Dynamic scheduling method for node energy load balance in wireless Internet of Things

Abstract

The invention discloses a dynamic scheduling method for node energy load balance in a wireless internet of things. The method comprises the following steps: S1, collecting energy state information of each node; s2, energy information is analyzed, a correlation model of network topology and node energy states is established, and key nodes and communication links are identified; s3, in combination with a topological phase change theory and an inverse game learning algorithm, predicting topological change, and generating a node task allocation and energy consumption strategy; s4, constructing a dynamic scheduling model, and adjusting task distribution in real time; s5, optimizing node scheduling by applying an anti-entropy driving algorithm; s6, executing a dynamic scheduling strategy, and adjusting a communication path and task distribution; s7, monitoring a scheduling execution condition, and optimizing a strategy; and S8, generating an energy load balance optimization report, and continuously optimizing the network. According to the invention, the problems of shortened network life and reduced reliability caused by unbalanced energy consumption and dynamic environment change in the prior art are effectively solved, and the overall performance and stability of the wireless Internet of Things system are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of wireless Internet of Things, and in particular to a dynamic scheduling method for node energy load balancing in a wireless Internet of Things. Background Art

[0002] In modern wireless IoT systems, node energy management and load balancing are key factors affecting network life and performance. With the rapid increase in the number of IoT devices and the increasing complexity of application scenarios, traditional energy management methods can no longer effectively deal with problems such as uneven node energy consumption, dynamic environmental changes, and frequent adjustments to topology structures. These problems directly lead to reduced network reliability and shortened overall life.

[0003] First, traditional node energy management and load balancing methods usually rely on predefined static strategies. These strategies are designed with certain fixed network conditions in mind, but are difficult to adapt to dynamic changes in actual operation. For example, the energy consumption rate of a node may change with changes in communication frequency, data transmission volume, and distance between nodes, but static strategies cannot be adjusted in real time, causing the energy of some nodes to be quickly exhausted, thus causing the entire network to fail. In addition, changes in the quality of communication links and network topology also make it impossible for static strategies to ensure balanced load distribution, further exacerbating the imbalance of energy consumption.

[0004] Secondly, although the existing dynamic scheduling algorithms have solved some energy management problems to some extent, they usually only focus on the optimization of a specific factor, such as data transmission rate or communication link quality, while ignoring the global energy load balance of the network. This single-dimensional optimization method cannot fully consider the mutual influence between nodes, especially when the network topology changes frequently, it is more likely to lead to a decline in the overall performance of the network. In addition, the existing scheduling algorithms usually do not consider the application of topological phase transition theory and game theory, lack the ability to accurately predict the node strategy behavior in a complex network environment, and it is difficult to achieve the optimal distribution of the global energy load.

[0005] Third, traditional distributed systems and network protocol designs show certain limitations when facing highly dynamic and uncertain network environments. Centralized scheduling methods can easily lead to overload or failure of central nodes, which in turn makes the scheduling of the entire network ineffective. Although distributed scheduling methods have certain flexibility, they lack sufficient collaborative optimization mechanisms when dealing with complex interdependencies between nodes, resulting in low efficiency in overall energy load distribution. In addition, existing energy management methods generally lack the use of anti-entropy driving mechanisms, making it difficult to achieve effective control of the entropy growth rate, and thus unable to maintain the energy load balance and stability of the network in a dynamic environment.

[0006] In summary, the existing wireless Internet of Things node energy management and load balancing methods have the following main defects when facing a dynamic and complex network environment: first, they lack the ability to adapt to dynamic changes in network topology; second, they cannot effectively balance energy loads globally; and third, the scheduling strategy lacks robustness in highly dynamic and uncertain environments. Therefore, how to provide a dynamic scheduling method for node energy load balancing in wireless Internet of Things is an urgent problem that technicians in this field need to solve. Summary of the invention

[0007] One object of the present invention is to propose a dynamic scheduling method for node energy load balancing in a wireless Internet of Things. The present invention makes full use of topological phase transition theory, inverse game learning algorithm and anti-entropy driving mechanism, and describes in detail the strategy of achieving node energy balance through dynamic scheduling, which has the advantages of strong adaptability, high energy utilization efficiency and strong network stability.

[0008] A dynamic scheduling method for node energy load balancing in a wireless Internet of Things according to an embodiment of the present invention comprises the following steps:

[0009] S1. Collect the energy status information of each node in the wireless Internet of Things, including the remaining energy of the node, energy consumption rate, communication link quality, network topology and distance between nodes;

[0010] S2. Analyze the collected energy status information, establish a correlation model between network topology and node energy status, and identify key nodes and communication links that affect energy balance;

[0011] S3. Combining topological phase transition theory with inverse game learning algorithm, predict the possible topological changes in the network, and generate node task allocation and energy consumption strategies by deducing the strategic behavior of each node in the topological change to achieve global balance of energy load.

[0012] S4. Based on the generated node task allocation strategy, a dynamic scheduling model is constructed to dynamically adjust the task allocation of each node to cope with real-time network changes and node energy status;

[0013] S5. Apply the anti-entropy driven dynamic scheduling algorithm to prioritize identifying and processing nodes and paths that may cause energy imbalance by calculating the overall entropy growth rate of the network;

[0014] S6. Execute the generated dynamic scheduling strategy, dynamically adjust the task allocation and communication path of each node according to the results of topological phase change and inverse game deduction, and anti-entropy driven calculation;

[0015] S7, monitor the execution of dynamic scheduling strategies in real time, evaluate node energy consumption and network topology changes through monitoring results, and continuously optimize scheduling strategies;

[0016] S8. Based on the execution evaluation results of the scheduling strategy, regularly generate energy load balancing optimization reports, analyze the energy utilization efficiency of the entire network, and continuously optimize the entire network.

