point-to-point achievable rate is:

[0055] R=Wlog 2 (1+ρλ max ); (2)

[0056] Where W is the bandwidth, ρ is the average signal-to-noise ratio (SNR) of each receiving antenna, and λ max is a matrix The maximum eigenvalue of Indicates H ij The conjugate transpose of .

[0057] However, due to limited network resources, complete feedback of CSI H is impractical. Therefore, this embodiment uses a predefined codebook for precoding. Assume that each node uses a b 1 Bit code book The receiver will select the optimal precoding vector c from C to achieve:

[0058]

[0059] Compared with the traditional single-antenna ad hoc network, the MIMO ad hoc network considered in the embodiment of the present invention can obtain better data transmission performance through precoding. However, it can be seen from formula (3) that the MIMO system requires accurate channel estimation and feedback to achieve perfect precoding design. In addition, the channel sensing distance also plays a decisive role in the routing quality of the MIMO ad hoc network. Therefore, the impact of channel sensing and feedback on system performance will be analyzed in detail.

[0060] This embodiment provides a schematic diagram of time resource allocation in a MIMO ad hoc network. Figure 2 As shown, assuming that the channel of the adhoc network is at time T 0 The period remains unchanged, which can be regarded as a complete time slot. Then, the entire time slot is divided into three parts, corresponding to the three stages, namely the sensing stage, the feedback stage and the transmission stage. The time occupied by the three stages is represented as T 1 , T 2 and T 0 -T 1 -T 2 .

[0061] Sensing stage: At each T 0 At the beginning, the channel sensing phase is introduced to collect channel information in a hop-by-hop manner. In this phase, when the sensing range is n, each node sends the preamble sequence and CSI of its n-1 hop neighbors through multiple broadcasts. The sensing results are the basis of MIMO precoding and multi-hop routing in adhoc networks.

[0062] Feedback phase: After the sensing phase, each node can estimate the exact channel matrix H from its 1-hop neighbors based on the received pilot sequence. When a node needs to receive data, it determines the optimal coding vector c based on the real-time channel H and the codebook C. The index of c will be fed back to the transmitter.

[0063] Transmission phase: Finally, in the data transmission phase, the source will determine the best route to the destination based on the CSI results in the sensing phase, and then send the data to the next hop based on the precoding index obtained in the feedback phase, as shown in formula (1).

[0064] from Figure 2 It can be seen that the sensing distance in the sensing phase affects the routing quality of multi-hop data transmission, while the feedback accuracy (codebook size) in the feedback phase affects the quality of point-to-point data transmission. However, increasing the CSI sensing distance and channel feedback accuracy will generate greater resource overhead, resulting in a reduction in transmission resources. Therefore, in resource-constrained ad hoc networks, it is necessary to jointly determine the optimal sensing range and feedback accuracy based on the network scale and transmission environment to obtain better transmission.

[0065] Therefore, this embodiment solves the impact of the channel state perception range and the channel state information feedback accuracy on the routing quality within the channel state perception range and the average achievable rate with feedback quantization loss, and adjusts the channel state perception range and the channel state information feedback accuracy with the goal of maximizing the product of the routing quality and the average achievable rate with feedback quantization loss, thereby obtaining better transmission.

[0066] It should be noted that, in this embodiment, the target communication parameters may include the channel information quantization length, the number of antennas for diversity transmission, the bandwidth and the average signal-to-noise ratio of the receiving antenna, the codebook information in the feedback phase, the node computing performance, and the like.

[0067] In step S102, the average achievable rate with feedback quantization loss is determined according to the channel state information feedback accuracy, the channel state perception range and the target communication parameters, including: determining the time overhead of the sensing phase according to the channel state perception range and the second target communication parameter; determining the time overhead of the feedback phase according to the channel state information feedback accuracy and the third target communication parameter; determining the average achievable rate with feedback quantization loss according to the time overhead of the sensing phase and the time overhead of the feedback phase.

