Energy efficiency optimization method for inter-cluster energy-carrying transmission of wireless sensor network

By constructing an energy efficiency optimization model and using the Lagrangian dual decomposition algorithm, the problem of transmission loss and resource allocation strategies in energy transport between clusters of wireless sensor networks is solved, and the optimization of transmission energy efficiency is achieved under the premise of meeting transmission quality. Supports low-cost, low-complexity clustered WSN high-energy-efficient transmission.

CN119012304APending Publication Date: 2024-11-22CHONGQING UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411114615.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-14
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

In the inter-cluster energy transport of wireless sensor networks, it is difficult for the prior art to minimize transmission loss while meeting transmission quality, and the discreteness of resource allocation strategies leads to increased optimization difficulty, making it difficult to converge to solve effective optimization solutions.

Method used

The energy efficiency optimization model is constructed, and the problem is converted into equivalent convex problems through dimension expansion and constraint relaxation. The solution is obtained by using the Lagrangian dual decomposition algorithm to obtain the energy efficiency optimization solution for data transmission between cluster heads of wireless sensor networks.

Benefits of technology

Convergence is achieved with fewer iterations, which maximizes transmission energy efficiency, balances transmission energy efficiency, transmit power and discrete power segmentation ratio, supports energy-carrying transmission nodes equipped with discrete power dividers, and is low-cost, The clustered WSN with low hardware complexity provides a high-energy-efficient transmission solution.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119012304A_ABST
    Figure CN119012304A_ABST
Patent Text Reader

Abstract

The invention relates to an energy efficiency optimization method for energy-carrying transmission between WSN (Wireless Sensor Network) clusters. The method comprises the following steps: establishing an energy efficiency optimization model suitable for carrying out energy-carrying transmission between WSN cluster heads by applying a discrete power divider; then, a dimension extension method for the model is provided, and the optimization problem with a discrete feasible solution set is converted into an equivalent convex problem easy to solve by means of convex conversion, variable relaxation and the like; afterwards, an algorithm for iteratively solving an optimization problem is provided, a variable closed expression in a Lagrange dual decomposition sub-algorithm is deduced, and solving of an energy efficiency optimization scheme is realized. Simulation results show that according to the method provided by the invention, the optimal solution can be obtained through convergence under the condition that the number of iterations is small, and the obtained energy efficiency optimization scheme can be balanced among the transmission energy efficiency, the transmitting power and the discrete power division ratio, so that the transmission energy efficiency is maximized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of wireless sensor network transmission, and in particular to an energy efficiency optimization method for inter-cluster energy transmission in a wireless sensor network. Background Art

[0002] Wireless Sensor Networks (WSNs) for large-scale industrial equipment, industrial manufacturing processes, and industrial environment status monitoring applications mostly adopt a clustered topology. In which, each cluster head collects the sensor data of member nodes and forms a transmission path to the sink node according to the routing rules. In this application context, the optimization of transmission energy efficiency between adjacent cluster heads has received continuous attention from researchers. The cooperative transmission method suitable for WSN introduces spatial diversity gain in the form of multi-node shared antenna resources, thereby improving the signal-to-noise ratio and increasing the transmission throughput.

[0003] On this basis, the introduction of SWIPT (Simultaneous Wireless Information and Power Transfer) technology solves the problem of relay energy limitation in single-hop cooperative transmission and enhances the flexibility and long-term effectiveness of cooperative diversity transmission schemes. Figure 1 As shown in Figure 1, in a cooperative energy transmission network, the SWIPT protocol is integrated with the relay transmission mode to form the Power Splitting Relaying (PSR) protocol. In this protocol, the receiving node of the wireless communication link is equipped with a wireless power distributor (see Figure 1). Figure 2 ). Therefore, the receiving end can use the proportionality coefficient ρ j E and ρ j I The received signal is divided into two power flows, which are then fed into the energy harvester and the information decoder respectively. A large number of researchers have explored the energy efficiency optimization problem of cooperative SWIPT networks in different applications by using the dynamic adjustment characteristics of the proportional coefficient of the power divider, and established convex optimization problems to solve the optimal node power, harvested wireless energy, and relay node resource allocation decisions.

[0004] The above studies are all based on continuous power allocation ratios. However, from the perspective of the electronic circuit hardware design technology studied, high-precision continuous power divider circuits are still complex. Considering node cost, hardware integration, and technical universality, general RF energy harvesting receivers suitable for WSN applications may only support a finite set of discrete power split ratios. Unfortunately, the discreteness of the wireless power allocation ratio leads to the discretization of the feasible solution set of the resource allocation strategy variables in the transmission energy efficiency optimization problem, which increases the difficulty of optimization and makes it difficult to converge on the solution to obtain an effective optimization solution.

[0005] Therefore, in the inter-cluster energy-carrying communication transmission of wireless sensor networks, how to minimize the transmission loss as much as possible while ensuring the transmission quality and optimizing the transmission energy efficiency as much as possible is still a new problem to be solved. Summary of the invention

[0006] In view of the deficiencies of the above-mentioned prior art, the present invention provides an energy efficiency optimization method for inter-cluster energy transmission in wireless sensor networks, which is used to obtain an energy efficiency optimization solution that minimizes transmission loss as much as possible and ensures that transmission energy efficiency is optimized as much as possible while meeting transmission quality, thereby solving the problem that it is difficult to find an effective solution to this problem in the prior art.

[0007] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0008] A method for optimizing energy efficiency of inter-cluster energy transmission in a wireless sensor network, comprising:

[0009] Construct an energy efficiency optimization model for data transmission between cluster heads in wireless sensor networks;

[0010] Expanding the dimension of the energy efficiency optimization model and relaxing the constraints, and converting the energy efficiency optimization model into an equivalent convex problem optimization model;

[0011] The Lagrangian dual decomposition sub-algorithm is used to solve the optimization model of the equivalent convex problem, and an energy efficiency optimization scheme for data transmission between cluster heads in wireless sensor networks is obtained.

