Dynamic network-oriented passive sensing information fusion method and system

Abstract

The invention provides a passive sensing information fusion method and system for a dynamic network. Aiming at the problem that the fusion performance is reduced based on dynamic change of environment energy collection and network topology in passive sensing monitoring, remote wireless connection is provided for passive Internet of Things edge equipment by utilizing an environment backscatter communication technology, and energy is collected; a Voronoi diagram is utilized to analyze the difference of the importance degree of a monitoring area, a dynamic network topology structure is considered, and the wake-up frequency of a passive sensing node is calculated based on the forwarding cost. Therefore, the energy collection efficiency of the environment backscattering device, the importance degree difference of the monitoring area and the dynamically changed network topology structure are comprehensively considered, the fusion weight of the sensor nodes is dynamically calculated, the information fusion performance is improved, and the passive sensing monitoring precision is improved.

Description

Technical Field

[0001] The present invention relates to the fields of passive Internet of Things, artificial intelligence, and in particular to a passive sensor information fusion method and system for dynamic networks. Background Art

[0002] The uncertainty of various physical environments causes the node fusion weight to change dynamically. Specifically, in many practical applications, the measurement data perceived by edge devices is real-time data (data stream), and the importance of different monitoring areas varies greatly, and the network topology changes dynamically, which causes the quality of sensor-perceived data to change dynamically over time, causing the fusion weight to change dynamically.

[0003] According to the literature: JK Devineni and HSDhillon. Non-Coherent Detection and Bit Error Rate for an Ambient Backscatter Link in Time-Selective Fading [J]. IEEE Transactions on Communications, 69 (1): 602-618, 2021, in order to broaden the application scope of passive Internet of Things and achieve continuous and stable operation, the use of environmental backscatter communication technology (such as Figure 1 As shown in the figure, it is a backscatter device internal structure diagram), which provides remote wireless connection for passive IoT edge devices, reduces the communication energy consumption of passive IoT edge devices, and can collect energy from the environment. However, in actual passive sensing monitoring, the dynamically changing network topology and the differences in the importance of the monitoring area are not considered. To address this problem, we can consider the differences in the importance of the monitoring area (such as Figure 2 As shown) and the dynamics of the network topology, the fusion weights of the passive sensor nodes are dynamically calculated.

[0004] From the perspective of a dynamically changing network environment, according to the literature: M. Cattani, A. Loukas, et al. Staffetta: Smart Duty-Cycling for Opportunistic Data Collection [C]. ACM Conference on Embedded Network Sensor Systems (ACM SenSys'16), 56-69, 2016, the wake-up frequency and weight of each edge device can be dynamically adjusted according to the current forwarding cost of the data packet, so that the nodes close to the Sink can be woken up more frequently. The activity gradient is used to automatically control the data packet in the correct direction: the probability of finding an active neighbor is highest in the direction of the Sink. This improves the performance of the opportunistic protocol in data transmission and fusion, while maintaining the ability to adapt to dynamically changing network topologies. However, it does not consider the impact of environmental backscattering and differences in the importance of the monitoring area on the fusion performance.

[0005] Therefore, how to dynamically update the weights of passive sensor nodes in information fusion in a dynamic and complex environment becomes an important scientific issue. Passive IoT has some unique characteristics, the most notable of which is the network topology based on environmental energy harvesting and dynamic changes. Therefore, how to perform efficient information fusion based on the particularity of passive sensor monitoring is a crucial issue. Summary of the invention

[0006] Aiming at the problem of reduced fusion performance based on environmental energy collection and dynamic changes in network topology in passive sensor monitoring, the environmental backscatter communication technology is used to provide remote wireless connection and energy collection for passive IoT edge devices; the Voronoi diagram is used to analyze the differences in the importance of the monitoring area, and the dynamic network topology is considered to calculate the wake-up frequency of the passive sensor node based on the forwarding cost. Therefore, the energy collection efficiency of the environmental backscatter device, the differences in the importance of the monitoring area, and the dynamically changing network topology are comprehensively considered, and the fusion weight of the sensor node is dynamically calculated to improve the information fusion performance, thereby improving the accuracy of passive sensor monitoring.

