CN109275099B - VOI-based multi-AUV (autonomous Underwater vehicle) efficient data collection method in underwater wireless sensor network - Google Patents

VOI-based multi-AUV (autonomous Underwater vehicle) efficient data collection method in underwater wireless sensor network Download PDF

Info

Publication number
CN109275099B
CN109275099B CN201811123847.3A CN201811123847A CN109275099B CN 109275099 B CN109275099 B CN 109275099B CN 201811123847 A CN201811123847 A CN 201811123847A CN 109275099 B CN109275099 B CN 109275099B
Authority
CN
China
Prior art keywords
auv
cluster
voi
data
data collection
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201811123847.3A
Other languages
Chinese (zh)
Other versions
CN109275099A (en
Inventor
韩光洁
唐正凯
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Changzhou Campus of Hohai University
Original Assignee
Changzhou Campus of Hohai University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Changzhou Campus of Hohai University filed Critical Changzhou Campus of Hohai University
Priority to CN201811123847.3A priority Critical patent/CN109275099B/en
Publication of CN109275099A publication Critical patent/CN109275099A/en
Application granted granted Critical
Publication of CN109275099B publication Critical patent/CN109275099B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/02—Services making use of location information
    • H04W4/021—Services related to particular areas, e.g. point of interest [POI] services, venue services or geofences
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B13/00—Transmission systems characterised by the medium used for transmission, not provided for in groups H04B3/00 - H04B11/00
    • H04B13/02—Transmission systems in which the medium consists of the earth or a large mass of water thereon, e.g. earth telegraphy
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/90—Services for handling of emergency or hazardous situations, e.g. earthquake and tsunami warning systems [ETWS]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W40/00—Communication routing or communication path finding
    • H04W40/02—Communication route or path selection, e.g. power-based or shortest path routing
    • H04W40/20—Communication route or path selection, e.g. power-based or shortest path routing based on geographic position or location
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W40/00—Communication routing or communication path finding
    • H04W40/24—Connectivity information management, e.g. connectivity discovery or connectivity update
    • H04W40/32—Connectivity information management, e.g. connectivity discovery or connectivity update for defining a routing cluster membership
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W84/00—Network topologies
    • H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Business, Economics & Management (AREA)
  • Health & Medical Sciences (AREA)
  • Emergency Management (AREA)
  • Environmental & Geological Engineering (AREA)
  • Public Health (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

The invention discloses a VOI-based multi-AUV (autonomous Underwater vehicle) efficient data collection method in an underwater wireless sensor network, which comprises the following steps: (1) the high VOI data packet is preferentially transmitted to the cluster head; (2) determining a moving Sink position of a three-dimensional underwater environment and dividing a space region; (3) path planning of the AUV in the sub-area; (4) multiple AUVs compete dynamically for handling urgent tasks. The invention can effectively solve the complex data collection task in the underwater wireless sensor network by utilizing the characteristics of the spatial distribution characteristic and the resource distribution of the multiple AUVs, the data collection mode of the multiple AUVs has better fault-tolerant capability and improves the robustness of the system, and the advantages of high reliability and low delay are not possessed by a single AUV system. The three-dimensional underwater region division is combined with AUV path planning, a multi-AUV dynamic competition mechanism is adopted to process emergency events, data in the network are dynamically collected, network energy consumption is balanced, and the service life of the network is prolonged.