[0012] Specifically, in the energy efficiency optimization model for data transmission between cluster heads in the wireless sensor network, the transmission energy efficiency optimization model between any two cluster heads is expressed as:

[0013]

[0014] Among them, P ij Indicates cluster head CH i To cluster head CH j The transmission power of the transmitted information; and Respectively represent the transmission power split to the cluster head CH j The power allocation ratio coefficient of the signal decoder and the energy harvester, and Indicated by P ij , Cluster head CH as a relevant variable i With cluster head CH j The network transmission energy efficiency value between (ij) and U TP(ij) They represent cluster heads CH iWith cluster head CH j The data throughput and total power consumption during the energy transmission process; and:

[0015]

[0016] U TP(ij) =εP ij +P C ;

[0017] Wherein, B is the baseband bandwidth, η is the energy conversion efficiency, and 0<η<1; Indicates cluster head CH i Communication link with relay R ir , relay R and cluster head CH j Communication link between rj and cluster head CH i With cluster head CH j Communication link between ij The normalized channel gains of the three wireless communication links, h ab represents the channel gains of the three wireless communication links, a,b∈{i,j,r}; λ ij Represents the equivalent signal-to-noise ratio of the energy transmission process; ε≥1 indicates cluster head CH i The conversion efficiency of the power amplifier; P C Indicates the total power consumption of the network circuit.

[0018] Specifically, the dimension expansion and constraint relaxation processing of the energy efficiency optimization model specifically include:

[0019] Firstly, the dimension of the energy efficiency optimization model of data transmission between cluster heads in wireless sensor networks is expanded to obtain a corresponding energy efficiency optimization expansion model, and then the constraint conditions of the energy efficiency optimization expansion model are relaxed to obtain a corresponding equivalent convex problem optimization model.

[0020] Specifically, the specific method of dimension expansion processing of the energy efficiency optimization model of data transmission between cluster heads in wireless sensor networks is as follows:

[0021] The transmission power is distributed to the cluster head CH j Power allocation ratio coefficient between signal decoder and energy harvester and Corresponding to several discrete gear values ​​of preset extension, that is, and The value of belongs to the preset extended discrete value space:

[0022]

[0023] in, and Respectively represent the power allocation ratio coefficient and At the nth value in the discrete value space, M represents the number of preset extended discrete gear values ​​contained in the discrete value space, n∈{1,2,…,M}, and

[0024] When the power allocation ratio coefficient of the signal decoder and the power distribution ratio coefficient of the energy harvester Select any discrete gear value and When , it corresponds to a receiving power allocation mode, and under this receiving power allocation mode, the cluster head CH i To the relay R and cluster head CH j The transmission power of the transmitted information is recorded as and Cluster Head CH j The energy gained is recorded as but:

[0025]

[0026] Where T is the transmission period, η is the energy conversion efficiency, and 0<η<1; h ij Indicates cluster head CH i With cluster head CH j Communication link between ij The channel gain of n and a n is the selection variable of the received power allocation mode, and s n ∈{0,1},a n ∈{0,1};

[0027] This achieves the dimensional expansion of the energy efficiency optimization model for data transmission between cluster heads in wireless sensor networks.

[0028] Specifically, the energy efficiency optimization extended model obtained by the dimension expansion processing of the energy efficiency optimization model includes the following objective functions and constraints:

[0029] The objective function of the energy efficiency optimization extended model is:

[0030]

[0031] in:

[0032]

[0033] The constraints of the energy efficiency optimization extended model include:

[0034]

[0035]

[0036] in, S M ={s n ,n=1,2,…,M} and AM ={a n , n=1,2,…,M} respectively represent the transmission power in the energy efficiency optimization extended model The allocation mode decision variable set and the selection variable s of the receiving power allocation mode n and a n The set of decision variables for the allocation mode; U eff(ij )(P M ,S M ) indicates that P M , S M Cluster head CH as a relevant variable i With cluster head CH j The network transmission energy efficiency value between (ij) (P M ,S M ) and U TP(ij) (P M ,S M ) represent P M , S M Cluster head CH as a relevant variable i With cluster head CH j Data throughput and total power consumption during energy transmission; Indicates cluster head CH j Energy requirement; P max Cluster head CH i The upper limit of the transmission power; α is the preference selection weight for the receiving power allocation mode selection; P ij Indicates cluster head CH i To cluster head CH j Transmission power of transmitted information; ε≥1 indicates cluster head CH i The conversion efficiency of the power amplifier; P C Indicates the total power consumption of the network circuit.

[0037] Specifically, the specific method of performing constraint relaxation processing on the energy efficiency optimization extended model is:

[0038] The variable s for selecting the receive power allocation mode n and a n The value range of is relaxed to a real number between 0 and 1, that is, the constraints C4 and C6 are relaxed to the following constraints:

[0039] C4′:s n ∈[0,1],n∈{1,2,…,M},

[0040] C6′:a n ∈[0,1],n∈{1,2,…,M};

[0041] Then, define the cluster head CH i To cluster head CH j Transmission power auxiliary variable for transmitting information Then we get the equivalent cluster head CH i Auxiliary variable of transmission power for transmitting information to relay R represents the equivalent signal-to-noise ratio of the energy transfer process, and Then, the constraint C1 is relaxed to the following constraint:

[0042]

[0043] Furthermore, the objective function of the energy efficiency optimization extended model will be transformed using the Dinkelbach method;

[0044] Thus, the constraint relaxation of the energy efficiency optimization extended model is completed.

[0045] Specifically, the equivalent convex problem optimization model obtained by performing constraint relaxation processing on the energy efficiency optimization extended model includes the following objective functions and constraints:

[0046] The objective function of the optimization model of the equivalent convex problem is:

[0047]

[0048] in,

[0049] The constraints of the optimization model for the equivalent convex problem include:

[0050]

[0051] Among them, F(P M ,S M ) indicates the head CH i With cluster head CH j The network transmission energy efficiency value U eff(ij) (P M ,S M ) equivalent convex problem optimization function; q * Transmit energy efficiency value U for the network eff( i j )(P M ,S M ) is the optimized value of .