[0007] The technical solution of the present invention is: a passive sensing information fusion method for dynamic networks, comprising the following steps:

[0008] Step 1: Use environmental backscatter communication technology to provide long-range wireless connectivity for passive IoT edge devices and calculate the energy that can be collected by the backscatter device in each node;

[0009] Step 2, according to the differences in the importance of the monitoring areas, the importance of the nodes is calculated using the Voronoi diagram;

[0010] Step 3, according to the dynamic network topology, the wake-up frequency of the passive sensor node is calculated based on the forwarding cost;

[0011] Step 4, calculating the wake-up probability of the passive sensor node by integrating the dynamic network topology, environmental backscatter energy collection and the differences in the importance of the monitoring area;

[0012] Step 5: Wake up the passive sensor nodes according to the wake-up probability calculated in step 4, and use it as the information fusion weight to perform information fusion, dynamically select appropriate nodes, improve information fusion accuracy and maximize network life.

[0013] Furthermore, the energy that can be collected by the backscattering device in step 1 is obtained by the following calculation formula:

[0014]

[0015] Among them, Λ n represents the energy that can be collected by the backscattering device in the nth node, l (0≤l≤1) is the energy collection efficiency of the scattering device, α n (0≤α n ≤1) is the power reflection coefficient of the backscattering device in the nth node; φ represents the matrix of subcarrier power allocation, φ n,k Indicates the power allocated to the kth subcarrier in the nth time slot; φ r,k represents the power allocated to the kth subcarrier in the rth time slot; η represents the backscattering time part vector, η n and η r They represent the backscatter device in the nth node and the time portion of the rth time slot, respectively, and the value range is [0,1]. n,k represents the forward channel response from the full-duplex access point to the nth backscatter device; M is the number of subcarriers sending OFDM signals; N represents the number of backscatter devices; k represents the kth subcarrier, r represents the rth time slot; n represents the backscatter device in the nth node.

[0016] Furthermore, the importance of the node in step 2 is obtained by the following calculation formula:

[0017]

[0018] in, represents the importance of the sensor node, N is the number of all passive sensor nodes in the monitoring area, θ is the adaptive parameter, Γ n (t) is the monitoring efficiency θ in the monitoring area ijThe sum of (t) represents the importance of the monitoring area, Ω is the total number of the maximum values ​​of each grid contained in π, π is a two-dimensional monitoring area, and the area π is divided into x×y grids. i and j are the row and column numbers of the grids.

[0019] Furthermore, Ω and The calculation formula is as follows:

[0020]

[0021]

[0022] Where π = {γ ij}, a ij Indicates the importance of each grid.

[0023] Furthermore, the calculation formula of the wake-up frequency in step 3 is as follows:

[0024]

[0025] in, refers to the node wake-up frequency, σ n (t) is the forwarding delay of node n at time t, is the maximum duty cycle.

[0026] Furthermore, the formula for calculating the wake-up probability of the passive sensor node in step 6 is as follows:

[0027]

[0028] in, Refers to the node's wake-up frequency, represents the importance of the sensor node, Λ n represents the energy that can be collected by the backscatter device in the nth node, is the wake-up probability, α+β+γ=1, α, β, γ are weight parameters, and their values ​​are all in [0,1]; For n Normalized result; Z i (t) represents the battery status of edge device i at time t, s is the state-aware energy consumption, tr For transmission energy consumption, represents the indicator function, O t Indicates the state, O t = n' means that the edge device senses and sends a new status update, O t =q indicates that the edge device senses the state update of the new transmission and the last transmission.

