A method for computing offloading and resource allocation in ocean monitoring networks

Abstract

The present invention discloses a method for computing offloading and resource allocation of an ocean monitoring network in the field of communication technology, the method comprising: constructing a weighted sum optimization problem of delay and energy consumption with the goal of minimizing the weighted sum of delay and energy consumption of an unmanned ship, jointly optimizing the transmission power of an underwater sensor node USN, the transmission power of an unmanned ship, bandwidth allocation, and offloading ratio; solving the weighted sum optimization problem of delay and energy consumption to obtain a computing offloading and resource allocation scheme for an ocean monitoring network; and performing computing offloading and resource allocation for an ocean monitoring network according to the computing offloading and resource allocation scheme for an ocean monitoring network. The method for computing offloading and resource allocation for an ocean monitoring network provided by the present invention takes the minimum sum of delay and energy consumption of an unmanned ship as the goal, and by jointly optimizing the transmission power of an underwater sensor, the transmission power of an unmanned ship, bandwidth allocation, and offloading ratio, the sum of delay and energy consumption can be significantly reduced, bringing a better experience to users.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a method for computing unloading and resource allocation in an ocean monitoring network. Background Art

[0002] Ocean Monitoring Networks (OMNs) are an important part of the smart ocean information network, which can provide ubiquitous Internet services for new computing-intensive maritime applications such as marine environment monitoring, marine resource exploration, disaster prevention, and assisted navigation. With the massive growth of various marine intelligent devices and services, users' demand for low latency and low energy consumption is difficult to meet. By introducing Mobile Edge Computing (MEC) technology into the ocean monitoring sensor network and sinking the data center to the edge of the network, local data processing and storage services can be provided.

[0003] In recent years, more and more people have begun to pay attention to drone-assisted ocean computing offload networks. As a mobile computing server, drones can be quickly deployed and flexibly offloaded in ocean communication networks, thereby improving the computing efficiency of ocean monitoring networks. At present, the study of applying drones to ocean monitoring scenarios and considering underwater sensor network data collection has just begun. In order to improve the experience of sea users, it is necessary to study the joint optimization of sea user latency and energy consumption. Summary of the invention

[0004] The purpose of the present invention is to overcome the deficiencies in the prior art and to provide a method for computing offloading and resource allocation in an ocean monitoring network, which can jointly optimize the latency and energy consumption of sea surface users.

[0005] To achieve the above object, the present invention is implemented by adopting the following technical solutions:

[0006] The present invention provides a method for computing offloading and resource allocation in an ocean monitoring network, the method comprising:

[0007] The weighted sum optimization problem of delay and energy consumption is constructed with the goal of minimizing the weighted sum of delay and energy consumption of the unmanned ship, and the transmission power of the USN, the transmission power of the unmanned ship, the bandwidth allocation and the unloading ratio are jointly optimized;

[0008] Solve the time delay and energy consumption weighted sum optimization problem to obtain the ocean monitoring network computing offloading and resource allocation solution;

[0009] According to the ocean monitoring network computing offloading and resource allocation scheme, ocean monitoring network computing offloading and resource allocation are performed.

[0010] Furthermore, the task completion delay of the kth unmanned ship is ;

[0011] in, is the data collection delay, I is the uploaded data set; for The upload data delay is , Upload data for the i-th unmanned ship of the k-th unmanned ship;

[0012] Data processing delay ,in, is the local delay of the kth unmanned ship, and ;

[0013] is the unloading ratio of the kth unmanned ship, is the number of CPU execution cycles required for the kth unmanned ship data, is the computing power of the kth unmanned ship; is the processing delay of the kth unmanned ship’s data offloading to the UAV, and ,in is the total number of bits of the kth unmanned ship data, is the transmission rate of the kth unmanned ship, The computing power allocated to the kth unmanned ship by the UAV;

[0014] Energy consumption of the kth unmanned ship ;

[0015] in, is the local computing energy consumption of the kth unmanned ship, and , is the energy consumption coefficient of the unmanned ship;

[0016] is the transmission energy consumption of the kth unmanned ship, and , is the transmission power of the kth unmanned ship;

[0017] Then the weighted sum formula of the time delay and energy consumption of the kth unmanned ship is:

[0018]

[0019] in, The weighting factor of the user task delay reflects the user's emphasis on delay; represents the task completion delay of the kth unmanned ship; The weighting factor representing the energy consumption of the user's task reflects the importance the user places on energy consumption; is the energy consumption of the kth unmanned ship; The USN set that sends data to the kth unmanned ship; The amount of data collected for the i-th USN; is the transmission rate of data uploaded from the i-th USN to the unmanned ship k; is the unloading ratio of the kth unmanned ship; The number of CPU execution cycles required for the kth unmanned ship data; is the computing power of the kth unmanned ship; is the total number of bits of the kth unmanned ship data; is the transmission rate of the kth unmanned ship; The computing power allocated to the kth unmanned ship by the UAV; is the energy consumption coefficient of the unmanned ship; is the transmission power of the kth unmanned ship;

[0020] The optimization problem expression of the weighted sum of delay and energy consumption is as follows:

[0021]

[0022] Among them, problem Z1 is an optimization problem, and C1.1-C1.5 are constraints of optimization problem Z1;

[0023] is the transmission power of the i-th USN uploading data to the k-th unmanned ship; Q represents the total weighted sum of the delay and energy consumption of the unmanned ship; Assemble for the unmanned ship; is the upper limit of the transmission power of the i-th USN; is the bandwidth allocated to the kth unmanned ship, The upper limit of bandwidth resources allocated to all unmanned ships; is the upper limit of the transmission power of the kth unmanned ship.

