Wireless sensor network and stochastic resonance enhancement method thereof

Abstract

The invention relates to a wireless sensor network and a stochastic resonance enhancement method of the wireless sensor network, and belongs to the technical field of wireless communication. According to the invention, a binary asymmetric channel is added between the last relay node of each transmission link and a fusion center to introduce stochastic resonance, and information transmitted to the fusion center by the last relay node is flipped, that is, all received information 0 is flipped to 1, and all information 1 is flipped to 0. Therefore, the accuracy of the received value of the fusion center can be improved. According to the invention, the resonance effect between the noise and the signal is utilized to find a balance between the signal and the noise, so that the detectability of the signal is improved, the transmission reliability of the wireless sensor network is further improved, and the wireless sensor network has higher robustness.

Description

Technical Field

[0001] The invention relates to a wireless sensor network and a random resonance enhancement method of the wireless sensor network, belonging to the technical field of wireless communications. Background Art

[0002] With the rapid development of "digitalization, networking, and intelligence", unmanned factories have gradually become an important trend in modern manufacturing. In unmanned factories, the intelligence and automation levels of equipment manufacturing equipment are constantly improving, and obtaining equipment parameter information through sensors has become an important means to achieve precise manufacturing and intelligent decision-making. In the equipment manufacturing process of unmanned factories, sensors are distributed in various equipment and production lines to collect parameters such as temperature, pressure, vibration, and speed in real time. These sensor nodes transmit the collected data to the central control system through wireless sensor networks for data analysis and intelligent decision-making. However, the equipment manufacturing environment is complex and changeable, wireless signals are easily interfered, and the reliability and accuracy of data transmission face huge challenges. Ensuring rapid perception of information and reliable data transmission in unmanned factories has become a key technology in the field of equipment manufacturing.

[0003] In order to improve the data transmission quality of wireless sensor networks, multi-hop multi-relay technology is widely used in data transmission between sensor nodes. The current wireless sensor network includes local sensors, relay nodes and fusion centers. The local sensors forward the information they receive to the fusion center through the transmission link formed by multiple relay nodes. The fusion center makes judgments and measurements, and gradually forwards the data through multiple relay nodes, thereby extending the transmission distance, enhancing the network coverage, and improving the overall transmission efficiency. However, in the case of severe channel noise and interference, especially when the network is attacked, the multi-hop multi-relay transmission method will produce a large number of errors in the received values ​​of the fusion center, making it difficult to ensure the reliability and stability of data transmission, resulting in data loss and transmission delay, affecting system performance.

[0004] In order to improve the overall transmission performance, the fusion center needs to accurately obtain the instantaneous CSI (channel state information) of each relay BSC (binary symmetric channel) in each link during the implementation of the system's optimal decision metric. It is difficult to accurately obtain the instantaneous CSI of each relay BSC in each link. Even if accurate instantaneous CSI information can be obtained, it will consume a lot of energy, which will affect the life and transmission efficiency of the entire wireless sensor network. Summary of the invention

[0005] The purpose of the present invention is to provide a wireless sensor network and a random resonance enhancement method for the wireless sensor network, so as to solve the problem that the data transmission reliability of the current wireless sensor network is relatively low when encountering a network attack.

[0006] In order to solve the above technical problems, the present invention provides a method for enhancing stochastic resonance of a wireless sensor network, which method comprises the following steps:

[0007] 1) When the channel quality of the equivalent channel between each transmission link from the information source to the last relay node in the wireless sensor network is poor, a binary asymmetric channel is added between the last relay node of each transmission link and the fusion center to introduce stochastic resonance, and the information transmitted from the last relay node to the fusion center is flipped;

[0008] 2) The fusion center fuses the received values ​​obtained from each transmission link to obtain the system fusion decision metric value, and obtains the final detection result based on the system fusion decision metric value.

[0009] Furthermore, in step 2), the decision method used to obtain the final detection result according to the system fusion decision metric value is:

[0010]

[0011] Wherein, Λ represents the system fusion decision metric value, M is the decision threshold, and M is equal to 0.