[0017] Optionally, S3 includes the following steps:

[0018] S31. Based on the association model between network topology and node energy state established in S2, define the network topology graph G(V,E), where V represents the node set, E represents the communication link between nodes, and the energy state of each node i is represented as E i (t), the communication link quality is represented by Q ij (t), the distance between nodes is denoted as d ij ;

[0019] S32. Apply the topological phase transition theory to analyze the critical state of the network and define a topological phase transition driving matrix M(t). The element m of M(t) is ij (t) represents the interaction strength between nodes i and j at time t, and the local phase change coefficient θ of each node is calculated i (t):

[0020]

[0021] Among them, θ i (t) is the local phase change coefficient of node i at time t, which measures the topological phase change potential around node i; V i is the set of neighbor nodes of node i, representing other nodes that communicate directly with node i; |V i | is the size of the neighbor node set, indicating the number of nodes directly connected to node i; m ij (t) is the interaction strength between node i and node j at time t; E i (t) and E j (t) is the residual energy of node i and node j at time t; ∈ is a tuning parameter used to control the impact of energy difference on the phase change coefficient. It is usually a small positive number and determines the sensitivity of energy difference to the local phase change coefficient. is the attenuation factor, which is used to describe the effect of the energy difference between two nodes on the node interaction. The greater the energy difference, the more obvious the attenuation.

[0022] S33, when the global phase change coefficient of the network Θ(t) = ∑ i∈V θ i (t) exceeds the critical value Θ c When , it indicates that the network is about to enter a topological phase transition state; at this time, the inverse game learning algorithm is applied to optimize the node strategy; the strategy optimization goal of node i is to minimize the node energy consumption and the change of the local phase transition coefficient:

[0023]

[0024] in, is the optimal strategy combination of node i, which represents the energy consumption and task allocation strategy adopted by the node under the current network state; is the strategy S of node i i The impact on the change of the local phase transition coefficient of neighbor node j indicates how the strategy selection changes the phase transition potential of the neighbor node; E i (t) is the residual energy of node i at time t; λ is the adjustment coefficient, which balances the relationship between energy consumption and phase change coefficient; is the rate of change of energy consumption, which is used to control the sensitivity of node strategy to energy consumption;

[0025] S34, after obtaining the optimal strategy of the node Finally, the energy load distribution equation is constructed to describe the dynamic changes of node task allocation and energy consumption:

[0026]

[0027] Among them, E i (t+1) is the remaining energy of node i at time t+1, indicating that the strategy The energy state of the next node; α is the energy consumption coefficient, which indicates the influence of task allocation on energy consumption; Q ij (t) is the quality of the communication link between node i and node j at time t; d ij is the physical distance between node i and node j; β is the adjustment coefficient, which controls the influence of phase change coefficient on energy consumption; is the ratio of phase change coefficients, describing the relative impact of the phase change potential of nodes i and j on energy consumption;

[0028] S35. Optimal strategy combination based on inverse game learning Calculate the global energy load L(t) of the network under various strategy combinations:

[0029]

[0030] Where L(t) is the energy load of the entire network at time t, which measures the uniformity of node energy distribution and topological stability in the network; E i (t+1) is the residual energy of node i at time t+1; is the average energy of all network nodes at time t+1; δ is the adjustment coefficient, which balances the energy distribution balance and the stability of topological phase transition; θ i (t) and θ j(t) is the local phase change coefficient of nodes i and j at time t, indicating their potential role in the phase change;

[0031] S36. Combine according to the optimal strategy Dynamically adjust node task allocation and communication paths;

[0032] S37, monitor the network status in real time, and if a new phase transition critical point or abnormal energy distribution is detected, reapply the inverse game learning algorithm to optimize the strategy;

[0033] S38: Input the finally optimized task allocation and energy consumption strategy into the subsequent scheduling execution.

[0034] Optionally, S4 includes the following steps:

[0035] S41. Node task allocation strategy based on generation Define the task load function T of each node i (t), describes the task allocation of node i at time t, where α i (t) is the load adjustment coefficient of node i at time t, which indicates the task load capacity of the node in the current state; is the optimal strategy combination of node i;

[0036] S42, considering the real-time changes of node energy status and network topology, by adjusting the coefficient α i (t) for the task load function T i (t) Dynamic adjustment; adjustment coefficient α i (t) is calculated as:

[0037]

[0038] Among them, E i (t) is the residual energy of node i at time t; V i is the set of neighbor nodes of node i; Q ij (t) is the quality of the communication link between node i and node j at time t;

[0039] S43. Construct a dynamic scheduling model to describe the dynamic allocation process of tasks among nodes. The dynamic scheduling model D(t) is expressed as:

[0040]

[0041] Where D(t) is the dynamic scheduling model at time t, which represents the distribution of tasks in the entire network; T i (t) is the task load function of node i at time t; |V i| is the number of neighbor nodes of node i;

[0042] S44, combining the dynamic scheduling model and the node task load function, real-time adjustment of the node task allocation and communication path, and dynamically optimizing the scheduling strategy P according to the node energy consumption changes and communication link quality fluctuations. * :

[0043]

[0044] Among them, P * (t) is the optimal scheduling strategy combination at time t, which represents the task and communication path allocation strategy under a given network state; γ is the adjustment coefficient, which balances the relationship between task allocation and communication link quality;

[0045] S45. In actual applications, nodes adjust task allocation and communication paths in real time based on a combination of dynamic scheduling models and optimal scheduling strategies to cope with changes in energy consumption and topology in the network and maintain network stability and energy load balance;

[0046] S46, real-time monitoring of node task allocation and energy consumption, and adjusting the adjustment coefficient α according to the monitoring results i (t) and the scheduling strategy P * (t) Perform adaptive adjustment.