[0024] Furthermore, the solution of the time delay and energy consumption weighted sum optimization problem to obtain the ocean monitoring network computing offloading and resource allocation scheme includes:

[0025] The delay and energy consumption weighted sum optimization problem is decoupled into the optimization problem of the data collection stage and the optimization problem of the data processing stage. The optimization problem of the data collection stage is constructed with the goal of minimizing the data collection delay of the unmanned ship, and the optimization variable is the transmission power of the USN. ; The optimization problem of the data processing stage is constructed with the goal of minimizing the weighted sum of the unmanned ship data processing delay and energy consumption. The optimization variable is the transmission power of the unmanned ship. , uninstall ratio and bandwidth ;

[0026] Solve the optimization problem in the data collection phase;

[0027] The optimization problem in the data processing stage is decomposed into the unmanned ship transmission power optimization sub-problem, the bandwidth allocation optimization sub-problem and the unloading ratio decision optimization sub-problem, and each optimization sub-problem is solved, and then the solution to the optimization problem in the data processing stage is obtained through alternating iterations;

[0028] According to the solutions of the optimization problems in the data collection phase and the data processing phase, the solution of the weighted optimization problem of delay and energy consumption is obtained, and then the computational offloading and resource allocation scheme of the ocean monitoring network is obtained.

[0029] Furthermore, the expression of the optimization problem in the data collection stage is:

[0030]

[0031] Among them, problem Z2 is the optimization problem in the data collection stage, and C2.1 is the constraint condition of problem Z2;

[0032] in, is the data collection delay of the kth unmanned ship;

[0033] and / or,

[0034] The expression of the optimization problem in the data processing stage is:

[0035]

[0036] Among them, problem Z3 is the optimization problem in the data processing stage, and C3.1-C3.4 are the constraints of problem Z3; is the processing delay of the kth unmanned ship.

[0037] Furthermore, solving the optimization problem in the data collection stage includes:

[0038] S2.2.1. The optimization problem in the data collection phase is transformed into the maximum delay minimization problem of the underwater sensor. The position vector of a particle represents a set of transmission power solutions. The maximum delay minimization problem of the underwater sensor is expressed as:

[0039]

[0040] Among them, problem Z4 is the maximum delay minimization problem of underwater sensors, and its corresponding constraint condition is C4.1; for Upload data delay; is the i-th USN within the coverage area of ​​the k-th unmanned ship;

[0041] S2.2.2. Initialize the number of individuals N in the population and the particle positions , particle speed ;

[0042] S2.2.3, according to the formula Calculate the fitness of individuals in the initial population, and obtain the individual historical optimal solution and the global historical optimal solution; among them, is a constant, ensuring that the fitness function value is positive; parameter is the penalty factor, which is a positive number. When the USN collects data and uploads it to the unmanned ship, if the maximum transmission power is exceeded, it will be punished by the penalty function, and the fitness of the particle will increase accordingly. Otherwise, the penalty function is 0.

[0043] S2.2.4. Update the position and velocity of the particles according to the following two formulas to generate a new population:

[0044]

[0045]

[0046] in, is the inertia weight; and is the learning factor; and is a random number between 0 and 1; is the historical optimal position of particle n; is the historical optimal position of the population;

[0047] S2.2.5. Calculate the fitness of each particle in the new population and update the individual historical optimal solution and the global historical optimal solution;

[0048] S2.2.6. Repeat S2.2.4 and S2.2.5 until the method converges and the solution to the optimization problem in the data collection phase is obtained.

[0049] Furthermore, the expression of the unmanned ship transmission power optimization sub-problem is:

[0050]

[0051] Among them, Z5 is the unmanned ship transmission power optimization sub-problem, and its corresponding constraint condition is C5.1; is the transmission energy consumption of uploading data from the kth unmanned ship to the UAV;

[0052] and / or,

[0053] The expression of the bandwidth allocation optimization subproblem is:

[0054]

[0055] Among them, Z6 is the bandwidth allocation optimization sub-problem, and its corresponding constraints are C6.1 and C6.2;

[0056] and / or,

[0057] The expression of the unloading ratio decision optimization sub-problem is:

[0058]

[0059] Among them, Z7 is the bandwidth allocation optimization sub-problem, and its corresponding constraint condition is C7.1.

[0060] Further, solving the transmit power optimization sub-problem includes:

[0061] S2.3.1.1. Under the current unloading ratio decision and signal-to-interference-noise ratio, calculate the minimum data unloading delay of the kth unmanned ship. , in , is the maximum transmission rate of the kth unmanned ship;

[0062] S2.3.1.2. Order ,in, is the processing delay and energy consumption of the kth unmanned ship, if Greater than Then execute S4.1.3, if Less than Then execute S4.1.4;

[0063] S2.3.1.3. Calculation The maximum transmission power of the kth unmanned ship If the partial derivative is greater than 0, there is a point where the partial derivative is 0. , The minimum value of , otherwise the minimum value is When obtained;

[0064] S2.3.1.4, let the local delay of the kth unmanned ship be equal to the processing delay of the data unloaded from the kth unmanned ship to the drone, and we get ,in, The local delay of the kth unmanned ship is equal to the power size of the processing delay unloaded to the UAV, , is the background noise power at the kth unmanned ship, is the channel gain of the kth unmanned ship, and is calculated exist The partial derivative at , if exist If the partial derivative at a point is greater than 0, then there is a point where the partial derivative is 0 , The minimum value of , otherwise the minimum value is Obtained when.

