A green power internet of things oriented feasibility perception relay scheduling and power division joint configuration method

Abstract

This invention provides a feasibility-aware relay scheduling and power partitioning joint configuration method for green power Internet of Things (IoT). Under a power partitioning-based wireless information and energy co-transmission architecture, this method constructs a three-source energy causal constraint model involving source node radio frequency signals, auxiliary signals from substation or line-side energy access points, and residual loop self-interference signals. Before scheduling execution, the feasible power partitioning intervals for each candidate relay are calculated based on this model, thereby pre-eliminating unschedulable relays that cannot meet the constraints of reliable information decoding, forwarding energy requirements, and self-energy recovery stability before scheduling decisions are made. Subsequently, within the set of schedulable relays, a scheduling utility function is constructed by integrating system throughput, scheduling margin, and the width of the feasible power partitioning interval. Based on this function, relay selection and power partitioning joint configuration are performed, thereby reducing the probability of scheduling congestion and improving the scheduling reliability and resource allocation effectiveness of the energy-constrained power IoT relay network.

Description

Technical Field

[0001] This invention relates to the field of power communication and power Internet of Things (IoT) network resource allocation technology, and in particular to a feasibility-aware relay scheduling and power partitioning joint configuration method for green power IoT. Background Technology

[0002] With the accelerated construction of new power systems, the power Internet of Things (IoT) is evolving towards massive terminal access, heterogeneous interconnection of multiple networks, and unmanned long-term operation. Across the entire power generation, transmission, transformation, distribution, and consumption chain, widely deployed power wireless sensor nodes, distribution automation terminals, and user-side intelligent data acquisition equipment must provide highly reliable, low-latency communication guarantees for services such as grid operation status perception, equipment status monitoring, and precise load control, all within limited spectrum resources and battery capacity. Especially in scenarios where power access is difficult, such as overhead lines, underground utility tunnels, and outdoor substations, the energy constraints of terminals have become a bottleneck limiting the wide-area coverage and long-term maintenance-free operation of the power IoT.

[0003] In power grid multi-terminal wireless access, traditional orthogonal multiple access (OMA) methods allocate independent time-frequency code resources to different terminals. In dense power distribution scenarios, this results in limited spectrum utilization, making it difficult to support large-scale concurrent connections. Non-orthogonal multiple access (NOA) technologies, through power domain multiplexing, allow multiple terminals to share the same resource block and utilize serial interference cancellation for multi-user signal detection, significantly improving spectrum utilization. Simultaneously, wireless information and energy transmission technologies enable power relay nodes or sensing terminals to collect energy from radio frequency signals while receiving information. This helps alleviate the energy replenishment challenges of low-power terminals under passive or semi-passive conditions, supporting long-term autonomous operation.

[0004] In power IoT relay networks supporting cooperative non-orthogonal multiple access, relay nodes must simultaneously complete information reception, energy harvesting, and signal forwarding while meeting the quality of service requirements of power services. For relay nodes employing a power-segmentation structure, the power segmentation ratio directly determines information decoding performance and energy harvesting capability: if the power allocated to the information decoding branch is too low, the relay may be unable to reliably decode critical information such as power grid telemetry and remote control; if the power allocated to the energy harvesting branch is too low, the relay will have insufficient energy available for forwarding, affecting the reliable delivery of downlink control commands. Therefore, there is a significant coupling relationship between power segmentation parameter configuration and relay scheduling decisions, directly affecting the quality of service assurance of power communication services.

[0005] Existing relay selection or resource allocation methods primarily optimize instantaneous rate, system throughput, outage probability, or link quality, typically assuming that candidate relays already meet basic communication requirements. However, in energy-constrained power IoT relay networks, the successful scheduling of candidate relays depends not only on wireless channel quality but also on multiple constraints, including the level of collectable energy, power service quality thresholds, energy causality, and system stability. Failure to pre-assess the feasibility of candidate relays before scheduling may result in the selected relay failing to reliably decode information, acquire sufficient forwarding energy, or meet the decoding requirements of the destination node. This can lead to scheduling congestion or a decline in service quality, impacting the accurate perception of grid operation status and the real-time issuance of control commands.

[0006] Full-duplex relays can simultaneously receive and transmit signals within the same frequency band, which is beneficial for improving the spectrum utilization efficiency of power communication, but it introduces loop self-interference. Traditional designs typically consider loop self-interference as a harmful factor that must be suppressed. In reality, residual loop self-interference signals may affect information decoding on the one hand, and contain radio frequency energy that can be utilized by energy harvesting branches on the other. In addition, energy access points or dedicated power beacons deployed in scenarios such as substations and transmission lines can provide external radio frequency energy to relay nodes. Existing methods generally lack a unified schedulability determination mechanism to determine whether candidate relays meet the constraints of power service quality, energy causality, and self-energy recovery stability, while simultaneously considering the source node radio frequency energy, external energy assistance, and self-interference energy recovery. This makes it difficult to guarantee the stable and reliable operation of energy-constrained power IoT relay networks. Summary of the Invention