[0029] Furthermore, the battery status of edge device i at time t+1 is expressed as follows:

[0030]

[0031]

[0032] Among them, i (t) represents the random energy collected and stored in the rechargeable battery of the edge device i at the end of each time t, For state-aware energy consumption, To transmit energy consumption; the collected energy is first stored in a rechargeable battery with limited capacity, Z max represents the capacity of the battery. At each time slot t, the edge device can take three possible actions: it can sense and send a new status update, namely O t =n', keep idle, that is, O t = m or resend the last sent status update, i.e. O t =q; if event C remains unchanged, the indicator function is equal to 1, otherwise it is 0; Formula (7) indicates that when the battery is full, energy is harvested and the battery overflows, while Formula (8) stipulates that if the battery does not have enough energy, no sensing or transmission operations are allowed.

[0033] Furthermore, in step 5, a suitable node is dynamically selected, that is, the node corresponding to the maximum wake-up probability calculated in step 4 is selected.

[0034] The present invention also provides a passive sensor information fusion system for dynamic networks, including a processor and a memory, the memory is used to store program instructions, and the processor is used to call the stored instructions in the memory to execute a passive sensor information fusion method for dynamic networks as described in the above technical solution.

[0035] In summary, the present invention proposes to use environmental backscatter communication technology to provide remote wireless connection, comprehensively consider whether the system can obtain sufficient energy from the environment, the difference in the importance of the monitoring area and the dynamically changing network topology, improve the performance of sensor network information fusion, and thus improve the accuracy of passive sensor monitoring. Its main innovation is that this method is based on environmental backscatter theory, combined with energy acquisition efficiency, the difference in the importance of the monitoring area and the dynamically changing network topology, calculates fusion weights, dynamically selects appropriate passive sensor nodes, and uses this to perform information fusion, which can truly reflect the dynamic network environment, while improving the information fusion performance of the passive Internet of Things and maximizing energy utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1Schematic diagram of the internal structure of the backscattering device in an embodiment of the present invention.

[0037] Figure 2 This is a difference diagram of the importance of monitoring areas in an embodiment of the present invention.

[0038] Figure 3 The figure is a flow chart of an embodiment of the present invention. DETAILED DESCRIPTION

[0039] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings.

[0040] like Figure 3 As shown, a passive sensing information fusion method for a dynamic network provided by an embodiment of the present invention includes the following steps:

[0041] Step 1: Leverage ambient backscatter communication technology to provide long-range wireless connectivity for passive IoT edge devices and calculate the energy that can be harvested by the backscatter device in each node.

[0042]

[0043] Among them, Λ n represents the energy that can be collected by the backscattering device in the nth node, e (0≤e≤1) is the energy collection efficiency of the scattering device, and α n (0≤α n ≤1) is the power reflection coefficient of the nth backscattering device. φ represents the matrix of subcarrier power allocation, φ n,k Indicates the power allocated to the kth subcarrier in the nth time slot; φ r,k represents the power allocated to the kth subcarrier in the rth time slot. η represents the backscattering time part vector, η n and η r They represent the time portion of the nth backscatter device and the rth time slot, respectively, and the value range is [0,1]. n,k represents the forward channel response from the full-duplex access point to the nth backscatter device. M is the number of subcarriers transmitting the OFDM signal; N is the number of backscatter devices; k is the kth subcarrier, r is the rth time slot; and n is the nth backscatter device.

[0044] Step 2: Considering the differences in the importance of the monitoring areas, the importance of the nodes is calculated using the Voronoi diagram.

[0045]

[0046] in, represents the importance of the sensor node, N is the number of all passive sensor nodes in the monitoring area, θ is the adaptive parameter, Γ n (t) is the monitoring efficiency in the monitoring area The sum of represents the importance of the monitoring area. Ω is the total number of the maximum values ​​of each grid contained in π, calculated as follows:

[0047]

[0048]

[0049] Where π is a two-dimensional monitoring area, and the area π is divided into a grid of x×y, that is, π={r ij}, i and j are the row and column numbers of the grid, a ij Indicates the importance of each grid;

[0050] Step 3: Considering the dynamic network topology, the wake-up frequency of passive sensor nodes is calculated based on the forwarding cost.