[0047] Optionally, S5 includes the following steps:

[0048] S51. Define the energy state matrix E(t) of the network, where each element E i (t) represents the residual energy of node i at time t. Based on the matrix E(t), the initial energy entropy H(t) of the entire network is calculated:

[0049]

[0050] Where H(t) is the energy entropy of the network at time t, which is used to quantify the uniformity of energy distribution in the network; E i (t) is the residual energy of node i at time t; V is the set of nodes in the entire network;

[0051] S52, based on the initial energy entropy H(t), defining an entropy growth rate ΔH(t) to characterize the change trend of the network in energy distribution;

[0052] S53, apply the anti-entropy driving mechanism to suppress the excessive growth of entropy growth rate ΔH(t) by adjusting the task allocation and energy consumption strategy of the nodes; the optimization strategy of node i The goal is to minimize the contribution to the global entropy growth rate:

[0053]

[0054] in, is the optimal scheduling strategy for node i, which represents the energy consumption and task allocation of the node in the current state; ΔH i (t) is the contribution of node i to the overall entropy growth rate of the network; is the impact of node i’s strategy on the entropy growth rate of neighbor node j; Q ij (t) is the quality of the communication link between node i and node j at time t; ζ adjustment coefficient, the influence of the control node strategy on the entropy growth rate; is the energy difference influence factor, which indicates the influence of energy difference on the interaction;

[0055] S54. Based on the above optimization strategy, an anti-entropy driven global scheduling model is constructed to describe the energy entropy change trend under node task allocation and communication path adjustment:

[0056]

[0057] Where ΔH total (t+1) is the global entropy growth rate at time t+1, which describes the change in the overall energy entropy of the network; is the optimal strategy combination of node i; It is the global entropy increase suppression factor, which is used to slow down the excessive growth of entropy growth rate;

[0058] S55. In practical applications, nodes adjust task allocation and communication paths in real time according to the anti-entropy driven scheduling strategy to cope with changes in energy consumption and network topology, and maintain the overall stability and energy load balance of the network;

[0059] S56, by real-time monitoring of the global entropy increase model of the network, identify the nodes and paths that may cause network instability, and adaptively adjust the scheduling strategy combination P according to the monitoring data * (t);

[0060] S57, Dynamic Optimization Scheduling Strategy Combination P * (t), without increasing the global entropy increase, the energy distribution balance of the network is improved, while suppressing the excessive growth of local energy consumption.

[0061] Optionally, the S6 includes the following steps:

[0062] S61. Based on the deduction results of topological phase transition and inverse game learning algorithm, as well as the anti-entropy driven scheduling strategy, define the node task allocation function F i (t), represents the task allocation strategy of node i at time t; the task allocation function is defined as:

[0063]

[0064] Among them, F i (t) is the task allocation function of node i at time t; is the optimal strategy combination of node i; Z i (t) is the normalization factor of node i; ΔH i (t) is the contribution of node i to the overall entropy growth rate of the network; H(t) is the energy entropy of the network at time t;

[0065] S62, combining the results of topological phase change prediction and inverse game learning algorithm, updating the communication path selection function P of each node i (t), function P i (t) describes the communication path chosen by node i at time t:

[0066]

[0067] Among them, P i (t) is the communication path selection function of node i at time t; Q ij (t) is the quality of the communication link between node i and node j; d ij is the physical distance between node i and node j; ΔH j (t) Contribution of node j to the overall entropy growth rate of the network;

[0068] S63, node-based task allocation function F i (t) and the communication path selection function P i (t), construct the global energy load model E total (t), describes the real-time distribution of energy in the network;

[0069] S64, executing the generated global energy load model E total (t), adjust the operation of each node in real time according to the node's task allocation and communication path selection;

[0070] S65. During the actual operation, the energy consumption and communication link quality of the node are monitored in real time, and the normalization factor Z is adjusted based on these monitoring data. i (t) and scheduling strategy combination

[0071] S66. Adaptively adjust the global energy load model and the task allocation and communication path of each node according to the real-time monitoring results to cope with the dynamically changing network environment and maintain the overall performance and energy load balance of the network;

[0072] S67. Regularly evaluate the energy load model and actual scheduling effect of the entire network, and generate optimization reports to provide data support and optimization direction for subsequent dynamic scheduling strategies.

[0073] The beneficial effects of the present invention are:

[0074] (1) The present invention proposes a dynamic scheduling method for balancing node energy load in a wireless Internet of Things. By combining the topological phase transition theory with the inverse game learning algorithm, the network topology changes are dynamically predicted, and the optimal task allocation strategy is generated based on the node energy state. This effectively achieves the global balance of energy load and significantly improves the stability and life of the network, especially when the node energy consumption is uneven and the network environment changes frequently.