[0068] Among them, the second communication parameters include the channel information quantization length, the number of antennas for diversity transmission, the bandwidth and the average signal-to-noise ratio of the receiving antenna. According to the channel state perception range and the second target communication parameters, the time overhead of the sensing phase is determined, including: determining the average broadcast overhead of the sensing phase according to the channel state perception range of each node and the channel information quantization length; determining the sensing transmission rate according to the number of antennas for diversity transmission, the bandwidth and the average signal-to-noise ratio of the receiving antenna; determining the time overhead of the sensing phase according to the average broadcast overhead and the sensing transmission rate of the sensing phase.

[0069] Specifically, without losing generality, we take node 1 as an example. In the sensing phase, each node in the network first broadcasts a pilot sequence, and node 1 can estimate the exact signal H based on the pilot sequence of its 1-hop neighbor j. 1,j Due to the multi-antenna system, the transmit and receive channels of node 1 are not completely reciprocal. Therefore, it is assumed that node 1 only obtains its own receive channel from its 1-hop neighbor through a single broadcast, and does not know the transmit channel to its 1-hop neighbor, that is, the current sensing range is 0.5 hop.

[0070] As shown in formula (2), λ max is the determining factor of the achievable rate. Therefore, each node will calculate all the channels obtained For example, at node 1, we get The maximum eigenvalue of , and then quantize them into b 2bits for further transmission. Then, all nodes will add the quantized feature values ​​of their 1-hop neighbors to the broadcast data packet for a second broadcast. In this way, node1 not only obtains the transmission channel information from itself to its 1-hop neighbors, but also obtains the receiving channel information of its 2-hop neighbors from its 1-hop neighbors. At this time, the sensing range of node 1 is 1.5 hops. The relationship between the broadcast quantity and the sensing distance is as follows Figure 3 As shown, multiple broadcasts result in a larger sensing range.

[0071] Since the overhead of broadcasting only the pilot sequence is small, the overhead of the first broadcast can be ignored. The amount of data required for the second broadcast depends on the amount of data per node A. 1 The number of 1-hop neighbors and the channel information quantization length b 2 , that is, the amount of data for the second broadcast is A 1 b 2 Similarly, the amount of data required for the (n+1)th broadcast is The corresponding perception range is n.5 steps. The relationship between the number of broadcasts, the relationship between the broadcast overhead and the perception range is shown in Table 1.

[0072] Table 1. Correspondence between broadcast number and overhead

[0073] Number of broadcasts Broadcast overhead Perception range 1 0 0.5 2 <![CDATA[A 1 b 2 ]]> 1.5 3 <![CDATA[2A 1 b 2 +A 2 b 2 ]]> 2.5 ... ... ... n+1 <![CDATA[2A 1 b 2 +2A 2 b 2 +…+A n b 2 ]]> n.5 ... ... ... K <![CDATA[2A 1 b 2 +2A 2 b 2 +…+A K-1 b 2 ]]> (K-1)

[0074] When the CSI sensing range of each node is n hops, the total overhead is O 1 (n) is the sum of the first n+1 broadcasts, that is:

[0075]

[0076] For analysis purposes, this embodiment calculates the average broadcast overhead, such as:

[0077]

[0078] The average number of single hops is Then, the average number of n hops is approximately:

[0079]

[0080] Assume that the transmission rate of the broadcast channel used for CSI sensing is R 1 , then the time consumed in the sensing phase is:

[0081]

[0082] The third communication parameter includes the codebook information and node computing performance of the feedback phase. The time overhead of the feedback phase is determined according to the channel state information feedback accuracy and the third target communication parameter, including: determining the time required to calculate the optimal precoding vector according to the codebook information and node computing performance of the feedback phase; determining the time required to transmit the index to the sender through the feedback channel according to the index of the optimal precoding vector and the channel state information feedback accuracy; determining the time overhead of the feedback phase according to the time required to calculate the optimal precoding vector and the time required to transmit the index to the sender through the feedback channel.