[0007] Purpose of the invention: This invention aims to provide a joint configuration method for feasibility-aware relay scheduling and power partitioning for green power Internet of Things, which solves the problem that existing relay selection methods mainly rely on instantaneous throughput, link quality or interruption probability for scheduling, lacking feasibility assessment before scheduling. This results in selected power relays failing to meet the low latency and high reliability requirements of power grid control services, and insufficient forwarding energy, thus causing scheduling congestion.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] A joint configuration method for feasibility-aware relay scheduling and power partitioning for the Internet of Things in green power includes the following steps: Step 1: Obtain network parameters, including the network controller obtaining channel state information, source node transmit power, energy access point transmit power or energy budget, energy conversion efficiency, non-orthogonal multiple access power allocation coefficient, and target signal-to-interference-plus-noise ratio threshold for each candidate relay. Step 2: For each candidate relay, establish a relay energy causal constraint model composed of the source node radio frequency signal, the energy access point auxiliary signal, and the residual loop self-interference signal to obtain the energy-supported relay transmit power. Step 3: Calculate the lower bound of the power split ratio of the candidate relay based on the first hop information decoding constraint and the self-energy recovery stability constraint; Step 4: Calculate the minimum relay transmit power that satisfies the quality of service requirements based on the non-orthogonal multiple access decoding constraints of the destination node; and calculate the upper bound of the power division ratio in conjunction with the energy-supported relay transmit power. Step 5: The schedulable power segmentation interval of the candidate relay is formed by the lower bound and the upper bound; the candidate relays whose schedulable power segmentation interval is empty or whose non-orthogonal multiple access decoding conditions are not met are determined to be unschedulable and removed, thus obtaining a set of schedulable relays; Step 6: Calculate the robust scheduling utility of each candidate relay in the schedulable power partitioning interval within the schedulable relay set, and select the candidate relay with the largest robust scheduling utility and its power partitioning ratio as the scheduling result.

[0010] As a preferred embodiment, in step 1, the power Internet of Things includes a source node, an energy access point, a candidate relay set consisting of multiple candidate relays, a remote destination node, and a near-end destination node; the source node sends non-orthogonal multiple access data packets containing source-end service flows directed to the remote destination node and source-end service flows directed to the near-end destination node; the selected candidate relay receives the source node signal in full-duplex mode and forwards it to the two destination nodes.

[0011] As a preferred embodiment, in step 2, the candidate relay adopts a power partitioning structure, and the power ratio allocated to the information decoding branch is: The power allocated to the energy harvesting branch is ; The contributions of the source node and the energy access point to the energy collection of the candidate relay are denoted as follows: :

[0012] in, The source node transmit power, The transmit power of the energy access point. The channel coefficients from the source node to the candidate relay are denoted as . The channel coefficients from the energy access point to the candidate relay; Power-supported relay transmission power The expression is:

[0013] in, For energy conversion efficiency, This represents the residual loop self-interference channel coefficient.

[0014] As a preferred embodiment, in step 3, the lower bound of the power division ratio is denoted as... The lower realm Find the lower bound of the first jump decoding and stability lower bound The maximum value in is expressed as:

[0015] in, , ; The channel coefficients from the source node to the candidate relay; The residual loop self-interference channel coefficient; This refers to the source node's transmit power. The target SINR threshold; This indicates the noise power at the relay point.

[0016] As a preferred embodiment, in step 4, the minimum relay transmit power is determined jointly based on the decoding constraints of the remote destination node on the service flow, the serial interference cancellation constraints of the near destination node on the service flow, and the decoding constraints of the near destination node on its own service flow, and its expression is:

[0017] in, Indicates the remote destination node decoding Required relay transmission power Indicates the near-end destination node decoding And perform the relay transmission power required for SIC pre-decoding. Indicates the near-end destination node decoding Required relay transmission power.

[0018] As a preferred embodiment, in step 4, the upper bound of the power segmentation ratio obtained from the energy causal constraint is denoted as... :

[0019] in, Minimum relay transmission power, For energy conversion efficiency, The contribution of source nodes and energy access points to energy collection of candidate relays. This represents the residual loop self-interference channel coefficient.

[0020] As a preferred embodiment, in step 5, the schedulable power segmentation interval is denoted as... It satisfies:

[0021] in, This is the lower bound of the power division ratio. This is the upper bound of the power division ratio. The proportion of power allocated to the information decoding branch.

[0022] As a preferred embodiment, in step 5, the necessary condition for the non-orthogonal multiple access decoding condition to be satisfied is:

[0023] in, and The power allocation coefficient for non-orthogonal multiple access; The target signal-to-interference-plus-noise ratio (SIN / N) threshold; If the non-orthogonal multiple access feasibility conditions are not met, or the upper bound is not met... Not positive, or the lower bound stated Greater than the upper bound If so, the candidate relay is determined to be unschedulable and removed before scheduling.