[0051]

[0052] in, refers to the node wake-up frequency, σ n (t) is the forwarding delay of node n at time t, is the maximum duty cycle.

[0053] Step 4: Considering the dynamic network topology, environmental backscatter energy collection, and differences in the importance of the monitoring area, the wake-up probability of the passive sensor node (edge ​​device) is calculated.

[0054]

[0055] in, is the wake-up probability, α+β+γ=1, α, β, γ are weight parameters, and their values ​​are all in [0,1]. For n After normalization, if the battery does not have sufficient energy, the wake-up probability is 0.

[0056] First, the battery model is quantitatively analyzed using the sensing energy consumption, transmission energy consumption, and battery capacity. The battery status of edge device i at time t+1 can be expressed as follows:

[0057]

[0058]

[0059] Among them, i(t) represents the random energy collected and stored in the rechargeable battery of the edge device i at the end of each time slot t. Energy consumption for state awareness. The collected energy is first stored in a rechargeable battery with limited capacity. max represents the capacity of the battery. At each time slot t, the edge device can take three possible actions: it can sense and send a new status update (O t =n), keep idle (O t = m) or resend the last sent status update (O t =q). If event C remains unchanged, the indicator function is equal to 1, otherwise it is 0. Formula (7) indicates that when the battery is full, energy is harvested and the battery overflows, while formula (8) stipulates that if the battery does not have enough energy, no sensing or transmission operations are allowed.

[0060] Step 5: Wake up the passive sensor nodes according to the wake-up probability calculated in step 4, and use it as the information fusion weight for information fusion, dynamically select the appropriate node (that is, select the node corresponding to the maximum wake-up probability calculated in step 4), improve the information fusion accuracy and maximize the network life.

[0061] Information fusion, also known as data fusion, sensor information fusion or multi-sensor information fusion, is an information processing process that associates, relates and integrates data and information obtained from single and multiple information sources to obtain accurate location and identity estimates, as well as a comprehensive and timely assessment of the situation and threat and their importance. This process is a continuous refinement process of its estimation, evaluation and evaluation of the need for additional information sources, and it is also a process of continuous self-correction of the information processing process to obtain improved results. Information fusion technology can be summarized as: the information processing process that uses computer technology to automatically analyze and comprehensively process the observation information of several sensors obtained in time series under certain criteria to complete the required decision-making and estimation tasks. According to this definition, the multi-sensor system is the hardware foundation of information fusion, multi-source information is the processing object of information fusion, and coordinated optimization and comprehensive processing are the core of information fusion.

[0062] On the other hand, an embodiment of the present invention also provides a passive sensor information fusion system for dynamic networks, including a processor and a memory, the memory is used to store program instructions, and the processor is used to call the stored instructions in the memory to execute a passive sensor information fusion method for dynamic networks as described in the above technical solution.

[0063] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A passive sensor information fusion method for dynamic networks, characterized in that: The steps include: Step 1: Use environmental backscatter communication technology to provide long-range wireless connectivity for passive IoT edge devices and calculate the energy that can be collected by the backscatter device in each node; Step 2, according to the differences in the importance of the monitoring areas, the importance of the nodes is calculated using the Voronoi diagram; Step 3, according to the dynamic network topology, the wake-up frequency of the passive sensor node is calculated based on the forwarding cost; Step 4, calculating the wake-up probability of the passive sensor node by integrating the dynamic network topology, environmental backscatter energy collection and the differences in the importance of the monitoring area; Step 5: Wake up the passive sensor nodes according to the wake-up probability calculated in step 4, and use it as the information fusion weight to perform information fusion, dynamically select appropriate nodes, improve information fusion accuracy and maximize network life.