[0012] Furthermore, when the channel quality of the equivalent channel between each transmission link from the source to the last relay node is poor, the fusion center obtains the system fusion decision metric value as follows:

[0013]

[0014] Where I is the number of local sensors and also the number of transmission links, y i is the information received by the fusion center from the i-th local sensor through the corresponding transmission link.

[0015] Furthermore, when the channel quality of the equivalent channel from the source to the last relay node is good, the fusion center obtains the system fusion decision metric value as follows:

[0016]

[0017] Where I is the number of local sensors, y i is the information received by the fusion center from the i-th local sensor through the corresponding transmission link.

[0018] Furthermore, the poor channel quality of the equivalent channel between each transmission link from the information source to the last relay node means that the wireless sensor network is attacked by an external network, resulting in the probability of error transfer of the equivalent channel from the information source to the last relay node being greater than 0.5, while the probability of correct transmission is less than 0.5.

[0019] Furthermore, the information transmitted on each transmission link is obtained by hard decision making of the data generated by the information source by the local sensor in the wireless sensor network using the maximum likelihood criterion.

[0020] Furthermore, the fusion center in step 2) uses the maximum likelihood criterion to obtain the system fusion decision metric value.

[0021] The present invention also provides a wireless sensor network, including local sensors, relay nodes and a fusion center. The local sensors forward the information they receive to the fusion center through a transmission link formed by multiple relay nodes. The wireless sensor network adopts the above-mentioned random resonance enhancement method of the wireless sensor network.

[0022] The beneficial effects of the present invention are as follows: as an improved invention, when the channel quality of the equivalent channel between each transmission link of the wireless sensor network from the source to the last relay node is poor, a large number of errors will be generated in the fusion center receiving value, so the present invention adds a binary asymmetric channel between the last relay node of each transmission link and the fusion center to introduce random resonance, and flips the information transmitted to the fusion center by the last relay node, that is, the received information 0 is all flipped to 1 and the information 1 is all flipped to 0, so that the accuracy of the fusion center receiving value can be improved. It can be seen that the present invention finds a balance between the signal and the noise by utilizing the resonance effect between the noise and the signal, improves the detectability of the signal, and then improves the reliability of the wireless sensor network transmission, and makes the wireless sensor network more robust. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a flow chart of a random resonance enhancement method for a wireless sensor network of the present invention;

[0024] Figure 2 is a schematic diagram of the relay channel structure from the sensor to the fusion center in an embodiment of the present invention;

[0025] Figure 3 is a schematic diagram of an equivalent channel structure from a source to a fusion center in an embodiment of the present invention;

[0026] Figure 4 is a schematic diagram of an equivalent channel structure from a source of stochastic resonance to a fusion center in an embodiment of the present invention;

[0027] Figure 5 is a flowchart of a wireless sensor network in an embodiment of the present invention;

[0028] Figure 6 Schematic diagram of SER performance comparison between the case where the local channel is ideal and the case where the number of routes changes when stochastic resonance enhancement is not used and when stochastic resonance enhancement is used in an embodiment of the present invention

[0029] Figure 7 It is a schematic diagram of SER performance comparison when stochastic resonance enhancement is not adopted and stochastic resonance enhancement is adopted when the number of relays changes in an embodiment of the present invention under the condition that the local channel is ideal;

[0030] Figure 8 It is a schematic diagram of SER performance comparison when the stochastic resonance strength is different under the condition that the local channel is ideal in an embodiment of the present invention;

[0031] Fig. 9 It is a schematic diagram comparing SER performance when random resonance enhancement is not adopted and when random resonance enhancement is adopted when the number of routes changes in the case of a small crossover probability of the relay channel in an embodiment of the present invention;

[0032] Fig.10 is a schematic diagram comparing SER performance when stochastic resonance enhancement is not used and when stochastic resonance enhancement is used when the number of relays changes in the case of a small crossover probability of the relay channel in an embodiment of the present invention;

[0033] Fig.11 is a schematic diagram of SER performance comparison when the stochastic resonance strength is different in the case of a small crossover probability in the relay channel according to an embodiment of the present invention;