[0024] As a preferred option, step 6 specifically includes: Within the schedulable power partition interval of each candidate relay in the set of schedulable relays, set There are 10 sampling points; for each sampling point, the robust scheduling utility is constructed by combining the system throughput, scheduling margin, and the width of the schedulable power segmentation interval, as shown in the following expression:

[0025] in, Represents the throughput term; This indicates the robustness margin of the run point relative to each constraint boundary; Indicates the width of the feasible interval; , , Non-negative weights; The network controller selects the combination with the highest utility from all schedulable relays and their sampled operating points, and outputs the selected relays and their power split ratio; if there are no schedulable candidate relays, it outputs the scheduling blocking status.

[0026] Furthermore, the present invention also proposes an electronic device comprising a processor and a memory storing computer program instructions; the processor, when executing the computer program instructions, implements the aforementioned joint configuration method for feasibility-aware relay scheduling and power partitioning for a green power Internet of Things.

[0027] Beneficial Effects: This patent application integrates source node radio frequency energy, EAP auxiliary energy, and residual loop self-interference recovery energy into energy causality constraints, forming a three-source energy harvesting model that can more accurately characterize the available forwarding capacity of energy-constrained IoT relay nodes. By calculating the lower and upper bounds of the candidate relay power segmentation ratio, this patent application can eliminate unschedulable relays that do not meet service quality, energy causality, or self-energy recovery stability constraints before scheduling, reducing invalid searches and congested access. By introducing system throughput, scheduling margin, and schedulable interval width into scheduling utility, this patent application avoids relay selection based solely on maximum throughput or single link quality, thereby improving the success rate and robustness of scheduling under service quality constraints. This patent application evaluates candidate operating points using a schedulable interval sampling method, showing that the algorithm complexity increases linearly with the number of candidate relays and the number of samples, facilitating periodic execution in practical power IoT controllers. Furthermore, this patent application can enhance the reliability of relay side operation while maintaining effective throughput performance, and can characterize the reliability trade-off between the benefits of self-energy recovery and the residual self-interference leakage of information decoding branches. The model is also compatible with EAP-assisted SWIPT networks without self-energy recovery, FD-SWIPT architectures with self-energy recovery without EAP assistance, and traditional FD-SWIPT relay models, and has good scalability and applicability. Attached Figure Description

[0028] Figure 1 This is a flowchart illustrating the overall design of the method of the present invention.

[0029] Figure 2 A schematic diagram of the EAP-assisted self-energy recovery FD-NOMA SWIPT power Internet of Things relay network model.

[0030] Figure 3 This is a schematic diagram of a power-division full-duplex relay structure with self-energy recovery capability.

[0031] Figure 4 This is a diagram showing the impact of the power split ratio on energy-supported transmit power, first-hop decoding capability, and feasible range.

[0032] Figure 5 The graph shows the system throughput performance under different EAP energy budgets.

[0033] Figure 6 This is a performance comparison chart for different combinations of energy sources.

[0034] Figure 7 This is a graph showing the scheduling performance and scalability analysis under varying candidate relay numbers.

[0035] Figure 8 A graph showing the success probability of QoS-constrained scheduling under different source transmit powers.

[0036] Figure 9 The diagram shows the system throughput under different source transmit powers. Detailed Implementation

[0037] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention. However, it will be apparent to those skilled in the art that the invention can be practiced without one or more of these details. In other instances, certain technical features well-known in the art have not been described in order to avoid obscuring the invention.

[0038] This embodiment discloses a joint configuration method for feasibility-aware relay scheduling and power partitioning for green power Internet of Things (IoT). This method is applicable to wireless power IoT networks assisted by energy access points, power cooperative relay communication networks, green low-power power wireless private networks, and scenarios involving simultaneous wireless information and energy transmission. It can be used to determine the schedulability of candidate relay nodes for power communication, configure power partitioning parameters, and select relays.

[0039] One objective of this invention is to establish a three-source energy causal constraint model for relay nodes in a full-duplex non-orthogonal multiple access (NOMA) power Internet of Things (IoT) relay network, comprehensively considering the source node's radio frequency energy, the energy access point's auxiliary energy, and the residual loop's self-interference recovery energy. Based on this, and combining first-hop information decoding constraints, destination node NOMA decoding constraints, serial interference cancellation constraints, and self-energy recovery stability constraints, the feasible power partitioning intervals for each candidate relay are determined. This allows for the accurate elimination of unschedulable relays before scheduling, avoiding the ineffective occupation of power communication resources. Secondly, within the set of schedulable relays, scheduling utility is constructed based on system throughput, scheduling margin, and the width of the feasible power partitioning interval. Based on this, serving relays and their power partitioning ratios are selected, avoiding the problems of congested access and insufficient adaptability to dynamic channel changes caused by selecting relays solely based on link quality or throughput. Through the above methods, the present invention can effectively reduce the invalid search overhead in the relay scheduling of dense power Internet of Things, increase the probability of successful scheduling of candidate relays in one go, reduce congested access under quality of service constraints, and significantly improve the long-term operational reliability of energy-constrained relay networks while ensuring the effective throughput performance of power services.