Description

VOI-based multi-AUV (autonomous Underwater vehicle) efficient data collection method in underwater wireless sensor network
Technical Field
The invention belongs to the technical field of multi-AUV data collection of an underwater wireless sensor network, and particularly relates to a VOI-based multi-AUV high-efficiency data collection method in the underwater wireless sensor network.
Background
Nodes in the underwater wireless sensor network are randomly deployed in a monitoring area interested by an application system, and the nodes form the underwater wireless sensor network in a self-organizing mode. The method is characterized in that a data collection task is executed in an underwater environment, and at least three modules, namely a common sensor node, a Sink node (Sink) and a task management center, are required. The common nodes can send the sensed data to the Sink through the multi-hop mode by using acoustic communication, and can also collect the data of the nodes in the cluster through a clustering mechanism by using cluster heads and send the data to the Sink through inter-cluster multi-hop. The Sink sends the gathered information to the task management center through wireless wave communication, so that the Sink can make a timely response strategy. The sensor nodes monitor dynamic information of the interest area in real time, and when the nodes in a certain area sense that a large amount of same data or a plurality of nodes overflow data and the like, the situation that an underwater emergency happens in the area is indicated.
The underwater wireless sensor network has the functions of being difficult to estimate in the aspects of regional monitoring, natural resource discovery, underwater target tracking, enemy investigation and the like. In the research of the underwater wireless sensor network, the method can be roughly divided into research directions of data collection, node positioning, network topology control, safety encryption, node charging and the like, and the data collection technology is taken as a key technology in basic research contents and has far-reaching significance in deep research of underwater environment characteristics. The underwater wireless sensor network has the following remarkable characteristics:
(1) the network size is large. Because the underwater environment is not a static environment, a specific area boundary cannot be planned, and the size of the underwater wireless sensor network is difficult to control. Meanwhile, the sensing range and the data transmission range of the common nodes are limited, and a large number of sensor nodes need to be deployed in the environment;
(2) the node energy is limited. The nodes deployed underwater are limited by cost and volume, the electric quantity of the nodes can only be provided by batteries, and battery replacement and electric quantity supplement are difficult. When the node is exhausted, the node dies, which easily causes an invisible routing hole and affects the network performance;
(3) underwater acoustic signals are severely attenuated. Due to the complex characteristics of the underwater environment and the absorption loss and the diffusion loss of the sound wave in the propagation process, the farther the acoustic signal is transmitted, the more serious the attenuation is. The selectable frequency of underwater communication is greatly limited by the characteristic, and the problems of high time delay and low transmission success rate are inevitably caused;
(4) and (4) node positioning problem. Different from a land sensor node, the underwater sensor node is influenced by water flow and sea wind, the position of the underwater sensor is dynamically changed and has uncertainty, positioning is difficult to achieve through a GPS, and the positioning mode based on acoustic signals has the problems of low precision, long time consumption and the like.
With the continuous development of science and technology, more advanced software and hardware equipment is introduced for the research of the underwater wireless sensor network. An Autonomous Underwater Vehicle (AUV) is one of Unmanned Underwater Vehicles (UUV), and artificial intelligence and other advanced computing technologies are integrated at present, so that the Autonomous Underwater Vehicle (AUV) plays an irreplaceable role in the field of underwater data collection. The AUV deployed underwater can cruise and access the common nodes according to a certain path due to the characteristics of sufficient energy, small influence of water flow and the like. The multiple AUVs can work cooperatively in a centralized and distributed mode, and the data collection efficiency is greatly improved. The network for deploying the AUV has strong expandability, and the tasks are often completed by setting different working modes in the face of different application environments and requirements.
The underwater wireless sensor network has high research value and very important significance for comprehensively researching ocean characteristics and developing and protecting ocean. In the recent international research progress, a new spherical node can control a signal mode to preset the depth, and an AUV (autonomous Underwater vehicle) can carry an underwater node to become an underwater mobile node and can continuously collect underwater information. The deep exploration of the marine environment in China has already started, but the research on the aspect of the underwater wireless sensor network is just started, and the main research institutions include acoustic research institute of Chinese academy of sciences, marine research institute of Chinese academy of sciences, automated research institute of Chinese academy of sciences, Harbin engineering university, Xiamen university and Chinese ocean university, and the like, and mainly perform research on underwater acoustic communication technology, networking protocols, system structures and the like.
Disclosure of Invention
In the randomly deployed underwater wireless sensor network, a data packet with high VOI is preferentially transmitted to a cluster head, and data of sub-regions are collected by using multiple AUVs through region division and path planning, so that network energy is balanced, and the service life of the network is prolonged.
The technical purpose is achieved, the technical effect is achieved, and the invention is realized through the following technical scheme:
a VOI-based multi-AUV efficient data collection method in an underwater wireless sensor network comprises the following steps:
(1) high VOI packet priority to cluster head transmission
The nodes of the underwater wireless sensor network are deployed randomly and clustered, after the source nodes in the cluster sense data, the information value VOI of the data packet is calculated at first, the VOI of the data packet stored by the node in the next hop is compared with the VOI of the received data packet, and the data packet with the high VOI is forwarded preferentially. Preferentially transmitting the data packet with higher VOI in the cluster to a cluster head, and waiting for AUV collection;
(2) mobile Sink position determination and space region division of three-dimensional underwater environment
