A smart reflective surface-assisted energy harvesting method for cognitive wireless sensor networks

Abstract

The present invention is a method for energy collection in a cognitive wireless sensor network assisted by an intelligent reflective surface. The method is characterized in that: a single-antenna convergence node located at the center of the network acts as a dedicated energy source, radiating RF signals outward at a fixed power P0 within a total energy collection time T; single-antenna cognitive wireless sensor network nodes distributed within the network area receive the RF signals radiated by the convergence node and use internal circuits with a conversion efficiency of α, 0<α<1 to convert the received RF signals into electrical energy stored in a battery; the intelligent reflective surface divided into sub-surfaces intelligently configures its reflection factor matrix to assist the cognitive wireless sensor network nodes in collecting energy simultaneously through direct links and cascaded reflection links between the nodes and the convergence node, thereby maximizing the sum of the energy collected by all cognitive wireless sensor network nodes; and by allowing reflection units within the same sub-surface to share the same reflection factor, the number of optimization variables is greatly reduced, thereby lowering the computational difficulty.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless communications, and in particular relates to an energy collection method for a cognitive wireless sensor network assisted by an intelligent reflective surface. Background Art

[0002] Cognitive Radio Sensor Networks (CRSNs) are a novel network architecture that intelligently integrates traditional wireless sensor networks with cognitive radio technology. They can effectively alleviate the spectrum resource shortages and other challenges faced by traditional wireless sensor networks. However, CRSN node batteries have limited capacity and cannot typically be replaced periodically. Furthermore, the additional energy required to perform cognitive functions can lead to rapid battery depletion and shorten the overall network lifespan. Energy harvesting (EH) technology allows nodes to supplement their limited battery capacity by harvesting energy from natural sources such as solar and wind power, or from RF sources. This technology offers the potential to ensure stable and sustainable operation of CRSNs and is a promising solution to their energy consumption challenges.

[0003] Compared to natural sources, the energy provided by RF sources is relatively stable and predictable, making RF EH technology more widely used. However, existing EH-CRSNs nodes only collect energy through direct links with the energy source. Severe path loss leads to low EH efficiency for nodes farther away from the RF source, limiting the potential of RF EH technology to extend the overall network life. Intelligent Reflecting Surfaces (IRS) achieve reconfiguration of the wireless propagation environment by intelligently adjusting their reflection factor matrix, thereby improving communication performance. Specifically, each passive reflective element on the IRS can independently change the amplitude and phase of the incident signal, allowing the signals from the direct link and the IRS cascade reflection link to superimpose in phase at the receiving end. Therefore, IRS has the potential to enhance the energy value collected by CRSNs nodes.

[0004] Research on IRS-assisted EH currently focuses primarily on wireless drive communication networks, wireless energy-carrying communication systems, and drone communication systems. However, unlike IRS-assisted wireless drive communication networks, where users rely entirely on energy harvested from downlinks for uplink data transmission, IRS-assisted EH-CRSNs nodes use energy harvested from downlinks to supplement node battery power. Unlike base stations in IRS-assisted wireless energy-carrying communication systems, which transmit both data and energy simultaneously via downlinks, EH-CRSNs nodes harvest energy from sink nodes via downlinks and transmit data to sink nodes via uplinks. Unlike drones, whose energy sources and data sinks may be constantly changing, CRSNs typically assume fixed energy sources and data sinks. Therefore, current research results on IRS-assisted EH are not applicable to CRSNs. Furthermore, IRS-assisted EH research typically determines the optimal reflection factor matrix configuration for the IRS by constructing and solving a non-convex optimization problem. The number of optimization variables solved is positively correlated with the number of reflection units and users. As the number of reflective units or users in an IRS increases, the time required to solve the optimization problem increases dramatically, and it can even be impossible to obtain a good solution in a short time, which is not conducive to the practical application of IRS. To achieve a compromise between network performance and the complexity of solving the optimization problem, a sub-surface IRS can be used. In this case, the number of variables in the optimization problem is positively correlated with the number of sub-surfaces, greatly reducing the difficulty of solving the problem. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to overcome the shortcomings of the existing technology and provide a scientific, reasonable, applicable and effective method for configuring the optimal reflection factor matrix of intelligent reflective surfaces, which can reduce the computational difficulty and maximize the sum of the energy collected by the nodes of the cognitive wireless sensor network.