[0052] Specifically, the specific process of solving the optimization model of the equivalent convex problem is:

[0053] S101, preset number of iterations I max and convergence threshold F Δ , and initialize the iteration number variable k=1, and initialize the iteration network transmission energy efficiency value q=0;

[0054] S102, let the optimized value of network transmission energy efficiency q * The value is the current iterative network transmission energy efficiency value q, that is, let q * =q, the Lagrangian dual decomposition algorithm is used to solve the objective function of the equivalent convex problem optimization model under the constraints, and the network resource allocation decision variable solution set {P M′ ,S M′ ,A M′} and the corresponding objective function solution F(P M′ ,S M′ ), where P M′ ,S M′ ,A M′ Respectively represent the transmission power after the current iteration update The allocation mode decision variable set and the selection variable s of the received power allocation mode n and a n The set of decision variables for the allocation mode;

[0055] S103, determine whether F(P M′ ,S M′ )<F Δ ; If satisfied, jump to step S105; otherwise, execute step S104;

[0056] S104: Determine whether k < I is satisfied at this time max If satisfied, then Let {P M ,S M ,A M}={P M′ ,S M′ ,A M′}Substitute into the equivalent convex problem optimization model, and increment the iteration number variable k by 1, then return to step S102; if not satisfied, execute step S105;

[0057] S105, using the current solution obtained for the network resource allocation decision variable solution set {P M′ ,S M′ ,A M′} is output as the optimization result of the network resource allocation decision variable.

[0058] Specifically, in step S102, the specific method of using the Lagrangian dual decomposition sub-algorithm to solve the objective function of the equivalent convex problem optimization model under the constraint condition is:

[0059] The Lagrangian dual decomposition method is used to convert the objective function solution problem of the equivalent convex problem optimization model into the corresponding dual problem. By solving the dual problem, the solution set of network resource allocation decision variables {P M′ ,S M′ ,A M′}; The transformation results in the dual problem being expressed as:

[0060]

[0061] in, are auxiliary variables of Lagrangian dual decomposition, and:

[0062]

[0063] β M ,δ,ζ,θ,ω,μ M and v M are the Lagrange multiplier constants corresponding to the constraints C1′, C2, C3, C5, C7, C8 and C9 of the optimization model of the equivalent convex problem.

[0064] Specifically, in the process of solving the dual problem, the parameter s n and a n The updated closed-form solutions in each iteration are:

[0065]

[0066] in, The parameters are s n and a n The updated value after iteration; X k and Y l They are s n and a n The maximum marginal benefit of α is the preference selection weight for receiving power allocation mode selection; β is the preference selection weight for receiving power allocation mode selection; n ∈β M , μ n ∈μ M 、v n ∈v M are the Lagrange multiplier constants β M , μ M 、v M The nth element in .

[0067] Compared with the prior art, the present invention has the following beneficial effects:

[0068] The present invention proposes an energy efficiency optimization method for inter-cluster energy transmission in wireless sensor networks.

[0069] An energy efficiency optimization model suitable for energy transmission between WSN cluster heads using discrete power dividers was established; then, we proposed a dimension expansion method for this model, and used convex transformation and variable relaxation to transform the optimization problem with discrete feasible solution sets into an equivalent convex problem that is easy to solve; then, an algorithm for iteratively solving the optimization problem was proposed, and the closed expression of variables in the Lagrangian dual decomposition sub-algorithm was derived to achieve the solution of the energy efficiency optimization scheme. The simulation results show that the method proposed in the present invention can converge to obtain the optimal solution with a small number of iterations, and the obtained energy efficiency optimization scheme can achieve a trade-off between transmission energy efficiency, transmission power and discrete power split ratio, thereby maximizing the transmission energy efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] In order to make the purpose, technical solution and advantages of the invention more clear, the present invention will be further described in detail below with reference to the accompanying drawings, in which:

[0071] Figure 1 This is an example diagram of an inter-cluster cooperative power transmission network (PSR);

[0072] Figure 2 This is a structural example diagram of a wireless energy transmission system;

[0073] Figure 3 The present invention is a flow chart of the energy efficiency optimization method for inter-cluster energy transmission in wireless sensor networks.

[0074] Figure 4 This is an example diagram of the cooperative energy-carrying (SWIPT) model for wireless sensor networks;

[0075] Figure 5 It is a convergence performance curve diagram of the method of the present invention under different noise powers in the embodiment;

[0076] Figure 6 This is an example diagram of the iterative process for solving the sub-problem OPT-2 in the embodiment;

[0077] Figure 7 A graph showing energy efficiency and power splitting resolution of different transmission protocols in the embodiment;

[0078] Figure 8 Graphs showing energy efficiency and maximum transmission power under different conditions in the embodiment. DETAILED DESCRIPTION

[0079] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. The components of the embodiments of the present invention generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the claimed invention, but only represents selected embodiments of the present invention. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the scope of protection of the present invention.

[0080] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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.

[0081] The present invention proposes an energy efficiency optimization method for inter-cluster energy transmission in wireless sensor networks. The specific process is as follows: Figure 3 As shown, including:

[0082] Construct an energy efficiency optimization model for data transmission between cluster heads in wireless sensor networks;

[0083] Expanding the dimension of the energy efficiency optimization model and relaxing the constraints, and converting the energy efficiency optimization model into an equivalent convex problem optimization model;

[0084] The Lagrangian dual decomposition sub-algorithm is used to solve the optimization model of the equivalent convex problem, and an energy efficiency optimization scheme for data transmission between cluster heads in wireless sensor networks is obtained.

[0085] The present invention is described in detail below in conjunction with the accompanying drawings and embodiments.

[0086] I. Energy efficiency optimization model for data transmission between cluster heads in wireless sensor networks

[0087] Consider an inter-cluster transmission subnetwork in a clustered wireless sensor network, which includes: a source cluster head CH i , a destination cluster head CH j and a relay R with decode-and-forward (DF) functionality, such as Figure 1 As shown. Among them, the relay R is from CH i and CH j The cluster member selected from among the cluster members can be i and CHj Wireless communication.