[0040] This method constructs a three-source energy causal constraint model under a power-segmented wireless information and energy co-transmission architecture, involving source node radio frequency signals, auxiliary signals from substation or line-side energy access points, and residual loop self-interference signals. Before scheduling execution, the feasible power segmentation intervals for each candidate relay are calculated based on this model, thereby pre-eliminating unschedulable relays that cannot meet the constraints of reliable information decoding, forwarding energy requirements, and self-energy recovery stability before scheduling decisions are made. Subsequently, within the set of schedulable relays, a scheduling utility function is constructed by integrating system throughput, scheduling margin, and the width of the feasible power segmentation interval. Based on this function, relay selection and power segmentation are jointly configured, thereby reducing the probability of scheduling congestion and improving the scheduling reliability and resource allocation effectiveness of the energy-constrained power IoT relay network.

[0041] The power Internet of Things (IoT) system includes source nodes (S), energy access points (E), and a set of candidate relays. Remote destination node and near-end destination node Source node S sends a service stream containing two source nodes. and Non-orthogonal multiple access data packets, among which Towards remote destination node , Targeting the near-end destination node Selected candidate relays As an admission and forwarding node for source service packets, it receives signals from the source node and forwards them to the two destination nodes in full-duplex mode.

[0042] The network controller acquires each candidate relay. Channel state information, source node transmit power Energy access point transmit power Or energy budget Q, energy conversion efficiency Non-orthogonal multiple access power allocation coefficient , and target SINR threshold Then, for each candidate relay The following processing is performed: A relay energy causal constraint model is established, consisting of the source node RF signal, the energy access point auxiliary signal, and the residual loop self-interference signal, to obtain the energy-supported relay transmit power; the power split ratio is calculated based on the first-hop information decoding constraint and the self-energy recovery stability constraint. The lower bound; based on the remote destination node. For business flow Decoding constraints, near-end destination node For business flow Serial interference cancellation constraints and near-end target node For its own business flow Decoding constraints are used to calculate the minimum relay transmit power required to meet service quality requirements, and the power split ratio is calculated in conjunction with the energy-supported relay transmit power. The upper bound; the lower bound and the upper bound form a candidate relay. The schedulable power segmentation interval is determined; candidate relays whose schedulable intervals are empty or whose non-orthogonal multiple access decoding conditions are not met are determined as unschedulable and eliminated; the robust scheduling utility is calculated within the schedulable power segmentation interval of the remaining candidate relays, and the candidate relay with the largest robust scheduling utility and its power segmentation ratio are selected as the scheduling result; if there are no schedulable candidate relays, the scheduling blocking status is output.

[0043] The specific implementation details of the method of this invention are disclosed below. This method is mainly used for candidate relay screening, power partitioning ratio configuration, congestion access determination, and robust operating point selection in energy-constrained power Internet of Things (IoT) relay networks. The configuration method proposed in this patent application does not use maximizing instantaneous throughput as the sole criterion. Instead, it first determines whether each candidate relay has a usable power partitioning interval based on service quality constraints, first-hop information decoding constraints, destination node NOMA decoding constraints, energy causality constraints, and self-energy recovery stability constraints. Then, it performs joint configuration within the set of schedulable relays by combining throughput, scheduling margin, and feasible interval width.

[0044] The overall implementation approach of this patent application is as follows: First, establish an EAP-assisted full-duplex NOMASWIPT relay network model, clarifying the signal transmission relationships between source nodes, energy access points, candidate relays, and destination nodes; second, establish a power-segmented relay architecture and a three-source energy harvesting model, uniformly incorporating source node RF energy, EAP-assisted energy, and residual loop self-interference recovery energy into the relay energy causal constraints; then, derive the lower and upper bounds of the power segmentation for candidate relays, forming a schedulable power segmentation interval; finally, perform sampling evaluation within the schedulable interval, outputting the relay with the highest scheduling utility and its power segmentation ratio. This process can eliminate infeasible relays before scheduling, reducing invalid searches and congested access. A detailed flowchart is shown below. Figure 1 As shown.

[0045] (1) Network architecture and channel model design The power Internet of Things (IoT) relay network designed in this patent application includes a source node S, a dedicated energy access point E, and a set of [unclear - likely referring to components or elements]. The set of candidate relays and two destination nodes and The system structure is as follows: Figure 2 As shown. Due to building obstruction, long-distance propagation, or congestion in complex industrial environments, it is assumed that the direct transmission links between the source node S and the two destination nodes are unavailable, and cooperative communication needs to be completed through selected relays. Each candidate relay... Equipped with two co-located antennas to support full-duplex operation, the source node, power access point, and destination node all employ single-antenna half-duplex mode. Indicates the remote destination node. Represent the near-end destination node and satisfy the following conditions: .