In a three-dimensional underwater environment, multiple AUVs are used for executing an underwater data collection task, the energy and flow load balance of the multiple AUVs is kept, and meanwhile, the AUVs are ensured to send collected data to a Sink in time. Through a region division algorithm, the vertical navigation position of the mobile Sink is firstly determined, and then attributes such as node number, node density and node depth are integrated to divide an underwater space region into a plurality of sub-regions. Each subarea consists of a plurality of clusters, and the same AUV is responsible for data collection tasks;
(3) path planning of AUV in sub-area
After the subareas are divided, each AUV is responsible for a data collection task of one subarea, and data collection of a plurality of clusters in the subareas is completed within a certain time. And setting a cluster which is closest to the moving Sink vertical navigation area in the sub-area as an AUV primary collection path termination access cluster, and randomly selecting an initial access cluster. The number and the required time of AUV access clusters are controlled by setting a dynamic reward function of a Q-learning algorithm, a sampling path is learned, and an access path with the highest return is dynamically established;
(4) Multi-AUV dynamic competition processing emergency task
When a cluster head of a certain sub-region has emergency tasks such as data overflow and the like, the cluster head sends data emergency collection request information to the mobile Sink in an inter-cluster multi-hop mode, the mobile Sink evaluates the current state of the AUV, the multi-AUV competes in aspects such as the distance from the emergency task, the residual energy and the task execution state, the competing and winning AUV obtains the processing right of the emergency task and immediately suspends the current data collection task, the AUV is switched to the emergency task processing state and goes to the emergency task place, and the AUV returns to an atomic region after the AUV is processed, and continues to complete the residual data collection task.
And (3) the information value VOI of the data packet in the step (1) consists of an event importance degree EIP and an information concentration degree ICN. The EIP describes the correlation and timeliness of the data collected by the nodes and the data required by the application, and the calculation formula is Ek,i(t)＝αkFk+(1-αk)f(t-tk,i) Wherein αkIs a constant between 0 and 1, t is the current time, tk,iThe time at which the packet is generated for node i,
Figure GDA0002451433860000031
α thereinkIs a constant between 0 and 1, t is the current time, FkThe correlation between the data packet K sensed by the node i and the data required by the application is shown, X is the physical signal of the data required by the application, and K is the physical signal sensed by the node. In EIP
Figure GDA0002451433860000032
The attenuation coefficient η of the attribute that represents that the information is valuable for decision only for a certain timekWith FkAnd are different. Because different nodes in the cluster generate the same data packet when sensing the same event, the correlation influences the value of the data packet, the ICN represents the concentration degree of the data packet k in the cluster, and the calculation formula is
Figure GDA0002451433860000041
β thereinkIs a constant between 0 and 1, TicIndicating the time required for the packet k to travel from node i to the cluster head, and n indicating the number of perceived identical packets k. t is tendIndicates the time point, t, of the last transmission of the packet to the cluster headstartIndicating the point in time at which the packet was first collected by the cluster head. The time information in the ICN is obtained by historical information, and the nodes in the cluster pass throughAnd interacting with the periodic command of the cluster head to obtain related information. The information value VOI is thus defined as V ═ gammakEk+(1-γk)Ik，γkIs constant between 0 and 1.
The area division algorithm in the step (2) maps the clusters in the network to particles on a two-dimensional plane according to the positions of the cluster heads. Using these particles to make two-dimensional Voronoi diagram, firstly, using formula
Figure GDA0002451433860000042
CHiE CH, i ═ 1,2,3.. n) to establish the specific position of the mobile Sink vertical movement. The Thiessen polygon in which each particle is located is called a mass region, where dSink,centerIs the distance between the center of the mass region where the sink is located and the center of the two-dimensional plane, di,centerNerbor (Sink) represents the number of the geological regions adjacent to the geological region where the Sink is moving, Nerbor (CH) being the distance between the geological region i and the center of the two-dimensional planei) Indicates the number of adjacent prime regions in the prime region, and n is the number of total prime regions. For a prime area of the Voronoi diagram, various attributes such as the number of nodes, the density of nodes in a cluster, the average depth of the nodes in the cluster and the like exist, and the attribute of each prime area i is expressed as P in a vector formi＝[pi1,pi2,…,pim]TFor the m-th attribute, the similarity of the sub-region s after the i-th prime region and the plurality of prime regions are combined is defined as fi(psm,pim) It can be obtained by Euclidean distance. With Sim (P)s,Pi) The similarity of multiple attributes is represented by the formula
Figure GDA0002451433860000043
λ1+λ2+···+λm1 is ═ 1; the growth criterion for calculating the attribute m for all the material regions in the Voronoi diagram is
Figure GDA0002451433860000044
Wherein
Figure GDA0002451433860000045
Figure GDA0002451433860000046
Adjustment parameter, p, for attribute jjmaxAnd pjminRepresenting the maximum and minimum values of the property j in all sub-regions. Randomly selecting a plurality of prime areas to carry out random synchronous growth on adjacent prime areas, and stopping the area division algorithm when each prime area is positioned in one sub-area s under the condition of meeting W.
In the path planning stage in the step (3), the reward matrix for AUV transfer between clusters is established, and in each sub-area, the reward matrix for AUV transfer between clusters is set to be
Figure GDA0002451433860000051
The Q matrix is initialized to 0. The choice of reward function will determine the rate and extent of convergence of the Q-learning algorithm, and in order for the AUV to traverse all clusters in a sub-region without repeatedly accessing the clusters, the reward for each cluster is set as follows
Figure GDA0002451433860000052
Wherein α is a constant between 0 and 1, C is a reward amplitude adjustment parameter, f is the number of accesses of the jth cluster, VjTotal VOI, d obtained for AUV going to cluster jij(f) Is the linear distance of cluster i and cluster j. The AUV learns in a random path mode until a stable Q matrix is obtained, and the calculation formula of the Q matrix is Q (s, a) ═ R + gamma maxa'Q (s ', a '), γ is the tuning parameter, s is the current state, a is the action performed in the current state, s ' is the next state into which action a is performed in state s, a ' is the action performed in state s '. The AUV learns using the algorithm described above, each experience being equivalent to one training. During each training, the agent explores the environment. And the AUV finds a route with the highest return to the terminal access cluster, and sends data to the mobile Sink.
