Hybrid Data Collection Method Considering Time Delay and AUV Energy Consumption in Underwater Dynamic Environment

Abstract

Hybrid data collection method considering time delay and AUV energy consumption in an underwater dynamic environment. During the hybrid data collection process using multi-hop routing transmission and AUV information collection methods, considering node movement caused by the underwater acoustic environment, time loss in the data transmission process, and energy loss during AUV travel, the selection of the next-hop node in the multi-hop process and AUV path planning are optimized. When selecting a route in multi-hop transmission, the next-hop node may be in the process of data transmission, and this situation is regarded as one of the influencing factors for selecting the next-hop node to reduce time loss; during the AUV path planning process, the mutual influence of AUVs between different layers is added to the genetic algorithm, enabling the lower-layer AUVs to transfer data to the upper-layer AUVs to avoid the situation where all AUVs need to reach the Sink node; the considered underwater scenario is dynamically changing, which is more in line with the actual marine environment. It can effectively reduce time delay and AUV energy consumption.

Description

Technical Field

[0001] The present invention relates to an underwater acoustic network, and more particularly to a hybrid data collection method considering delay and AUV energy consumption in an underwater dynamic environment. Background Art

[0002] With the continuous deepening of human exploration of the ocean, establishing an underwater communication network to obtain information in the ocean has important research significance. Due to the advantage of slow attenuation of acoustic signals in the ocean environment, establishing an underwater acoustic sensor network (UASNs) for data collection and signal transmission is an effective implementation approach. When conducting underwater data collection, there are two common methods: multi-hop routing transmission and autonomous underwater vehicle (AUV) collection. The multi-hop routing transmission method has the problem of unbalanced energy consumption, while the AUV data collection has the problem of long traversal time and is not timely enough for the transmission of emergency data.

[0003] Designing an efficient underwater data collection method is crucial for accurately obtaining marine information. In the past few years, S.K. Durandher et al. (S.K. Durandher, M.S. Obaidat, S. Goel, et al, “Optimizing energy through parabola based routing in underwater sensor networks,” in Proc. IEEE Global Telecommun. Conf., Houston, TX, USA, 2011) proposed a parabola-based routing algorithm that ensures the optimality of the selected nodes by obtaining node location information, taking into account node energy consumption and end-to-end delay. However, the problem of unbalanced energy consumption caused by multi-hop transmission has not been solved. G. Han et al. (G. Han, X. Long, C. Zhu, et al, “An AUV location prediction-based data collection scheme for underwater wireless sensor networks,” IEEE Trans. Veh. Technol., vol. 68, no. 6, pp. 6037–6049, Jun. 2019) proposed a data collection scheme using AUVs, predefined the trajectories of AUVs, nodes close to the trajectories directly send data to the AUVs, while other nodes send data to neighbors closer to the trajectories, and avoid the problem of energy imbalance by periodically adjusting the trajectories, effectively extending the network lifetime. However, the path selection of this method is based on energy consumption and does not consider the problem of time loss. Z. Liu et al. (Z. Liu, X. Meng, Y. Liu, et al, “AUV-Aided Hybrid Data Collection Scheme Based on Value of Information for Internet of Underwater Things,” IEEE Internet of things journal, vol. 9, no. 9, pp. 6944-6955, May. 2022) proposed an AUV-aided hybrid data collection scheme HDCS, which collects emergency data through multi-hop transmission, while for time-insensitive data, it is collected and transmitted by a slower AUV, effectively reducing the time loss of emergency data and alleviating the energy consumption imbalance caused by multi-hop. However, the selection of the next-hop node in multi-hop transmission in this method does not consider the time loss caused by node status, and the energy consumption is also large during AUV collection, and the scenario considered in this method is static and does not conform to the actual underwater scenario.

[0004] It can be seen that although there are already relevant literatures combining the AUV data collection method and the multi-hop transmission method, the designed scenario does not conform to the actual state of underwater nodes. And there is still room for further optimization in the selection of the next-hop node and the AUV path loss, so as to reduce the time loss and path loss. Therefore, the present invention will further optimize on the basis of this hybrid data collection scheme. Summary of the Invention

[0005] The purpose of the present invention is to provide a hybrid data collection method that can reduce the time loss during multi-hop transmission and effectively reduce the AUV path loss during AUV collection, considering delay and AUV energy consumption in the underwater dynamic environment.

