UWSNs resource allocation method based on hybrid AC and MI communication

Abstract

The invention relates to a UWSNs resource allocation method based on hybrid AC and MI communication, and belongs to the technical field of wireless communication. According to the method, under the condition that constraint conditions of transmitting power, bandwidth, transmission rate and transmission time are met at the same time, a joint allocation problem of the transmitting power and channel resources is modeled as an optimization problem, so that the goal of minimizing energy consumption of completing a total task of a system is achieved. According to the method, the total energy consumption of the system can be minimized by optimizing the UWSNs resource allocation strategy based on hybrid AC and MI communication.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communication, and particularly relates to a resource allocation method for UWSNs based on hybrid AC and MI communication. Background Art

[0002] With the development of ocean exploration, underwater wireless sensor networks (UWSNs) have received increasing attention. Due to their flexible deployment, ability to communicate and collect data in real time and efficiently, and prevention of potential dangers, they are widely used in the monitoring of underwater pipelines, offshore platforms, and underwater structures. Due to the particularity of the underwater environment, electromagnetic waves will be strongly attenuated, and optical communication will also become difficult to maintain due to the complex environment, so neither of them is generally applicable. Even though the transmission rate of underwater acoustic communication (AC) is limited and the transmission delay is large, it is still the best choice for achieving kilometer-level transmission underwater. At the same time, since underwater nodes are generally powered by batteries and have limited energy, designing a reasonable resource allocation scheme can extend the lifespan of UWSNs. Therefore, studying the resource allocation strategy of UWSNs based on AC is of great significance.

[0003] However, most of the existing studies usually adopt a single AC method. Although AC can achieve long-distance underwater communication, the AC channel has severe multipath fading and Doppler effects, resulting in a low transmission rate. When the data volume is large, it will cause a large delay in information transmission, which will have a serious impact on some information with high timeliness. Magnetic induction (MI) communication, as a promising communication technology, has the advantages of a stable channel, low propagation delay, and high transmission rate, etc., but MI communication also has the limitation of a short transmission distance. If the AC and MI communication methods are considered collaboratively in UWSNs, it will help improve the performance of the system. In addition, traditional single-coil MI communication can obtain good transmission performance only when the transceiver coils are coaxially parallel, which is difficult to control in the underwater environment. We will consider using an omnidirectional antenna as the transceiver coil, which can effectively reduce the influence of coil offset on transmission performance. At the same time, most of the existing studies on SNs are located on the seabed, and the three-dimensional situation in seawater is less considered. Therefore, we will study the resource allocation strategy of UWSNs based on hybrid AC and MI communication to improve the system performance. Summary of the Invention

[0004] The object of the present invention is to solve the problem of low data rate in traditional AC-based UWSNs, and provide a resource allocation method for UWSNs based on hybrid AC (Acoustic Communication, AC) and MI (Magnetic Induction, MI) communication. This method models the joint allocation problem of transmission power and channel resources as an optimization problem under the constraints of transmission power, bandwidth, transmission rate, and transmission time, so as to achieve the goal of minimizing the total energy consumption for task completion of the system. The present invention can minimize the total energy consumption of the system by optimizing the resource allocation strategy for UWSNs based on hybrid AC and MI communication.

[0005] To achieve the above object, the technical solution of the present invention is: A resource allocation method for UWSNs based on hybrid AC and MI communication, including:

[0006] S1. Model the network structure of UWSNs with hybrid AC and MI communication;

[0007] S2. Model the equivalent circuit structure of MI communication between SNs and CHs using omnidirectional coils;

[0008] S3. Model the channel gain of the AC link between CHs and the water surface BS;

[0009] S4. Model the energy consumed by SNs to send data to CHs;

[0010] S5. Model the energy consumed by CHs to aggregate and forward data to the BS;

[0011] S6. Model the total energy consumed by the MI communication link and the AC link during transmission;

[0012] S7. Model the limiting conditions of the transmission power and transmission rate of SNs, the transmission power, transmission rate, and channel resources of CHs, and the transmission time of the entire system;

[0013] S8. Model the optimization model for minimizing the total energy consumption of the system;

[0014] S9. Use the SL-PSO algorithm to solve the optimization model for minimizing the total energy consumption of the system.

[0015] In an embodiment of the present invention, the step S1 is specifically:

[0016] Step S11: Model the network structure of UWSNs that hybridizes AC and MI communication, including a surface base station BS and N sensor nodes SNs. First, divide the SNs into K clusters, and select the SNs with the highest data volume in each cluster as cluster heads CHs. Each SNs contains an omnidirectional coil to perform MI communication with the CHs. The CHs transmit information to the BS through hybrid acoustic communication AC. Assume that the BS is located on the sea surface at a height of 0 with a fixed position.