[0088] Figure 4 The cooperative energy transfer (SWIPT) process model of the wireless sensor network is presented, describing the cluster head CH i With the help of relay R, the cluster head CH j The process of transmitting information and energy. T is a transmission period. In phase 1 (the first phase) (duration T / 2), the cluster head CH i With transmission power P ij Broadcast information. Cluster head CH j and relay R are equipped with Figure 2 The discrete power divider shown is used to divide the received signal power flow to the signal decoder and the energy harvester (respectively according to the power distribution ratio coefficient ρ I and ρ E In phase 2 (duration T / 2), relay R decodes the information received in phase 1 and uses the harvested energy to forward the information to cluster head CH j Cluster head CH j The maximum ratio combining (MRC) method is used to combine the node cluster heads CH i And the signal sent by relay R.

[0089] Among them, cluster head CH i Use relay R to cluster head CH j Data throughput (wireless communication capacity, i.e. maximum transmission rate) R ij for:

[0090]

[0091] Where B is the baseband bandwidth, η is the energy conversion efficiency, and 0<η<1. Figure 1 In the cluster head CH i Communication link with relay R ir , relay R and cluster head CH j Communication link between rj and cluster head CH i With cluster head CH j Communication link between ij These three wireless communication links assume that the receiver can perceive perfect channel state information (CSI). Then, the actual normalized channel gains of the three wireless communication links are defined as a,b∈{i,j,r}; where represents the channel gains of the three wireless communication links, describing the small-scale Rayleigh fading characteristics of the wireless channel, L abis the channel transmission distance, and σ is the path loss coefficient.

[0092] Cluster Head CH j Energy gained E ij It can be expressed as follows:

[0093]

[0094] Moreover, the power allocation ratio coefficient of the signal decoder in relay R is It can be expressed as (3), the power distribution ratio coefficient of the energy harvester is

[0095]

[0096] The total power consumption of wireless sensor networks during cooperative energy transmission is defined as:

[0097] U TP(ij) =εP ij +P C (4)

[0098] Among them, P C is the total power consumption of the network circuit; ε≥1 indicates that the cluster head CH i The conversion efficiency of the power amplifier. Then, define the cluster head CH i With cluster head CH j The data throughput during the energy transmission process is U (ij) =R ij Therefore, the cluster head CH i With cluster head CH j The network transmission energy efficiency value between is defined as:

[0099]

[0100] Thus, the energy efficiency optimization model of data transmission between cluster heads in wireless sensor networks is constructed, where the transmission energy efficiency optimization model between any two cluster heads is expressed as:

[0101]

[0102] Among them, P ij Indicates cluster head CH i To cluster head CH j The transmission power of the transmitted information; and Respectively represent the transmission power split to the cluster head CH j The power allocation ratio coefficient of the signal decoder and the energy harvester, and Indicated by P ij , Cluster head CH as a relevant variablei With cluster head CH j The network transmission energy efficiency value between them.

[0103] II. Optimization Problem Modeling

[0104] In order to maximize the network transmission energy efficiency value U eff(ij) In the transmission energy efficiency optimization model, U eff(ij) Set as the objective function to be maximized, and set P ij , and Set as optimization variable. However, the non-convex U eff(ij) The discrete power allocation ratio hinders the optimization work. In order to avoid traversal optimization, the method of the present invention introduces a variable dimension expansion method, firstly, the dimension expansion processing of the energy efficiency optimization model of data transmission between cluster heads of the wireless sensor network is carried out to obtain the corresponding energy efficiency optimization expansion model; then, the constraint relaxation processing of the energy efficiency optimization expansion model is considered to obtain the corresponding equivalent convex problem optimization model for solving.

[0105] Specifically, the specific method of dimension expansion processing of the energy efficiency optimization model of data transmission between cluster heads in wireless sensor networks is as follows:

[0106] The transmission power is distributed to the cluster head CH j Power allocation ratio coefficient between signal decoder and energy harvester and Corresponding to several discrete gear values ​​of preset extension, that is, and The value of belongs to the preset extended discrete value space:

[0107]

[0108] in, and Respectively represent the power allocation ratio coefficient and At the nth value in the discrete value space, M represents the number of preset extended discrete gear values ​​contained in the discrete value space, n∈{1,2,…,M}, and

[0109] In this way, the cluster head CH j The power allocation ratio coefficient and They can only take discrete values ​​from the set shown in formula (7).

[0110] The power allocation ratio coefficient of relay R is and Similarly, its value is calculated according to formula (3) and the closest discrete gear value is selected.

[0111] Specifically, when the power allocation ratio coefficient of the signal decoder and the power distribution ratio coefficient of the energy harvester Select any discrete gear value and When , it corresponds to a receiving power allocation mode, and under this receiving power allocation mode, the cluster head CH i To the relay R and cluster head CH j The transmission power of the transmitted information is recorded as and Cluster Head CH j The energy gained is recorded as but:

[0112]

[0113] Where T is the transmission period, η is the energy conversion efficiency, and 0<η<1; h ij Indicates cluster head CH i With cluster head CH j Communication link between ij The channel gain of n and a n is the selection variable of the received power allocation mode, and s n ∈{0,1},a n ∈{0,1}.

[0114] Correspondingly, the dimensionally expanded U (ij) and U TP(ij) They can be expressed as (9) and (10) respectively, where α defines the tendency of selecting the receiving power allocation mode.

[0115]

[0116]

[0117] Therefore, the above variables are expanded from one dimension to M dimensions. In addition, the selection variable s of the receiving power allocation mode is set n and a n To achieve the selection of a unique receiving power allocation mode.

[0118] This achieves the dimensional expansion of the energy efficiency optimization model for data transmission between cluster heads in wireless sensor networks.

[0119] According to the above method, the maximization problem of the energy efficiency optimization model is defined, and the energy efficiency optimization extended model OPT-1 is obtained by dimension expansion as follows:

[0120] The objective function of the energy efficiency optimization extended model is:

[0121]

[0122] The constraints of the energy efficiency optimization extended model include:

[0123]

[0124] in, S M ={s n ,n=1,2,…,M} and A M ={a n , n=1,2,…,M} respectively represent the transmission power in the energy efficiency optimization extended model The allocation mode decision variable set and the selection variable s of the receiving power allocation mode n and a n The set of decision variables for the allocation mode; U eff(ij) (P M ,S M ) indicates that P M , S M Cluster head CH as a relevant variable i With cluster head CH j The network transmission energy efficiency value between (ij) (P M ,S M ) and U TP(ij) (P M ,S M ) represent P M , S M Cluster head CH as a relevant variable i With cluster head CH j Data throughput and total power consumption during energy transmission; Indicates cluster head CH j energy demand; α is the preference selection weight for receiving power allocation mode selection; P ij Indicates cluster head CH i To cluster head CH j Transmission power of transmitted information; ε≥1 indicates cluster head CH i The conversion efficiency of the power amplifier; P C Indicates the total power consumption of the network circuit.