[0083] Specifically, the time overhead of the feedback phase consists of two parts. The first part is the time T required for the node to calculate the optimal precoding vector based on the codebook and the real-time channel H. c , and the other part is the time T required for the index of the vector to be transmitted to the sender through the feedback channel t .

[0084] As mentioned above, b 1 The bit is used for CSI feedback in the feedback phase. The feedback transmission rate R is used. 2 The transmission time can be calculated by In addition, a larger codebook will introduce more computation and selection time for each node. Since the dimension of matrix H is M×M, the dimension of codebook á is So the calculation time T c It can be quantified as:

[0085]

[0086] Where μ is the time it takes for a node to calculate a complex operation. Then the total time cost of the feedback phase is:

[0087] T 2 =T t +T c (9)

[0088] For the CSI sensing transmission rate R 1 and feedback transmission rate R 2 , this embodiment uses the following method to solve:

[0089] In order to achieve stable transmission of sensing information and feedback information, this embodiment adopts diversity transmission in the sensing stage and the feedback stage. By evenly distributing the transmission power to M antennas, the signal received by node j from node i can be expressed as:

[0090]

[0091] Among them, X T is the transpose of X, so the average SNR can be expressed as:

[0092]

[0093] Among them, ||X|| F is the Frobenius norm of the matrix X. Since h ij ~CN(0,1), we have:

[0094]

[0095] Therefore, the CSI sensing transmission rate R 1 and feedback transmission rate R 2 It can be calculated as follows:

[0096] R 1 =R 2 =Wlog 2 (1+Mρ); (13)

[0097] The above-mentioned time cost of the sensing stage and the time cost of the feedback stage are obtained. According to the time cost of the above two stages, the average achievable rate with feedback quantization loss is determined, specifically:

[0098] According to formula (2), when a singular value decomposition-based precoding scheme is adopted, the achievable rate of complete CSI feedback can be given by the following formula, taking into account the resource overhead of the sensing stage and the feedback stage:

[0099]

[0100] Where T0 represents the start time, T1 represents the time consumed in the sensing phase, T2 represents the total time overhead in the feedback phase, W represents the bandwidth, ρ represents the average signal-to-noise ratio (SNR) of each receiving antenna, and λ max Representation Matrix The maximum eigenvalue of .

[0101] With limited feedback resources 1 In the case of , the average achievable rate with feedback quantization loss can be expressed as:

[0102]

[0103] Among them, α is the coefficient, that is:

[0104]

[0105] In step S103, the routing quality within the channel state perception range is determined based on the current channel state perception range and the target communication parameters. The specific process includes: determining the average number of sensed nodes based on the current channel state perception range; determining the average weight of the best path selection within the channel state perception range based on the average number of sensed nodes and the first target communication parameters; determining the routing quality within the channel state perception range based on the average weight of the best route selection.

[0106] Specifically, in an Ad hoc network, the quality of multi-hop routing depends on the sensing range of the node. In this embodiment, the As an approximation of the delay of each hop. For an N-dimensional square matrix H whose elements obey a circularly symmetric complex Gaussian distribution, The maximum eigenvalue λ max It follows Tracy-Widom (TW) type 2 distribution, with mean u M =4M, the variance is Right now:

[0107]

[0108] Among them, TW 2 is a random variable following TW 2 distribution. Therefore, formula (17) can be rewritten as:

[0109] λ max ≈u M +δ M ·TW 2 ; (18)

[0110] Then, the single-hop transmission delay can be expressed as:

[0111]

[0112] The total amount of information to be sent is set to 1. Taylor expansion at , formula (19) can be approximated as:

[0113]

[0114] Ignoring the second term on the right, it can be approximated as a random variable following an exponential distribution with parameter 4M, that is: The probability density function (PDF) is:

[0115]

[0116] Here, let the set of all paths between source and destination be:

[0117] Φ={φ 1 ,φ2 ,…,φ 8 ,…,φ K-1}; (twenty two)