[0075] (2) The present invention introduces an anti-entropy driven dynamic scheduling algorithm, which calculates the entropy growth rate of the entire network, prioritizes the identification and processing of nodes and paths that may cause energy imbalance, and thus optimizes the node energy scheduling strategy. Through the anti-entropy driving mechanism, the rapid growth of global energy entropy is suppressed in real time, ensuring that the balance of energy distribution is maintained in a complex dynamic network environment, significantly improving the energy utilization efficiency and robustness of the network.

[0076] (3) The present invention monitors the execution of dynamic scheduling strategies in real time and continuously optimizes the scheduling strategies based on the monitoring results, thereby ensuring the dynamic adjustment of task allocation and communication paths of each node, adapting to different network topology changes and energy consumption conditions, and enhancing the system's adaptive capabilities, thereby effectively maintaining the performance and stability of the network in a changing IoT environment.

[0077] (4) The present invention constructs a global energy load model, and through the synergy of the dynamic scheduling model and the anti-entropy driving algorithm, dynamically adjusts the task allocation and communication path of each node, and realizes the real-time balance of energy load in the network. It not only effectively avoids network failure caused by node energy exhaustion, but also improves the overall energy utilization efficiency of the network and the collaborative working ability between nodes through multi-level dynamic scheduling strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0079] Figure 1 A flow chart of a dynamic scheduling method for node energy load balancing in a wireless Internet of Things proposed by the present invention;

[0080] Figure 2A flowchart for combining the topological phase transition theory with the inverse game learning algorithm to predict network topology changes and generate a task allocation strategy in the present invention;

[0081] Figure 3 Schematic diagram of applying the anti-entropy driven dynamic scheduling algorithm to optimize node energy scheduling in the present invention. DETAILED DESCRIPTION

[0082] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0083] refer to Figure 1-3 , a dynamic scheduling method for node energy load balancing in a wireless Internet of Things, comprising the following steps:

[0084] S1. Collect the energy status information of each node in the wireless Internet of Things, including the remaining energy of the node, energy consumption rate, communication link quality, network topology and distance between nodes;

[0085] S2. Analyze the collected energy status information, establish a correlation model between network topology and node energy status, and identify key nodes and communication links that affect energy balance;

[0086] S3. Combining topological phase transition theory with inverse game learning algorithm, predict the possible topological changes in the network, and generate node task allocation and energy consumption strategies by deducing the strategic behavior of each node in the topological change to achieve global balance of energy load.

[0087] S4. Based on the generated node task allocation strategy, a dynamic scheduling model is constructed to dynamically adjust the task allocation of each node to cope with real-time network changes and node energy status;

[0088] S5. Apply the anti-entropy driven dynamic scheduling algorithm to prioritize identifying and processing nodes and paths that may cause energy imbalance by calculating the overall entropy growth rate of the network;

[0089] S6. Execute the generated dynamic scheduling strategy, dynamically adjust the task allocation and communication path of each node according to the results of topological phase change and inverse game deduction, and anti-entropy driven calculation;

[0090] S7, monitor the execution of dynamic scheduling strategies in real time, evaluate node energy consumption and network topology changes through monitoring results, and continuously optimize scheduling strategies;

[0091] S8. Based on the execution evaluation results of the scheduling strategy, regularly generate energy load balancing optimization reports, analyze the energy utilization efficiency of the entire network, and continuously optimize the entire network.

[0092] In this implementation, S3 includes the following steps:

[0093] S31. Based on the association model between network topology and node energy state established in S2, define the network topology graph G(V,E), where V represents the node set, E represents the communication link between nodes, and the energy state of each node i is represented as E i (t), the communication link quality is represented by Q ij (t), the distance between nodes is denoted as d ij ;

[0094] S32. Apply the topological phase transition theory to analyze the critical state of the network and define a topological phase transition driving matrix M(t). The element m of M(t) is ij (t) represents the interaction strength between nodes i and j at time t, and the local phase change coefficient θ of each node is calculated i (t):

[0095]

[0096] Among them, θ i (t) is the local phase change coefficient of node i at time t, which measures the topological phase change potential around node i; V i is the set of neighbor nodes of node i, representing other nodes that communicate directly with node i; |V i | is the size of the neighbor node set, indicating the number of nodes directly connected to node i; m ij (t) is the interaction strength between node i and node j at time t; E i (t) and E j (t) is the residual energy of node i and node j at time t; ∈ is a tuning parameter used to control the impact of energy difference on the phase change coefficient. It is usually a small positive number and determines the sensitivity of energy difference to the local phase change coefficient. is the attenuation factor, which is used to describe the effect of the energy difference between two nodes on the node interaction. The greater the energy difference, the more obvious the attenuation.

[0097] S33, when the global phase change coefficient of the network Θ(t) = ∑ i∈V θ i (t) exceeds the critical value Θ c When , it indicates that the network is about to enter a topological phase transition state; at this time, the inverse game learning algorithm is applied to optimize the node strategy; the strategy optimization goal of node i is to minimize the node energy consumption and the change of the local phase transition coefficient:

[0098]

[0099] in, is the optimal strategy combination of node i, which represents the energy consumption and task allocation strategy adopted by the node under the current network state; is the strategy S of node i i The impact on the change of the local phase transition coefficient of neighbor node j indicates how the strategy selection changes the phase transition potential of the neighbor node; E i (t) is the residual energy of node i at time t; λ is the adjustment coefficient, which balances the relationship between energy consumption and phase change coefficient; is the rate of change of energy consumption, which is used to control the sensitivity of node strategy to energy consumption;