[0125] It can be seen that the optimization objective function is U eff(ij) , which is defined by equation (5). The optimization variables include: Dimensionally expanded variables S M ={s n ,n=1,2,…,M} and A M ={a n,n=1,2,…,M}. In the constraints of the energy efficiency optimization extended model,

[0126] Constraint C1 states that the sum of the energy harvested in all modes should satisfy the cluster head CH j Energy requirements Constraint C2 gives CH i The upper limit of the transmission power P max ; Constraint C3 gives the lower limit R of the throughput of the transmission network min ; Constraints C4 to C7 ensure and The uniqueness of the choice; constraints C8 and C9 indicate that no additional energy will be generated or obtained during the power allocation process.

[0127] III. Conversion and solution of optimization problems

[0128] The challenges of solving the maximization problem OPT-1 of the energy efficiency optimization extended model come from the non-convex objective function, the coupled optimization variables in the constraints, and the integer constraints in the constraints C4 and C6. In order to make OPT-1 easier to handle, the method of the present invention further implements the following transformation processing.

[0129] A. Constraint relaxation of the objective function based on the transformation method

[0130] First, the Dinkelbach method is used to transform the objective function U eff(ij) The fractional form of is converted into the subtraction form shown in formula (12):

[0131]

[0132] Here, the problem described by equation (12) is expressed as OPT-2. The above changes in OPT-2 are equivalent to OPT-1, and OPT-2 can be iteratively solved using the Discrete SWIPT Network Resource Allocation Algorithm (DSNRA) shown in Table 1.

[0133] The specific process of solving the problem shown in Table 1 can be expressed as follows:

[0134] S101, preset number of iterations I max and convergence threshold F Δ , and initialize the iteration number variable k=1, and initialize the iteration network transmission energy efficiency value q=0;

[0135] S102, let the optimized value of network transmission energy efficiency q * The value is the current iterative network transmission energy efficiency value q, that is, let q *=q, the Lagrangian dual decomposition algorithm is used to solve the objective function of the equivalent convex problem optimization model under the constraints, and the network resource allocation decision variable solution set {P M′ ,S M′ ,A M′} and the corresponding objective function solution F(P M′ ,S M′ ), where P M′ ,S M′ ,A M′ Respectively represent the transmission power after the current iteration update The allocation mode decision variable set and the selection variable s of the received power allocation mode n and a n The set of decision variables for the allocation mode;

[0136] S103, determine whether F(P M′ ,S M′ )<F Δ ; If satisfied, jump to step S105; otherwise, execute step S104;

[0137] S104: Determine whether k < I is satisfied at this time max If satisfied, then Let {P M ,S M ,A M}={P M′ ,S M′ ,A M′}Substitute into the equivalent convex problem optimization model, and increment the iteration number variable k by 1, then return to step S102; if not satisfied, execute step S105;

[0138] S105, using the current solution obtained for the network resource allocation decision variable solution set {P M′ ,S M′ ,A M′} is output as the optimization result of the network resource allocation decision variable.

[0139] Table 1

[0140]

[0141] Through the above process, by solving OPT-2 in each iteration, the DSNRA algorithm can converge to the optimal energy efficiency.

[0142] B. Transformation and solution of sub-problems

[0143] In order to convert OPT-2 into a convex problem, it is necessary to further relax its constraints. The specific method is:

[0144] First, the selection variable s of the receiving power allocation mode in the “integer constraints” C4 and C6 is set to n and a n The value range of is relaxed to a real number between 0 and 1, that is, the constraints C4 and C6 are relaxed to the following constraints:

[0145] C4′:s n ∈[0,1],n∈{1,2,…,M},

[0146] C6′:a n ∈[0,1],n∈{1,2,…,M}; (13)

[0147] Then, define the cluster head CH i To cluster head CH j Transmission power auxiliary variable for transmitting information Then we get the equivalent cluster head CH i Auxiliary variable of transmission power for transmitting information to relay R represents the equivalent signal-to-noise ratio of the energy transfer process, and Then, the constraint C1 is relaxed to the following constraint:

[0148]

[0149] Replacing the constraint C1 with C1′ shown in equation (14) can decouple and It is worth noting that although a n The relaxation process is performed, but the a obtained in the following text is n It is still in binary form, which shows that the above replacements are equivalent.

[0150] After converting the objective function of the energy efficiency optimization extended model using the Dinkelbach method, combined with the above-mentioned constraint relaxation processing, the equivalent convex problem optimization model is obtained, including the following objective function and constraints:

[0151] The objective function of the optimization model of the equivalent convex problem is:

[0152] OPT-2:

[0153] in,

[0154] The constraints of the optimization model for the equivalent convex problem include:

[0155]

[0156]

[0157] Among them, F(P M ,S M ) indicates the head CH i With cluster head CH j The network transmission energy efficiency value U eff(ij) (P M ,S M ) equivalent convex problem optimization function; q * Transmit energy efficiency value U for the network eff(ij) (P M ,S M ) is the optimized value of .

[0158] After the above transformation and processing, OPT-2 is about optimizing variables s n and a n First, using the second derivative and perspective transformation, we can prove that U (ij) and About and n is a concave function, and U TP(ij) is about a n Second, it is obvious that constraint C3 describes a convex set, while constraints C1′, C2, C3-C9 are all about the optimization variable {P M ,S M ,A M In summary, it can be proved that OPT-2 is a convex problem.