[0034] Fig.12 It is a schematic diagram comparing SER performance when random resonance enhancement is not adopted and when random resonance enhancement is adopted when the number of routes changes in the case of a large crossover probability of the relay channel in an embodiment of the present invention;

[0035] Fig.13 It is a schematic diagram of SER performance comparison between not using stochastic resonance enhancement and using stochastic resonance enhancement when the number of relays changes in the case of a large crossover probability of the relay channel in an embodiment of the present invention;

[0036] Fig.14 It is a schematic diagram of SER performance comparison when the stochastic resonance strength is different under the condition of large crossover probability of the relay channel in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] The specific implementation of the present invention will be further described below in conjunction with the accompanying drawings.

[0038] Embodiment of the stochastic resonance enhancement method for wireless sensor networks

[0039] The present invention adds a binary asymmetric channel between the last relay node of each transmission link and the fusion center to introduce random resonance, and flips the information transmitted by the last relay node to the fusion center, that is, all the received information 0 is flipped to 1 and all the information 1 is flipped to 0, so that the accuracy of the received value of the fusion center can be improved. The specific implementation process of this method is as follows Figure 1As shown, the detailed description is given below.

[0040] 1. The information received by each local sensor is forwarded to the integration center through the corresponding relay transmission channel formed by multiple relay nodes.

[0041] like Figure 5 As shown, the wireless sensor network in this embodiment includes I local sensors, and the signal source generates a binary sequence H 0 and H 1 , is received by I local sensors through a binary channel, and the local sensors make hard decisions using the maximum likelihood criterion, and then forward the hard decision results to the relay nodes, which are then forwarded by multiple relay nodes and sent to the fusion center. The specific implementation process is:

[0042] like Figure 3 As shown, the i-th local sensor receives the binary data H sent by the source. 0 or H 1 , the result after hard decision based on the maximum likelihood criterion is x i , x i ∈{0,1}; data x i After being forwarded through the transmission link composed of multiple relay nodes, it is sent to the fusion center. The sequence received by the fusion center is y i In this embodiment, the relay node adopts an amplification and forwarding strategy with an amplification factor of 1, that is, the local sensor sends data x i , x i After being transmitted on the first BSC (binary symmetric channel), it reaches the first relay node. The data received by this relay node is Since the amplification factor is 1, the relay node does not do any processing and directly Just send it again. After the second BSC transmits, it reaches the second relay node, and the data received by this relay node is The subsequent processing is similar, and finally when the data reaches the fusion center, formula (1) is obtained.

[0043]

[0044] Among them, e i,j represents the channel transmission error corresponding to the jth BSC of the ith link. Each local sensor corresponds to a transmission link with the fusion center. Therefore, I represents the total number of transmission links, that is, the number of local sensors. J represents the total number of BSCs (binary symmetric channels) on the transmission link from the local sensor to the fusion center. Represents modulo 2 addition operation.

[0045] For the transmission link from local sensors to the fusion center, Figure 2 As shown in the figure, the channel of J BSCs connected in series can be equivalent to one BSC, and its error transfer probability is:

[0046]

[0047] Among them, ε i,j =P r (e i,j =1) is the error transfer probability of the jth BSC in the ith link.

[0048] 2. When the channel quality of the equivalent channel from the source to the last relay node of each transmission link is poor, a binary asymmetric channel (BAC) is added between the last relay node of each transmission link and the fusion center to introduce stochastic resonance and flip the information transmitted from the last relay node to the fusion center.

[0049] When the channel quality of the equivalent channel between each transmission link from the source to the last relay node is very poor, the fusion center receiving value will produce a large number of errors. Therefore, the present invention considers enhancing by stochastic resonance, such as Figure 4 As shown, all the received values ​​of the fusion center are flipped, that is, all the received information 0 is flipped to 1 and all the information 1 is flipped to 0. The very poor transmission link channel condition here means that the wireless sensor network has been attacked from the outside, resulting in the error transfer probability of the equivalent channel from the source to the last relay node greater than 0.5, while the probability of correct transmission is less than 0.5, that is, the probability of channel transmission error is greater than the probability of correct transmission, for example, the probability of error transmission is 0.7. The feasibility of this solution is explained below.