[0065] Furthermore, solving the bandwidth allocation optimization sub-problem includes:

[0066] S2.3.2.1. Input the current UAV launch power and unloading strategy, and prove that problem Z6 is a convex optimization problem;

[0067] S2.3.2.2. Under the condition of limited total bandwidth resources, use the CVX toolbox provided by MATLAB to solve problem Z6.

[0068] Furthermore, solving the unloading ratio decision optimization sub-problem includes:

[0069] S2.3.3.1. Number of individuals in the input population , population evolution times , mutation probability , crossover probability and the number of individuals eliminated in each round ;

[0070] S2.3.3.2. Initialize the gene values ​​of the individuals in the population and calculate the initial fitness function according to the following formula:

[0071]

[0072] in, Is a constant, ensuring that the fitness function value is positive;

[0073] S2.3.3.3. Perform mutation operation on each individual in the population, and generate a random number between 0 and 1 for each gene in the individual ,like Less than the mutation probability , then the gene is mutated; the mutation position is recorded as , the variation expression is ,in, Indicates the current number of evolutions. Represents a random number between [0,1];

[0074] S2.3.3.4. Perform a crossover operation on each individual in the population and randomly generate a random number between 0 and 1 , if the random number Less than the crossover probability , then perform a crossover operation on the individual and randomly select the gene value on the individual, and record its position as , and then randomly select a position in the population as The genes that ;

[0075] S2.3.3.5. Recalculate individual fitness, find a new optimal individual and replace the stored optimal individual, and replace the previous individual with the minimum current fitness function. Individuals are replaced by the best individuals;

[0076] S2.3.3.6. Repeat S2.3.3.3, S2.3.3.4 and S2.3.3.5 until the number of evolutions is greater than or equal to the number of population evolutions. .

[0077] In a second aspect, the present invention provides a computer-readable storage medium having a computer program / instruction stored thereon, which, when executed by a processor, implements the steps of the ocean monitoring network computational offloading and resource allocation method.

[0078] Compared with the prior art, the present invention has the following beneficial effects:

[0079] The ocean monitoring network computing offloading and resource allocation method provided by the present invention constructs a UAV-assisted ocean monitoring network multi-address computing offloading model, which consists of two stages: data acquisition and data processing. A joint optimization problem with the weighted sum minimization of surface user task completion delay and energy consumption is established, and the optimization problem is decoupled into two stages: data acquisition stage and data processing stage, which are solved in sequence.

[0080] In the data collection stage, considering the large attenuation of the underwater acoustic channel, non-orthogonal multiple access (NOMA) is used to upload data to improve the utilization of the acoustic channel. Since the unmanned ship processes data only after collecting all sensor data, the problem of minimizing the data collection delay is transformed into the problem of minimizing the maximum delay of the underwater sensor. When communication resources are limited, we control the transmission power to ensure that the upload time of each sensor data is equal, thereby minimizing the data collection delay of the unmanned ship;

[0081] In the data processing stage, the unmanned ship offloads part of the task to the drone through frequency division multiple access (FDMA), and the unmanned ship and the drone jointly complete the data processing task. Since the optimization problem in the data processing stage is still a non-convex problem, it is further decomposed into three optimization sub-problems: the transmission power of the unmanned ship, the offloading ratio decision, and the bandwidth allocation. Among them, the gradient descent method is used to solve the optimal transmission power of the unmanned ship; considering that there is a maximum value function in the objective function of the offloading ratio decision sub-problem, it cannot be differentiated, and the genetic algorithm is used to solve it; since the bandwidth allocation sub-problem is a convex optimization problem, CVX is used to solve it, thereby solving the joint optimization problem with the weighted sum of the surface user task completion delay and energy consumption minimization as the goal, and jointly optimizing the delay and energy consumption of sea surface users. BRIEF DESCRIPTION OF THE DRAWINGS

[0082] Figure 1 It is a flow chart of a method for offloading computation and allocating resources in an ocean monitoring network provided by an embodiment of the present invention;

[0083] Figure 2 A schematic diagram of a system model corresponding to the ocean monitoring network computing offloading and resource allocation method provided in an embodiment of the present invention;

[0084] Figure 3 A schematic diagram of performance simulation of the ocean monitoring network computing offloading and resource allocation method provided by an embodiment of the present invention under different numbers of users;

[0085] Figure 4 A schematic diagram of performance simulation of the ocean monitoring network computing offloading and resource allocation method provided by an embodiment of the present invention under different cache capacities;

[0086] Figure 5 A schematic diagram of simulating the total system delay performance corresponding to changing the size of the recommendation list provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0087] The present invention will be further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0088] In the present invention, the underwater sensor node (USN) transmits the collected data to the unmanned surface vehicle (USV) on the water by non-orthogonal multiple access. After receiving the data sent by the USN, the unmanned surface vehicle unloads part of the data to the drone by frequency division multiple access. The unmanned aerial vehicles (UAV) and the unmanned ship collaborate to complete the data processing task.