[0159] At this point, the method described in steps S101 to S105 above can be used to solve the maximization problem OPT-2 of the equivalent convex problem optimization model. During the solution process, in step S102, it is necessary to actually perform a solution process under constraints on the objective function OPT-2 of the equivalent convex problem optimization model. Convexity indicates that the converted OPT-2 problem satisfies the Slater condition. Since the optimal dual difference is zero, the present invention uses the Lagrangian dual decomposition method to solve the converted OPT-2 problem, and the specific method is:

[0160] The Lagrangian dual decomposition method is used to convert the objective function solution problem of the equivalent convex problem optimization model into the corresponding dual problem. By solving the dual problem, the solution set of network resource allocation decision variables {P M′ ,S M′ ,A M′}; The transformation obtains the dual problem expressed as the Lagrangian function as follows:

[0161]

[0162] in, are auxiliary variables of Lagrangian dual decomposition, and:

[0163]

[0164] β M ,δ,ζ,θ,ω,μ M and ν M are the Lagrange multiplier constants corresponding to the constraints C1′, C2, C3, C5, C7, C8 and C9 of the optimization model of the equivalent convex problem.

[0165] This transformation leads to the dual problem, which is solved iteratively at two levels. At the lower level, the internal maximization subproblem is solved while keeping the Lagrange multiplier constant. At the higher level, the multiplier is updated using the solution to the minimization subproblem. At the lower level, when q is known, equations (17), (18), and (19) give s n and a n The closed-form solution of in each iteration. As shown in equations (18) and (19), s n and a n is a Boolean quantity (satisfying binary logic), whose values ​​are determined by their maximum marginal benefits X k and Y l Decide.

[0166]

[0167]

[0168] in, The parameters are s n and a n The updated value after iteration; X k and Y l They are s n and a n The maximum marginal benefit of α is the preference selection weight for receiving power allocation mode selection; β is the preference selection weight for receiving power allocation mode selection; n ∈β M , μ n ∈μ M 、v n ∈v M are the Lagrange multiplier constants β M , μ M 、v M The nth element in .

[0169] At a high level, we use the general gradient descent method to obtain the update formula, which is not explained in detail here. By alternating the above strategies, the objective function value of the transformed OPT-2 will eventually converge to the optimal value due to its convexity.

[0170] In summary, the present invention focuses on the energy efficiency optimization of data transmission between cluster heads equipped with discrete power dividers, and proposes an energy efficiency optimization method for inter-cluster energy transmission in wireless sensor networks. This method establishes an energy efficiency optimization model suitable for energy transmission between WSN cluster heads using discrete power dividers; then, we propose a dimension expansion method for this model, and use convex transformation and variable relaxation to convert the optimization problem with a discrete feasible solution set into an equivalent convex problem that is easy to solve; then, an algorithm for iteratively solving the optimization problem is proposed, and the closed expression of variables in the Lagrangian dual decomposition sub-algorithm is derived, realizing the solution of the energy efficiency optimization scheme. The simulation results show that the method proposed by the present invention can converge to obtain the optimal solution with a small number of iterations, and the obtained energy efficiency optimization scheme can achieve a trade-off between transmission energy efficiency, transmission power and discrete power split ratio, so as to maximize the transmission energy efficiency. The method of the present invention can support energy transmission nodes equipped with discrete power dividers, and provides a technical reference for high-energy-efficiency transmission of clustered WSNs with low cost and low hardware complexity.

[0171] Example:

[0172] To verify the effectiveness of the method of the present invention, Figure 2 The analysis is carried out taking the wireless sensor network structure schematic diagram shown as an example.

[0173] A. Simulation environment and parameter settings

[0174] set up Figure 1 In the network shown, the maximum communication distance of the cluster head node is 20 meters, while the maximum communication distance of the cluster member nodes is 10 meters, and the nodes can communicate with each other. In addition, it is assumed that the small-scale fading of the communication link between network nodes obeys Rayleigh fading, and the channel gain is The coefficient α = 2, It obeys the standard complex Gaussian distribution CN(0,1). In the following experiments, unless otherwise specified, we set the parameters as follows: B = 0.2MHz, T = 10s, η = 0.7, ε = 2, P max =0.3W,P C =0.01W,R min =0.1Mbps,α=1. Let the power allocation ratio coefficient of the node be and The value is a number in the set {0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0}. Generally, the following experimental result values ​​are obtained by averaging 10,000 independent experimental results under different channel conditions.

[0175] B. Verification of algorithm convergence

[0176] We first verify the convergence of the DSNRA algorithm. The convergence process of the algorithm to solve the optimal subnet transmission energy efficiency is as follows: Figure 5 As shown. Assume that the noise power parameters of each channel are: Visible: In When different values ​​are taken, the DSNRA algorithm can converge within about 5 iterations. The horizontal axis variable in the figure is the number of iterations of the main loop in the algorithm, and does not include the number of iterations of the dual decomposition method used each time to solve the subproblem OPT-2. Figure 6 The convergence process of the objective function when solving OPT-2 using the dual decomposition method under the condition of given q is demonstrated. Figure 5 In the figure, the curve named "case 1" is the optimal transmission energy efficiency of the cooperative SWIPT inter-cluster network obtained for the continuous power divider. The curve named "case 2" shows the optimization result obtained by the algorithm proposed in the present invention when the discrete power divider is applied. It can be seen that the optimal energy efficiency obtained by the DSNRA algorithm when the discrete power divider is applied (case 2) is 96% of the optimal value in Case 1. This proves the feasibility of the algorithm in discrete power divider engineering.

[0177] C. Performance Verification under the Influence of Power Splitting Resolution

[0178] Figure 7 The power splitting resolution ρ of different transmission protocols is shown in Res The best transmission energy efficiency under the influence of Res Defined as the difference between two adjacent discrete power split values. Figure 7 The curve marked as RT in the figure is the optimization result under the PSR transmission protocol, and the curve marked as DT is the optimization result under the scenario of direct transmission between cluster heads without relays. Obviously, the application of the PSR protocol can effectively improve the optimal transmission energy efficiency. When ρ Res ∈[0.125,0.25], the performance of the DSNRA algorithm is better and can provide a reference for engineering applications.