[0046] Source node S sends a service stream containing two source ends. and NOMA packets, in which Towards remote destination node , Targeting the near-end destination node Selected relay The relay receives service packets from the source at the first hop and forwards the two service flows from the source at the second hop, without sending its own local service flow. In other words, the relay in this patent application performs the functions of admitting and forwarding service flows from the source, rather than generating new local service flows.

[0047] All wireless channels are modeled as quasi-static Rayleigh fading, and the channel coefficients from node i to node j are denoted as follows. Its average power gain is related to the node distance. and path loss index Related. The additive white Gaussian noise at node j is denoted as... The noise power is ,in Indicates the noise power at the relay point. Indicates the destination node Noise power at the location. The residual loop self-interference channel is denoted as... Used to describe relay Self-interference leakage components that remain after RF isolation, analog cancellation, and digital cancellation during full-duplex transmission and reception.

[0048] (2) Design of source-end NOMA signal and relay forwarding model The source node transmits two service streams simultaneously using power-domain NOMA. The superimposed signal at the source end can be represented as: (1) in and The NOMA power allocation factor satisfies: (2) because For remote destination nodes, a larger power ratio is typically allocated to remote service flows, i.e. > Selected relay After decoding the source NOMA data packet, the same group of source traffic flows is forwarded at the second hop, and the forwarding signal is as follows: (3) This formula indicates that the relay forwards the source service flow. and It is a superimposed signal, rather than a superimposed local information signal.

[0049] Within a transport block, the EAP sends a dedicated energy signal to assist relay energy harvesting. The EAP energy budget is denoted as Q= T; when the transport block duration At that time, Q can be used with It is uniformly used as the energy or power budget for EAP. The EAP signal is a deterministic energy-carrying signal, which can be predicted by the relay or transmitted on an energy waveform orthogonal to the information decoding branch. Therefore, its main contribution enters the energy harvesting model and does not enter the first-hop information decoding SINR as an unknown information interference term.

[0050] (3) Power split relay structure design Selected relay Employing a power-split SWIPT architecture, such as Figure 3 As shown. The radio frequency signal received by the relay is divided into two parts after passing through a power divider: one part enters the information decoding branch, and the other part enters the energy harvesting branch. Let... This represents the proportion of power allocated to the information decoding branch, then 1- This represents the proportion of power allocated to the energy harvesting branch, and satisfies: (4) when When the signal is small, the relay can provide more power for energy harvesting, but the signal in the information decoding branch is weak, which may prevent reliable decoding of source NOMA packets; when When the power split ratio is larger, the information decoding capability is enhanced, but the collectable energy for subsequent forwarding is reduced. Therefore, the choice of power split ratio directly determines whether the candidate relay is schedulable.

[0051] (4) Design of first-hop information decoding constraints For candidate relays The first-hop information decoding SINR can be represented as: (5) in This represents the source node's transmit power. If the target SINR threshold is... Then, the relay reliable decoding source node service packet needs to meet the following requirements: (6) This allows us to obtain the first-hop decoding lower bound for the power split ratio: (7) This lower bound indicates that when the channel gain from the source node to the relay is low, the source transmit power is low, or the target SINR threshold is high, a larger proportion of the received power needs to be allocated to the information decoding branch; otherwise, the candidate relay cannot complete the first hop decoding.

[0052] (5) Design of three-source energy harvesting and energy causal constraints relay The energy harvesting sources consist of three parts: first, the radio frequency signal emitted by the source node S; second, the dedicated energy signal emitted by the EAP; and third, the recoverable radio frequency energy from the residual loop self-interference signal generated during full-duplex operation. Let: (8) in Let be the EAP transmit power. Assume the energy conversion efficiency is . Considering the self-interference recovery of the residual loop, the transmit power supported by the relay energy can be expressed as: (9) The denominator in the formula reflects the feedback gain of self-energy recovery. As long as the feedback loop remains stable, the residual loop self-interference energy can increase the available relay forwarding power; however, if the feedback gain is too large, it may lead to an unbounded energy support power expression. Therefore, this patent application introduces a self-energy recovery stability constraint: (10) From this constraint, a lower bound on stability can be obtained: (11) when When the residual loop self-interference recovery benefit is small, stability constraints are usually not a major limitation; when While self-energy recovery can improve forwarding capacity when increased, excessive loop feedback can compress the feasible power partitioning range. Therefore, this patent application incorporates the benefits of self-energy recovery and system reliability into the scheduling decision through a stability lower bound.