[0050] Based on the maximum likelihood criterion, the decision metric of the fusion center can be expressed as:

[0051]

[0052] Where y=[y 1 ,y 2 ...,y I ] is a binary vector received at FC, containing the local sensor information from I transmission links after J relays. i ) indicates that the fusion center receives data y based on the i-th transmission link i The extracted decision metric value of the i-th transmission link.

[0053] Similarly, based on the maximum likelihood criterion, the decision metric value of the i-th transmission link can be given:

[0054]

[0055] Combining (3) and (4), we can get

[0056]

[0057] Among them, H 0 and H 1 Indicates the state of the source, H 0 and H 1 Represent binary signals 0 and 1 respectively. The binary bit signal generated by the source is transmitted to I sensors, which independently observe these signals; the detection probability P di and false alarm probability P fi It is used to describe the quality of data transmission from the source to the i-th sensor, P di represents the probability that the hard decision result of the i-th sensor is 1 when the source sends a binary signal 1, P fi represents the probability that the hard decision result of the i-th sensor is 1 when the source sends a binary signal 0. i is the result of hard decision of the source data by the i-th sensor through the ML (maximum likelihood) criterion, x i ∈{0,1},x i After being forwarded to FC (fusion center) through multiple relays, the data received by FC is recorded as a i , where a i ∈{0,1}; add noise to FC through stochastic resonance, where α i It means that a i =1 changes to y i =0 probability, β i It means that a i =0 changes to y i = 1. 1 ,y 2 ...,y I ] is a binary vector received at FC, containing the local sensor information from I branches after J, relays. i ) indicates that the fusion center receives data y based on the i-th channel i The decision metric value of the i-th transmission link is extracted. Finally, all received values ​​are fused to obtain the log-likelihood ratio.

[0058] The derivation process is as follows:

[0059]

[0060] When i =1, then:

[0061]

[0062] When i =0, then:

[0063]

[0064] Substituting formula (7) and formula (8) into formula (3), we can obtain formula (5).

[0065] In the process of realizing the optimal decision metric value of the system, since the fusion center needs to obtain the instantaneous CSI (channel state information) of each relay BSC in each link, even if accurate instantaneous CSI information can be obtained, it will consume a lot of energy. Therefore, three simplified extraction methods of decision metrics and their theoretical derivation process are considered.

[0066] 1) When the local channel is an ideal channel

[0067] The local channel is an ideal channel, which refers to the channel from the source to the sensor. When the channel from the source to the sensor is ideal, P di =1 and P fi =0, then formula (5) can be simplified to:

[0068]

[0069] in:

[0070]

[0071] If the crossover probability of BAC is α i Equal to β i , the original BAC becomes BSC, that is, noise is added to the original FC reception value through BSC. From (10), we can get

[0072]

[0073] When the channel conditions of each route are the same, that is, the equivalent BSC crossover probability ε of the relay channel before adding stochastic resonance (SR) i , the crossover probability of BAC α i and β i do not change with i, and are denoted as ε, α, and β respectively. From (11), we can get:

[0074]

[0075] When the channel quality of the equivalent channel from the source to the last relay node is good, that is, the probability of channel transmission error is less than the probability of correct transmission, for example, the probability of correct transmission is 0.7, stochastic resonance has no effect, that is, α = β = 0, and formula (12) becomes:

[0076]

[0077] Therefore, formula (13) can be further simplified as:

[0078]

[0079] When the decision threshold of the fusion center is 0, deleting the items greater than zero in formula (15) does not affect the final decision result. Therefore, the fusion metric value when the channel condition is good is:

[0080]

[0081] When the channel quality of the equivalent channel from the source to the last relay node is very poor, a large number of errors will occur in the fusion center receiving value. Therefore, the present invention sets α=β=1 to flip all the information of the fusion center receiving value during random resonance, that is, all the received information 0 is flipped to 1 and all the information 1 is flipped to 0. At this time, the decision metric value becomes (17):