[0089] Example 1

[0090] This embodiment provides a method for offloading computation and allocating resources in an ocean monitoring network. Its application model is as follows: Figure 2 As shown, for each unmanned ship in the target area, when each unmanned ship starts to collect data sent by USN, the following steps are performed:

[0091] Step A: construct an optimization problem with the goal of minimizing the weighted sum of the delay and energy consumption of the unmanned ship, jointly optimize the transmission power of the USN, the transmission power of the unmanned ship, the bandwidth allocation and the unloading ratio, and then enter step B;

[0092] Step B: Decouple the delay and energy consumption weighted sum optimization problem into two stages: data collection stage and data processing stage, and solve them in turn. The data collection stage optimization problem is constructed with the goal of minimizing the data collection delay of the unmanned ship, and the optimization variable is the transmission power of the USN. ; The data processing phase optimization problem is constructed with the goal of minimizing the weighted sum of the unmanned ship data processing delay and energy consumption, and the optimization variable is the transmission power of the unmanned ship , uninstall ratio and bandwidth , then proceed to step C;

[0093] Step C: transform the optimization problem in the data collection phase into the maximum delay minimization problem of the underwater sensor. Under the condition of limited communication resources, the data upload delay of each sensor is equalized by controlling the transmission power, and then proceed to step D.

[0094] Step D: Decompose the optimization problem in the data processing stage into three optimization sub-problems: the transmission power of the unmanned ship, the unloading ratio decision, and the bandwidth allocation. Use the gradient descent method, genetic algorithm, and CVX to solve them respectively, and then obtain the solution to the optimization problem in the data processing stage through alternating iterations.

[0095] The application model of the ocean monitoring network computing offloading and resource allocation method provided in this embodiment assumes that the position of each USN is random, and under the constraints of maximum transmission power and fixed unmanned ship channel bandwidth, the delay and energy consumption of the unmanned ship are minimized by jointly optimizing the transmission power of the USN, the transmission power of the unmanned ship, bandwidth allocation and offloading ratio.

[0096] The application model of the ocean monitoring network computing offloading and resource allocation method provided in this embodiment is composed of M USNs, K unmanned watercrafts and a single UAV. The set of USNs is denoted as , the set of unmanned ship users is , divide the M USNs into K clusters, each cluster of USNs corresponds to an unmanned ship, denoted as The USN is deployed on the seabed to detect the marine environment and send the detection data to the user unmanned ship on the sea surface. The USN set of the kth unmanned ship is , the kth unmanned ship Received collection After receiving all the data sent by USN, the collected data needs to be processed immediately. The data processing task of the kth unmanned ship is ,in is the number of CPU execution cycles required for the kth unmanned ship data, is the total number of bits of the kth unmanned ship data, is the task complexity, The relationship between , The delay that can be tolerated to complete the task, The energy consumption limit for processing tasks. represents the unloading ratio of the kth unmanned ship, that is, the amount of data processed by the kth unmanned ship is , the amount of data unloaded is .

[0097] In this embodiment, the data collected by the USN is uploaded to the unmanned ship through the underwater acoustic channel, and the NOMA method is adopted to improve the utilization rate of the acoustic channel. and The data between them is transmitted through the underwater acoustic channel, and the model of underwater acoustic communication is expressed as:

[0098]

[0099] in, is the attenuation of the underwater acoustic signal, parameter is the expansion factor, parameter is the center frequency of the underwater acoustic signal.

[0100] The acoustic channel gain between the kth unmanned ship and the kth unmanned ship can be expressed as:

[0101]

[0102] in, for The bandwidth of Ocean noise, unlike traditional white noise, usually has four main components: turbulence 、Unmanned Ship ,storm and thermal noise .

[0103] Due to the use of NOMA technology, USNs within the same unmanned ship coverage area share the same channel, so USNs within the same USV coverage area will interfere with each other. The USN transmission signals are sorted by the receiving power:

[0104]

[0105] when Towards When sending data, Upload data to The transmission rate It can be expressed as:

[0106]

[0107] in, for The channel width, is the transmission power of the i-th USN uploading data to the k-th unmanned ship, represents the additive white Gaussian noise power at the sensor.

[0108] The unmanned ship offloads part of the data to the drone for processing in the FDMA mode. There is no channel interference between the unmanned ships. The communication link between the drone and the unmanned ship is modeled as a line-of-sight (LoS) link. The channel gain can be described by the free space path loss model as follows: , where the parameter is the channel power gain at unit distance (i.e. 1m). is the path loss exponent. Data rate for uplink transmission to UAV It can be expressed as:

[0109]

[0110] in, To assign to The channel bandwidth, for The transmission power, for The background noise power at .