[0100] S34, after obtaining the optimal strategy of the node Finally, the energy load distribution equation is constructed to describe the dynamic changes of node task allocation and energy consumption:

[0101]

[0102] Among them, E i (t+1) is the remaining energy of node i at time t+1, indicating that the strategy The energy state of the next node; α is the energy consumption coefficient, which indicates the influence of task allocation on energy consumption; Q ij (t is the quality of the communication link between node i and node j at time t; d ij is the physical distance between node i and node j; β is the adjustment coefficient, which controls the influence of phase change coefficient on energy consumption; is the ratio of phase change coefficients, describing the relative impact of the phase change potential of nodes i and j on energy consumption;

[0103] S35. Optimal strategy combination based on inverse game learning Calculate the global energy load L(t) of the network under various strategy combinations:

[0104]

[0105] Where L(t) is the energy load of the entire network at time t, which measures the uniformity of node energy distribution and topological stability in the network; E i (t+1) is the residual energy of node i at time t+1; is the average energy of all network nodes at time t+1; δ is the adjustment coefficient, which balances the energy distribution balance and the stability of topological phase transition; θ i (t) and θ j (t) is the local phase change coefficient of nodes i and j at time t, indicating their potential role in the phase change;

[0106] S36. Combine according to the optimal strategy Dynamically adjust node task allocation and communication paths;

[0107] S37, monitor the network status in real time, and if a new phase transition critical point or abnormal energy distribution is detected, reapply the inverse game learning algorithm to optimize the strategy;

[0108] S38: Input the finally optimized task allocation and energy consumption strategy into the subsequent scheduling execution.

[0109] In this implementation, S4 includes the following steps:

[0110] S41. Node task allocation strategy based on generation Define the task load function T of each node i (t), describes the task allocation of node i at time t, where α i (t) is the load adjustment coefficient of node i at time t, which indicates the task load capacity of the node in the current state; is the optimal strategy combination of node i;

[0111] S42, considering the real-time changes of node energy status and network topology, by adjusting the coefficient α i (t) for the task load function T i (t) Dynamic adjustment; adjustment coefficient α i (t) is calculated as:

[0112]

[0113] Among them, E i (t) is the residual energy of node i at time t; V i is the set of neighbor nodes of node i; Q ij (t) is the quality of the communication link between node i and node j at time t;

[0114] S43. Construct a dynamic scheduling model to describe the dynamic allocation process of tasks among nodes. The dynamic scheduling model D(t) is expressed as:

[0115]

[0116] Where D(t) is the dynamic scheduling model at time t, which represents the distribution of tasks in the entire network; T i (t) is the task load function of node i at time t; |V i | is the number of neighbor nodes of node i;

[0117] S44, combining the dynamic scheduling model and the node task load function, real-time adjustment of the node task allocation and communication path, and dynamically optimizing the scheduling strategy P according to the node energy consumption changes and communication link quality fluctuations. * :

[0118]

[0119] Among them, P * (t) is the optimal scheduling strategy combination at time t, which represents the task and communication path allocation strategy under a given network state; γ is the adjustment coefficient, which balances the relationship between task allocation and communication link quality;

[0120] S45. In actual applications, nodes adjust task allocation and communication paths in real time based on a combination of dynamic scheduling models and optimal scheduling strategies to cope with changes in energy consumption and topology in the network and maintain network stability and energy load balance;

[0121] S46, real-time monitoring of node task allocation and energy consumption, and adjusting the adjustment coefficient α according to the monitoring results i (t) and the scheduling strategy P * (t) Perform adaptive adjustment.

[0122] In this implementation, S5 includes the following steps:

[0123] S51. Define the energy state matrix E(t) of the network, where each element E i 9t) represents the residual energy of node i at time t. Based on the matrix E9t), the initial energy entropy H(t) of the entire network is calculated:

[0124]

[0125] Where H(t) is the energy entropy of the network at time t, which is used to quantify the uniformity of energy distribution in the network; E i (t) is the residual energy of node i at time t; V is the set of nodes in the entire network;

[0126] S52, based on the initial energy entropy H(t), defining an entropy growth rate ΔH(t) to characterize the change trend of the network in energy distribution;

[0127] S53, apply the anti-entropy driving mechanism to suppress the excessive growth of entropy growth rate ΔH(t) by adjusting the task allocation and energy consumption strategy of the nodes; the optimization strategy of node i The goal is to minimize the contribution to the global entropy growth rate:

[0128]

[0129] in, is the optimal scheduling strategy for node i, which represents the energy consumption and task allocation of the node in the current state; ΔH i (t) is the contribution of node i to the overall entropy growth rate of the network; is the impact of node i’s strategy on the entropy growth rate of neighbor node j; Q ij (t) is the quality of the communication link between node i and node j at time t; ζ adjustment coefficient, the influence of the control node strategy on the entropy growth rate; is the energy difference influence factor, which indicates the influence of energy difference on the interaction;

[0130] S54. Based on the above optimization strategy, an anti-entropy driven global scheduling model is constructed to describe the energy entropy change trend under node task allocation and communication path adjustment:

[0131]

[0132] Where ΔH total (t+1) is the global entropy growth rate at time t+1, which describes the change in the overall energy entropy of the network; is the optimal strategy combination of node i; It is the global entropy increase suppression factor, which is used to slow down the excessive growth of entropy growth rate;

[0133] S55. In practical applications, nodes adjust task allocation and communication paths in real time according to the anti-entropy driven scheduling strategy to cope with changes in energy consumption and network topology, and maintain the overall stability and energy load balance of the network;

[0134] S56, by real-time monitoring of the global entropy increase model of the network, identify the nodes and paths that may cause network instability, and adaptively adjust the scheduling strategy combination P according to the monitoring data * (t);

[0135] S57, Dynamic Optimization Scheduling Strategy Combination P * (t), without increasing the global entropy increase, the energy distribution balance of the network is improved, while suppressing the excessive growth of local energy consumption.