[0118] Among them, φ s is a set of paths with s hops. Since the weight (delay) of each hop follows an exponential distribution, the weight γ of any s-hop path is s ∈φ s It obeys the gamma distribution with parameters s and 4M, that is:

[0119] τ(γ s )~Γ(s,4M); (23)

[0120] Among them, τ(γ s ) represents the path γ s The total multi-hop weight of s ) can be expressed as:

[0121]

[0122] where x! is the factorial of x. is the path with the minimum total multi-hop weight, that is, So, The CDF of can be calculated as:

[0123]

[0124] Among them, V s =|φ s | is the set φ s The cardinality of .

[0125] Define Z as the path from source to destination with the minimum multi-hop weight, which can be expressed as:

[0126] τ(Z)=min{τ(Φ)}; (26)

[0127] Therefore, the CDF of τ(Z) is:

[0128]

[0129] In order to avoid the influence of the source and destination, it is necessary to obtain the average multi-hop weight of the network, which can be done by:

[0130]

[0131] Since it is difficult to calculate the formula (28) about V s , this embodiment is approximately:

[0132]

[0133] When there is an s-hop path from the source to the destination, this means that there are s-1 sequentially connected nodes as relay nodes among the remaining K-2 nodes, and s links sequentially connect them. Therefore, |V s |=x is:

[0134]

[0135] Obviously, V s It follows a binomial distribution with parameters and p m The expected value can be obtained by the following formula:

[0136]

[0137] Then, formula (29) can be rewritten as:

[0138]

[0139] In practical applications, it is difficult to obtain the global CSI of all nodes, and the source can only select the local optimal path based on the CSI within the sensing range n obtained in the sensing stage. The optimal path selection for the sensing range n under the conditions represented in this embodiment is Q(n). According to formula (6), when the sensing range is n, the average number of sensed nodes is approximately Where [x] represents the rounding function. ω(n) nodes can form at most ω(n) hop paths, assuming that these paths are within the sensing range of the source. By following the same logic as in formula (32), the average weight of Q(n) is:

[0140]

[0141] In this embodiment, since the weight of the total multi-hop path represents the transmission delay, it is hoped to find the path with the minimum weight. In order to combine the routing performance with the point-to-point transmission performance, the routing quality with a sensing range of n is defined as:

[0142]

[0143] In steps S104 and S105, the goal of this embodiment is to obtain the best CSI feedback accuracy b 1 and sensing range n. It is worth noting that since few bits are sufficient to accurately reflect the CSI of the link, the sensing overhead b 2 Therefore, it can be considered that b 2 is a constant. The optimal problem can be written as:

[0144] P 1 :

[0145] stn∈N + ; (34a)

[0146] n≤K-1, (34b)

[0147] b 1 ∈N + , (34c)

[0148] T 1 +T 2 <T 0 , (34d)

[0149] Among them, the optimal problem P 1 is to obtain the optimal n and b in terms of the cost and reward in the sensing stage and the feedback stage 1 Constraints (34a) and (34b) ensure that n is a positive integer describing the number of hops, and that the maximum number of hops does not exceed the network size. Constraint (34c) ensures that b 1 is an integer number of bits. Finally, constraint (34d) means that the time overhead of the sensing phase and the feedback phase cannot exceed the stable period of the channel.

[0150] Since the optimization problem P 1 It is difficult to solve, so b can be updated sequentially in an iterative manner 1 and n. According to formula (34), given b 2 and n can be obtained from the following optimization problem:

[0151] P 2 : Similarly, given n, the optimal b 1 It can be determined by the following optimization problem:

[0152] P 3 :

[0153] This embodiment provides a multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing networks. It analyzes the impact of the sensing range and channel state information feedback accuracy on the transmission performance, and constructs an optimization problem based on the impact of the sensing range and channel state information feedback accuracy on the transmission performance, determines the optimal sensing range and feedback accuracy for multi-hop data transmission, and adjusts the self-organizing network with the optimal sensing range and feedback accuracy, which can obtain a lower average transmission delay, reduce resource overhead, improve transmission efficiency, and ensure the transmission advantages of the multi-antenna system. Simulation results Figure 4 It is shown that, compared with the traditional sensing and feedback method, the average transmission delay of the method proposed in this embodiment is significantly improved.