[0006] The present invention includes the following steps:

[0007] 1) Divide the underwater sensor nodes SN into N layers according to depth, with M SNs in each layer; there is one AUV in each layer to collect data from the cluster head CH, and plan the AUV path through the genetic algorithm; after the data collection is completed, the AUV patrols to the sink node Sink to transmit data information; when considering the influence of ocean current movement, it is assumed that the movement of the sensor nodes is mainly in the horizontal direction, with small fluctuations in the vertical direction, but it has no impact on the layering result;

[0008] 2) Select the data transmission method of the cluster head CH according to the importance of the data: use the multi-hop method to transmit important data; ordinary data is collected by the AUV; since the energy consumption of the lower-layer AUV for round trips between the data collection layer and the sink node Sink is relatively large, when transmitting data, it can be determined according to the state of the upper-layer AUV whether the data is directly transmitted to the sink node Sink or relayed by the upper-layer AUV to the upper-layer AUV and then finally transmitted to the sink node Sink;

[0009] 3) When collecting data with a high degree of importance, select the multi-hop method for data transmission; when selecting the next-hop node CH i select the next-hop node CH j when, CH i knows the remaining energy, location information, and status of the CHs within the communication range. Define the ratio of the remaining energy of CH j to its battery capacity as the relative remaining energy of CH j and, together with the distance between CH i and CH j and the distance between CH j and Sink, as the influencing factors for selecting the next-hop node;

[0010] At the same time, when the state of CH j is in the data collection or transmission stage, CH iThe arrival data needs to wait for a period of time until CH j is in the idle state; therefore, the parameter T W is defined as:

[0011] T W = T b - 2·D i,j / v sound

[0012] where T b represents the estimated waiting time information feedback from CH i to CH j when CH j communicates with CH i , D i,j represents the distance between CH i and CHj, and v sound represents the speed of sound;

[0013] 4) The member nodes within the cluster to which the cluster head CH j belongs transmit data to CH j , and CH j propagates data to the next-hop node CH g , all through acoustic waves; when the propagation distance is D, the overall attenuation of the underwater acoustic signal is:

[0014] A(D,f) = D k a(f) D

[0015] where k represents the diffusion factor and a(f) is the absorption coefficient; according to the Thorp formula, when the frequency of the acoustic signal is f, the absorption coefficient is:

[0016]

[0017] When the transmit power is P, the signal-to-noise ratio is:

[0018]

[0019] where is the bandwidth and N(f) is the power spectral density of the ambient noise in the ocean. Therefore, the rate of SN j transmitting to SN g can be expressed as:

[0020] R j,g = Δflog2(1 + SNR(D j,g ,f))

[0021] Then the time taken for SN to transmit L bits of data is:

[0022]

[0023] 5) When CH j is in the idle state, that is, when CH j is not collecting data from member nodes and not transmitting data to the next hop, set T W to 0; when CH j is in the stage of member nodes transmitting data to the cluster head, it is divided into two cases: if the importance of the collected data is not high, then calculate the time when the cluster head finishes collecting data as T b ; if the importance of the collected data is high, then it is necessary to calculate the time when the cluster head finishes collecting data and the time when the transmission to the next hop by CH j is completed, as the T j of CH b ; when CH j is in the stage of transmitting to the next hop, use the remaining transmission time of CH j as T b . In the above three cases, when Tb < 2·D i,j / v sound , define T W as 0; if T b > 2·D i,j / v sound , then calculate according to the definition in step 3): T W = T b - 2·D i,j / v sound ;

[0024] 6) At the same time, use T W as an influencing factor for selecting the next-hop node, and the selection probability P j of the next-hop node can be defined as:

[0025]

[0026] where E j , E0 respectively represent the remaining energy and battery capacity of CHj, D i,sink and D j,sink respectively represent the distances from CH i and CH j to Sink, and λ, η, β, γ represent the weights of each influencing factor;

[0027] 7) When general data is collected, the AUV traverses the CHs in the current layer and collects data; use the genetic algorithm to plan the path of the AUV. Since the path of the lower-layer AUV directly transmitting data to Sink consumes a large amount of energy, a suitable upper-layer AUV is selected as a relay;