[0136] In this implementation, S6 includes the following steps:

[0137] S61. Based on the deduction results of topological phase transition and inverse game learning algorithm, as well as the anti-entropy driven scheduling strategy, define the node task allocation function F i (t), represents the task allocation strategy of node i at time t; the task allocation function is defined as:

[0138]

[0139] Among them, Fi (t) is the task allocation function of node i at time t; is the optimal strategy combination of node i; Z i (t) is the normalization factor of node i; ΔH i (t) is the contribution of node i to the overall entropy growth rate of the network; H(t is the energy entropy of the network at time t;

[0140] S62, combining the results of topological phase change prediction and inverse game learning algorithm, updating the communication path selection function P of each node i (t), function P i (t) describes the communication path chosen by node i at time t:

[0141]

[0142] Among them, P i (t) is the communication path selection function of node i at time t; Q ij (t) is the quality of the communication link between node i and node j; d ij is the physical distance between node i and node j; ΔH j (t) Contribution of node j to the overall entropy growth rate of the network;

[0143] S63, node-based task allocation function F i (t) and the communication path selection function P i (t), construct the global energy load model E total (t), describes the real-time distribution of energy in the network;

[0144] S64, executing the generated global energy load model E total (t), adjust the operation of each node in real time according to the node's task allocation and communication path selection;

[0145] S65. During the actual operation, the energy consumption and communication link quality of the node are monitored in real time, and the normalization factor Z is adjusted based on these monitoring data. i (t) and scheduling strategy combination

[0146] S66. Adaptively adjust the global energy load model and the task allocation and communication path of each node according to the real-time monitoring results to cope with the dynamically changing network environment and maintain the overall performance and energy load balance of the network;

[0147] S67. Regularly evaluate the energy load model and actual scheduling effect of the entire network, and generate optimization reports to provide data support and optimization direction for subsequent dynamic scheduling strategies.

[0148] Embodiment 1:

[0149] In a large-scale intelligent industrial park, thousands of wireless IoT nodes are deployed to monitor and control the status of production equipment, environmental parameters, and safety assurance systems. Due to the complexity and dynamic changes of the industrial environment, the energy consumption of each node varies greatly, and the quality of the communication link changes at any time. Previous energy management methods cannot effectively adapt to this complex environment, causing some nodes to quickly run out of energy in a short period of time, seriously affecting the stability of the entire industrial park network and the normal operation of equipment.

[0150] In the wireless Internet of Things system of this intelligent industrial park, the dynamic scheduling method based on topological phase change theory, inverse game learning algorithm and anti-entropy driving mechanism proposed in the present invention is adopted to solve the problems caused by uneven node energy consumption and dynamic environmental changes.

[0151] First, the system collects the energy status information of each node in real time, including the remaining energy of the node, energy consumption rate, communication link quality, network topology, and distance between nodes. By analyzing this information, the system establishes a correlation model between network topology and node energy status, and identifies key nodes and communication links that consume energy quickly and may have a significant impact on network stability.

[0152] Then, the system combined the topological phase change theory with the inverse game learning algorithm to predict the possible changes in network topology in different time periods. By deducing the strategic behavior of each node under different topological structures, the optimal node task allocation and energy consumption strategy was generated. On this basis, a dynamic scheduling model was constructed that can respond to various dynamic changes in the industrial park in real time, including changes in the energy state of nodes, fluctuations in the quality of communication links, etc., to ensure that the task load of each node is reasonably distributed.

[0153] In addition, the system also applies an anti-entropy driven dynamic scheduling algorithm. By calculating the overall entropy growth rate of the network, it prioritizes the identification and processing of nodes and paths that may cause energy imbalance, thereby optimizing the node energy scheduling strategy. In the entire industrial park, the system can monitor the task allocation and communication path of each node in real time, and continuously optimize the scheduling strategy based on the monitoring results to ensure that the network's energy load is always maintained in a relatively balanced state.

[0154] Concrete data demonstrating beneficial effects

[0155] In Area A, Area B and Area C of the smart industrial park, a week-long monitoring and data collection was carried out to evaluate the effect of the present invention in different environments. The test results show that after adopting the dynamic scheduling method of the present invention, the node energy consumption is more balanced, and the network stability and communication quality are significantly improved.

[0156] Table 1 Data table

[0157]

[0158]

[0159] Through these data, we can more intuitively see the effectiveness of the dynamic scheduling method of the present invention in different environments, which significantly improves the energy utilization efficiency, network stability and communication quality of the system, and successfully solves the problems of uneven energy consumption and dynamic environmental changes that traditional methods cannot cope with. These data also prove the adaptability and superiority of the present invention in complex industrial Internet of Things scenarios.