[0053] (6) Design of NOMA decoding constraints for the destination node In the second hop, the remote destination node Decoding business flow and will Considered interference; near-end target node Decode first Then, decoding is performed by eliminating serial interference. To ensure that both remote and near-end service flows meet the target SINR threshold, separate considerations are needed. decoding , decoding as well as decoding The constraints. A necessary condition for NOMA decoding to be feasible is: (12) When this condition is met, the minimum required relay transmit power can be calculated based on the decoding constraints of the three destinations, and the maximum value can be taken as the candidate relay. Minimum forwarding power requirement: (13) in Indicates the remote destination node decoding Required relay transmission power Indicates the near-end destination node decoding And perform the relay transmission power required for SIC pre-decoding. Indicates the near-end destination node decoding The required relay transmit power. Only when the energy supports a transmit power no less than this minimum requirement can a candidate relay meet the quality of service requirements of both destination nodes.

[0054] (7) Construction of feasible power segmentation intervals For each candidate relay First, the lower bound of the power partition is constructed using the first-hop decoding constraint and the self-energy recovery stability constraint: (14) in: (15) (16) Subsequently, bound by energy causality The upper bound of power division is obtained: (17) Therefore, candidate relays The schedulable power partition is as follows: (18) If the feasibility conditions for NOMA are not met, or The upper bound is not positive, or the lower bound of the power division is greater than the upper bound. Then candidate relay If a candidate relay is not schedulable under the current channel conditions, EAP energy budget, and QoS threshold, it should be eliminated before scheduling. This feasible interval determination mechanism differs from the traditional method of selecting a relay first and then adjusting its power. Its core idea is to first determine whether there is an operating point for the candidate relay that satisfies all constraints, and then compare the scheduling utility.

[0055] Figure 4 The impact of power splitting ratio on the feasibility of relay operation is presented. With... Increasing the power output enhances the first-hop decoding capability; however, due to the reduced power entering the energy harvesting branch, the energy-supported relay transmission power decreases. Therefore, the schedulable operating point can only be located within the interval jointly defined by the lower and upper bounds.

[0056] (8) Robust scheduling utility design After obtaining the schedulable power partitioning intervals, the network controller does not directly select the point with the highest instantaneous throughput, but instead sets [the appropriate parameters] within each feasible interval. Each sampling point. For candidate relays. Each sampling point Calculate the corresponding energy-supported transmit power, end-to-end throughput, constraint margin, and feasible interval width, and construct the scheduling utility: (19) in Represents the throughput term. This represents the robustness margin of the run point relative to each constraint boundary. Indicates the width of the feasible interval. The weights are non-negative. The throughput term is used to ensure system transmission efficiency, the robustness margin term is used to avoid the operating point being too close to the QoS or energy boundary, and the feasible interval width term is used to characterize the candidate relay's tolerance to channel fluctuations and energy fluctuations.

[0057] Ultimately, the network controller selects the combination with the highest utility among all schedulable relays and their sampled operating points: (20) And output the selected relay Power split ratio The system outputs the energy support for transmission power and the corresponding system throughput. If all candidate relays are unschedulable, the output shows the blocking access result.

[0058] (9) Scheduling Algorithm The schedulability configuration algorithm is described as follows: Step 1: The network controller obtains the candidate relay set R and the source node transmit power. EAP transmit power or energy budget Energy conversion efficiency NOMA power allocation factor and Target SINR threshold Information on the channel status of each link and information on self-interference channels in the residual loop.

[0059] Step 2: For each candidate relay Calculate the contributions of source nodes and EAP to its energy harvesting. And calculate based on the first-hop decoding constraints .

[0060] Step 3: Calculate the lower bound of stability based on the residual loop self-interference channel and energy conversion efficiency. To determine whether the self-energy recovery feedback meets the stability requirements.

[0061] Step 4: Based on the destination node and NOMA decoding constraints for calculating minimum relay transmit power requirements .

[0062] Step 5: Calculate the upper bound of power segmentation based on energy-supported power constraints. And form a schedulable power partitioning interval. .

[0063] Step 6: If the schedulable interval of a candidate relay is empty, or the NOMA decoding condition is not met, then mark the candidate relay as unschedulable; otherwise, add it to the set of schedulable relays.

[0064] Step 7: Set within each schedulable interval For each sampling point, the throughput, robustness margin, interval width, and scheduling utility are calculated point by point.

[0065] Step 8: Select the relay with the highest scheduling utility and its power division ratio as the final output; if the schedulable set is empty, then the output is blocked.

[0066] Since each candidate relay only needs to undergo closed-range determination once, and within the feasible interval... Therefore, the algorithm complexity is: (Number of samples evaluated) This complexity increases with the number of candidate relays. Number of sampling points Linear growth facilitates periodic execution in practical power IoT controllers.

[0067] To verify the effectiveness of the proposed method, the number of candidate relays is set to a default value. Energy conversion efficiency NOMA power allocation factor Target SINR threshold =1.25, residual loop self-interference power gain Noise power is normalized to 1, path loss exponent Typical distance parameters are set as follows: , , , The EAP energy budget Q varies depending on different simulation scenarios; some comparisons are taken from... .

[0068] The comparative schemes include the maximum and minimum link quality selection method, the throughput-first selection method, the scheme without EAP assistance, the scheme without residual self-interference energy recovery, and the scalability scheme under different numbers of candidate relays. Through these comparisons, the impact of schedulable interval pre-screening, EAP external energy assistance, self-energy recovery, and robust scheduling utility design on system performance can be verified.