[0028] Before the AUV at layer m collects data, it broadcasts the layer index and time information to the lower-layer AUVs. The time information includes the current time and the data collection times of the previous two rounds. Define the time when the AUV at layer m is expected to finish collecting data as:

[0029] t estm = θt m [r - 1]+(1 - θ)t m [r - 2]

[0030] where θ is a weighting factor, t m [r - 1] and t m [r - 2] are the data collection times of the previous round and the round before that, respectively. After receiving the information, the lower-layer AUV calculates the difference between the time t self it needs to collect data and t estm . When |t self - t estm | ≤ b, it is considered that the lower-layer AUV can use the AUV at layer m as a relay; otherwise, it does not use the AUV at layer m as a relay. The lower-layer AUV feeds back to the AUV at layer m whether it chooses the AUV at layer m as a relay and the final position information of its genetic algorithm planned path. The AUV at layer m then plans the path according to the feedback information;

[0031] 8) Suppose the AUV at layer n is located at the point (x ln , y ln , z ln ) after path planning. If the AUV at layer m receives s feedbacks using it as a relay, then in the design of the objective function of the genetic algorithm, the influence of the final positions of the s-layer AUVs on the AUV at layer m is added, and the objective function is designed as:

[0032]

[0033] where x i is the i-th individual, and the individual is defined as a sequence of different permutations and combinations of AUVs visiting CHs; κ is the set of all AUVs that need to use the AUV at layer m as a relay; D ln,lm is the distance between the final positions of the AUV at layer n and the AUV at layer m in path planning, and D lm,sink is the distance between the final position of the AUV at layer m and the Sink, is a weighting factor;

[0034] Define the fitness function as the reciprocal of the objective function, then the fitness function is:

[0035] fitness(x i ) = 1 / F(x i )

[0036] The AUV will select the path with the maximum fitness function for data collection.

[0037] The present invention is a hybrid data collection method considering time delay and AUV energy consumption in an underwater dynamic environment, which can further reduce the time loss in the multi-hop process and the path loss in the AUV collection process. Therefore, in the underwater data collection scheme of the present invention, the selection basis of the next-hop node is designed, and the consideration of the possible time loss caused by selecting the node when the next-hop node is in the data collection and transmission stage is added; moreover, in the AUV data collection stage, the method that the upper AUV can be used as a relay for the lower AUV is adopted, and the influence of the states of other AUVs is added when planning the AUV path by the genetic algorithm, so as to achieve the purpose of reducing the AUV path loss under the condition of less time loss.

[0038] The present invention takes into account that the sensor node has two tasks of its own data transmission and acting as a relay node, and also takes into account the problem of the too long travel path of the AUV during AUV data collection. In the two data collection methods of multi-hop transmission and data transmission using AUV, the methods of classifying the node state into busy and idle, and using the upper AUV as a relay after the AUV collection is completed are respectively adopted to further reduce the time loss during multi-hop transmission and the path loss during AUV collection.

[0039] The present invention has the following outstanding advantages:

[0040] 1) Considering that the underwater sensor node has two tasks of its own data transmission and acting as a relay node, the state of the next-hop node is divided into two cases of busy and idle, and the node state is used as a factor affecting the selection of the next-hop node to avoid time loss caused by waiting for node data transmission;

[0041] 2) Using the upper AUV as a relay and adding the mutual influence between AUVs in the genetic algorithm to avoid the path loss caused by all AUVs reaching the Sink to transmit data;

[0042] 3) Considering that the underwater sensor nodes are moving, when conducting computer simulation verification, the position information of the nodes is dynamically changing in different data collection rounds. Brief Description of the Drawings

[0043] Figure 1 It is a scenario diagram of the hybrid data collection method considering time delay and AUV energy consumption in the underwater dynamic environment of the present invention.

[0044] Figure 2 It is a flowchart of the hybrid data collection method considering time delay and AUV energy consumption in the underwater dynamic environment of the present invention.

[0045] Figure 3Time loss comparison chart of the hybrid data collection method considering time delay and AUV energy consumption in the underwater dynamic environment of the present invention.

[0046] Figure 4 Fourth-layer AUV path diagram planned by the genetic algorithm of the hybrid data collection method considering time delay and AUV energy consumption in the underwater dynamic environment of the present invention.