[0017] Step S12: In the communication network of the UWSNs network structure that hybridizes AC and MI communication, use a three-dimensional Cartesian coordinate system to describe the positions of the autonomous underwater vehicle AUV and SNs. For the k-th cluster, the number of SNs within the cluster is denoted as N k , where k ∈ [1, K]. Then, the CHs receive the data collected by the SNs within the cluster through magnetic induction MI communication. After being aggregated by the CHs, the data is forwarded to the BS through AC. The data volume collected by the SNs is represented by I n,k , and the data collected by the CHs is represented by I k . Assume that for the k-th cluster, the position of CH k is (x k , y k , z k ), and the coordinates of the n-th node SN n,k in the cluster are (x n,k , y n,k , z n,k ), where n ∈ [1, N k .

[0018] In an embodiment of the present invention, the specific content of step S2 is as follows:

[0019] The omnidirectional coil is formed by connecting three sub-coils in series, and the sub-coils are orthogonal to each other in pairs. Assume that the center of the transmitting coil is located at the origin O(0, 0, 0) of the Cartesian coordinate system, and one of its normal vectors coincides with the Z-axis. The center of the receiving antenna is located at Q(x, y, z), then the distance between the centers of the transmitting and receiving coils is denoted as The normalized normal vector of a certain sub-coil of the receiving coil is n 1 = (cosα 1 , cosβ 1 , cosγ 1 ), where α 1 , β 1 , γ 1 are the angles between n 1 and the yoz, xoz, xoy planes respectively. Since the three sub-coils are orthogonal to each other in pairs, the normal vectors n 1 , n 2 , n 3A cross matrix R can be formed. According to the knowledge of linear algebra, we have where c = (100)-n 1 , n 1 ≠(100); thus, the normal vectors n 2 = (cosα 2 , cosβ 2 , cosγ 2 ) and n 3 = (cosα 3 , cosβ 3 , cosγ 3 ) of the other two sub-coils are obtained;

[0020] The mutual inductance calculation formula between two omnidirectional coils with spatial distribution is In the formula, μ represents the magnetic permeability of free space; N t and N r represent the number of turns of the transmitting coil and the receiving coil respectively; a t and a r represent the radii of the transmitting coil and the receiving coil respectively; A 1 = 3xz + 3xy + 2x 2 - y 2 - z 2 ; A 2 = 3xy + 3yz + 2y 2 - x 2 - z 2 ; A 3 = 3xz + 3yz + 2z 2 - x 2 - y 2 ; δ 1 = cosα 1 + cosα 2 + cosα 3 ; δ 2 = cosβ 1 + cosβ 2 + cosβ 3 ; δ 3 = cosγ 1 + cosγ 2 + cosγ 3 ;

[0021] If the working angular frequency of the system is ω, the self-impedance Z t of the transmitting coil and the self-impedance Z r of the receiving coil are respectively where R t , R r are the resistances at the transmitting end and the receiving end respectively; L t , L rThe inductors at the transmitting end and the receiving end respectively; C t , C r The capacitors at the transmitting end and the receiving end respectively;

[0022] According to Kirchhoff's voltage law, the loop equations are listed as U in = i 1 Z t + i 2 ·jωM o , 0 = i 2 Z r + i 1 ·jωM o , and the solution is Among them, U in is the input voltage; i 1 , i 2 are the currents in the transmitting coil and receiving coil loops respectively;

[0023] Then the transmission power P t S of SNs and the receiving power P r C of CHs are expressed as

[0024] Therefore, the path loss of the omnidirectional transceiver coil is expressed as

[0025] In an embodiment of the present invention, the step S3 is specifically as follows:

[0026] For the k-th cluster, the coordinates of the cluster head are (x k , y k , z k ), and the coordinates of the BS are (0, 0, 0). Therefore, the distance between the cluster head and the base station is: Assume that the frequency of the acoustic signal is f C , then the path loss of the underwater acoustic channel is expressed as Among them, A 0 is the unit normalization constant, q is the diffusion factor, is the absorption coefficient, which is given by the Thorp empirical formula: So the channel gain of the underwater acoustic channel is

[0027] In an embodiment of the present invention, the step S4 is specifically as follows:

[0028] For the k-th cluster, the SN n,k within the cluster wants to send information to the CH k through MI communication. According to the Shannon formula, the transmission rate of the SN n,k is:

[0029]

[0030] Among them, B is CH k of the system bandwidth, and it is assumed that all CHs have the same bandwidth, P n,k represents the transmission power of SN n,k and N 0 represents the noise power of MI communication; the noise of the underwater MI system comes from the thermal noise of the resistor elements in the circuit, and the thermal noise power is modeled as N 0 =K 0 TB, where K 0 is the Boltzmann constant and T is the temperature; I n,k represents the amount of data collected by SN n,k , then its information transmission time is expressed as:

[0031]

[0032] Then the energy consumed by all SNs to transmit information in the MI communication stage is:

[0033]

[0034] The total time spent in the MI communication stage is:

[0035]

[0036] In an embodiment of the present invention, the step S5 is specifically:

[0037] In the AC stage, the CHs forward the aggregated information to the BS in a frequency-division multiple access mode, then the transmission rate of CH k is:

[0038]

[0039] Among them, B k represents the sub-channel resource allocated by the BS to CH k , P k represents the transmission power of CH k , η k represents the channel gain, represents the noise power of AC communication, and is modeled as Let v s represent the propagation speed of the acoustic signal, I k represents the amount of data collected by CH k itself, then its transmission time is:

[0040]

[0041] Then the energy consumed by all CHs in forwarding data during the AC phase is:

[0042]

[0043] In an embodiment of the present invention, in step S6, the total energy consumed by the MI communication link and the AC link transmission, ignoring the receiving power of CHs and the BS, is expressed as:

[0044] E = E SN + E CH .

[0045] In an embodiment of the present invention, step S7 is specifically:

[0046] The transmission power limit condition of SNs is:

[0047] The transmission rate limit condition of SNs is:

[0048] The transmission power limit condition of CHs is:

[0049] The transmission rate limit condition of CHs is:

[0050] The channel resource limit condition of CHs is:

[0051] The transmission time limit condition of the system is:

[0052] Among them, P n,k represents the transmission power of SN n,k , R n,k represents the transmission rate of SN n,k , P k represents the transmission power of CH k , B k represents the sub-channel resource allocated by the BS to CH k , represents the noise power of the AC communication, η k represents the channel gain of the underwater acoustic channel, T MI represents the total time spent in the MI communication phase, T k represents the transmission time of CH k ; and respectively represent the maximum transmission powers of SNs and CHs; and respectively represent the minimum transmission rates of SNs and CHs; B max represents the maximum bandwidth resource of CHs in the AC; t maxRepresents the minimum time limit for system transmission.

[0053] In an embodiment of the present invention, step S8 is specifically as follows: Under the constraint conditions of relevant transmission power, transmission rate, channel resources, and transmission time, with the goal of minimizing the total energy consumption for system task completion, determine the optimal resource allocation strategy and model the optimization model for minimizing the total energy consumption of the system, that is Wherein, Represents the transmission power of SNs; Represents the transmission power of CHs; Represents the bandwidth resource allocated by the BS to CHs; E represents the total energy consumption of the system.

[0054] In an embodiment of the present invention, step S9 is specifically as follows:

[0055] S91. For the optimization problem with constraint conditions, use the penalty function method to transform the problem with inequality constraints into an unconstrained problem, and construct a fitness function composed of an objective function and a penalty function, which is given by the following formula:

[0056] f i (P SN , P CH , B CH ) = f o (P SN , P CH , B CH ) + ζ(t)f p (P SN , P CH , B CH )

[0057] Wherein, f o (P SN , P CH , B CH ) is the objective function, ζ(t) is the penalty factor, and f p (P SN , P CH , B CH ) is the penalty function, which includes the following six expressions:

[0058]

[0059] Wherein, R k Represents the transmission rate of CH k ,

[0060] S92. Solve the fitness function in step S91 using the SL - PSO algorithm, specifically: Initialize the number of iterations \(t\), learning probability factor \(\lambda\), control parameter \(\tau\), the number of population particles \(m\), and social learning factor \(\xi\); In the SL - PSO algorithm, the particle swarm is sorted according to the fitness of particle \(i\), and \(X\) i (t) is the particle position; The particle updates its behavior vector using the social learning strategy, as shown in the following formula:

[0061]

[0062] X i,j (t) is the behavior vector of particle \(i\) in the \(j\)-th dimension at the \(t\)-th iteration, \(P\) i L is the learning probability, \(\Delta X\) i,j (t + 1) is the behavior correction. When the random probability \(p\) i (t) satisfies \(p\) i (t)\(\leq P\) i L , particle \(i\) will learn and correct its behavior vector; \(\Delta X\) i,j (t + 1) is further expressed as:

[0063] \(\Delta X\) i,j (t + 1)=r 1 (t)\(\Delta X\) i,j (t)+r 2 (t)I i,j (t)+r 3 (t)\(\xi C\) i,j (t)

[0064] where \(r\) 1 (t), \(r\) 2 (t) and \(r\) 3 (t) are random coefficients within \((0, 1)\), \(I\) i,j (t) is the imitation component, and \(C\) i,j (t) is the social influence component, expressed as:

[0065]

[0066] where \(i\lt k\leq j\), \(I\) i,j (t) means that the particle learns from any better particle in the population, is the average behavior of all particles in the population, and \(C\) i,j (t) means that the particle learns from the collective behavior of the population; is expressed as:

[0067]

[0068] Except for the best particles, all particles will be updated by learning any particle with better fitness; these particles do not learn from the historical best positions, but learn from any better particle in the current particle swarm; by continuously iterating until the optimization goal converges, the minimum system energy consumption can be obtained.