[0160] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A dynamic scheduling method for node energy load balancing in a wireless Internet of Things, characterized in that: The steps include: S1. Collect the energy status information of each node in the wireless Internet of Things, including the remaining energy of the node, energy consumption rate, communication link quality, network topology and distance between nodes; S2. Analyze the collected energy status information, establish a correlation model between network topology and node energy status, and identify key nodes and communication links that affect energy balance; S3. Combining topological phase transition theory with inverse game learning algorithm, predict the possible topological changes in the network, and generate node task allocation and energy consumption strategies by deducing the strategic behavior of each node in the topological change to achieve global balance of energy load. S4. Based on the generated node task allocation strategy, a dynamic scheduling model is constructed to dynamically adjust the task allocation of each node to cope with real-time network changes and node energy status; S5. Apply the anti-entropy driven dynamic scheduling algorithm to prioritize identifying and processing nodes and paths that may cause energy imbalance by calculating the overall entropy growth rate of the network; S6. Execute the generated dynamic scheduling strategy, dynamically adjust the task allocation and communication path of each node according to the results of topological phase change and inverse game deduction, and anti-entropy driven calculation; S7, monitor the execution of dynamic scheduling strategies in real time, evaluate node energy consumption and network topology changes through monitoring results, and continuously optimize scheduling strategies; S8. Based on the execution evaluation results of the scheduling strategy, regularly generate energy load balancing optimization reports, analyze the energy utilization efficiency of the entire network, and continuously optimize the entire network.

2. A dynamic scheduling method for node energy load balancing in a wireless Internet of Things according to claim 1, characterized in that: The S3 comprises the following steps: S31. Based on the association model between network topology and node energy state established in S2, define the network topology graph G(V,E), where V represents the node set, E represents the communication link between nodes, and the energy state of each node i is represented as E i (t), the communication link quality is represented by Q ij (t), the distance between nodes is denoted as d ij ; S32. Apply the topological phase transition theory to analyze the critical state of the network and define a topological phase transition driving matrix M(t). The element m of M(t) is ij (t) represents the interaction strength between nodes i and j at time t, and the local phase change coefficient θ of each node is calculated i (t): Among them, θ i (t) is the local phase change coefficient of node i at time t, which measures the topological phase change potential around node i; V i is the set of neighbor nodes of node i, representing other nodes that communicate directly with node i; |V i | is the size of the neighbor node set, indicating the number of nodes directly connected to node i; m ij (t) is the interaction strength between node i and node j at time t; E i (t) and E j (t) is the residual energy of node i and node j at time t; ∈ is a tuning parameter used to control the impact of energy difference on the phase change coefficient. It is usually a small positive number and determines the sensitivity of energy difference to the local phase change coefficient. is the attenuation factor, which is used to describe the effect of the energy difference between two nodes on the node interaction. The greater the energy difference, the more obvious the attenuation. S33, when the global phase change coefficient of the network Θ(t) = ∑ i∈V θ i (t) exceeds the critical value Θ c When , it indicates that the network is about to enter a topological phase transition state; at this time, the inverse game learning algorithm is applied to optimize the node strategy; the strategy optimization goal of node i is to minimize the node energy consumption and the change of the local phase transition coefficient: in, is the optimal strategy combination of node i, which represents the energy consumption and task allocation strategy adopted by the node under the current network state; is the strategy S of node i i The impact on the change of the local phase transition coefficient of neighbor node j indicates how the strategy selection changes the phase transition potential of the neighbor node; E i (t) is the residual energy of node i at time t; λ is the adjustment coefficient, which balances the relationship between energy consumption and phase change coefficient; is the rate of change of energy consumption, which is used to control the sensitivity of node strategy to energy consumption; S34, after obtaining the optimal strategy of the node Finally, the energy load distribution equation is constructed to describe the dynamic changes of node task allocation and energy consumption: Among them, E i (t+1) is the remaining energy of node i at time t+1, indicating that the strategy The energy state of the next node; α is the energy consumption coefficient, which indicates the influence of task allocation on energy consumption; Q ij (t) is the quality of the communication link between node i and node j at time t; d ij is the physical distance between node i and node j; β is the adjustment coefficient, which controls the influence of phase change coefficient on energy consumption; is the ratio of phase change coefficients, describing the relative impact of the phase change potential of nodes i and j on energy consumption; S35. Optimal strategy combination based on inverse game learning Calculate the global energy load L(t) of the network under various strategy combinations: Where L(t) is the energy load of the entire network at time t, which measures the uniformity of node energy distribution and topological stability in the network; E i (t+1) is the residual energy of node i at time t+1; is the average energy of all network nodes at time t+1; δ is the adjustment coefficient, which balances the energy distribution balance and the stability of topological phase transition; θ i (t) and θ j (t) is the local phase change coefficient of nodes i and j at time t, indicating their potential role in the phase change; S36. Combine according to the optimal strategy Dynamically adjust node task allocation and communication paths; S37, monitor the network status in real time, and if a new phase transition critical point or abnormal energy distribution is detected, reapply the inverse game learning algorithm to optimize the strategy; S38: Input the finally optimized task allocation and energy consumption strategy into the subsequent scheduling execution.