[0179] D. Performance verification under the influence of maximum transmission power

[0180] exist Figure 8 In the figure, we plotted the maximum transmit power P under different scenarios. max The optimal energy efficiency diagram under the influence ofmax When it is less than 20dBm, the optimal transmission efficiency increases with P max For comparison, we replace the optimization objective of OPT-2 with the maximum transmission throughput. We then plot the relevant numerical results. Figure 6 It can be observed that increasing P max The benefit to transmission energy efficiency is limited. In addition, the performance of the DSNRA algorithm is better when the network chooses the PSR protocol instead of the DT protocol.

[0181] VI. Conclusion

[0182] In the method of the present invention, the transmission energy efficiency maximization of the "inter-cluster network using discrete power allocators" is considered, and the corresponding optimization model is established using the dimension expansion method; then, a discrete energy transmission network resource allocation (DSNRA) algorithm is proposed, which uses convex transformation and variable relaxation methods to solve non-convex discrete optimization problems. The simulation results show that the method proposed in the present invention can significantly optimize the energy efficiency of cooperative energy transmission between WSN clusters, and can support energy transmission nodes equipped with discrete power allocators, providing a reference for high-energy-efficiency transmission of clustered WSNs with low cost and low hardware complexity.

[0183] It is worth noting that the method of the present invention is based on the "discrete power allocator" as a technical premise, taking into account the maximization of the transmission energy efficiency of the "inter-cluster network using a discrete power allocator", and using the dimension expansion method to establish a corresponding optimization model to achieve the energy efficiency optimization solution for the inter-cluster energy transmission of wireless sensor networks. The main contributions of the present invention can be summarized as follows:

[0184] 1. A dimension expansion method is introduced in the establishment of the energy efficiency maximization model, which effectively reduces the difficulty of solving discrete problems;

[0185] 2. Different discrete power split ratio modes (gears) selection parameters (optimization parameters) are designed for energy harvesting and information decoding, so as to have greater flexibility in practical applications;

[0186] 3. A resource allocation optimization algorithm is proposed for inter-cluster energy transmission in wireless sensor networks equipped with discrete power allocators. The simulation results show that the method proposed in the present invention can converge to the optimal solution with a small number of iterations, and the resulting energy efficiency optimization scheme can achieve the maximum system energy efficiency among transmission energy efficiency, transmission power and discrete power split ratio.

[0187] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit the technical solution. Those skilled in the art should understand that those modifications or equivalent substitutions of the technical solution of the present invention that do not depart from the purpose and scope of the technical solution should be included in the scope of the claims of the present invention.

Claims

1. A method for optimizing energy efficiency of inter-cluster energy transmission in a wireless sensor network, characterized in that: include: Construct an energy efficiency optimization model for data transmission between cluster heads in wireless sensor networks; Expanding the dimension of the energy efficiency optimization model and relaxing the constraints, and converting the energy efficiency optimization model into an equivalent convex problem optimization model; The Lagrangian dual decomposition sub-algorithm is used to solve the optimization model of the equivalent convex problem, and an energy efficiency optimization scheme for data transmission between cluster heads in wireless sensor networks is obtained.

2. The energy efficiency optimization method for inter-cluster energy transmission in wireless sensor networks according to claim 1 is characterized in that: In the energy efficiency optimization model for data transmission between cluster heads in the wireless sensor network, the transmission energy efficiency optimization model between any two cluster heads is expressed as: Among them, P ij Indicates cluster head CH i To cluster head CH j The transmission power of the transmitted information; and Respectively represent the transmission power split to the cluster head CH j The power allocation ratio coefficient of the signal decoder and the energy harvester, and Indicated by P ij , Cluster head CH as a relevant variable i With cluster head CH j The network transmission energy efficiency value between (ij) and U TP(ij) They represent cluster heads CH i With cluster head CH j The data throughput and total power consumption during the energy transmission process; and: U TP(ij) =εP ij +P C ; Wherein, B is the baseband bandwidth, η is the energy conversion efficiency, and 0<η<1; Indicates cluster head CH i Communication link with relay R ir , relay R and cluster head CH j Communication link between rj and cluster head CH i With cluster head CH j Communication link between ij The normalized channel gains of the three wireless communication links, h ab represents the channel gains of the three wireless communication links, a,b∈{i,j,r}; λ ij Represents the equivalent signal-to-noise ratio of the energy transmission process; ε≥1 indicates cluster head CH i The conversion efficiency of the power amplifier; P C Indicates the total power consumption of the network circuit.

3. The energy efficiency optimization method for inter-cluster energy transmission in wireless sensor networks according to claim 1 is characterized in that: The dimension expansion and constraint relaxation processing of the energy efficiency optimization model specifically include: Firstly, the dimension of the energy efficiency optimization model of data transmission between cluster heads in wireless sensor networks is expanded to obtain a corresponding energy efficiency optimization expansion model, and then the constraint conditions of the energy efficiency optimization expansion model are relaxed to obtain a corresponding equivalent convex problem optimization model.

4. The energy efficiency optimization method for inter-cluster energy transmission in wireless sensor networks according to claim 3 is characterized in that: The specific method of dimension expansion processing for the energy efficiency optimization model of data transmission between cluster heads in wireless sensor networks is as follows: The transmission power is distributed to the cluster head CH j Power allocation ratio coefficient between signal decoder and energy harvester and Corresponding to several discrete gear values ​​of preset extension, that is, and The value of belongs to the preset extended discrete value space: in, and Respectively represent the power allocation ratio coefficient and At the nth value in the discrete value space, M represents the number of preset extended discrete gear values ​​contained in the discrete value space, n∈{1,2,…,M}, and When the power allocation ratio coefficient of the signal decoder and the power distribution ratio coefficient of the energy harvester Select any discrete gear value and When , it corresponds to a receiving power allocation mode, and in this receiving power allocation mode, the cluster head CH i To the relay R and cluster head CH j The transmission power of the transmitted information is recorded as and Cluster Head CH j The energy gained is recorded as but: Where T is the transmission period, η is the energy conversion efficiency, and 0<η<1; h ij Indicates cluster head CH i With cluster head CH j Communication link between ij The channel gain of n and a n is the selection variable of the received power allocation mode, and s n ∈{0,1},a n ∈{0,1}; This achieves the dimensional expansion of the energy efficiency optimization model for data transmission between cluster heads in wireless sensor networks.