[0006] The solution to the technical problem is a method for energy harvesting in a cognitive wireless sensor network assisted by an intelligent reflective surface, which is characterized in that the method comprises the following steps:

[0007] 1) A single-antenna aggregation node located at the center of the network acts as a dedicated energy source, radiating RF signals with a fixed power P0 within the total energy collection time T;

[0008] 2) Single-antenna cognitive wireless sensor network nodes distributed within the network area receive RF signals radiated by the sink node and use internal circuits with a conversion efficiency of α, 0 < α < 1 to convert the received RF signals into electrical energy stored in the battery;

[0009] 3) The intelligent reflective surface divided into sub-surfaces intelligently configures its reflection factor matrix to assist the cognitive wireless sensor network nodes in collecting energy through direct links and cascaded reflection links between the aggregation nodes, so as to maximize the sum of the energy collected by all cognitive wireless sensor network nodes.

[0010] Furthermore, in step 3), the intelligent reflective surface is composed of N, N=L×M passive reflective units regularly arranged in L rows and M columns, and the size of each reflective unit is d x ×d y , ignoring the intervals between adjacent reflective units, N reflective units are divided into B, B = L / l × M / m sub-surfaces, each sub-surface consists of l × m reflective units sharing the same reflection factor; the smart reflective surface is deployed at position R just above the aggregation node to ensure that all cognitive wireless sensor network nodes can fully utilize the array gain brought by the smart reflective surface.

[0011] Furthermore, in step 3), the optimal reflection factor matrix configuration of the smart reflective surface is obtained by constructing and solving the problem of maximizing the sum of the energy collected by all cognitive wireless sensor network nodes within the total energy collection time T. The problem of maximizing the sum of the energy collected by all cognitive wireless sensor network nodes within the total energy collection time T is constructed as:

[0012]

[0013]

[0014] Where K is the total number of cognitive wireless sensor network nodes distributed in the network area; B is the set of all sub-surfaces on the smart reflective surface; is the equivalent composite channel between the sink node and the smart reflective surface, r s_b is the equivalent channel between the sink node and the sub-surface b of the smart reflective surface, b = 1, 2, ..., B, and the Euclidean distance between the two closely related, is the equivalent composite channel between the smart reflective surface and the cognitive wireless sensor network node k, k = 1, 2, ..., K, is the equivalent channel between the sub-surface b of the smart reflective surface and the node k of the cognitive wireless sensor network, and the Euclidean distance between the two closely related, is the diagonal reflection factor matrix of the smart reflective surface, is the reflection factor of subsurface b, β b and Ω b are the reflection amplitude and reflection phase shift of sub-surface b respectively. To maximize the energy and energy collected by all nodes, take β b =1, is the equivalent channel between the sink node and the cognitive wireless sensor network node k, and the Euclidean distance between the two Closely related.

[0015] Furthermore, the problem of maximizing the energy collected by all cognitive wireless sensor network nodes within the total energy collection time T and constructing (P1) can be simplified to:

[0016]

[0017]

[0018] make where μ=[μ1,...,μ b ,...,μ B ] H , In addition, the phase shift constraint in Equation (4) can be equivalently converted into the unit mode constraint |μ b |=1, then (P2) can be rewritten as:

[0019]

[0020]

[0021] Among them, the objective function is a non-concave function of μ, and the constraint is a non-convex constraint. Therefore, the optimization problem is a non-convex optimization problem and cannot be optimally solved. Using the continuous convex approximation algorithm, the objective function in formula (5) can be expanded as follows:

[0022]

[0023] make According to the first-order Taylor expansion of the convex function, the global lower bound of the first-order Taylor expansion of W at a certain feasible point is restricted, and a local feasible point μ satisfying Equation (6) is randomly generated. The lower bound of the objective function is:

[0024]

[0025] Then the optimal reflection factor configuration is:

[0026]

[0027] If the average error between μ* and μ is greater than the required accuracy ε, set μ = μ* and repeat the calculation process of formula (9) until the accuracy requirement is met; otherwise, output the optimal reflection factor matrix configuration μ*.

[0028] Compared with existing achievements, the intelligent reflective surface-assisted cognitive wireless sensor network energy harvesting method of the present invention has the following beneficial effects:

[0029] 1. Introducing smart reflective surfaces into auxiliary nodes of cognitive wireless sensor networks for energy harvesting can improve energy utilization;

[0030] 2. The sub-surface partitioning approach can reduce the complexity of solving additional optimization problems caused by the increase in the number of reflective units on the smart reflective surface or the number of nodes in the cognitive wireless sensor network;

[0031] 3. By constructing and solving a non-convex optimization problem aimed at maximizing the sum of the energy collected by all cognitive wireless sensor network nodes, we obtain the optimal reflectivity matrix configuration for the smart reflective surface, achieving an effective trade-off between energy harvesting performance and computational complexity.