[0069] Compared with the prior art, the present invention has the following beneficial effects:

[0070] 1. The present invention adopts a hybrid AC and MI communication method to solve the problem of low data transmission rate of traditional acoustic communication.

[0071] 2. Considering the underwater SNs data transmission problem, the present invention uses an omnidirectional coil for MI communication, effectively alleviating the influence of coil bias on the transmission performance, and providing an idea for underwater MI communication problems.

[0072] 3. Under the constraints of transmit power, bandwidth, transmission rate, and transmission time, the present invention jointly optimizes the transmit power of SNs and CHs and the bandwidth resources of CHs to minimize the total system energy consumption. At the same time, the SL-PSO algorithm adopted by the present invention has a faster convergence speed and better global convergence ability compared with the benchmark algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 is the network structure diagram of UWSNs with hybrid AC and MI communication in an embodiment of the present invention;

[0074] Figure 2 is the flowchart of the resource allocation algorithm based on the SL-PSO algorithm in an embodiment of the present invention;

[0075] Figure 3 is the flowchart of the method of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0076] The technical solution of the present invention will be specifically described below with reference to the drawings.

[0077] The present invention provides a UWSNs resource allocation method based on hybrid AC and MI communication, including:

[0078] S1. Modeling the network structure of UWSNs with hybrid AC and MI communication;

[0079] S2. Modeling the equivalent circuit structure of SNs and CHs using an omnidirectional coil for MI communication;

[0080] S3. Modeling the AC link channel gain between CHs and the water surface BS;

[0081] S4. Modeling the energy consumed by SNs to send data to CHs;

[0082] S5. Model the energy consumed by CHs to aggregate and forward data to the BS;

[0083] S6. Model the total energy consumed by the MI communication link and the AC link transmission;

[0084] S7. Model the constraints on the transmission power and rate of SNs, the transmission power, rate and channel resources of CHs, and the transmission time of the entire system;

[0085] S8. Model the optimization model for minimizing the total energy consumption of the system;

[0086] S9. Use the SL-PSO algorithm to solve the optimization model for minimizing the total energy consumption of the system.

[0087] The following is the specific implementation process of the present invention.

[0088] Aiming at the shortcoming of low data rate in traditional AC-based UWSNs, the present invention overcomes the above shortcomings by means of a method of hybrid AC and MI communication. We propose a resource allocation strategy for UWSNs based on hybrid AC and MI communication to minimize the total energy consumption of the system. Specifically, first, the nodes at different depths underwater are clustered. The SNs within the cluster use omnidirectional coils to transmit data to the cluster heads CHs through MI communication. After aggregating the data, the CHs forward the data to the surface BS using AC. By jointly optimizing the transmission power of SNs, the transmission power and bandwidth resources of CHs, the problem of joint allocation of transmission power and bandwidth is modeled as an optimization problem, and the SL-PSO algorithm is used for iterative solution to achieve the goal of minimizing the total energy consumption of the system. Compared with other solutions, this strategy can more effectively reduce the system energy consumption.

[0089] I. Network Model of UWSNs with Hybrid AC and MI Communication

[0090] In this embodiment, a resource allocation strategy for UWSNs based on hybrid AC and MI communication is proposed. The network structure diagram of UWSNs with hybrid AC and MI communication is as Figure 1 shown. In this system, a network structure of UWSNs with hybrid AC and MI communication is constructed, including a surface BS and N SNs. First, the SNs are divided into K clusters, and the SNs with the highest data volume in each cluster are selected as CHs. Each SNs includes an omnidirectional coil for MI communication with CHs. The CHs transmit information to the BS through AC. It is assumed that the BS is located at the fixed sea level with a height of 0.

[0091] II. Establish a Resource Allocation Model for UWSNs with Hybrid AC and MI Communication

[0092] Please refer to Figure 3, A resource allocation strategy for UWSNs based on hybrid AC and MI communication proposed in this implementation case includes the following steps:

[0093] S1: Model the network structure of UWSNs with hybrid AC and MI communication

[0094] The system network structure mainly includes SNs, CHs, and BS. In this embodiment, a three-dimensional Cartesian coordinate system is used to describe the positions of AUVs and SNs. For the k-th cluster, the number of SNs within the cluster is denoted as N k , where k ∈ [1, K]. Then, CHs receive the data collected by SNs within the cluster through MI communication, and after summarization by CHs, forward it to the surface BS through AC. The amount of data collected by SNs is denoted by I n,k , and the data collected by CHs is denoted by I k . Assume that for the k-th cluster, the position of CH k is (x k , y k , z k ), and the coordinates of the n-th node SN n,k in the cluster are (x n,k , y n,k , z n,k ), where n ∈ [1, N k .