[0047] Figure 5 Third-layer AUV path diagram planned by the genetic algorithm of the hybrid data collection method considering time delay and AUV energy consumption in the underwater dynamic environment of the present invention.

[0048] Figure 6 Third-layer AUV path diagram as a relay planned by the genetic algorithm of the hybrid data collection method considering time delay and AUV energy consumption in the underwater dynamic environment of the present invention.

[0049] Figure 7 Path loss diagram under different AUV relays of the hybrid data collection method considering time delay and AUV energy consumption in the underwater dynamic environment of the present invention. Detailed implementation mode

[0050] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0051] 1) As Figure 1 shown, the underwater sensor nodes SN are divided into N = 4 layers according to depth. There are M = 16 CHs in each layer, and there are 10 to 20 sensor nodes SNs in each cluster to which each CH belongs. There is one AUV in each layer to collect data from the cluster heads CH, and the AUV path is planned through the genetic algorithm. After the data collection is completed, the AUV patrols to the sink node Sink to transmit data information. When considering the influence of ocean current movement, it is assumed that the movement of the sensor nodes is mainly in the horizontal direction, with small fluctuations in the vertical direction, but it has no impact on the layering result.

[0052] 2) Select the data transmission method of the cluster head CH according to the importance of the data: for important data, use the multi-hop method for transmission; ordinary data is collected by the AUV. Since the energy consumption of the lower-layer AUV traveling back and forth between the data collection layer and the sink node Sink is relatively large compared with the upper layer, when transmitting data, it can be determined according to the status of the upper-layer AUV whether the data is directly transmitted to the sink node Sink, or the upper-layer AUV is used as a relay to transmit the data to the upper-layer AUV and finally transmitted to the sink node Sink.

[0053] 3) When collecting data with a high degree of importance, select the multi-hop method for data transmission. When the node CH i selects the next-hop node CH j , CH i knows the remaining energy, location information, and status of the CHs within the communication range. Define CHj The ratio of the remaining energy to the battery capacity is CH j The relative remaining energy, compared with CH i and CH j The distance to CH j Together with the distance between CH and the Sink, they are used as influencing factors for selecting the next-hop node.

[0054] Meanwhile, when CH j is in the data collection or transmission phase, the data arriving at CH i needs to wait for a period of time until CH j is in the idle state. Therefore, the parameter T W is defined as:

[0055] T W = T b - 2·D i,j / v sound

[0056] Wherein, T b represents the estimated waiting time information feedback from CH i to CH j when CH j communicates with CH i D i,j represents the distance between CH i and CHj, and v sound represents the speed of sound.

[0057] 4) The member nodes within the cluster to which the cluster head CH j belongs transmit data to CH j and CH j propagates data to the next-hop node CH g are all propagated through sound waves. When the propagation distance is D, the overall attenuation of the underwater acoustic signal is:

[0058] A(D,f) = D k a(f) D

[0059] Wherein, k represents the diffusion factor, and a(f) is the absorption coefficient. According to the Thorp formula, when the frequency of the acoustic signal is f, the absorption coefficient is:

[0060]

[0061] When the transmission power is P, the signal-to-noise ratio is:

[0062]

[0063] Wherein, is the bandwidth, and N(f) is the power spectral density of the ambient noise in the ocean. Therefore, SNj To SN g The transmission rate can be expressed as:

[0064] R j,g = Δf log2(1 + SNR(D j,g , f))

[0065] Then the time taken for SN to transmit L bits of data is:

[0066]

[0067] 5) When the state of CH j is idle, i.e., when CH j is not collecting data from member nodes and is not transmitting data to the next hop, set T W to 0; when CH j is in the stage of member nodes transmitting data to the cluster head, it is divided into two cases: if the importance of the collected data is not high, then calculate the time when the cluster head finishes collecting data as T b ; if the importance of the collected data is high, then it is necessary to calculate the time when the cluster head finishes collecting data and the time when CH j completes transmission to the next hop as the T j of CH b . When CH j is in the stage of transmitting to the next hop, use the remaining transmission time of CH j as T b . In the above three cases, when Tb < 2·D i,j / v sound , define T W as 0; if T b > 2·D i,j / v sound , then calculate according to the definition in step 3): T W = T b - 2·D i,j / v sound .