[0154] The present application also provides an electronic device, such as Figure 5 As shown, a processor 501 and a memory 502, wherein the processor 501 and the memory 502 may be connected via a bus or other means.

[0155] The processor 501 may be a central processing unit (CPU). The processor 501 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or a combination of the above chips.

[0156] The memory 502 is a non-transitory computer-readable storage medium that can be used to store non-transitory software programs, non-transitory computer executable programs and modules, such as program instructions / modules corresponding to the multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing networks in the embodiment of the present invention. The processor executes various functional applications and data processing of the processor by running the non-transitory software programs, instructions and modules stored in the memory.

[0157] The memory 502 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the processor, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory 502 may optionally include a memory remotely arranged relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0158] The one or more modules are stored in the memory 502, and when executed by the processor 501, the following is performed: Figure 1 A multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing networks in the illustrated embodiment.

[0159] For details of the above electronic equipment, please refer to Figure 1 The corresponding related descriptions and effects in the illustrated embodiments can be understood and will not be repeated here.

[0160] This embodiment also provides a computer storage medium, which stores computer executable instructions, and the computer executable instructions can execute the multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing network in any of the above method embodiments. Among them, the storage medium can be a disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), a flash memory, a hard disk (HDD) or a solid-state drive (SSD), etc.; the storage medium can also include a combination of the above types of memory.

[0161] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made in form and details without departing from the scope defined by the claims of the present invention.

Claims

1. A multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing networks, characterized in that: include: Obtain the current channel state perception range, channel state information feedback accuracy, and target communication parameters; Determine an average achievable rate with feedback quantization loss based on channel state information feedback accuracy, channel state perception range, and target communication parameters; Determine the routing quality within the channel state perception range according to the current channel state perception range and the target communication parameters; Determine whether the product of the routing quality under the current channel state perception range and the average achievable rate with feedback quantization loss is maximized under the constraints. If not, update the channel state perception range and channel state information feedback accuracy with the goal of maximizing the product of the routing quality and the average achievable rate with feedback quantization loss under the constraints.

2. According to claim 1, a multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing network is characterized in that: The target communication parameter includes a first target communication parameter, and determining the routing quality within the channel state perception range according to the current channel state perception range and the target communication parameter includes: Determine the average number of sensed nodes according to the current channel state sensing range; Determine an average weight of the best path selection within the channel state sensing range according to the average number of sensed nodes and the first target communication parameter; The routing quality within the channel state perception range is determined according to the average weight of the best route selection.

3. The multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing network according to claim 1 is characterized in that: The target communication parameters include a second target communication parameter and a third target communication parameter. According to the channel state information feedback accuracy, the channel state perception range, and the target communication parameters, an average achievable rate with feedback quantization loss is determined, including: Determine the time cost of the sensing phase according to the channel state sensing range and the second target communication parameter; Determining the time overhead of the feedback phase according to the channel state information feedback accuracy and the third target communication parameter; Based on the time overhead of the sensing phase and the time overhead of the feedback phase, the average achievable rate with feedback quantization loss is determined.

4. The multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing network according to claim 3 is characterized in that: The second communication parameters include the channel information quantization length, the number of antennas for diversity transmission, the bandwidth, and the average signal-to-noise ratio of the receiving antenna. According to the channel state perception range and the second target communication parameters, the time overhead of the sensing phase is determined, including: According to the channel state perception range of each node and the quantization length of the channel information, the average broadcast overhead of the sensing phase is determined; Determine the sensing transmission rate according to the number of antennas for diversity transmission, bandwidth, and average signal-to-noise ratio of the receiving antennas; The time cost of the sensing phase is determined according to the average broadcast cost and the sensing transmission rate in the sensing phase.