[0032] 4. The method is scientific and reasonable, highly applicable and has good effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a flow chart of a method for energy harvesting in a cognitive wireless sensor network assisted by an intelligent reflective surface according to the present invention;

[0034] Figure 2 Network architecture diagram of cognitive wireless sensor network assisted by smart reflective surface;

[0035] Figure 3 Schematic diagram of the intelligent reflective surface structure divided into sub-surfaces;

[0036] Figure 4 Flowchart of the solution to the problem of maximizing the energy collected by all nodes in a cognitive wireless sensor network. DETAILED DESCRIPTION

[0037] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0038] like Figure 1 As shown, the present invention provides a method for energy collection in a cognitive wireless sensor network assisted by an intelligent reflective surface, comprising the following steps:

[0039] 1) The single-antenna aggregation node located at the center of the network acts as a dedicated energy source, radiating RF signals with a fixed power P0 within the total energy collection time T, as shown in Figure 2 As shown, the coordinates of the sink node are (x0, y0, 0).

[0040] 2) K single-antenna cognitive wireless sensor network nodes distributed in the network area receive the RF signal radiated by the sink node and use internal circuits with a conversion efficiency of α, 0<α<1 to convert the received RF signal into electrical energy stored in the battery. The coordinates of the cognitive wireless sensor network node k, k = 1, 2, ..., K are (x k ,y k ,0), therefore, the Euclidean distance between node k and the sink node can be expressed as

[0041] 3) The intelligent reflective surface divided into sub-surfaces intelligently configures its reflection factor matrix to assist the cognitive wireless sensor network nodes to collect energy through direct links and cascade reflection links between the sink nodes, so as to maximize the sum of the energy collected by all cognitive wireless sensor network nodes. Figure 2 As shown in Figure 1, the intelligent reflective surface is deployed at position R just above the aggregation node to ensure that all cognitive wireless sensor network nodes can fully utilize the array gain brought by the intelligent reflective surface. The coordinates of its center point are (x0, y0, R). Figure 3 As shown, the smart reflective surface consists of N, N = L × M reflective units arranged regularly in L rows and M columns. The size of each reflective unit is d x ×d y Ignoring the intervals between adjacent reflective units, the N reflective units are divided into B, B = L / l × M / m sub-surfaces, each sub-surface consisting of l × m reflective units sharing the same reflection factor. According to the coordinates of the center of the smart reflective surface, the size of each reflective unit and sub-surface, the center coordinates of the sub-surface b, b = 1, 2, ..., B can be expressed as (x0-(Mm)d x / 2+m×(j-1)d x ,y0+(Ll)d y / 2-l×(i-1)d y ,R), where i∈[1,L / l], j∈[1,M / m]. According to the coordinates of the sink node and the cognitive wireless sensor network node, the Euclidean distances between the sink node and the cognitive wireless sensor network node k and the center of the sub-surface b can be determined as follows:

[0042]

[0043] The above distance and channel type jointly determine the equivalent channel r between the sink node and sub-surface b, sub-surface b and cognitive wireless sensor network node k, and the sink node and cognitive wireless sensor network node k. s_b , and At this time, the equivalent composite channels between the sink node and the smart reflective surface, and between the smart reflective surface and the cognitive sensor network node k can be expressed as and In the intelligent reflective surface-assisted cognitive wireless sensor network energy harvesting method, the total energy collected by all cognitive wireless sensor network nodes within the total energy harvesting time T is:

[0044]

[0045] in, is the diagonal reflection factor matrix of the smart reflection surface divided into sub-surfaces, is the reflection factor of subsurface b, β b and Ω b are the reflection amplitude and reflection phase shift of sub-surface b respectively. In order to maximize the energy collected by all cognitive wireless sensor network nodes, take β b =1.

[0046] The problem of maximizing the energy collected by all cognitive wireless sensor network nodes within the total energy collection time T is formulated as:

[0047]

[0048]

[0049] Wherein, B is the set of all sub-surfaces on the smart reflective surface.

[0050] Under the conditions that the total energy collection time T, the transmission power P0 of the sink node, and the conversion efficiency α of the energy collection circuit are constant, the optimization problem (P1) can be simplified as follows:

[0051]

[0052]

[0053] make where μ=[μ1,...,μ b ,...,μ B ] H , In addition, the phase shift constraint in Equation (5) can be equivalently converted into the unit mode constraint |μ b |=1, then (P2) can be rewritten as:

[0054]

[0055]

[0056] Among them, the objective function is a non-concave function of μ, and the constraint is a non-convex constraint. Therefore, this optimization problem is a non-convex optimization problem and generally cannot be optimally solved. Figure 4As shown, the present invention uses a continuous convex approximation algorithm to solve the constructed non-convex optimization problem. Specifically, the objective function shown in formula (6) can be expanded as:

[0057]

[0058] make Then formula (8) is transformed into:

[0059]

[0060] According to the first-order Taylor expansion of the convex function, the first-order Taylor expansion of W at a certain feasible point is globally limited. Randomly generate a local feasible point μ that satisfies Equation (7), and the quadratic term μ H Wμ satisfies the following conditions:

[0061] μ H Wμ≥2Re{μ H Wμ}-μ H Wμ (10)

[0062] Where Re represents μ H The real part of the Wμ term. The lower bound of the objective function is:

[0063]

[0064] Then the optimal reflection factor configuration is:

[0065]

[0066] If the average error between μ* and μ is greater than the required accuracy ε, set μ = μ* and repeat the calculation process of formula (9) until the accuracy requirement is met; otherwise, output the optimal reflection factor matrix configuration μ*.