2. The method for passive sensor information fusion for dynamic networks as claimed in claim 1, characterized in that: The energy that the backscatter device can collect in step 1 is obtained by the following calculation formula: Among them, Λ n represents the energy that can be collected by the backscattering device in the nth node, l (0≤l≤1) is the energy collection efficiency of the scattering device, α n (0≤α n ≤1) is the power reflection coefficient of the backscattering device in the nth node; φ represents the matrix of subcarrier power allocation, φ n,k Indicates the power allocated to the kth subcarrier in the nth time slot; φ r,k represents the power allocated to the kth subcarrier in the rth time slot; η represents the backscattering time part vector, η n and η r They represent the backscatter device in the nth node and the time portion of the rth time slot, respectively, and the value range is [0,1]. n,k represents the forward channel response from the full-duplex access point to the nth backscatter device; M is the number of subcarriers sending OFDM signals; N represents the number of backscatter devices; k represents the kth subcarrier, r represents the rth time slot; n represents the backscatter device in the nth node.

3. The passive sensing information fusion method for dynamic networks as claimed in claim 1, characterized in that: The importance of the node in step 2 is obtained by the following calculation formula: in, represents the importance of the sensor node, N is the number of all passive sensor nodes in the monitoring area, θ is the adaptive parameter, Γ n (t) is the monitoring efficiency in the monitoring area The sum of represents the importance of the monitoring area, Ω is the total number of the maximum values ​​of each grid contained in π, π is a two-dimensional monitoring area, and the area π is divided into x×y grids. i and j are the row and column numbers of the grid.

4. The passive sensing information fusion method for dynamic networks as claimed in claim 3, characterized in that: Ω and The calculation formula is as follows: Where π = {γ ij }, a ij Indicates the importance of each grid.

5. The method for passive sensor information fusion for dynamic networks as claimed in claim 1, characterized in that: The calculation formula for the wake-up frequency in step 3 is as follows: in, refers to the node wake-up frequency, σ n (t) is the forwarding delay of node n at time t, is the maximum duty cycle.

6. The method for passive sensor information fusion for dynamic networks as claimed in claim 1, characterized in that: The formula for calculating the wake-up probability of the passive sensor node in step 6 is as follows: in, Refers to the node's wake-up frequency, represents the importance of the sensor node, Λ n represents the energy that can be collected by the backscatter device in the nth node, is the wake-up probability, α+β+γ=1, α, β, γ are weight parameters, and their values ​​are all in [0,1]; For n Normalized result; Z i (t) represents the battery status of edge device i at time t, s is the state-aware energy consumption, tr For transmission energy consumption, represents the indicator function, O t Indicates the state, O t = n' means that the edge device senses and sends a new status update, O t =q indicates that the edge device senses the state update of the new transmission and the last transmission.

7. The method for passive sensor information fusion for dynamic networks as claimed in claim 6, characterized in that: The battery status of edge device i at time t+1 is expressed as follows: Among them, i (t) represents the random energy collected and stored in the rechargeable battery of the edge device i at the end of each time t, For state-aware energy consumption, To transmit energy consumption; the collected energy is first stored in a rechargeable battery with limited capacity, Z max represents the capacity of the battery. At each time slot t, the edge device can take three possible actions: it can sense and send a new status update, namely O t =n', keep idle, that is, O t = m or resend the last sent status update, i.e. O t =q; if event C remains unchanged, the indicator function is equal to 1, otherwise it is 0; Formula (7) indicates that when the battery is full, energy is harvested and the battery overflows, while Formula (8) stipulates that if the battery does not have enough energy, no sensing or transmission operations are allowed.

8. The method for passive sensor information fusion for dynamic networks as claimed in claim 1, characterized in that: In step 5, a suitable node is dynamically selected, that is, the node corresponding to the maximum wake-up probability calculated in step 4 is selected.

9. A passive sensor information fusion system for dynamic networks, characterized by: It comprises a processor and a memory, the memory is used to store program instructions, and the processor is used to call the stored instructions in the memory to execute a passive sensing information fusion method for dynamic networks as described in any one of claims 1 to 8.

Images