5. The energy efficiency optimization method for inter-cluster energy transmission in wireless sensor networks according to claim 4 is characterized in that: The energy efficiency optimization model obtained by dimension expansion processing includes the following objective functions and constraints: The objective function of the energy efficiency optimization extended model is: in: The constraints of the energy efficiency optimization extended model include: in, S M ={s n ,n=1,2,…,M} and A M ={a n , n=1,2,…,M} respectively represent the transmission power in the energy efficiency optimization extended model The allocation mode decision variable set and the selection variable s of the receiving power allocation mode n and a n The set of decision variables for the allocation mode; U eff(ij) (P M ,S M ) indicates that P M , S M Cluster head CH as the relevant variable i With cluster head CH j The network transmission energy efficiency value between (ij) (P M ,S M ) and U TP(ij) (P M ,S M ) represent P M , S M Cluster head CH as the relevant variable i With cluster head CH j Data throughput and total power consumption during energy transmission; Indicates cluster head CH j Energy requirement; P max The cluster head CH i The upper limit of the transmission power; α is the preference selection weight for the receiving power allocation mode selection; P ij Indicates cluster head CH i To cluster head CH j Transmission power of transmitted information; ε≥1 indicates cluster head CH i The conversion efficiency of the power amplifier; P C Indicates the total power consumption of the network circuit.

6. The energy efficiency optimization method for inter-cluster energy transmission in wireless sensor networks according to claim 5, characterized in that: The specific method of performing constraint relaxation processing on the energy efficiency optimization extended model is as follows: The variable s for selecting the receiving power allocation mode is n and a n The value range of is relaxed to a real number between 0 and 1, that is, the constraints C4 and C6 are relaxed to the following constraints: C4′:s n ∈[0,1[,n∈{1,2,…,M}, C6′:a n ∈[0,1],n∈{1,2,…,M}; Then, define the cluster head CH i To cluster head CH j Transmission power auxiliary variable for transmitting information Then we get the equivalent cluster head CH i Auxiliary variable of transmission power for transmitting information to relay R represents the equivalent signal-to-noise ratio of the energy transfer process, and Then, the constraint C1 is relaxed to the following constraint: C1′: Furthermore, the objective function of the energy efficiency optimization extended model will be transformed using the Dinkelbach method; Thus, the constraint relaxation of the energy efficiency optimization extended model is completed.

7. The energy efficiency optimization method for inter-cluster energy transmission in wireless sensor networks according to claim 6 is characterized in that: The equivalent convex problem optimization model obtained by performing constraint relaxation processing on the energy efficiency optimization extended model includes the following objective functions and constraints: The objective function of the optimization model of the equivalent convex problem is: in, The constraints of the optimization model for the equivalent convex problem include: C1′: C2: C3: C4′:s n ∈[0,1],n∈[1,2,…,M}, C5: C6′:a n ∈[0,1],n∈{1,2,…,M}, C7: C8: C9: Among them, F(P M ,S M ) indicates the head CH i With cluster head CH j The network transmission energy efficiency value U eff(ij) (P M ,S M ) equivalent convex problem optimization function; q * Transmit energy efficiency value U for the network eff(ij) (P M ,S M ) is the optimized value of .

8. The energy efficiency optimization method for inter-cluster energy transmission in wireless sensor networks according to claim 7, characterized in that: The specific process of solving the optimization model of the equivalent convex problem is as follows: S101, preset number of iterations I max and convergence threshold F Δ , and initialize the iteration number variable k=1, and initialize the iteration network transmission energy efficiency value q=0; S102, let the optimized value of network transmission energy efficiency q * The value is the current iterative network transmission energy efficiency value q, that is, let q * =q, the Lagrangian dual decomposition algorithm is used to solve the objective function of the equivalent convex problem optimization model under the constraints, and the network resource allocation decision variable solution set {P M′ ,S M′ ,A M′ } and the corresponding objective function solution F(P M′ ,S M′ ); Among them, P M′ ,S M′ ,A M′ Respectively represent the transmission power after the current iteration update The allocation mode decision variable set and the selection variable s of the received power allocation mode n and a n The set of decision variables for the allocation mode; S103, determine whether F(P M′ ,S M′ )<F Δ ; If satisfied, jump to step S105; otherwise, execute step S104; S104: Determine whether k < I is satisfied at this time max If satisfied, then Let {P M ,S M ,A M }={P M′ ,S M′ ,A M′ }Substitute into the equivalent convex problem optimization model, and increment the iteration number variable k by 1, then return to step S102; if not satisfied, execute step S105; S105, using the current solution obtained for the network resource allocation decision variable solution set {P M′ ,S M′ ,A M′ } is output as the optimization result of the network resource allocation decision variable.

9. The method for optimizing energy efficiency of inter-cluster energy transmission in wireless sensor networks according to claim 8, characterized in that: In step S102, the specific method of using the Lagrangian dual decomposition sub-algorithm to solve the objective function of the equivalent convex problem optimization model under the constraint condition is: The Lagrangian dual decomposition method is used to convert the objective function solution problem of the equivalent convex problem optimization model into the corresponding dual problem. By solving the dual problem, the solution set of network resource allocation decision variables {P M′ ,S M′ ,A M′ }; The transformation results in the dual problem being expressed as: in, are auxiliary variables of Lagrangian dual decomposition, and: β M ,δ,ζ,θ,ω,μ M and ν M are the Lagrange multiplier constants corresponding to the constraints C1′, C2, C3, C5, C7, C8 and C9 of the optimization model of the equivalent convex problem.

10. The method for optimizing energy efficiency of inter-cluster energy transmission in wireless sensor networks according to claim 9, characterized in that: In the process of solving the dual problem, the parameter s n and a n The updated closed-form solutions in each iteration are: in, The parameters are s n and a n The updated value after iteration; X k and Y l They are s n and a n The maximum marginal benefit of α is the preference selection weight for receiving power allocation mode selection; β is the preference selection weight for receiving power allocation mode selection; n ∈β M , μ n ∈μ M 、v n ∈v M are the Lagrange multiplier constants β M , μ M 、v M The nth element in .