[0068] 6) At the same time, use T W as a factor affecting the selection of the next-hop node. The probability P j of selecting the next-hop node can be defined as:

[0069]

[0070] where E j , E0 respectively represent the remaining energy and battery capacity of CHj, D i,sink and D j,sink respectively represent CH i and CH jThe distance to the Sink, and λ, η, β, γ represent the weights of various influencing factors.

[0071] 7) When general data is collected, the AUV traverses the CHs in the current layer and collects data. The genetic algorithm is used to plan the path of the AUV. Since the energy consumption of the direct path for the lower-layer AUV to transmit data to the Sink is relatively large, a suitable upper-layer AUV is selected as a relay.

[0072] Before the AUV in the m-th layer collects data, it broadcasts the layer index and time information to the lower-layer AUVs. Among them, the time information includes the current time and the data collection times of the previous two rounds. Define the time when the AUV in the m-th layer is expected to finish collecting data as:

[0073] t estm = θt m [r - 1]+(1 - θ)t m [r - 2]

[0074] where θ is a weighting factor, t m [r - 1] and t m [r - 2] are the data collection times of the previous round and the round before the previous round respectively. After receiving the information, the lower-layer AUV calculates the difference between the time t self it needs to collect data and t estm . When |t self -t estm |≤b, it is considered that the lower-layer AUV can use the AUV in the m-th layer as a relay; otherwise, it does not use the AUV in the m-th layer as a relay. The lower-layer AUV feeds back to the AUV in the m-th layer whether it selects the AUV in the m-th layer as a relay and the final position information of its own genetic algorithm planned path. The AUV in the m-th layer then plans the path according to the feedback information.

[0075] 8) Assume that the AUV in the n-th layer is located at the point (x ln , y ln , z ln ) after path planning. If the AUV in the m-th layer receives s feedbacks using it as a relay, then in the design of the objective function of the genetic algorithm, the influence of the final positions of the s-layer AUVs on the AUV in the m-th layer is added, and the objective function is designed as:

[0076]

[0077] where x i is the i-th individual, and the individual is defined as a sequence of different permutations and combinations of AUVs visiting CHs. κ is the set of all AUVs that need to use the AUV in the m-th layer as a relay. D ln,lm is the distance between the final position of the path planning of the AUV in the n-th layer and the AUV in the m-th layer, and D lm,sinkis the distance between the final position of the m-th layer AUV and the Sink. is the weighting factor.

[0078] Define the fitness function as the reciprocal of the objective function, then the fitness function is:

[0079] fitness(x i ) = 1 / F(x i )

[0080] The AUV will select the path with the maximum fitness function for data collection.

[0081] The above processes for next-hop node selection and AUV relay selection are as Figure 2 shown.

[0082] Next, the feasibility of the method described in the present invention is verified by computer simulation.

[0083] Set the simulation scenario of the present invention as a three-dimensional underwater scenario of 600 * 600 * 600 m. Divide this scenario into four layers with a depth of 150 m. The number of CHs in each layer is 16, and the number of member nodes of each CH is from 10 to 20. The data packet size is 1024 bits, the carrier frequency f = 20 kHz, and the bandwidth is 1 KHz. Define the initial energy of the CH q0 = 50 J, and the sound speed vsound = 1500 m / s. λ, η, β, γ are 0.5, 0.2, 0.1, 0.4 respectively, which are 0.6, 0.3, 0.1 respectively. In the genetic algorithm, the population size G = 200, the number of iterations C = 500, the elimination acceleration index s = 2, the crossover probability Pc = 0.5, and the mutation probability Pm = 0.2.

[0084] Assume that the sensor nodes in each layer have random positions due to obvious horizontal movement, and are affected by relatively small fluctuations in the vertical direction, resulting in the node positions fluctuating up and down by about 5 m in each layer. Set that 6 - 18 CHs have the transmission task of emergency data. As can be seen from Figure 3 , compared with the comparative scheme HDCS that also adopts the hybrid data collection scheme, the proposed scheme in the present invention has obvious optimization in terms of time loss during the data collection and transmission process. When the number of CHs with emergency data is 6 - 8, since the number of CHs using the multi-hop transmission method is small, and the number of relay nodes with their own multi-hop transmission tasks on the multi-hop path is also small, that is, the probability of having busy nodes on the path is low. Therefore, the results show that the data transmission time losses of the two schemes are the same; and as the number of CHs with emergency data increases, the probability of having busy nodes on the multi-hop path also increases. In summary, the proposed scheme in the present invention has less time loss compared with HDCS, effectively improving the timeliness of information.