[0069] Figure 5 The impact of EAP energy budget on system throughput is demonstrated. As the EAP energy budget increases, the external RF energy that the relay can collect increases, leading to higher relay transmit power and enhanced second-hop transmission capability. Therefore, in the low to medium EAP energy range, system throughput significantly improves with increasing EAP energy budget. However, as EAP energy continues to increase, system performance gradually becomes limited by first-hop information decoding capability, NOMA decoding conditions, and the feasible range of power splitting, and the throughput gain tends to plateau. These results indicate that EAP energy supplementation can effectively alleviate relay energy constraints, but it still needs to be configured in conjunction with power splitting ratios and QoS constraints.

[0070] Figure 6 The performance differences under different energy source combinations are demonstrated. Relying solely on source node RF energy, the available relay forwarding power is limited, especially when the source node transmit power is low or the source-to-relay link is weak, making candidate relays prone to becoming unschedulable due to insufficient energy. Introducing EAP assistance increases the available relay energy, improving both the scheduling success probability and system throughput. Further introducing residual loop self-interference energy recovery further enhances the available forwarding capability on the relay side, providing additional performance benefits when external energy is insufficient or relay energy constraints are tight. These results validate the necessity of unified modeling of source node RF energy, EAP-assisted energy, and residual loop self-interference recovery energy.

[0071] Figure 7The impact of the number of candidate relays on system performance and algorithm operation is demonstrated. As the number of candidate relays K increases, the network controller has more selectable relay nodes, making it easier to find relays that simultaneously satisfy information decoding, energy causality, NOMA decoding, and stability constraints. Therefore, both system throughput and scheduling success probability are improved. Furthermore, since the algorithm in this patent application only performs closed-range decision-making and finite-sampling evaluation for each candidate relay, the algorithm complexity increases linearly with K, avoiding unacceptable computational overhead due to the increase in the number of candidate relays.

[0072] Figure 8 The success probability of QoS-constrained scheduling under different source transmit powers is demonstrated. Compared with methods that select relays solely based on throughput or maximum / minimum link quality, the method in this patent application eliminates infeasible relays before scheduling and selects operating points with greater robustness margins within the schedulable range, thus significantly reducing the probability of congested access. Especially when the source node power is low, energy availability and first-hop decoding constraints are more likely to become performance bottlenecks, making the success scheduling probability advantage of the method in this patent application even more pronounced.

[0073] Figure 9 The paper demonstrates the trend of system throughput variation under different source transmit powers. As the source transmit power increases, both the first-hop decoding capability and the RF energy contribution of the source node are improved, leading to a corresponding increase in system throughput. Because the method in this patent application considers scheduling margin and feasible interval width in addition to throughput, its throughput can remain comparable to that of throughput-first methods, while avoiding the selection of vulnerable operating points near constraint boundaries, thereby improving scheduling reliability.

[0074] In summary, this patent application achieves schedulable configuration in energy-constrained FD-NOMA power IoT relay networks through three-source energy causal constraints, power segmentation feasible interval determination, and robust scheduling utility design. Compared with traditional relay selection methods, this method can identify infeasible relays before scheduling, reducing the probability of invalid searches and congested access; it can utilize EAP auxiliary energy and residual loop self-interference to recover energy and enhance relay forwarding capabilities; and it can improve scheduling reliability under QoS constraints while maintaining comparable throughput. It is suitable for green power IoT, wireless energy transmission, and low-power cooperative communication scenarios.

[0075] The logical ideas behind the methods disclosed in the above embodiments can be implemented, in whole or in part, through software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs.

[0076] When the computer instructions or computer program are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium may be a solid-state drive (SSD).

[0077] It should be noted that the above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A joint configuration method for feasibility-aware relay scheduling and power partitioning for the green power Internet of Things, characterized in that, Includes the following steps: Step 1: Obtain network parameters, including the network controller obtaining channel state information, source node transmit power, energy access point transmit power or energy budget, energy conversion efficiency, non-orthogonal multiple access power allocation coefficient, and target signal-to-interference-plus-noise ratio threshold for each candidate relay. Step 2: For each candidate relay, establish a relay energy causal constraint model composed of the source node radio frequency signal, the energy access point auxiliary signal, and the residual loop self-interference signal to obtain the energy-supported relay transmit power. Step 3: Calculate the lower bound of the power split ratio of the candidate relay based on the first hop information decoding constraint and the self-energy recovery stability constraint; Step 4: Calculate the minimum relay transmit power that satisfies the quality of service requirements based on the non-orthogonal multiple access decoding constraints of the destination node; and calculate the upper bound of the power division ratio in conjunction with the energy-supported relay transmit power. Step 5: The schedulable power segmentation interval of the candidate relay is formed by the lower bound and the upper bound; the candidate relays whose schedulable power segmentation interval is empty or whose non-orthogonal multiple access decoding conditions are not met are determined to be unschedulable and removed, thus obtaining a set of schedulable relays; Step 6: Calculate the robust scheduling utility of each candidate relay in the schedulable power partitioning interval within the schedulable relay set, and select the candidate relay with the largest robust scheduling utility and its power partitioning ratio as the scheduling result.