[0111] The kth unmanned ship Task completion delay It is expressed as:

[0112]

[0113] in, Data collection delay , I is the total number of USNs; for The upload data delay is , is the data uploaded from the i-th USN to the k-th unmanned ship, for The number of bits of data, for The transmission rate;

[0114] , is the processing delay of the kth unmanned ship;

[0115] in, is the local delay of the kth unmanned ship, and ; is the unloading ratio of the kth unmanned ship, is the number of CPU execution cycles required for the kth unmanned ship data, is the computing power of the kth unmanned ship;

[0116] is the processing delay of the kth unmanned ship’s data offloading to the UAV, and ,in is the total number of bits of the kth unmanned ship data, The computing power allocated to the kth unmanned ship by the UAV;

[0117] Data collection delay of the kth unmanned ship , is the time delay for the i-th USN to upload data to the k-th unmanned ship, , ;

[0118] Energy consumption of the kth unmanned ship ;

[0119] in, is the local computing energy consumption of the kth unmanned ship, and , is the energy consumption coefficient of the unmanned ship;

[0120] is the transmission energy consumption of the kth unmanned ship, and , is the transmission power of the kth unmanned ship;

[0121] Then the weighted sum formula of the time delay and energy consumption of the kth unmanned ship is:

[0122]

[0123] The optimization problem constructed according to the weighted sum formula of the time delay and energy consumption of the kth unmanned ship is:

[0124]

[0125] Among them, problem Z1 is an optimization problem; Q represents the total weighted sum of the delay and energy consumption of the unmanned ship, The weighting factor of the user task delay reflects the user's emphasis on delay; The weighting factor representing the energy consumption of the user's task reflects the importance the user places on energy consumption; for Upload data delay; is the i-th USN within the coverage area of ​​the k-th unmanned ship; is the energy consumption of the kth unmanned ship;

[0126]

[0127] C1.1-C1.5 are the constraints of the optimization problem Z1, K is the set of unmanned ships, where:

[0128] for The transmission power, is the upper limit of USN transmission power;

[0129] is the bandwidth allocated to the kth unmanned ship, The upper limit of bandwidth resources allocated to all unmanned ships;

[0130] is the unloading ratio of the kth unmanned ship;

[0131] is the transmission power of the kth unmanned ship; is the upper limit of the transmission power of the kth unmanned ship.

[0132] According to the weighted sum formula of delay and energy consumption, the mission completion delay of the unmanned ship is composed of data acquisition delay and data processing delay, and the mission energy consumption is composed of local computing energy consumption and transmission energy consumption. The data acquisition delay is related to the transmission power of USN, the local computing energy consumption is related to the unloading ratio, and the data processing delay and transmission energy consumption are related to the transmission power, bandwidth allocation and unloading ratio of the unmanned ship. Therefore, combined with the process in step B-step D, the weighted sum of delay and energy consumption of the unmanned ship is minimized by jointly optimizing the transmission power of USN, the transmission power of the unmanned ship, the bandwidth allocation and the unloading ratio.

[0133] In this embodiment, the optimization problem in the data collection phase is:

[0134]

[0135] Among them, problem Z2 is the optimization problem in the data collection stage. is the data collection delay of the kth unmanned ship; is the transmission power of the i-th USN;

[0136] C2.1 is the constraint condition of problem Z2.

[0137] The optimization problem in the data processing stage is:

[0138]

[0139] Among them, problem Z3 is the optimization problem in the data processing stage; is the processing delay of the kth unmanned ship;

[0140] C3.1-C3.4 are the constraints of problem Z3.

[0141] This embodiment uses a particle swarm optimization algorithm to solve the optimization problem in the data collection phase, including:

[0142] S2.1. The optimization problem in the data collection phase is transformed into the maximum delay minimization problem of the underwater sensor. The position vector of a particle represents a set of transmission power solutions. The maximum delay minimization problem of the underwater sensor is expressed as:

[0143]

[0144] Among them, problem Z4 is the maximum delay minimization problem of underwater sensors, and its corresponding constraint condition is C4.1;

[0145] S2.2.2. Initialize the number of individuals N in the population and the particle positions , particle speed ;

[0146] S2.2.3, according to the formula Calculate the fitness of individuals in the initial population, and obtain the individual historical optimal solution and the global historical optimal solution; among them, is a constant, ensuring that the fitness function value is positive; parameter is the penalty factor, which is a positive number. When the USN collects data and uploads it to the unmanned ship, if the maximum transmission power is exceeded, it will be punished by the penalty function, and the fitness of the particle will increase accordingly. Otherwise, the penalty function is 0.

[0147] S2.2.4. Update the position and velocity of the particles according to the following two formulas to generate a new population:

[0148]

[0149]

[0150] in, is the inertia weight; and is the learning factor; and is a random number between 0 and 1; is the historical optimal position of particle n; is the historical optimal position of the population;

[0151] S2.2.5. Calculate the fitness of each particle in the new population and update the individual historical optimal solution and the global historical optimal solution;

[0152] S2.2.6. Repeat S2.2.4 and S2.2.5 until the method converges and the solution to the optimization problem in the data collection phase is obtained.

[0153] In this embodiment, the unmanned ship transmission power optimization sub-problem is expressed as:

[0154]

[0155] Among them, Z5 is the unmanned ship transmission power optimization sub-problem, and its corresponding constraint condition is C5.1; is the transmission energy consumption of uploading data from the kth unmanned ship to the UAV;

[0156] The bandwidth allocation optimization subproblem is expressed as:

[0157]

[0158] Among them, Z6 is the bandwidth allocation optimization sub-problem, and its corresponding constraints are C6.1 and C6.2; is the transmission rate between the kth unmanned ship and the UAV;

[0159] The unloading ratio decision optimization sub-problem is expressed as:

[0160]

[0161] Among them, Z7 is the bandwidth allocation optimization sub-problem, and its corresponding constraint condition is C7.1.