5. The multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing network according to claim 4 is characterized in that: The third communication parameter includes codebook information and node computing performance in the feedback phase. The time overhead of the feedback phase is determined according to the channel state information feedback accuracy and the third target communication parameter, including: Determine the time required to calculate the optimal precoding vector based on the codebook information in the feedback phase and the node computing performance; Determine the time required for the index to be transmitted to the sender through the feedback channel according to the index of the optimal precoding vector and the feedback accuracy of the channel state information; The time overhead of the feedback phase is determined based on the time required to calculate the optimal precoding vector and the time required to transmit the index to the sender through the feedback channel.

6. The multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing network according to claim 2 is characterized in that: The routing quality formula under the channel state perception range is as follows: in, E[τ(Q(n))] represents the average weight of Q(n), Q(n) represents the optimal path selection for the sensing range n, ω(n) represents the average number of sensed nodes, n represents the channel state sensing range, M represents the number of antennas, and s represents the number of path hops from the source to the destination. and p s denote the binomial distribution parameters of the cardinality of a set of s-hop paths.

7. The multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing network according to claim 3 is characterized in that: The time cost calculation formula of the sensing phase is: in, E[A n ] represents the average number of n hops, K represents the number of nodes in the ad hoc network, p represents the probability of connectivity between any two nodes, n represents the channel state perception range, b2 represents the channel information quantization length, R1=Wlog2(1+Mρ), W represents the bandwidth, ρ represents the average signal-to-noise ratio (SNR) of each receiving antenna, and M represents the number of antennas.

8. The multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing network according to claim 3 is characterized in that: The time cost of the feedback phase includes: T2=T t +T c ; Among them, T c represents the time required to calculate the optimal precoding vector, b1 represents the accuracy of channel state information feedback, μ represents the node computing performance, that is, the time it takes for the node to calculate a complex operation, T t It represents the time required for the index to be transmitted to the sender through the feedback channel. R2=Wlog2(1+Mρ), W represents bandwidth, ρ represents the average signal-to-noise ratio SNR of each receiving antenna, and M represents the number of antennas.

9. A multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing networks according to claim 7 or 8, characterized in that: Based on the time overhead of the sensing phase and the time overhead of the feedback phase, the average achievable rate with feedback quantization loss is determined, including: in, T0 represents the start time, T1 represents the time consumed in the sensing phase, T2 represents the total time overhead in the feedback phase, W represents the bandwidth, ρ represents the average signal-to-noise ratio (SNR) of each receiving antenna, and λ max Representation Matrix b1 represents the maximum eigenvalue of , b1 represents the channel state information feedback accuracy, and M represents the number of antennas.

10. The multi-antenna channel transmission method for large-scale unmanned aerial vehicle self-organizing network according to claim 1, characterized in that: With the goal of maximizing the product of routing quality and average achievable rate with feedback quantization loss under constraints, the channel state perception range and channel state information feedback accuracy are updated, including: The objective is to maximize the product of routing quality and average achievable rate with feedback quantization loss under constraints including: s.t.n∈N + ; n≤K-1, b1∈N + , T1+T2<T0, Among them, R real represents the average achievable rate with feedback quantization loss, R routing represents the routing quality, n represents the channel state perception range, b1 represents the channel state information feedback accuracy, N + represents a positive integer, K represents the number of nodes in the ad hoc network, T0 represents the starting time, T1 represents the time consumed in the sensing phase, and T2 represents the total time consumption in the feedback phase; The product of maximizing the routing quality and the average achievable rate with feedback quantization loss is converted into the first optimization objective and the second optimization objective, including: Among them, E[τ(Q(n))] represents the average weight of Q(n), and Q(n) represents the optimal path selection within the sensing range n.

Images