[0067] The above embodiments are provided for illustrative purposes only and are not intended to limit the embodiments. Those skilled in the art will appreciate that other variations or modifications based on the above descriptions are possible. It is not necessary and impossible to enumerate all embodiments here. Obvious variations or modifications arising therefrom remain within the scope of protection of the present invention.

Claims

1. A method for energy harvesting in a cognitive wireless sensor network assisted by an intelligent reflective surface, characterized in that: The method comprises the following steps: 1) A single-antenna aggregation node located at the center of the network acts as a dedicated energy source, radiating RF signals with a fixed power P0 within the total energy collection time T; 2) Single-antenna cognitive wireless sensor network nodes distributed within the network area receive RF signals radiated by the sink node and use internal circuits with a conversion efficiency of α, 0 < α < 1 to convert the received RF signals into electrical energy stored in the battery; 3) The intelligent reflective surface, which is divided into sub-surfaces, intelligently configures its reflection factor matrix to assist the cognitive wireless sensor network nodes in collecting energy through both direct links and cascaded reflective links with the sink node, thereby maximizing the sum of the energy collected by all cognitive wireless sensor network nodes; In step 3), the smart reflective surface is composed of N, N=L×M passive reflective units arranged regularly in L rows and M columns, and the size of each reflective unit is d x ×d y , ignoring the intervals between adjacent reflective units, N reflective units are divided into B, B = L / l × M / m sub-surfaces, each sub-surface consists of l × m reflective units sharing the same reflection factor; the smart reflective surface is deployed at position R just above the aggregation node to ensure that all cognitive wireless sensor network nodes can fully utilize the array gain brought by the smart reflective surface.

2. The method for energy harvesting in a cognitive wireless sensor network assisted by an intelligent reflective surface according to claim 1, wherein: In step 3), the optimal reflection factor matrix configuration of the smart reflective surface is obtained by constructing and solving the problem of maximizing the sum of the energy collected by all cognitive wireless sensor network nodes within the total energy collection time T. The problem of maximizing the sum of the energy collected by all cognitive wireless sensor network nodes within the total energy collection time T is constructed as: Where K is the total number of cognitive wireless sensor network nodes distributed in the network area; B is the set of all sub-surfaces on the smart reflective surface; is the equivalent composite channel between the sink node and the smart reflective surface, r s_b is the equivalent channel between the sink node and the sub-surface b of the smart reflective surface, b = 1, 2, ..., B, and the Euclidean distance between the two closely related, is the equivalent composite channel between the smart reflective surface and the cognitive wireless sensor network node k, k = 1, 2, ..., K, is the equivalent channel between the sub-surface b of the smart reflective surface and the node k of the cognitive wireless sensor network, and the Euclidean distance between the two closely related, is the diagonal reflection factor matrix of the smart reflective surface, is the reflection factor of subsurface b, β b and Ω b are the reflection amplitude and reflection phase shift of sub-surface b respectively. To maximize the energy collected by all nodes, take β b =1, is the equivalent channel between the sink node and the cognitive wireless sensor network node k, and the Euclidean distance between the two Closely related.

3. The method for energy harvesting in a cognitive wireless sensor network assisted by an intelligent reflective surface according to claim 2, wherein: The problem of maximizing the energy collected by all cognitive wireless sensor network nodes within the total energy collection time T and constructing (P1) can be simplified to: make where μ=[μ1,...,μ b ,...,μ B ] H , In addition, the phase shift constraint in Equation (4) can be equivalently converted into the unit mode constraint |μ b |=1, then (P2) can be rewritten as: Among them, the objective function is a non-concave function of μ, and the constraint is a non-convex constraint. Therefore, this problem is a non-convex optimization problem and cannot be optimally solved. Using the continuous convex approximation algorithm, the objective function in formula (5) can be expanded as follows: make According to the first-order Taylor expansion of the convex function, the global lower bound of the first-order Taylor expansion of W at a certain feasible point is limited, and a local feasible point satisfying formula (6) is randomly generated The lower bound of the objective function is: Then the optimal reflection factor configuration is: If μ* and If the average error of is greater than the required accuracy ε, then let Repeat the calculation process of formula (9) until the accuracy requirement is met; otherwise, output the optimal reflection factor matrix configuration μ*.

Images