[0162] This embodiment uses the gradient descent method to solve the transmit power optimization sub-problem, including:

[0163] S2.3.1.1. Under the current unloading ratio decision and signal-to-interference-noise ratio, calculate the minimum data unloading delay of the kth unmanned ship. , in , is the maximum transmission rate of the kth unmanned ship;

[0164] S2.3.1.2. Order ,in, is the processing delay and energy consumption of the kth unmanned ship, if Greater than Then execute S4.1.3, if Less than Then execute S4.1.4;

[0165] S2.3.1.3. Calculation The maximum transmission power of the kth unmanned ship If the partial derivative is greater than 0, there is a point where the partial derivative is 0. , The minimum value of , otherwise the minimum value is When obtained;

[0166] S2.3.1.4, let the local delay of the kth unmanned ship be equal to the processing delay of the data unloaded from the kth unmanned ship to the drone, and we get ,in, The local delay of the kth unmanned ship is equal to the power size of the processing delay unloaded to the UAV, , is the background noise power at the kth unmanned ship, is the channel gain of the kth unmanned ship, and is calculated exist The partial derivative at , if exist If the partial derivative at a point is greater than 0, then there is a point where the partial derivative is 0 , The minimum value of , otherwise the minimum value is Obtained when.

[0167] In this embodiment, the bandwidth allocation optimization sub-problem is solved using the convex optimization CVX toolbox, including:

[0168] S2.3.2.1. Input the current UAV launch power and unloading strategy, and prove that problem Z6 is a convex optimization problem;

[0169] S2.3.2.2. Under the condition of limited total bandwidth resources, use the CVX toolbox provided by MATLAB to solve problem Z6.

[0170] In this embodiment, a genetic algorithm is used to solve the unloading ratio decision optimization sub-problem, including:

[0171] S2.3.3.1. Number of individuals in the input population , population evolution times , mutation probability , crossover probability and the number of individuals eliminated in each round ;

[0172] S2.3.3.2. Initialize the gene values ​​of the individuals in the population and calculate the initial fitness function according to the following formula:

[0173]

[0174] in, Is a constant, ensuring that the fitness function value is positive;

[0175] S2.3.3.3. Perform mutation operation on each individual in the population, and generate a random number between 0 and 1 for each gene in the individual ,like Less than the mutation probability , then the gene is mutated; the mutation position is recorded as , the variation expression is ,in, Indicates the current number of evolutions. Represents a random number between [0,1];

[0176] S2.3.3.4. Perform a crossover operation on each individual in the population and randomly generate a random number between 0 and 1 , if the random number Less than the crossover probability , then perform a crossover operation on the individual and randomly select the gene value on the individual, and record its position as , and then randomly select a position in the population as The genes that ;

[0177] S2.3.3.5. Recalculate individual fitness, find a new optimal individual and replace the stored optimal individual, and replace the previous individual with the minimum current fitness function. Individuals are replaced by the best individuals;

[0178] S2.3.3.6. Repeat S2.3.3.3, S2.3.3.4 and S2.3.3.5 until the number of evolutions is greater than or equal to the number of population evolutions. .

[0179] Aiming at the scenario of computing offloading of multi-unmanned ship ocean monitoring network, this paper proposes a joint optimization method for unmanned ship delay and energy consumption. Taking the transmission power of USN, the transmission power of unmanned ship, bandwidth allocation and offloading ratio as optimization variables, under the constraints of factors such as maximum transmission power, maximum tolerable delay, maximum tolerable energy consumption and bandwidth, the optimization problem of minimizing the delay and energy consumption of unmanned ship is constructed, and it is decomposed into two stages of data acquisition and data processing to solve them in turn.

[0180] Figure 3 The performance comparison diagram of the method of the present invention and other methods under different numbers of unmanned ships is given. Figure 3 The horizontal axis represents the number of unmanned ships, and the vertical axis represents the weighted sum of the delay and energy consumption of the unmanned ships. Among them, Algorithm 1 is an even distribution of bandwidth resources for unmanned ships, and the optimization of other parameters is the same as the optimization method of the present invention; Algorithm 2 takes the upper limit of the transmission power of the unmanned ship, and the optimization of other parameters is the same as the optimization method of the present invention; Algorithm 3 takes the fixed value of 0.5 as the unloading ratio, and the optimization of other parameters is the same as the optimization method of the present invention; Algorithm 4 takes the upper limit of the transmission power of the USN, and the optimization of other parameters is the same as the optimization method of the present invention; Algorithm 5 optimizes the unloading ratio when the user bandwidth is evenly distributed and the transmission power is fixed. Figure 2 It can be seen that the performance of the present invention is the best under all numbers of unmanned ships compared with other methods.

[0181] Figure 4 The influence of the number of USNs on the weighted sum of delay and energy consumption is given. Figure 4 The horizontal axis represents the number of unmanned ships, and the vertical axis represents the weighted sum of the delay and energy consumption of the unmanned ships. Figure 4 It can be seen that with the increase in the number of USNs, the weighted increase in the latency and energy consumption of unmanned ships is because with the increase in the number of USNs, the amount of data that unmanned ships and drones need to process increases. However, the communication and computing resources are limited, which increases the data upload latency, computing latency, and energy consumption.