3. A dynamic scheduling method for node energy load balancing in a wireless Internet of Things according to claim 1, characterized in that: The S4 comprises the following steps: S41. Node task allocation strategy based on generation Define the task load function T of each node i (t), describes the task allocation of node i at time t, where α i (t) is the load adjustment coefficient of node i at time t, which indicates the task load capacity of the node in the current state; is the optimal strategy combination of node i; S42, considering the real-time changes of node energy status and network topology, by adjusting the coefficient α i (t) for the task load function T i (t) Dynamic adjustment; adjustment coefficient α i (t) is calculated as: Among them, E i (t) is the residual energy of node i at time t; V i is the set of neighbor nodes of node i; Q ij (t) is the quality of the communication link between node i and node j at time t; S43. Construct a dynamic scheduling model to describe the dynamic allocation process of tasks among nodes. The dynamic scheduling model D(t) is expressed as: Where D(t) is the dynamic scheduling model at time t, which represents the distribution of tasks in the entire network; T i (t) is the task load function of node i at time t; |V i | is the number of neighbor nodes of node i; S44, combining the dynamic scheduling model and the node task load function, real-time adjustment of the node task allocation and communication path, and dynamically optimizing the scheduling strategy P according to the node energy consumption changes and communication link quality fluctuations. * : Among them, P * (t) is the optimal scheduling strategy combination at time t, which represents the task and communication path allocation strategy under a given network state; γ is the adjustment coefficient, which balances the relationship between task allocation and communication link quality; S45. In actual applications, nodes adjust task allocation and communication paths in real time based on a combination of dynamic scheduling models and optimal scheduling strategies to cope with changes in energy consumption and topology in the network and maintain network stability and energy load balance; S46, real-time monitoring of node task allocation and energy consumption, and adjusting the adjustment coefficient α according to the monitoring results i (t) and the scheduling strategy P * (t) Perform adaptive adjustment.

4. A dynamic scheduling method for node energy load balancing in a wireless Internet of Things according to claim 1, characterized in that: The S5 comprises the following steps: S51. Define the energy state matrix E(t) of the network, where each element E i (t) represents the residual energy of node i at time t. Based on the matrix E(t), the initial energy entropy H(t) of the entire network is calculated: Where H(t) is the energy entropy of the network at time t, which is used to quantify the uniformity of energy distribution in the network; E i (t) is the residual energy of node i at time t; V is the set of nodes in the entire network; S52, based on the initial energy entropy H(t), defining an entropy growth rate ΔH(t) to characterize the change trend of the network in energy distribution; S53, apply the anti-entropy driving mechanism to suppress the excessive growth of entropy growth rate ΔH(t) by adjusting the task allocation and energy consumption strategy of the nodes; the optimization strategy of node i The goal is to minimize the contribution to the global entropy growth rate: in, is the optimal scheduling strategy for node i, which represents the energy consumption and task allocation of the node in the current state; ΔH i (t) is the contribution of node i to the overall entropy growth rate of the network; is the impact of node i’s strategy on the entropy growth rate of neighbor node j; Q ij (t) is the quality of the communication link between node i and node j at time t; ζ adjustment coefficient, the influence of the control node strategy on the entropy growth rate; is the energy difference influence factor, which indicates the influence of energy difference on the interaction; S54. Based on the above optimization strategy, an anti-entropy driven global scheduling model is constructed to describe the energy entropy change trend under node task allocation and communication path adjustment: Where ΔH total (t+1) is the global entropy growth rate at time t+1, which describes the change in the overall energy entropy of the network; is the optimal strategy combination of node i; It is the global entropy increase suppression factor, which is used to slow down the excessive growth of entropy growth rate; S55. In practical applications, nodes adjust task allocation and communication paths in real time according to the anti-entropy driven scheduling strategy to cope with changes in energy consumption and network topology, and maintain the overall stability and energy load balance of the network; S56, by real-time monitoring of the global entropy increase model of the network, identify the nodes and paths that may cause network instability, and adaptively adjust the scheduling strategy combination P according to the monitoring data * (t); S57, Dynamic Optimization Scheduling Strategy Combination P * (t), without increasing the global entropy increase, the energy distribution balance of the network is improved, while suppressing the excessive growth of local energy consumption.

5. A dynamic scheduling method for node energy load balancing in a wireless Internet of Things according to claim 1, characterized in that: The S6 comprises the following steps: S61. Based on the deduction results of topological phase transition and inverse game learning algorithm, as well as the anti-entropy driven scheduling strategy, define the node task allocation function F i (t), represents the task allocation strategy of node i at time t; the task allocation function is defined as: Among them, F i (t) is the task allocation function of node i at time t; is the optimal strategy combination of node i; Z i (t) is the normalization factor of node i; ΔH i (t) is the contribution of node i to the overall entropy growth rate of the network; H(t) is the energy entropy of the network at time t; S62, combining the results of topological phase change prediction and inverse game learning algorithm, updating the communication path selection function P of each node i (t), function P i (t) describes the communication path chosen by node i at time t: Among them, P i (t) is the communication path selection function of node i at time t; Q ij (t) is the quality of the communication link between node i and node j; d ij is the physical distance between node i and node j; ΔH j (t) Contribution of node j to the overall entropy growth rate of the network; S63, based on the node task allocation function F i (t) and the communication path selection function P i (t), construct the global energy load model E total (t), describes the real-time distribution of energy in the network; S64, executing the generated global energy load model E total (t), adjust the operation of each node in real time according to the node's task allocation and communication path selection; S65. During the actual operation, the energy consumption and communication link quality of the node are monitored in real time, and the normalization factor Z is adjusted based on these monitoring data. i (t) and scheduling strategy combination S66. Adaptively adjust the global energy load model and the task allocation and communication path of each node according to the real-time monitoring results to cope with the dynamically changing network environment and maintain the overall performance and energy load balance of the network; S67. Regularly evaluate the energy load model and actual scheduling effect of the entire network, and generate optimization reports to provide data support and optimization direction for subsequent dynamic scheduling strategies.

Images