2. The joint configuration method for feasibility-aware relay scheduling and power partitioning for a green power Internet of Things according to claim 1, characterized in that, In step 1, the power Internet of Things includes a source node, an energy access point, a candidate relay set consisting of multiple candidate relays, a remote destination node, and a near destination node; the source node sends non-orthogonal multiple access data packets containing source-end service flows directed to the remote destination node and source-end service flows directed to the near destination node. The selected candidate relay receives the source node signal in full-duplex mode and forwards it to the two destination nodes.

3. The method for joint configuration of feasibility-aware relay scheduling and power partitioning for green power Internet of Things according to claim 1, characterized in that, In step 2, the candidate relay adopts a power partitioning structure, and the power ratio allocated to the information decoding branch is: The power allocated to the energy harvesting branch is ; The contributions of the source node and the energy access point to the energy collection of the candidate relay are denoted as follows: : in, The source node transmit power, The transmit power of the energy access point. The channel coefficients from the source node to the candidate relay are denoted as . The channel coefficients from the energy access point to the candidate relay; Power-supported relay transmission power The expression is: in, For energy conversion efficiency, This represents the residual loop self-interference channel coefficient.

4. The method for joint configuration of feasibility-aware relay scheduling and power partitioning for green power Internet of Things according to claim 1, characterized in that, In step 3, the lower bound of the power division ratio is denoted as... The lower realm Find the lower bound of the first jump decoding and stability lower bound The maximum value in is expressed as: in, , ; The channel coefficients from the source node to the candidate relay; The residual loop self-interference channel coefficient; This refers to the source node's transmit power. The target SINR threshold; This indicates the noise power at the relay point.

5. The method for joint configuration of feasibility-aware relay scheduling and power partitioning for green power Internet of Things according to claim 1, characterized in that, In step 4, the minimum relay transmit power is determined jointly based on the decoding constraints of the remote destination node on the service flow, the serial interference cancellation constraints of the near destination node on the service flow, and the decoding constraints of the near destination node on its own service flow. Its expression is: in, Indicates the remote destination node decoding Required relay transmission power, Indicates the near-end destination node decoding And perform the relay transmission power required for SIC pre-decoding. Indicates the near-end destination node decoding Required relay transmission power.

6. The joint configuration method for feasibility-aware relay scheduling and power partitioning for green power Internet of Things according to claim 5, characterized in that, In step 4, the upper bound of the power segmentation ratio obtained from the energy causality constraint is denoted as... : in, Minimum relay transmission power, For energy conversion efficiency, The contribution of source nodes and energy access points to energy collection of candidate relays. This represents the residual loop self-interference channel coefficient.

7. The method for joint configuration of feasibility-aware relay scheduling and power partitioning for green power Internet of Things according to claim 1, characterized in that, In step 5, the schedulable power segmentation interval is denoted as It satisfies: in, This is the lower bound of the power division ratio. This is the upper bound of the power division ratio. The proportion of power allocated to the information decoding branch.

8. A joint configuration method for feasibility-aware relay scheduling and power partitioning for a green power Internet of Things according to claim 7, characterized in that, In step 5, the necessary condition for the non-orthogonal multiple access decoding condition to be satisfied is: in, and The power allocation coefficient for non-orthogonal multiple access; The target signal-to-interference-plus-noise ratio (SIN / N) threshold; If the non-orthogonal multiple access feasibility conditions are not met, or the upper bound is not met... Not positive, or the lower bound stated Greater than the upper bound If so, the candidate relay is determined to be unschedulable and removed before scheduling.

9. A joint configuration method for feasibility-aware relay scheduling and power partitioning for a green power Internet of Things according to claim 1, characterized in that, Step 6 specifically includes: Within the schedulable power partition interval of each candidate relay in the set of schedulable relays, set There are 10 sampling points; for each sampling point, the robust scheduling utility is constructed by combining the system throughput, scheduling margin, and the width of the schedulable power segmentation interval, as shown in the following expression: in, Represents the throughput term; This indicates the robustness margin of the run point relative to each constraint boundary; Indicates the width of the feasible interval; , , Non-negative weights; The network controller selects the combination with the highest utility from all schedulable relays and their sampled operating points, and outputs the selected relays and their power split ratio; if there are no schedulable candidate relays, it outputs the scheduling blocking status.

10. An electronic device, characterized in that, include: Processor and memory storing computer program instructions; When the processor executes the computer program instructions, it implements a joint configuration method for feasibility-aware relay scheduling and power partitioning for green power Internet of Things as described in any one of claims 1 to 9.

Images