[0182] Figure 5 The influence of the total bandwidth of the unmanned ship on the performance of the method is given. Figure 5 The horizontal axis represents the number of unmanned ships, and the vertical axis represents the weighted sum of the delay and energy consumption of the unmanned ships. Figure 5 It can be seen that the larger the total bandwidth, the smaller the weighted sum of delay and energy consumption. This is because when the total bandwidth increases, the transmission rate will also increase, and the data upload delay and transmission energy consumption will decrease accordingly.

[0183] Although the present invention has been disclosed as above with preferred embodiments, it is not intended to limit the present invention. A person with ordinary knowledge in the technical field to which the present invention belongs may make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be determined by the definition of the claims.

[0184] Example 2

[0185] This embodiment provides a computer-readable storage medium on which a computer program / instruction is stored. When the computer program / instruction is executed by a processor, the steps of the ocean monitoring network computing offloading and resource allocation method described in Example 1 are implemented.

[0186] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0187] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0188] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0189] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0190] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the purpose of the present invention and the claims, which all fall within the protection of the present invention.

Claims

1. A method for computing offloading and resource allocation in ocean monitoring networks, It is characterized in that The method comprises: With the goal of minimizing the weighted sum of delay and energy consumption of the unmanned ship, a delay and energy consumption weighted sum optimization problem is constructed to jointly optimize the transmit power of the underwater sensor node USN, the transmit power of the unmanned ship, the bandwidth allocation and the unloading ratio; Solve the time delay and energy consumption weighted sum optimization problem to obtain the ocean monitoring network computing offloading and resource allocation solution; Perform ocean monitoring network computing offloading and resource allocation according to the ocean monitoring network computing offloading and resource allocation plan; The expression of the weighted sum of delay and energy consumption is as follows: Among them, w 1 The weighting factor of the user task delay reflects the user's emphasis on delay; t k represents the task completion delay of the kth unmanned ship; w 2 The weighting factor of the user's task energy consumption reflects the user's emphasis on energy consumption; E k is the energy consumption of the kth unmanned ship; is the USN set that sends data to the kth unmanned ship; R k,i The amount of data collected for the i-th USN; r k,i is the transmission rate of data uploaded from the i-th USN to the unmanned ship k; β k is the unloading ratio of the kth unmanned ship; C k is the number of CPU execution cycles required for the kth unmanned ship data; F k is the computing power of the kth unmanned ship; R k is the total number of bits of the kth unmanned ship data; r k is the transmission rate of the kth unmanned ship; f k is the computing power allocated by the UAV to the kth unmanned ship; α is the energy consumption coefficient of the unmanned ship; p k is the transmission power of the kth unmanned ship; The optimization problem expression of the weighted sum of delay and energy consumption is as follows: Among them, problem Z1 is an optimization problem, and C1.1-C1.5 are constraints of optimization problem Z1; P k,i is the transmission power of the i-th USN uploading data to the k-th unmanned ship; Q represents the total weighted sum of the delay and energy consumption of the unmanned ship; Assemble for unmanned ships; P i max is the upper limit of the transmission power of the i-th USN; B k is the bandwidth allocated to the kth unmanned ship, B total The upper limit of bandwidth resources allocated to all unmanned ships; is the upper limit of the transmission power of the kth unmanned ship; The method of solving the weighted and optimization problem of delay and energy consumption to obtain the ocean monitoring network computing offloading and resource allocation scheme includes: The optimization problem of the weighted sum of delay and energy consumption is decoupled into the optimization problem of the data collection stage and the optimization problem of the data processing stage. The optimization problem of the data collection stage is constructed with the goal of minimizing the data collection delay of the unmanned ship, and the optimization variable is the transmission power P of the USN; the optimization problem of the data processing stage is constructed with the goal of minimizing the weighted sum of the data processing delay and energy consumption of the unmanned ship, and the optimization variables are the transmission power p, unloading ratio β and bandwidth B of the unmanned ship; Solve the optimization problem in the data collection phase; The optimization problem in the data processing stage is decomposed into the unmanned ship transmission power optimization sub-problem, the bandwidth allocation optimization sub-problem and the unloading ratio decision optimization sub-problem, and each optimization sub-problem is solved, and then the solution to the optimization problem in the data processing stage is obtained through alternating iterations; According to the solutions of the optimization problems in the data collection phase and the data processing phase, the solutions of the weighted sum optimization problems of delay and energy consumption are obtained, and then the computational offloading and resource allocation scheme of the ocean monitoring network is obtained. The expression of the optimization problem in the data collection stage is: Among them, problem Z2 is the optimization problem in the data collection stage, and C2.1 is the constraint condition of problem Z2; in, is the data collection delay of the kth unmanned ship; and / or, The expression of the optimization problem in the data processing stage is: Among them, problem Z3 is the optimization problem in the data processing stage, and C3.1-C3.4 are the constraints of problem Z3; is the processing delay of the kth unmanned ship; Wherein, solving the optimization problem in the data collection stage includes: S2.2.