When a plurality of emergency events such as data overflow occur in the step (4), the mobile Sink evaluates the plurality of events and preferentially processes the data collection request information with high data overflow degree. Handling of current emergency by multiple AUVsThe weights are competitive, and the mobile Sink calculates the competitive win value of AUV in the network according to the following formula
Figure GDA0002451433860000053
Wherein α is a constant between 0 and 1, M is a boundary of the network, d is a distance between the AUV and the emergency site, E is a remaining capacity of the AUV, E is an energy consumed by the AUV for a unit distance traveled, and v is an average velocity of the AUV
Figure GDA0002451433860000054
Is calculated by the formula
Figure GDA0002451433860000055
Wherein, tnowRepresenting the current time, tiIndicating the time required by the AUV to handle the current emergency, dnext(i) Is the linear distance from the current emergency event to an event. The winning AUV can obtain processing power and go immediately to the relevant location for data collection.
The invention has the beneficial effects that:
the invention can effectively solve the complex data collection task in the underwater wireless sensor network by utilizing the characteristics of the spatial distribution characteristic and the resource distribution of the multiple AUVs, the data collection mode of the multiple AUVs has better fault-tolerant capability and improves the robustness of the system, and the advantages of high reliability and low delay are not possessed by a single AUV system. The three-dimensional underwater region division is combined with AUV path planning, a multi-AUV dynamic competition mechanism is adopted to process emergency events, data in the network are dynamically collected, network energy consumption is balanced, and the service life of the network is prolonged.
Drawings
FIG. 1 is a diagram of a network model according to an embodiment of the present invention;
FIG. 2 is a Voronoi model diagram of an embodiment of the present invention;
FIG. 3 is a zone division diagram of one embodiment of the present invention;
fig. 4 is a diagram illustrating data transmission from the AUV to the mobile sink according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The following detailed description of the principles of the invention is provided in connection with the accompanying drawings.
When an underwater wireless sensor network is deployed, the distribution situation of the nodes is random and uneven in density due to the complex characteristic of the underwater environment. In order to reduce the energy consumption of the nodes and prolong the service life of the network, a specific mechanism is adopted for clustering, and multiple AUVs send collected cluster head data to the mobile Sink. This network is subjected to a modeling process as shown in fig. 1.
Isomorphic nodes are randomly deployed in an interest area, the energy and communication range is limited, and the geographic position is known. The nodes are clustered according to a specific mechanism. Due to the fact that the acoustic signal transmission speed is low, the bandwidth is low, and the influence of underwater routing holes is caused, the application of multiple AUVs provides possibility for data collection work of the general situation. And moving Sink as a convergence center to make vertical motion with constant speed at a specific position of the region. The application system can predict and estimate the total data flow of the network, and determine the deployment number of the multiple AUVs according to the size of the distribution area of the network and the total data flow of the network. Therefore, the invention provides a VOI-based multi-AUV efficient data collection method in an underwater wireless sensor network, which comprises the following steps:
step one, high VOI data packet is preferentially transmitted to cluster head
The nodes of the underwater wireless sensor network are deployed randomly and clustered, after the source nodes in the cluster sense data, the information value VOI of the data packet is calculated at first, the VOI of the data packet stored by the node in the next hop is compared with the VOI of the received data packet, and the data packet with the high VOI is forwarded preferentially. And the data packet with higher VOI in the cluster is preferentially transmitted to the cluster head to wait for the collection of AUV. The information value VOI of a data packet consists of an event importance level EIP and an information concentration ICN. EIP describes the data collected by a node and responsesThe calculation formula is E according to the correlation and the timeliness of the required datak,i(t)＝αkFk+(1-αk)f(t-tk,i) Wherein αkIs a constant between 0 and 1, t is the current time, tk,iThe time at which the packet is generated for node i,
Figure GDA0002451433860000071
Fkthe correlation between the data packet K sensed by the node i and the data required by the application is shown, X is the physical signal of the data required by the application, and K is the physical signal sensed by the node. In EIP
Figure GDA0002451433860000072
The attenuation coefficient η of the attribute that represents that the information is valuable for decision only for a certain timekWith FkAnd are different. Because different nodes in the cluster generate the same data packet when sensing the same event, the correlation influences the value of the data packet, the ICN represents the concentration degree of the data packet k in the cluster, and the calculation formula is
Figure GDA0002451433860000073
β thereinkIs a constant between 0 and 1, TicIndicating the time required for the packet k to travel from node i to the cluster head, and n indicating the number of perceived identical packets k. t is tendIndicates the time point, t, of the last transmission of the packet to the cluster headstartIndicating the point in time at which the packet was first collected by the cluster head. The time information in the ICN is obtained by historical information, and the nodes in the cluster acquire related information through regular command interaction with the cluster head. The information value VOI is thus defined as V ═ gammakEk+(1-γk)Ik，γkIs constant between 0 and 1.
Step two, mobile Sink position determination and space region division of three-dimensional underwater environment
In a three-dimensional underwater environment, multiple AUVs are used for executing an underwater data collection task, the energy and flow load balance of the multiple AUVs is kept, and meanwhile, the AUVs are ensured to send collected data to a Sink in time. Inventive passing zoneThe domain division algorithm comprises the steps of firstly determining the vertical navigation position of the mobile Sink, then integrating the attributes such as the number of nodes, the density of the nodes and the depth of the nodes, and dividing an underwater space region into a plurality of sub-regions. Each sub-region is composed of a plurality of clusters, and the same AUV is responsible for data collection tasks. The region partitioning algorithm first maps clusters in the network to particles on a two-dimensional plane with the locations of the cluster heads, as shown in fig. 2. Using these particles to make two-dimensional Voronoi diagram, firstly, using formula
Figure GDA0002451433860000074