[0085] When using AUVs for data collection, taking the third-layer AUV and the fourth-layer AUV as examples for analysis, the results are as follows Figures 4 - 6 as shown. Among them, Figure 4 is the genetic algorithm planned path of the fourth-layer AUV. When the third-layer AUV does not meet the conditions to become a relay, this AUV performs genetic algorithm path planning according to the shortest path, and the results are as follows Figure 5 shown, and the final path is 2.2555 km; when the AUV meets the relay conditions, the final position of the fourth-layer AUV is added as an influencing factor for the third-layer AUV to perform path planning, and the results are as follows Figure 6 shown. It can be seen that the re-planned AUV path is 2.4182 km, compared with Figure 5 the path has increased, but the final position after planning is close to the final position of the third-layer AUV, and there is no need to reach the Sink to transmit data. Therefore, the overall path of the AUV is significantly reduced. As can be seen in Figure 7 as the number of AUVs that meet the relay conditions increases, the total distance traveled by all AUVs is significantly reduced. Therefore, the present invention can reduce the path loss of AUVs.

[0086] The present invention adopts a hybrid data collection method, and adopts a multi-hop transmission method or an AUV collection method according to the urgency of the data. When adopting the multi-hop transmission method, considering that other sensor nodes on the transmission path may have their own transmission tasks, that is, the nodes are in a busy state. Therefore, when selecting the next-hop node, the consideration of the node state is added, and the remaining energy of the node, the distance between nodes, the distance between the node and the Sink, and the node waiting time are comprehensively used as the selection probability to reduce the time loss while saving energy consumption; when adopting the AUV collection method, according to the time information and the final position information after the AUV collects the data, it is judged whether the AUV can be used as a relay node for the lower-layer AUV, and the AUV that meets the conditions is used as a relay to receive the data information of the lower-layer AUV, avoiding the path loss caused by all AUVs reaching the Sink to transmit data. The hybrid data collection method considering the node state and the AUV path loss proposed by the present invention can reduce the time loss of data transmission and the AUV path loss.

Claims

1. Hybrid data collection method considering time delay and AUV energy consumption in underwater dynamic environment, characterized in that Including the following steps: 1) Divide the underwater sensor nodes SN into N layers according to depth, with M SNs in each layer; there is an AUV in each layer to collect data from the cluster head CH, and the AUV path is planned through a genetic algorithm; after data collection is completed, the AUV travels to the sink node Sink to transmit data information; when considering the influence of ocean current movement, it is assumed that the movement of the sensor nodes is mainly in the horizontal direction, with small fluctuations in the vertical direction, but it has no impact on the layering result; 2) Select the data transmission method of the cluster head CH according to the importance of the data: multi-hop transmission is used for important data; ordinary data is collected by the AUV; since the energy consumption of the lower-layer AUV for round trips between the data collection layer and the sink node Sink is relatively larger than that of the upper layer, when transmitting data, it is determined according to the status of the upper-layer AUV whether the data is directly transmitted to the sink node Sink or the upper-layer AUV is used as a relay to transmit the data to the upper-layer AUV and then finally transmitted to the sink node Sink; 3) When important data is collected, a multi-hop method is selected for data transmission; at node CH i Select the next-hop node CH j When, CH i The remaining energy, location information, and status of CHs within the communication range are known; define the ratio of the remaining energy of CH j to the battery capacity as the relative remaining energy of CH j , together with the distance between CH i and CH j , and the distance between CH j and Sink, are used as influencing factors for selecting the next-hop node; Meanwhile, when the state of CH j is in the data collection or transmission phase, the arriving data of CH i needs to wait for a period of time until CH j is in the idle state; Therefore, define the parameter T W : T W = T b - 2·D i,j / v sound Among them, T b represents the expected waiting time information that CH i feeds back to CH j when communicating with CH j ; D i represents the distance between CH i,j and CHj, and v i represents the speed of sound; sound ​ 4) When the state of CH j is idle, that is, when CH j collects no data from member nodes and transmits no data to the next hop, set T W to 0; when CH j is in the stage of member nodes transmitting data to the cluster head, it is divided into two cases: if the importance of the collected data is not high, then calculate the time when the cluster head finishes collecting data as T b ; if the importance of the collected data is high, then it is necessary to calculate the time when the cluster head finishes collecting data and the time when CH j completes the transmission to the next hop as the T j of CH b ; when CH j is in the stage of transmitting to the next hop, use the remaining transmission time of CH j as T b ; in the above three cases, when Tb < 2·D i,j / v sound , define T W as 0; if T b > 2·D i,j / v sound , then calculate according to the definition in step 3): T W = T b - 2·D i,j / v sound . 5) Take T W as an influencing factor for selecting the next-hop node, and define the next-hop node selection probability P j as follows: Among them, E j , and E0 respectively represent the remaining energy and battery capacity of CHj, D i,sink and D j,sink respectively represent the distances from CH i and CH j to the sink, and λ, η, β, γ represent the weights of various influencing factors.