[0082]

[0083] Similarly, when the decision threshold of the fusion center is 0, delete the terms greater than zero in formula (17): It does not affect the final judgment result, so formula (17) can be simplified as:

[0084]

[0085] From the above formula, we can find that when the channel quality of the equivalent channel from the source to the last relay node is very poor, the random resonance enhancement effect α=β=1 is used to flip all the information received by the fusion center. The difference in the measurement value is a negative sign compared to when the channel quality of the equivalent channel is good. At this time, SR enhancement can enable the system to achieve better results.

[0086] 2) When the error transfer probability of the relay transmission channel is small

[0087] The low probability of error transfer in the relay transmission channel means that the probability of relay BSC crossover is low, that is, ε i →0

[0088] When i =1, formula (4) can be simplified to:

[0089]

[0090] When i =0, formula (4) can be simplified to:

[0091]

[0092] Substituting (19) and (20) into (5), we can obtain the decision metric:

[0093]

[0094] Because P di +P fi =1, for the convenience of representation, P is used uniformly di It means that (21) can be simplified to

[0095]

[0096] When the channel conditions of each route are the same, that is, the equivalent BSC crossover probability ε of the relay channel before adding SR i , the crossover probability of BAC α i and β i They do not change with i and are denoted as ε, α and β respectively. Formula (22) can be changed to:

[0097]

[0098] When the channel quality of the equivalent channel from the source to the last relay node is good, stochastic resonance has no effect, that is, α = β = 0, and formula (23) can be changed to:

[0099]

[0100] Because 0.5<P d ≤1, so we have Right now Since the decision threshold used by the subsequent fusion center when making a decision is 0, It will not affect the decision result of the fusion center, so formula (24) can be simplified as:

[0101]

[0102] When the channel quality of the equivalent channel from the source to the last relay node is very poor, the fusion center receiving value will produce a large number of errors. Therefore, it is considered to enhance α = β = 1 through random resonance to flip all the information of the fusion center receiving value, that is, all the received information 0 is flipped to 1 and all the information 1 is flipped to 0. At this time, the metric value in the above formula will become

[0103]

[0104] Because 0.5<P d ≤1, so we have Right now (26) can be simplified as:

[0105]

[0106] 3) When the probability of error transfer in the relay transmission channel is large

[0107] The lower probability of error transfer in the relay transmission channel means that the relay BSC crossover probability is higher, i.e., ε i →0.5.

[0108] When i =1, formula (4) can be simplified to:

[0109]

[0110] When x→0, log(1+x)→x, formula (28) can be simplified to:

[0111]

[0112] Similarly, when y i = 0, the fusion statistic Λ(y i ) can be simplified to:

[0113]

[0114] Substituting the obtained formula (29) and formula (30) into formula (5), we can get the fusion decision metric value as:

[0115]

[0116] Among them, P d -P f > 0, so the common item (P d -P f ) does not affect the judgment result, so formula (31) can be further simplified as:

[0117]

[0118] Considering the EGC form of the fusion statistic Λ(y) is:

[0119]

[0120] If the local sensors of each channel are the same, that is, the detection probability of the sensor is P di and false alarm probability P fi The crossover probability of BAC does not change with i. i and β i Does not change with i, we can get:

[0121]

[0122] When the channel quality of the equivalent channel from the source to the last relay node is good, the fusion center reception value will only produce very few errors, and stochastic resonance does not need to have an effect at this time. Therefore, α=β=0 can be used at this time, and the fusion metric value at this time can be obtained as:

[0123]

[0124] When the channel quality of the equivalent channel from the source to the last relay node is very poor, the fusion center reception value generates a large number of errors. Therefore, we consider using random resonance to enhance α=β=1 to flip all the information received by the fusion center, that is, all the received information 0 is flipped to 1 and all the information 1 is flipped to 0. At this time, the metric value of the above formula becomes:

[0125]

[0126] From the above formula, we can find that when the channel quality of the equivalent channel from the source to the last relay node is very poor, the random resonance enhancement α=β=1 is used to flip all the information received by the fusion center. The difference in the measurement value is a negative sign compared to when the channel quality of the equivalent channel is good. At this time, SR can enhance the decision fusion performance of the system.