CHiE CH, i ═ 1,2,3.. n) to establish the specific position of the mobile Sink vertical movement. The Thiessen polygon in which each particle is located is called a mass region, where dSink,centerIs the distance between the center of the mass region where the sink is located and the center of the two-dimensional plane, di,centerNerbor (Sink) represents the number of the geological regions adjacent to the geological region where the Sink is moving, Nerbor (CH) being the distance between the geological region i and the center of the two-dimensional planei) Indicates the number of adjacent prime regions in the prime region, and n is the number of total prime regions. For a prime area of the Voronoi diagram, various attributes such as the number of nodes, the density of nodes in a cluster, the average depth of the nodes in the cluster and the like exist, and the attribute of each prime area i is expressed as P in a vector formi＝[pi1,pi2,…,pim]TFor the m-th attribute, the similarity of the sub-region s after the i-th prime region and the plurality of prime regions are combined is defined as fi(psm,pim) It can be obtained by Euclidean distance. With Sim (P)s,Pi) The similarity of multiple attributes is represented by the formula
Figure GDA0002451433860000081
λ1+λ2+···+λm1 is ═ 1; the growth criterion for calculating the attribute m for all the material regions in the Voronoi diagram is
Figure GDA0002451433860000082
Wherein
Figure GDA0002451433860000083
Figure GDA0002451433860000084
Adjustment parameter, p, for attribute jjmaxAnd pjminRepresenting the maximum and minimum values of the property j in all sub-regions. As shown in fig. 2, a plurality of quality areas are randomly selected to perform random synchronous growth to adjacent quality areas, and in the case of satisfying W, when each quality area is located in one sub-area s, the area division algorithm is stopped.
Step three, planning the path of the AUV in the sub-area
After the subareas are divided, each AUV is responsible for a data collection task of one subarea, and data collection of a plurality of clusters in the subareas is completed within a certain time. And setting a cluster which is closest to the moving Sink vertical navigation area in the sub-area as an AUV primary collection path termination access cluster, and randomly selecting an initial access cluster. The number and the required time of AUV access clusters are controlled by setting a dynamic reward function of a Q-learning algorithm, a sampling path is learned, and an access path with the highest return is dynamically established. In the path planning stage, an incentive matrix for AUV transfer between clusters is established, and in each sub-area, the incentive matrix for AUV transfer between clusters is set as
Figure GDA0002451433860000085
The Q matrix is initialized to 0. The choice of reward function will determine the rate and extent of convergence of the Q-learning algorithm, and in order for the AUV to traverse all clusters in a sub-region without repeatedly accessing the clusters, the reward for each cluster is set as follows
Figure GDA0002451433860000091
Wherein α is a constant between 0 and 1, C is a reward amplitude adjustment parameter, f is the number of accesses of the jth cluster, VjTotal VOI, d obtained for AUV going to cluster jij(f) Is the linear distance of cluster i and cluster j. The AUV learns in a random path mode until a stable Q matrix is obtained, and the calculation formula of the Q matrix is Q (s, a) ═ R + gamma maxa'Q (s ', a'); gamma is regulating parameter, s is current state, and a is current stateThe action performed, s ', is the next state into which action a is performed in state s, and a ' is the action performed in state s '. The AUV learns using the algorithm described above, each experience being equivalent to one training. During each training, the agent explores the environment. The purpose of training is to enhance the AUV brain, the more training results will result in a better Q matrix, the AUV will find the route with the highest return to the terminating access cluster, and send the data to the mobile Sink, as shown in fig. 4.
Step four, multiple AUV dynamic competition processing emergency tasks
When a cluster head of a certain sub-region has emergency tasks such as data overflow and the like, the cluster head sends data emergency collection request information to the mobile Sink in an inter-cluster multi-hop mode, the mobile Sink evaluates the current state of the AUV, the multi-AUV competes in aspects such as the distance from the emergency task, the residual energy and the task execution state, the competing and winning AUV obtains the processing right of the emergency task and immediately suspends the current data collection task, the AUV is switched to the emergency task processing state and goes to the emergency task place, and the AUV returns to an atomic region after the AUV is processed, and continues to complete the residual data collection task. When a plurality of emergency events such as data overflow occur, the mobile Sink evaluates the plurality of events and preferentially processes data collection request information with high data overflow degree. A plurality of AUVs compete for the processing right of the current emergency, and the mobile Sink calculates the competitive win value of the AUVs in the network according to the following formula
Figure GDA0002451433860000092
Wherein α is a constant between 0 and 1, M is a boundary of the network, d is a distance between the AUV and the emergency site, E is a remaining capacity of the AUV, E is an energy consumed by the AUV for a unit distance traveled, and v is an average velocity of the AUV
Figure GDA0002451433860000093
Is calculated by the formula
Figure GDA0002451433860000094
Wherein, tnowTo representCurrent time, tiIndicating the time required by the AUV to handle the current emergency, dnext(i) Is the linear distance from the current emergency event to an event. The winning AUV can obtain processing power and go immediately to the relevant location for data collection.
In summary, the following steps:
the invention discloses a VOI-based multi-AUV (autonomous Underwater vehicle) efficient data collection method in an underwater wireless sensor network. And then, determining the vertical moving position of the Sink by using a Voronoi diagram and dividing an underwater space region according to multiple attribute characteristics of different clusters in a three-dimensional underwater environment, so that the load balance of energy and flow is achieved when the AUV executes an underwater data collection task of a sub-region. The size of the VOI in the cluster and the distance between the clusters are main influence factors of the path planning of the sub-area of the AUV, so that the method controls the number of AUV access clusters and the required time by setting a dynamic reward function of a Q-learning algorithm, can adapt to the requirements of different applications, and accelerates the data collection efficiency. And finally, when emergency events such as data overflow in the cluster occur, the multiple AUVs obtain event processing permission through a dynamic competition mechanism, and after the event processing is finished, the AUVs return to the atomic region to continue data collection.
The foregoing shows and describes the general principles, essential features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims (5)