[0095] S2: Model the equivalent circuit structure of MI communication between SNs and CHs using omnidirectional coils

[0096] The omnidirectional coil is formed by connecting three sub-coils in series, and the sub-coils are orthogonal to each other in pairs. Assume that the center of the transmitting coil is located at the origin O(0, 0, 0) of the Cartesian coordinate system, and one of its normal vectors coincides with the Z-axis. Assume that the center of the receiving antenna is located at Q(x, y, z), then the distance between the centers of the transmitting and receiving coils can be denoted as Assume that the normalized normal vector of a sub-coil of the receiving coil is n 1 = (cosα 1 , cosβ 1 , cosγ 1 ), where α 1 , β 1 , γ 1 are the angles between n 1 and the yoz, xoz, and xoy planes respectively. Since the three sub-coils are orthogonal to each other in pairs, the normal vectors n 1 , n 2 , n 3 corresponding to the sub-coils can form an intersection matrix R. According to the knowledge of linear algebra, there is where c = (100) - n 1 , n 1≠(100). From this, the normal vectors n of the other two sub-coils can be obtained. 2 =(cosα 2 , cosβ 2 , cosγ 2 ), n 3 =(cosα 3 , cosβ 3 , cosγ 3 ).

[0097] The mutual inductance calculation formula between two omnidirectional coils with spatial distribution is where μ represents the magnetic permeability of free space; N t and N r represent the number of turns of the transmitting coil and the receiving coil respectively; a t and a r represent the radii of the transmitting coil and the receiving coil respectively; A 1 =3xz + 3xy + 2x 2 -y 2 -z 2 ; A 2 =3xy + 3yz + 2y 2 -x 2 -z 2 ; A 3 =3xz + 3yz + 2z 2 -x 2 -y 2 ; δ 1 =cosα 1 +cosα 2 +cosα 3 ; δ 2 =cosβ 1 +cosβ 2 +cosβ 3 ; δ 3 =cosγ 1 +cosγ 2 +cosγ 3 .

[0098] If the system operating angular frequency is ω, then the self-impedance Z t of the transmitting coil and the self-impedance Z r of the receiving coil are respectively where R t , R r are the resistances of the transmitting end and the receiving end respectively; L t , L r are the inductances of the transmitting end and the receiving end respectively; C t , C r are the capacitances of the transmitting end and the receiving end respectively.

[0099] According to Kirchhoff's voltage law, the loop equation is listed as U in = i 1 Z t + i 2 ·jωM o , 0 = i 2 Z r + i 1 ·jωM o , and the solution is where U in is the input voltage; i 1 , i 2 are the currents of the transmitting coil and receiving coil circuits respectively.

[0100] Then the transmitting power P t S of SNs and the receiving power P r C of CHs can be expressed as

[0101] Therefore, the path loss of the omnidirectional transceiver coil can be expressed as

[0102] S3: Modeling the AC link channel gain between CHs and the water surface BS

[0103] For the k-th cluster, the coordinates of the cluster head are (x k , y k , z k ), and the coordinates of the water surface base station are (0, 0, 0). Therefore, the distance between the cluster head and the base station is:[[]] Assume the frequency of the acoustic signal is f C . Then the path loss of the underwater acoustic channel can be expressed as where A 0 is the unit normalization constant, q is the diffusion factor, is the absorption coefficient, which can be given by the Thorp empirical formula:[[]] So the channel gain of the underwater acoustic channel is

[0104] S4: Modeling the energy consumed by SNs to send data to CHs

[0105] For the k-th cluster, the intra-cluster SN n,k needs to send information to the CH k through MI communication. According to the Shannon formula, the transmission rate of SN n,k is:[[]]

[0106]

[0107] where B is the CH ksystem bandwidth, and assume that all CHs have the same bandwidth, P n,k represents the transmit power of SN n,k N 0 represents the noise power of MI communication. The noise of the underwater MI system mainly comes from the thermal noise of the resistor elements in the circuit, and the thermal noise power can be modeled as N 0 = K 0 TB, where K 0 is the Boltzmann constant and T is the temperature. I n,k represents the amount of data collected by SN n,k then its information transmission time is expressed as:

[0108]

[0109] Then the energy consumed by all SNs to transmit information in the MI communication phase is:

[0110]

[0111] The total time spent in the MI communication phase is:

[0112]

[0113] S5: Model the energy consumed by CHs to aggregate and forward data to the BS

[0114] In the AC phase, CHs forward the aggregated information to the surface BS in a frequency-division multiple access mode, then the transmission rate of CH k is:

[0115]

[0116] where B k represents the sub-channel resource allocated by the BS to CH k P k represents the transmit power of CH k η k represents the channel gain, represents the noise power of AC communication, which can be modeled as Let v s represent the propagation speed of the acoustic signal, I k represent the amount of data collected by CH k itself, then its transmission time is:

[0117]

[0118] Then the energy consumed by all CHs to forward data in the AC phase is:

[0119]