[0127] 3. The fusion center fuses the received values ​​obtained from each transmission link to obtain the system fusion decision metric value, and obtains the final detection result based on the system fusion decision metric value.

[0128] The fusion center can obtain the system decision metric value based on the maximum likelihood criterion according to the received value obtained from each transmission link, and compare it with the decision threshold to obtain the final detection result. The decision formula used in this embodiment is:

[0129]

[0130] in, It represents the decision result obtained by the fusion center, 0 is the decision threshold, and Λ represents the fusion decision metric value.

[0131] Wireless sensor network implementation

[0132] The wireless sensor network in this embodiment includes local sensors, relay nodes and a fusion center. The local sensors forward the information they receive to the fusion center through a transmission link formed by multiple relay nodes. It is characterized in that the wireless sensor network adopts the above-mentioned random resonance enhancement method of the wireless sensor network. The random resonance enhancement method has been described in detail in the previous embodiment and will not be repeated here.

[0133] In order to further verify the effect of the present invention, a simulation verification of the present invention is carried out below.

[0134] When the local channel is ideal, the system bit error rate (SER) performance comparison between using stochastic resonance enhancement and not using stochastic resonance enhancement is shown in the figure below. Figure 6 and Figure 7 Given in Figure 6 The simulation conditions are set as follows: the number of routes is I = 1, I = 3, and I = 5, and the number of relays is J = 3. Figure 7 The simulation conditions are set to the number of routes I = 5, the number of relays J = 1, J = 3 and J = 5, and the horizontal axis represents the change in the detection performance of the local sensor. It can be found that the performance of the method using stochastic resonance enhancement is significantly better than that without stochastic resonance enhancement. When the local channel is ideal, the SER performance comparison with different stochastic resonance strengths is Figure 8 It is given in , the stochastic resonance intensity is set to α = 0.7, α = 0.8, α = 0.9, α = 0.93, α = 0.96 and α = 0.99, it can be found that the greater the stochastic resonance intensity, the better the system performance. This means that when the local channel is ideal, by introducing the stochastic resonance enhancement effect in the system, the system performance can be greatly improved under poor channel conditions, and the greater the stochastic resonance enhancement, the more obvious the system performance improvement.

[0135] In the case of a small crossover probability in the relay channel, the system bit error rate (SER) performance comparison between using stochastic resonance enhancement and not using stochastic resonance enhancement is shown in the figure below. Fig. 9 and Fig.10 Given in Fig. 9 The simulation conditions are set as follows: the number of routes is I = 5, I = 7, and I = 9, the number of relays is J = 4, Fig.10 The simulation conditions are set to the number of routes I = 7, the number of relays J = 2, J = 4 and J = 6. The horizontal axis represents the change in the detection performance of the local sensor. It can be found that the performance of the method using stochastic resonance enhancement is significantly better than that without stochastic resonance enhancement. In the case of a small crossover probability in the relay channel, the SER performance comparison with different stochastic resonance strengths is Fig.11 It is given in Figure 1 that the stochastic resonance intensity is set to α = 0.7, α = 0.8, α = 0.9, α = 0.93, α = 0.96 and α = 0.99. It can be found that the greater the stochastic resonance intensity, the better the system performance. This means that when the relay channel has a small crossover probability, by introducing the stochastic resonance enhancement effect in the system, the system performance can be greatly improved under poor channel conditions, and the greater the stochastic resonance enhancement, the more obvious the system performance improvement.