1. A VOI-based multi-AUV efficient data collection method in an underwater wireless sensor network is characterized by comprising the following steps:
(1) high VOI packet priority to cluster head transmission
The method comprises the following steps that nodes of an underwater wireless sensor network are deployed randomly and clustered, after an in-cluster source node senses data, the information value VOI of a data packet is calculated at first, a next-hop node compares the VOI of the data packet stored by the next-hop node with the VOI of a received data packet, and the data packet with a high VOI is forwarded preferentially; preferentially transmitting the data packet with higher VOI in the cluster to a cluster head, and waiting for AUV collection;
(2) mobile Sink position determination and space region division of three-dimensional underwater environment
In a three-dimensional underwater environment, multiple AUVs are used for executing an underwater data collection task, the energy and flow load balance of the multiple AUVs is kept, and meanwhile, the AUVs are ensured to send collected data to a Sink in time; through a region division algorithm, firstly determining the vertical navigation position of the mobile Sink, then integrating the attributes of the number of nodes, the density of the nodes and the depth of the nodes, and dividing an underwater space region into a plurality of sub-regions; each subarea consists of a plurality of clusters, and the same AUV is responsible for data collection tasks;
(3) path planning of AUV in sub-area
After the subareas are divided, each AUV is responsible for a data collection task of one subarea, and data collection of a plurality of clusters in the subareas is completed within a certain time; setting a cluster which is closest to a vertical navigation area of the mobile Sink in the sub-area as an access termination cluster of an AUV (autonomous Underwater vehicle) primary collection path, and randomly selecting an initial access cluster; the number and the required time of AUV access clusters are controlled by setting a dynamic reward function of a Q-learning algorithm, a sampling path is learned, and an access path with the highest return is dynamically established;
(4) Multi-AUV dynamic competition processing emergency task
When a cluster head of a certain sub-region has an emergency task, the cluster head sends data emergency collection request information to a mobile Sink in an inter-cluster multi-hop mode, the mobile Sink evaluates the current state of the AUV, the multi-AUV competes in the aspects of the distance from the emergency task, the residual energy and the task execution state, the competing and winning AUV obtains the processing right of the emergency task and immediately suspends the current data collection task, the AUV is switched to the emergency task processing state and goes to the emergency task place, and the AUV returns to an atomic region after the processing is completed, and continues to complete the residual data collection task.
2. A multi-AUV efficient data collection method based on VOI in an underwater wireless sensor network according to claim 1, characterized in that: the information value VOI of the data packet in the step (1) consists of an event importance degree EIP and an information concentration degree ICN; the EIP describes the correlation and timeliness of the data collected by the nodes and the data required by the application, and the calculation formula is Ek,i(t)＝αkFk+(1-αk)f(t-tk,i) Wherein αkA constant between 0 and 1, t is the current time,
Figure FDA0002459955470000021
Fkthe correlation between a data packet K sensed by a node i and data required by an application is represented, X represents a physical signal of the data required by the application, and K is the physical signal sensed by the node; in EIP
Figure FDA0002459955470000022
The attenuation coefficient η of the attribute that represents that the information is valuable for decision only for a certain timekWith FkDifferent from each other; ICN represents the concentration degree of the data packet k in the cluster, and the calculation formula is
Figure FDA0002459955470000023
TicThe time required for the data packet k to be transmitted from the node i to the cluster head is shown, and n represents the number of the same data packets k; t is tendIndicates the time point, t, of the last transmission of the packet to the cluster headstartIndicating a point in time when the packet was first collected by the cluster head; time information in the ICN is obtained by historical information, and the nodes in the cluster acquire related information through regular command interaction with the cluster head; the information value VOI is defined as V ═ γkEk+(1-γk)Ik，γkIs constant between 0 and 1.