[0120] S6: Model the total energy consumed by the MI communication link and the AC link transmission

[0121] Since the received powers of CHs and BS are relatively small and can be neglected, the total system energy consumption can be expressed as:

[0122] E = E SN + E CH

[0123] S7: Model the preset constraints on the transmission power and rate of SNs, the transmission power, rate, and channel resources of CHs, and the transmission time of the entire system

[0124] The transmission power limit condition for SNs is:

[0125] The transmission rate limit condition for SNs is:

[0126] The transmission power limit condition for CHs is:

[0127] The transmission rate limit condition for CHs is:

[0128] The channel resource limit condition for CHs is:

[0129] The transmission time limit condition for the system is:

[0130] Among them, and represent the maximum transmission powers of SNs and CHs respectively; and represent the minimum transmission rates of SNs and CHs respectively; B max represents the maximum bandwidth resource of CHs in the AC; t max represents the minimum time limit for system transmission.

[0131] S8: Model the optimization model for minimizing the total system energy consumption

[0132] Under the constraint conditions of relevant transmission power, rate, channel resources, and transmission time, with the goal of minimizing the total energy consumption for system task completion, determine the optimal resource allocation strategy, that is Among them, represents the transmission power of SNs; represents the transmission power of CHs; represents the bandwidth resource allocated by the BS to CHs; E represents the total system energy consumption.

[0133] III. Solving the Optimization Model of Resource Allocation Based on the SL-PSO Algorithm

[0134] Reference Figure 3 , in this embodiment, an iterative optimization algorithm based on the SL-PSO algorithm is adopted to solve the optimization model for minimizing the total task energy consumption of the system. Specifically:

[0135] Step 1: First, for the optimization problem with constraints, the inequality constraint problem is transformed into an unconstrained problem by the penalty function method, and a fitness function composed of an objective function and a penalty function is constructed. It is given by the following formula:

[0136] f i (P SN ,P CH ,B CH ) = f o (P SN ,P CH ,B CH ) + ζ(t)f p (P SN ,P CH ,B CH )

[0137] Among them, f o (P SN ,P CH ,B CH ) is the objective function, ζ(t) is the penalty factor, and f p (P SN ,P CH ,B CH ) is the penalty function, which includes the following six formulas:

[0138]

[0139] Step 2: Next, the basic process of the SL-PSO algorithm is introduced. First, the number of iterations t, the learning probability factor λ, the control parameter τ, the number of population particles m, and the social learning factor ξ are initialized. In the SL-PSO algorithm, the particle swarm is sorted according to the fitness of particle i, and X i (t) is the particle position. The particle updates its behavior vector using the social learning strategy, as shown in the following formula:

[0140]

[0141] X i,j (t) is the behavior vector of particle i in the jth dimension at the tth iteration, P i L is the learning probability, ΔX i,j (t + 1) is the behavior correction. When the random probability p i (t) satisfies pi (t) ≤ P i L Only when the particle i will learn and correct its behavior vector. ΔX i,j (t + 1) can be further expressed as:

[0142] ΔX i,j (t + 1) = r 1 (t)ΔX i,j (t) + r 2 (t)I i,j (t) + r 3 (t)ξC i,j (t)

[0143] Among them, r 1 (t), r 2 (t) and r 3 (t) are random coefficients within (0, 1), I i,j (t) is the imitation component, C i,j (t) is the social influence component and can be expressed as:

[0144]

[0145] Among them, i < k ≤ j, I i,j (t) represents that the particle learns from any better particle in the population, is the average behavior of all particles in the population, C i,j (t) represents that the particle learns from the collective behavior of the population. It can be expressed as:

[0146]

[0147] Except for the best particle, all particles will be updated by learning any particle with better fitness. These particles do not learn from the historical best positions but from any better particle in the current particle swarm. By continuously iterating until the optimization goal converges, the minimum system energy consumption can be obtained.

[0148] The above is the preferred embodiment of the present invention. All changes made according to the technical solution of the present invention and whose functional effects do not exceed the scope of the technical solution of the present invention shall fall within the protection scope of the present invention.

Claims

1. A UWSNs resource allocation method based on hybrid AC and MI communication, characterized in that: include: S1, Modeling the network structure of UWSNs with hybrid AC and MI communication; S2, modeling the equivalent circuit structure of MI communication between SNs and CHs using omnidirectional coils; S3, modeling the AC link channel gain between CHs and the surface BS; S4, modeling the energy consumed by SNs sending data to CHs; S5, modeling the energy consumed by CHs to aggregate and forward data to the BS; S6, the total energy consumed by modeling MI communication link and AC link transmission; S7, modeling the transmission power and transmission rate of SNs, the transmission power, transmission rate and channel resources of CHs, and the transmission time constraints of the entire system; S8, optimization model for minimizing the total energy consumption of the modeling system; S9. Use the SL-PSO algorithm to solve the optimization model for minimizing the total energy consumption of the system.