2. The hybrid data collection method considering time delay and AUV energy consumption in the underwater dynamic environment as described in claim 1, the method further includes: Cluster Head CH j The member nodes within the cluster to which it belongs transmit data to CH j and CH j propagates data to the next-hop node CH g both through acoustic waves; when the propagation distance is D, the overall attenuation of the underwater acoustic signal is: A(D,f) = D k a(f) D Where k represents the diffusion factor and a(f) is the absorption coefficient; according to the Thorp formula, when the acoustic signal frequency is f, the absorption coefficient is: When the transmission power is P, the signal-to-noise ratio is: where is the bandwidth and N(f) is the power spectral density of ambient noise in the ocean; thus, the rate of transmission to SN j to SN g is expressed as: R j,g = Δflog2(1 + SNR(D j,g , f)) Then the time taken for the SN to transmit L bits of data is:

3. The hybrid data collection method considering time delay and AUV energy consumption in the underwater dynamic environment as described in claim 1, the method further includes: When general data is collected, the AUV traverses the CHs in the layer where it is located and collects data; The genetic algorithm is used to plan the path of the AUV; since the energy consumption of the path for the lower-layer AUV to directly transmit data to the Sink is relatively large, a suitable upper-layer AUV is selected as a relay; Before the AUV in the m-th layer collects data, it broadcasts the layer index and time information to the lower-layer AUV; where the time information includes the current time and the data collection times of the previous two rounds; define the time when the AUV in the m-th layer is expected to complete data collection as: t estm = θt m [r - 1]+(1 - θ)t m [r - 2] where θ is a weighting factor, t m [r - 1] and t m [r - 2] are the data collection times of the previous round and the round before the previous round respectively; after receiving the information, the lower - layer AUV calculates the time difference t self and t estm between them; when |t self - t estm | ≤ b, it is considered that the lower - layer AUV takes the m - layer AUV as a relay; otherwise, it does not take the m - layer AUV as a relay; the lower - layer AUV feeds back to the m - layer AUV whether it selects the m - layer AUV as a relay and the final position information of its own genetic - algorithm - planned path; the m - layer AUV then plans the path according to the feedback information.

4. The hybrid data collection method considering time delay and AUV energy consumption in the underwater dynamic environment as described in claim 1, the method further includes: Set the nth layer AUV to be located at the point (x ln , y ln , z ln ) after path planning. If the mth layer AUV receives a total of s feedbacks regarding it as a relay, then in the design of the objective function of the genetic algorithm, the influence of the final positions of the s-layer AUVs on the mth layer AUV is added, and the objective function is designed as follows: where x i is the i-th individual, and the individual is defined as a sequence of different permutations and combinations of AUVs accessing CHs; κ is the set of all AUVs that need to use the AUV at the m-th layer as a relay; D ln,lm is the distance between the AUV at the n-th layer and the final position of the path planning of the AUV at the m-th layer, and D lm,sink is the distance between the final position of the AUV at the m-th layer and the Sink, and δ, ε, are weighting factors; Define the fitness function as the reciprocal of the objective function, then the fitness function is: fitness(x i ) = 1 / F(x i ) The AUV will select the path with the maximum fitness function for data collection.

Images