3. A VOI-based multi-AUV high-efficiency data collection method in an underwater wireless sensor network according to claim 1, characterized in that: the region division algorithm in the step (2) firstly maps clusters in the network into particles on a two-dimensional plane according to the positions of cluster heads; using these particles to make two-dimensional Voronoi diagram, firstly, using formula
Figure FDA0002459955470000024
CHiDetermining the specific position of the vertical movement of the mobile Sink by epsilon CH, i-n (1,2,3.. n); the Thiessen polygon in which each particle is located is called a mass region, where dSink,centerIs the distance between the center of the mass region where the sink is located and the center of the two-dimensional plane, di,centerDistance of the texture region i from the center of the two-dimensional plane, Nerbor (CH)i) Representing the number of adjacent prime zones of the prime zone; for a prime area of the Voronoi diagram, various attributes such as the number of nodes, the density of nodes in a cluster, the average depth of the nodes in the cluster and the like exist, and the attribute of each prime area i is expressed as P in a vector formi＝[pi1,pi2,…,pim]TFor the m-th attribute, the similarity of the sub-region s after the i-th prime region and the plurality of prime regions are combined is defined as fi(psm,pim) By Fs,iA similarity matrix F representing a sub-region s obtained by combining the ith prime region with a plurality of prime regionss,i＝[fi(ps1,pi1),fi(ps2,pi2),…,fi(psm,pim)]TUsing Sim (P)s,Pi) The similarity of multiple attributes is represented by the formula
Figure FDA0002459955470000031
The growth criterion for calculating the attribute m for all the material regions in the Voronoi diagram is
Figure FDA0002459955470000032
Wherein
Figure FDA0002459955470000033
Randomly selecting a plurality of prime areas to carry out random synchronous growth on adjacent prime areas, and stopping the area division algorithm when each prime area is positioned in one sub-area s under the condition of meeting W.
4. A VOI-based multi-AUV high-efficiency data collection method in an underwater wireless sensor network according to claim 1, characterized in that: in the path planning stage in the step (3), an incentive matrix for AUV to transfer among clusters is established, and in each sub-area, the incentive matrix for AUV to transfer among clusters is set as
Figure FDA0002459955470000034
Initializing a Q matrix to be 0; the prize for each cluster is set as follows
Figure FDA0002459955470000035
Wherein α is a constant between 0 and 1, C is a reward amplitude adjustment parameter, f is the number of accesses of the jth cluster, VjTotal VOI, d obtained for AUV going to cluster jij(f) Is the linear distance of cluster i and cluster j; the AUV learns in a random path mode until a stable Q matrix is obtained, and the calculation formula of the Q matrix is Q (s, a) ═ R + gamma maxa'Q (s ', a'), gamma is a regulating parameter, and s is the current state; AUV learns by using the above algorithm, each experience is equivalent to one training; in each training, the agent explores the environment; and the AUV finds a route with the highest return to the terminal access cluster, and sends data to the mobile Sink.
5. A VOI-based multi-AUV high-efficiency data collection method in an underwater wireless sensor network according to claim 1, characterized in that: when a plurality of emergency events are generated in the step (4), the mobile Sink evaluates the plurality of events and preferentially processes the data collection request information with high data overflow degree; a plurality of AUVs compete for the processing right of the current emergency, and the mobile Sink calculates the competition for the AUVs in the network according to the following formulaWin value
Figure FDA0002459955470000036
Wherein α is a constant between 0 and 1, M is a boundary of the network, d is a distance between the AUV and the emergency site, E is a remaining capacity of the AUV, E is an energy consumed by the AUV for navigating the unit distance, and v is an average speed of the AUV, and when the AUV executes an emergency processing task, a calculation formula of a competition win value phi is
Figure FDA0002459955470000041
Wherein, tiIndicating the time required by the AUV to handle the current emergency, dnext(i) A linear distance from a current emergency event to an event; the winning AUV can obtain processing power and go immediately to the relevant location for data collection.
CN201811123847.3A 2018-09-26 2018-09-26 VOI-based multi-AUV (autonomous Underwater vehicle) efficient data collection method in underwater wireless sensor network Active CN109275099B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811123847.3A CN109275099B (en) 2018-09-26 2018-09-26 VOI-based multi-AUV (autonomous Underwater vehicle) efficient data collection method in underwater wireless sensor network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811123847.3A CN109275099B (en) 2018-09-26 2018-09-26 VOI-based multi-AUV (autonomous Underwater vehicle) efficient data collection method in underwater wireless sensor network