2. A UWSNs resource allocation method based on hybrid AC and MI communication according to claim 1, characterized in that: The step S1 is specifically as follows: Step S11, modeling the UWSNs network structure of hybrid AC and MI communication, including a surface base station BS and N sensor nodes SNs; firstly, the SNs are divided into K clusters, and the SNs with the highest data volume in each cluster are selected as cluster heads CHs; each SNs contains an omnidirectional coil for MI communication with CHs; CHs transmits information with BS through hybrid acoustic communication AC; it is assumed that the BS is located at a fixed position on the sea level with a height of 0; Step S12: In the communication network of the UWSNs network structure of mixed AC and MI communication, a three-dimensional Cartesian coordinate system is used to describe the positions of the autonomous underwater vehicle AUV and SNs; for the kth cluster, the number of SNs in the cluster is expressed as N k , where k∈[1,K]; then CHs receives the data collected by SNs in the cluster through magnetic induction MI communication, aggregates them through CHs and forwards them to BS through AC; the amount of data collected by SNs is expressed as I n,k Indicates that the data collected by CHs are expressed as I k Indicates; Assume that the kth cluster, CH k The position of (x k ,y k ,z k ), the nth node SN in the cluster n,k The coordinates of (x n,k ,y n,k ,z n,k ), where n∈[1,N k ].

3. A UWSNs resource allocation method based on hybrid AC and MI communication according to claim 2, characterized in that: The step S2 is specifically as follows: The omnidirectional coil is composed of three sub-coils connected in series, and the sub-coils are orthogonal to each other. Assume that the center of the transmitting coil is located at the origin of the Cartesian coordinate system O(0,0,0), and one of the normal vectors coincides with the Z axis; the center of the receiving antenna is located at Q(x,y,z), then the distance between the centers of the transmitting and receiving coils is recorded as The normalized normal vector of a sub-coil of the receiving coil is n1 = (cosα1, cosβ1, cosγ1), where α1, β1, γ1 are the angles between n1 and the yoz, xoz, xoy planes respectively; since the three sub-coils are mutually orthogonal, the normal vectors n1, n2, n3 corresponding to the sub-coils can form an intersection matrix R. According to the knowledge of linear algebra, we have Wherein, c = (100) - n1, n1 ≠ (100); thus, the normal vectors of the other two sub-coils are obtained as n2 = (cos α2, cos β2, cos γ2), n3 = (cos α3, cos β3, cos γ3); The formula for calculating the mutual inductance between two spatially distributed omnidirectional coils is: Where μ represents the magnetic permeability of free space; N t and N r Respectively represent the number of turns of the transmitting coil and the receiving coil; a t and a r Respectively represent the radius of the transmitting coil and the receiving coil; A1=3xz+3xy+2x 2 -y 2 -z 2 ; A2=3xy+3yz+2y 2 -x 2 -z 2 ; A3=3xz+3yz+2z 2 -x 2 -y 2 ; δ1=cosα1+cosα2+cosα3; δ2=cosβ1+cosβ2+cosβ3; δ3=cosγ1+cosγ2+cosγ3; If the system operating angular frequency is ω, the self-impedance Z of the transmitting coil is t and the self-impedance Z of the receiving coil r They are Among them, R t , R r are the resistances of the transmitting end and the receiving end respectively; L t , L r are the inductance of the transmitting end and the receiving end respectively; C t , C r are the capacitances of the transmitting and receiving ends respectively; According to Kirchhoff's voltage law, the circuit equation is listed as U in =i1Z t +i2·jωM o , 0 = i2Z r +i1·jωM o , and solve Among them, U in is the input voltage; i1 and i2 are the currents of the transmitting coil and receiving coil loops respectively; Then the transmission power of SNs is and the received power of CHs Expressed as Therefore, the path loss of the omnidirectional transceiver coil is expressed as 4. A UWSNs resource allocation method based on hybrid AC and MI communication according to claim 2, characterized in that: The step S3 is specifically as follows: For the kth cluster, the coordinates of the cluster head are (x k ,y k ,z k ), the coordinates of BS are (0,0,0), so the distance between the cluster head and the base station is: Assume that the frequency of the sound signal is f C , then the path loss of the underwater acoustic channel is expressed as Where A0 is the unit normalization constant, q is the diffusion factor, is the absorption coefficient, given by Thorp's empirical formula: So the channel gain of the underwater acoustic channel is 5. The UWSNs resource allocation method based on hybrid AC and MI communication according to claim 3, characterized in that: The step S4 is specifically as follows: For the kth cluster, the SN within the cluster n,k To send information to CH via MI communication k According to Shannon's formula, SN n,k The transmission rate is: Where B is CH k system bandwidth, and assuming that all CHs have the same bandwidth, P n,k Indicates SN n,k The transmission power of MI communication is represented by N0, and the noise power of MI communication is represented by N0. The noise of underwater MI system comes from the thermal noise of the resistor in the circuit. The thermal noise power is modeled as N0=K0TB, where K0 is the Boltzmann constant and T is the temperature. n,k Indicates SN n,k The amount of data collected, the information transmission time is expressed as: The energy consumed by all SNs transmitting information during the MI communication phase is: The total time spent in the MI communication phase is:

6. A UWSNs resource allocation method based on hybrid AC and MI communication according to claim 5, characterized in that: The step S5 is specifically as follows: In the AC phase, CHs forwards the aggregated information to the BS in frequency division multiplexing mode, so CH k The transmission rate is: Among them, B k Indicates that BS is assigned to CH k Sub-channel resources, P k Indicates CH k The transmission power, η k represents the channel gain, represents the noise power of AC communication, which is modeled as Let v s Represents the propagation speed of the sound signal, I k Indicates CH k The amount of data collected by itself, the transmission time is: Then the energy consumed by all CHs forwarding data in the AC stage is:

7. A UWSNs resource allocation method based on hybrid AC and MI communication according to claim 6, characterized in that: In step S6, the total energy consumed by the MI communication link and the AC link transmission, ignoring the receiving power of CHs and BS, is expressed as: E=E SN +E CH 。 8. A UWSNs resource allocation method based on hybrid AC and MI communication according to claim 2, characterized in that: The step S7 is specifically as follows: The transmission power limit conditions of SNs are: The transmission rate limit conditions of SNs are: The transmit power limit conditions of CHs are: The transmission rate limit of CHs is: The channel resource constraints of CHs are: The transmission time constraints of the system are: Among them, P n,k Indicates SN n,k The transmission power, R n,k Indicates SN n,k The transmission rate, P k Indicates CH k The transmission power, B k Indicates that BS is assigned to CH k Sub-channel resources, represents the noise power of AC communication, η k represents the channel gain of the underwater acoustic channel, T MI Represents the total time spent in the MI communication phase, T k Indicates CH k The transmission time, and denote the maximum transmit power of SNs and CHs respectively; and Respectively represent the minimum transmission rate of SNs and CHs; B max Indicates the maximum bandwidth resource of CHs in AC; t max Indicates the minimum time limit for system transmission.

9. The UWSNs resource allocation method based on hybrid AC and MI communication according to claim 1, characterized in that: The step S8 is specifically as follows: under the constraints of relevant transmission power, transmission rate, channel resources and transmission time, with the goal of minimizing the energy consumption of the system to complete the total task, determine the optimal resource allocation strategy, and build an optimization model for minimizing the total energy consumption of the system, that is, in, Indicates the transmit power of SNs; Indicates the transmit power of CHs; represents the bandwidth resources allocated by BS to CHs; E represents the total energy consumption of the system.

10. A UWSNs resource allocation method based on hybrid AC and MI communication according to claim 8, characterized in that: The step S9 is specifically as follows: S91. For optimization problems with constraints, the penalty function method is used to transform the inequality constraint problem into an unconstrained problem, and a fitness function consisting of an objective function and a penalty function is constructed, which is given by the following formula: f i (P SN ,P CH ,B CH )=f o (P SN ,P CH ,B CH )+ζ(t)f p (P SN ,P CH ,B CH ) Among them, f o (P SN ,P CH ,B CH ) is the objective function, ζ(t) is the penalty factor, f p (P SN ,P CH ,B CH ) is the penalty function, which contains the following 6 formulas: Among them, R k Indicates CH k The transmission rate, S92, using the SL-PSO algorithm to solve the fitness function of step S91, specifically: initializing the number of iterations t, the learning probability factor λ, the control parameter τ, the number of particles in the population m and the social learning factor ξ; in the SL-PSO algorithm, the particle swarm is sorted according to the fitness of particle i, X i (t) is the particle position; the particle uses the social learning strategy to update its behavior vector, as shown in the following formula: X i,j (t) is the behavior vector of particle i in the jth dimension at the tth iteration, is the learning probability, ΔX i,j (t+1) is behavior modification, when the random probability p i (t) Satisfaction When ΔX i,j (t+1) is further expressed as: ΔX i,j (t+1)=r1(t)ΔX i,j (t)+r2(t)I i,j (t)+r3(t)ξC i,j (t) Among them, r1(t), r2(t) and r3(t) are random coefficients in (0,1), I i,j (t) is the imitation component, C i,j (t) is the social impact component, expressed as: Among them, i<k≤j, I i,j (t) indicates that the particle learns from any better particle in the population, is the average behavior of all particles in the population, C i,j (t) indicates that particles learn from the collective behavior of the population; It is expressed as: Except for the best particle, all particles will be updated by learning any particle with better fitness; these particles do not learn from the best position in the history, but learn from any better particle in the current particle group; by continuously iterating until the optimization target converges, the minimum system energy consumption can be obtained.

Images