[0136] In the case of a large crossover probability in the relay channel, the system bit error rate (SER) performance comparison between using stochastic resonance enhancement and not using stochastic resonance enhancement is shown in the figure below. Fig. 9 and Fig.10 Given in Fig.12 The simulation conditions are set as follows: the number of routes is I = 3, I = 5, and I = 7, the number of relays is J = 3, Fig.13 The simulation conditions are set to the number of routes I = 5, the number of relays J = 1, J = 3 and J = 5, and the horizontal axis represents the change in the detection performance of the local sensor. It can be found that the performance of the method using stochastic resonance enhancement is significantly better than that without stochastic resonance enhancement; in the case of a large crossover probability of the relay channel, the SER performance comparison with different stochastic resonance strengths is Fig.14 It is given in Figure 1 that the stochastic resonance intensity is set to α = 0.7, α = 0.8, α = 0.9, α = 0.93, α = 0.96 and α = 0.99. It can be found that the greater the stochastic resonance intensity, the better the system performance. This means that when the relay channel has a large crossover probability, by introducing the stochastic resonance enhancement effect in the system, the system performance can be greatly improved under poor channel conditions, and the greater the stochastic resonance enhancement, the more obvious the system performance improvement.

[0137] In summary, the present invention can improve the local channel under the three conditions of ideal, small relay channel cross probability and large relay channel cross probability. It has the characteristics of high reliability and strong robustness, and can improve the real-time monitoring capability of the operating status of equipment manufacturing equipment in the unmanned factory where the wireless sensor network is located, provide more accurate data support for fault prediction and preventive maintenance, and further promote the efficient and intelligent development of unmanned factories.

Claims

1. A method for enhancing stochastic resonance in a wireless sensor network, characterized in that: The method comprises the following steps: 1) When the channel quality of the equivalent channel between each transmission link from the information source to the last relay node in the wireless sensor network is poor, a binary asymmetric channel is added between the last relay node of each transmission link and the fusion center to introduce stochastic resonance, and the information transmitted from the last relay node to the fusion center is flipped; 2) The fusion center fuses the received values ​​obtained from each transmission link to obtain the system fusion decision metric value, and obtains the final detection result based on the system fusion decision metric value.

2. The method for enhancing random resonance of a wireless sensor network according to claim 1, characterized in that: In step 2), the final detection result is obtained according to the system fusion decision metric value. The decision method used is: Wherein, Λ represents the system fusion decision metric value, M is the decision threshold, and M is equal to 0.

3. The method for enhancing random resonance of a wireless sensor network according to claim 2, characterized in that: When the channel quality of the equivalent channel between each transmission link from the source to the last relay node is poor, the fusion center obtains the system fusion decision metric value as follows: Where I is the number of local sensors and also the number of transmission links, y i is the information received by the fusion center from the i-th local sensor through the corresponding transmission link.

4. The method for enhancing random resonance of a wireless sensor network according to claim 2, characterized in that: When the channel quality of the equivalent channel from the source to the last relay node is good, the fusion center obtains the system fusion decision metric value as follows: Where I is the number of local sensors, y i is the information received by the fusion center from the i-th local sensor through the corresponding transmission link.

5. The method for enhancing random resonance of a wireless sensor network according to any one of claims 1 to 4, characterized in that: The poor channel quality of the equivalent channels of each transmission link means that the wireless sensor network is attacked by an external network, resulting in an error transfer probability of the equivalent channel from the source to the last relay node greater than 0.5, while the probability of correct transmission is less than 0.

5.

6. The method for enhancing random resonance of a wireless sensor network according to any one of claims 1 to 4, characterized in that: The information transmitted on each transmission link is obtained by the local sensor in the wireless sensor network using the maximum likelihood criterion to make hard decisions on the data generated by the source.

7. The method for enhancing random resonance of a wireless sensor network according to any one of claims 1 to 4, characterized in that: The fusion center in step 2) uses the maximum likelihood criterion to obtain the system fusion decision metric value.

8. A wireless sensor network, comprising a local sensor, a relay node and a fusion center, wherein the local sensor forwards the information it receives to the fusion center via a transmission link formed by a plurality of relay nodes, characterized in that: The wireless sensor network adopts the random resonance enhancement method for the wireless sensor network described in any one of claims 1 to 7.

Images