1. The optimization problem in the data collection phase is transformed into the maximum delay minimization problem of the underwater sensor. The position vector of a particle represents a set of transmission power solutions. The maximum delay minimization problem of the underwater sensor is expressed as: Among them, problem Z4 is the maximum delay minimization problem of underwater sensors, and its corresponding constraint condition is C4.1; t k,i For USN k,i Upload data delay; USN k,i is the i-th USN within the coverage area of ​​the k-th unmanned ship; S2.2.

2. Initialize the number of individuals N, particle position X, and particle velocity v in the population; S2.2.3, according to the formula Calculate the fitness of individuals in the initial population, and obtain the individual historical optimal solution and the global historical optimal solution; O is a constant to ensure that the fitness function value is positive; parameter λ is the penalty factor, which is a positive number. During the process of USN collecting data and uploading it to the unmanned ship, if the maximum transmission power is exceeded, it will be punished through the penalty function, and the fitness of the particle will increase accordingly. Otherwise, the penalty function is 0; S2.2.

4. Update the position and velocity of the particles according to the following two formulas to generate a new population: Among them, ω is the inertia weight; c 1 and c 2 is the learning factor; r 1 and r 2 is a random number between 0 and 1; is the historical optimal position of particle n; xbest t is the historical optimal position of the population; S2.2.

5. Calculate the fitness of each particle in the new population and update the individual historical optimal solution and the global historical optimal solution; S2.2.6, repeat S2.2.4 and S2.2.5 until the method converges and the solution to the optimization problem in the data collection phase is obtained; The expression of the unmanned ship transmission power optimization sub-problem is: Among them, Z5 is the unmanned ship transmission power optimization sub-problem, and its corresponding constraint condition is C5.1; in, is the transmission energy consumption of uploading data from the kth unmanned ship to the UAV; and / or, The expression of the bandwidth allocation optimization subproblem is: Among them, Z6 is the bandwidth allocation optimization sub-problem, and its corresponding constraints are C6.1 and C6.2; and / or, The expression of the unloading ratio decision optimization sub-problem is: Among them, Z7 is the bandwidth allocation optimization sub-problem, and its corresponding constraint condition is C7.1; Wherein, solving the transmit power optimization sub-problem includes: S2.3.1.

1. Under the current unloading ratio decision and signal-to-interference-noise ratio, calculate the minimum data unloading delay of the kth unmanned ship. in is the maximum transmission rate of the kth unmanned ship; S2.3.1.

2. Order Among them, Q k is the processing delay and energy consumption of the kth unmanned ship, if Greater than Then execute S4.1.3, if Less than Then execute S4.1.4; S2.3.1.

3. Calculation of Q k The maximum transmission power of the kth unmanned ship If the partial derivative is greater than 0, there is a point where the partial derivative is 0. Q k The minimum value of , otherwise the minimum value is When obtained; S2.3.1.4, let the local delay of the kth unmanned ship be equal to the processing delay of the data unloaded from the kth unmanned ship to the drone, and we get in, The local delay of the kth unmanned ship is equal to the power size of the processing delay unloaded to the UAV, N k is the background noise power at the kth unmanned ship, H k is the channel gain of the kth unmanned ship, calculate Q k exist The partial derivative at Q k exist If the partial derivative at a point is greater than 0, then there is a point where the partial derivative is 0 Q k The minimum value of , otherwise the minimum value is When obtained; Wherein, solving the bandwidth allocation optimization sub-problem includes: S2.3.2.

1. Input the current UAV launch power and unloading strategy to prove that problem Z6 is a convex optimization problem. S2.3.2.2, under the condition of limited total bandwidth resources, use the CVX toolbox provided by MATLAB to solve problem Z6; Wherein, solving the unloading ratio decision optimization sub-problem includes: S2.3.3.

1. Number of individuals in the input population Number of population evolutions Mutation probability Crossover probability and the number of individuals eliminated in each round S2.3.3.

2. Initialize the gene values ​​of the individuals in the population and calculate the initial fitness function according to the following formula: Among them, T is a constant, ensuring that the fitness function value is positive; S2.3.3.

3. Perform mutation operation on each individual in the population, and generate a random number rand between 0 and 1 for each gene in the individual. m , if rand m Less than the mutation probability Then the gene is mutated; the mutation position is (u, k), and the mutation expression is Among them, tau represents the current evolution times, and rand represents a random number between [0,1]; S2.3.3.

4. Perform a crossover operation on each individual in the population and randomly generate a random number rand between 0 and 1 a , if the random number rand a Less than the crossover probability Then perform a crossover operation on the individual, randomly select the gene value on the individual, record its position as (u, k), and then randomly select the position (u a ,k a ) gene, let β′(u,k)=β(u a ,k a ); S2.3.3.

5. Recalculate individual fitness, find a new optimal individual and replace the stored optimal individual, and replace the previous individual with the minimum current fitness function. Individuals are replaced by the best individuals; S2.3.3.

6. Repeat S2.3.3.3, S2.3.3.4 and S2.3.3.5 until the number of evolutions is greater than or equal to the number of population evolutions.

2. A computer-readable storage medium having a computer program / instructions stored thereon, It is characterized in that When the computer program / instruction is executed by a processor, the steps of the ocean monitoring network computing offloading and resource allocation method described in claim 1 are implemented.

Images