Publications (2)

Publication Number Publication Date
CN109275099A CN109275099A (en) 2019-01-25
CN109275099B true CN109275099B (en) 2020-06-09

Family

ID=65198151

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811123847.3A Active CN109275099B (en) 2018-09-26 2018-09-26 VOI-based multi-AUV (autonomous Underwater vehicle) efficient data collection method in underwater wireless sensor network

Country Status (1)

Country Link
CN (1) CN109275099B (en)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110366226B (en) * 2019-06-06 2021-06-15 中国船舶工业***工程研究院 Underwater wireless sensor network routing method based on reinforcement learning
CN110430547B (en) * 2019-07-24 2022-07-15 河海大学常州校区 Q-learning-based multi-AUV cooperative data collection method in UASNs
CN110933680B (en) * 2019-10-31 2021-10-15 中国矿业大学 Underwater acoustic-magnetic heterogeneous network rapid networking method based on sounding-communication integration
CN111307153B (en) * 2020-02-26 2023-03-21 河海大学 Multi-AUV task allocation and path planning method based on hexagonal grid map
CN111542020B (en) * 2020-05-06 2023-06-13 河海大学常州校区 Multi-AUV cooperative data collection method based on region division in underwater acoustic sensor network
CN115855226B (en) * 2023-02-24 2023-05-30 青岛科技大学 Multi-AUV cooperative underwater data acquisition method based on DQN and matrix completion
CN116819025B (en) * 2023-07-03 2024-01-23 中国水利水电科学研究院 Water quality monitoring system and method based on Internet of things

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107105466A (en) * 2017-03-14 2017-08-29 南京邮电大学 A kind of mobile Sink methods of data capture based on enhancing learning algorithm
CN107276684A (en) * 2017-07-19 2017-10-20 河海大学常州校区 Method of data capture based on AUV position predictions in underwater sensor network
CN107994948A (en) * 2017-12-30 2018-05-04 山东省科学院海洋仪器仪表研究所 A kind of mobile Sink paths planning methods for underwater heterogeneous sensor network
CN108011981A (en) * 2018-01-11 2018-05-08 河海大学常州校区 High Availabitity method of data capture based on more AUV in underwater sensor network

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10081416B2 (en) * 2016-11-07 2018-09-25 Raytheon Company Autonomous underwater vehicle for transport of payloads

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107105466A (en) * 2017-03-14 2017-08-29 南京邮电大学 A kind of mobile Sink methods of data capture based on enhancing learning algorithm
CN107276684A (en) * 2017-07-19 2017-10-20 河海大学常州校区 Method of data capture based on AUV position predictions in underwater sensor network
CN107994948A (en) * 2017-12-30 2018-05-04 山东省科学院海洋仪器仪表研究所 A kind of mobile Sink paths planning methods for underwater heterogeneous sensor network
CN108011981A (en) * 2018-01-11 2018-05-08 河海大学常州校区 High Availabitity method of data capture based on more AUV in underwater sensor network

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Energy-Efficient Data Collection Over AUV-Assisted;Jing Yan等;《IEEE SYSTEMS JOURNAL》;20180118;第12卷(第4期);全文 *
UWSN中基于信息价值的数据传输方案;白钢华等;《网络与信息安全》;20150206;第41卷(第2期);全文 *

Also Published As

Publication number Publication date
CN109275099A (en) 2019-01-25

Similar Documents

Publication Publication Date Title
CN109275099B (en) VOI-based multi-AUV (autonomous Underwater vehicle) efficient data collection method in underwater wireless sensor network
Han et al. District partition-based data collection algorithm with event dynamic competition in underwater acoustic sensor networks
Hu et al. Cooperative internet of UAVs: Distributed trajectory design by multi-agent deep reinforcement learning
Han et al. Multi-AUV collaborative data collection algorithm based on Q-learning in underwater acoustic sensor networks
Wang et al. Hierarchical deployment optimization for wireless sensor networks
Lin et al. A path planning scheme for AUV flock-based Internet-of-Underwater-Things systems to enable transparent and smart ocean
CN108684005B (en) SOM-based multi-AUV efficient data collection method in underwater sensor network
Su et al. Localization and data collection in AUV-aided underwater sensor networks: Challenges and opportunities
CN111542020B (en) Multi-AUV cooperative data collection method based on region division in underwater acoustic sensor network
Liu et al. AUV-aided hybrid data collection scheme based on value of information for Internet of Underwater Things
CN110430547B (en) Q-learning-based multi-AUV cooperative data collection method in UASNs
CN114625151A (en) Underwater robot obstacle avoidance path planning method based on reinforcement learning
Han et al. Probabilistic neighborhood location-point covering set-based data collection algorithm with obstacle avoidance for three-dimensional underwater acoustic sensor networks
Gupta et al. HEPSO: an efficient sensor node redeployment strategy based on hybrid optimization algorithm in UWASN
Xia et al. AI-driven and MEC-empowered confident information coverage hole recovery in 6G-enabled IoT
Han et al. AUV-aided data importance based scheme for protecting location privacy in smart ocean
CN112866911A (en) Underwater data collection method assisted by autonomous underwater vehicle based on Q learning
Wang et al. Task scheduling for distributed AUV network target hunting and searching: An energy-efficient AoI-aware DMAPPO approach
He et al. State prediction-based data collection algorithm in underwater acoustic sensor networks
Lin et al. Fast node clustering based on an improved birch algorithm for data collection towards software-defined underwater acoustic sensor networks
Jiang et al. Underwater searching and multi-round data collection via AUV swarms: An energy-efficient AoI-aware MAPPO approach
CN116501094B (en) Unmanned aerial vehicle cluster control method based on self-organizing model
Han et al. Online Path Planning for AUV in Dynamic Ocean Scenarios: A Lightweight Neural Dynamics Network Approach
Zhao et al. Adaptive multi-UAV trajectory planning leveraging digital twin technology for urban IIoT applications
CN115465125A (en) AUV cluster underwater energy rescue method based on wireless charging technology

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant