CN116827992A - Edge computing and unloading method based on Internet of things scene - Google Patents
Edge computing and unloading method based on Internet of things scene Download PDFInfo
- Publication number
- CN116827992A CN116827992A CN202310792578.4A CN202310792578A CN116827992A CN 116827992 A CN116827992 A CN 116827992A CN 202310792578 A CN202310792578 A CN 202310792578A CN 116827992 A CN116827992 A CN 116827992A
- Authority
- CN
- China
- Prior art keywords
- edge
- sparrow
- task
- server
- unloading
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 24
- 241000287127 Passeridae Species 0.000 claims abstract description 63
- 230000008901 benefit Effects 0.000 claims abstract description 55
- 238000005265 energy consumption Methods 0.000 claims abstract description 35
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 34
- 238000005295 random walk Methods 0.000 claims abstract description 13
- 230000005540 biological transmission Effects 0.000 claims description 19
- 238000005457 optimization Methods 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 8
- 230000006870 function Effects 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 4
- SLXKOJJOQWFEFD-UHFFFAOYSA-N 6-aminohexanoic acid Chemical compound NCCCCCC(O)=O SLXKOJJOQWFEFD-UHFFFAOYSA-N 0.000 claims description 3
- 238000009826 distribution Methods 0.000 claims description 3
- 238000009827 uniform distribution Methods 0.000 claims description 3
- 238000010845 search algorithm Methods 0.000 abstract description 9
- 238000004088 simulation Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 230000007423 decrease Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 230000003190 augmentative effect Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000002431 foraging effect Effects 0.000 description 1
- 230000005021 gait Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000009191 jumping Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Landscapes
- Mobile Radio Communication Systems (AREA)
Abstract
The application discloses an edge computing and unloading method based on an Internet of things scene, which belongs to the technical field of Internet of things task unloading and comprises the steps of constructing an Internet of things edge computing system model with cloud edge coordination; initializing related parameters of an edge computing system model; establishing a time delay model and an energy consumption model aiming at a cloud end and an edge server; constructing a system benefit function according to different configurations of the unloading mode of the mobile equipment; and using an improved sparrow search algorithm to continuously and iteratively update the sparrow position and obtain the optimal unloading strategy and system benefit value. The method introduces sine and cosine algorithm ideas to improve the capability of the algorithm to jump out of a local optimal solution. The Levy flight strategy and the random walk strategy are introduced to comprehensively explore in a search space, so that each user equipment in the system model can unload tasks to a cloud or edge server, the system benefit is improved, and the task discarding rate is reduced.
Description
Technical Field
The application belongs to the technical field of computing and unloading of the Internet of things, and particularly relates to a computing and unloading method for mobile edge computing in an Internet of things scene.
Background
In recent years, with the continuous popularization of the fifth generation mobile communication technology (5G), the number of internet of things (IoT) devices is also continuously increasing, which will have a severe requirement on the cloud infrastructure and the computing performance of the devices themselves. A large number of intensive computing requests tends to accelerate the energy consumption of the mobile device and shorten its useful life. Offloading tasks to cloud servers also presents a significant challenge due to the stringent requirements for high accuracy and low latency. And due to the problems of overlarge time delay, high network load and the like caused by the factors of limited frequency spectrum resources, longer backhaul link distance, poor channel quality and the like, the user experience and the network performance are reduced.
Mobile Edge Computing (MEC) has therefore emerged as an emerging technology that can effectively overcome the serious power consumption of mobile devices and the high latency that cloud servers may incur. In contrast to cloud servers, edge servers are implemented directly at a cellular Base Station (BS) or a local wireless Access Point (AP) using a general purpose computing platform. By virtue of this feature, the MEC allows applications to be executed in the vicinity of the terminal device, thereby greatly reducing end-to-end time delay and reducing the burden on the backhaul network. In view of this, offloading computing tasks using edge servers can result in ultra-low latency and flexible computing for intensive computing requests from mobile users. With the advent of MEC, the ability of resource-constrained terminal devices to offload computing tasks to edge servers is expected to be widely used in the fields of augmented reality, virtual reality, and unmanned. At present, most of researches on edge computing offloading aim at optimizing a certain index of time delay or energy consumption, and in an actual scene, the time delay directly affects the reliability of user equipment, and the energy consumption is also an important factor considered by enterprises, and lower energy consumption means lower cost. Therefore, the application optimizes the two indexes simultaneously to obtain the optimal unloading strategy which maximizes the network benefit.
Often, the solution to the computational offload strategy requires conversion to an integer nonlinear optimization problem, and thus it is difficult to solve directly to obtain the best decision. Heuristic algorithms are a better solution for solving the problem. The heuristic algorithm has the advantage that it is more efficient than the general search algorithm, and that an optimal offloading strategy can be obtained with a shorter number of iterations. However, the biggest disadvantage of the heuristic search algorithm is that the convergence speed and the network benefit value of the final result cannot be ensured due to the change of the step length in each iteration, and the Sparrow Search Algorithm (SSA) is a novel group intelligent optimization algorithm, which takes the foraging and anti-predation behaviors of the sparrows as inspiration, and finds the optimal solution of the objective function by simulating the process of searching food by the sparrows and comparing the fitness values. The sparrow search algorithm has the advantages of few control parameters, high solving precision, easiness in combination with other algorithms and the like. Compared with most group intelligent algorithms and heuristic algorithms, the sparrow search algorithm has certain advantages in optimizing most problems, but still has the problems of low convergence accuracy and difficulty in jumping out of local extremum.
Disclosure of Invention
Aiming at the problems, the application provides an edge computing and unloading method based on an Internet of things scene, which comprises the steps of firstly constructing an edge computing network model of the Internet of things with cloud edge cooperated, respectively obtaining an edge server computing model and a cloud computing model according to the established network model, further obtaining a network benefit function considering time delay and energy consumption simultaneously, then adopting an improved sparrow searching algorithm to compute and unload each Internet of things device, finding out an optimal computing and unloading strategy to enable the network benefit function to be maximum, thereby ensuring that each Internet of things device in the system can reasonably unload tasks to the cloud or the edge server for processing, improving the system benefit and reducing the task discarding rate.
The technical scheme adopted for solving the technical problems is as follows:
an edge computing and unloading method in an internet of things scene comprises the following steps:
s1, constructing a cloud edge collaborative edge computing system model of the Internet of things: the model comprises a group of mobile devices U, a group of edge servers S and a cloud server M with computing power;
s2, initializing related parameters of an edge computing system model;
s3, establishing a time delay model and an energy consumption model aiming at the cloud and the edge server;
s4, when the edge server of a certain area is idle, the mobile equipment of the area directly unloads the task to the edge server; if the edge server of a certain area is not idle, the mobile equipment can select to unload the task to the cloud server M, and a system benefit function is constructed according to the task;
s5, sine and cosine SCA algorithm ideas, levy flight strategies and random walk strategies are respectively introduced into the sparrow position updating strategy to improve the sparrow position updating strategy, and the improved sparrow searching algorithm is used for continuously and iteratively updating the sparrow position and obtaining the optimal unloading strategy and the optimal system benefit value.
Further, the step S2 specifically includes:
s21, the set of mobile devices and edge servers are denoted by u= {1,2, … U } and s= {1,2, … S } respectively, each mobile device U e U generating one computing task L at a time u =<c L ,d L ,T a ,T b>, wherein cL Representing the workload of completing a computational task, d L Representing the amount of input data required to transfer program execution from the mobile device to the server, T a Respectively represent ideal time delay, T b Representing the maximum tolerable delay;
s22, defining unloading strategy according to network modelWherein Q represents a mobile device generated set of computing tasks, wherein +.>Is a binary variable +.>Indicating that task L is offloaded to s.epsilon.S, +.>Indicating that task L is offloaded to cloud M;
s23, the position of the mobile device is determined byIndicating that the location of the edge server is defined by +.>Representation, where x u ,y u Respectively representing the abscissa and the ordinate of the mobile device, and assuming that the height of the mobile device is 0; x is x s ,y s Respectively representing the abscissa and ordinate of the edge server, and H represents the height of the edge server.
Further, the step S3 specifically includes:
s31, the channel gain from the mobile U e U to the MEC server S e S in the uplink can be expressed as:
in the formula h0 The channel power gain when the transmission power is 1W and the transmission distance is 1m is shown;
s32, assuming that the locations of the mobile device and the edge server do not change during a time interval, the transmission rate from the mobile device to the edge server may be expressed as:
in the formula, u' noteq, B and sigma 2 Power, p= { p, representing system bandwidth and noise of uplink, respectively us U e U, S e S represents the distance from the mobile U e U to the edge server S e SA transmit power level;
s33, respectively obtaining time delay and energy consumption of the edge server and the cloud server M through the network model established in the S1.
Further, in step S33, the specific contents of the delay and the energy consumption of the edge server and the cloud server M include:
the time delay of task offloading to the edge server can be expressed as:
wherein Rus (t) represents an uplink transmission rate; c (C) us Representing computing power of the edge server;
s332, the time delay of task offloading to the cloud server may be expressed as:
in the formula dL Representing an amount of input data required to transfer program execution from the mobile device to the server; the relative distance between the cloud server and the mobile device is approximately constant, and the rate R of unloading the mobile user to the cloud is equal to the rate R um For a fixed value, C um Representing computing power of the cloud;
s333, the energy consumption of task offloading to the cloud server or the edge server may be expressed as:
the energy consumption of task unloading comprises the energy consumption of task transmission and the energy consumption of task processing, and the task transmission energy consumption of the mobile equipment U epsilon U is defined as:
the energy consumption of the edge server or cloud server processing task is expressed as:
wherein ps Representing average power consumption of edge servers, p m Representing the average power consumption of the cloud server.
Further, the step S4 specifically includes:
s41, when the mobile devices in the same area compete for the edge server in the area to finish task unloading within ideal time delay, defining the efficiency eta of the edge system L The method comprises the following steps:
wherein :T a and Tb Respectively represents ideal time delay and maximum tolerable time delay, T avg Representing the average time delay;
s42, when a mobile device in a certain area is processing a task, if the mobile device in the area selects to offload the task to the cloud server M, the offload cost is defined as:
o L =θ·η L +(1-θ)E pro
wherein θ represents a weighting coefficient, 0.5;
s43, constructing a calculation unloading system model network benefit under the scene of the Internet of things according to S3 and S4, wherein the calculation unloading system model network benefit is as follows:
wherein eL Representing the normalized energy consumption of the device,
further, the step S5 specifically includes:
s51, setting a sparrow population in an algorithm and initializing related parameters;
and S52, continuously updating the sparrow position in each optimizing through multiple iterations, finally finding out the sparrow position with the optimal fitness value, taking the sparrow position into the calculated unloading system model built in the step S1, and obtaining network benefits according to the step S43.
Further, in the step S52, updating the sparrow position in each optimizing specifically includes:
s521, a sine and cosine SCA algorithm idea is adopted in a position updating strategy of the sparrow finder, and a nonlinear sine learning factor is introduced; the learning factor formula and the improved finder position formula are as follows:
ω=ω min +(ω max -ω min )·sin(tπ/iter max )
in the formula ,r1 Is [0,2 pi ]]Random number in r 2 Is [0,2 ]]The random number in the memory, omega is a learning factor,when the iteration number is t, the position of the ith sparrow in the jth dimension is represented by R 2 ∈[0,1]And ST e [0.5,1 ]]Respectively representing an early warning value and a safety value;
s522, combining the Levy flight strategy and the random walk strategy, comprehensively exploring in a search space, and simultaneously keeping searching for a local optimal solution;
levy flight is introduced into the sparrow adder update formula, and the improved formula is as follows:
where d is the dimension of the vector,
the calculation formula of the levy flight strategy is as follows:
where Γ (x) = (x-1) +.! Q is a random number obeying normal distribution, r 3 ,r 4 ∈[0,1]Random numbers in the range, and value of xi is 1.5;
the sparrow jointer position updating formula after the random walk strategy is introduced is as follows:
wherein , and />Respectively representing two random solutions when the iteration times are t, wherein tau is a scaling factor and is subject to uniform distribution of (0, 1);
s523, obtaining the current maximum network benefit value through the sparrow position of the wheel, and recording the corresponding sparrow position;
s524, comparing the maximum network benefit value obtained in the round with the global optimal network benefit value, if the network benefit value in the round is better than the global optimal benefit value, updating the sparrow position, and updating the global optimal network benefit value into the current network benefit value;
s525, judging whether the iteration times of the round reach the maximum iteration times, if not, carrying out S52 again, otherwise, outputting the optimal network benefit value and the corresponding sparrow position.
The technical scheme of the application can produce the following technical effects:
(1) According to the edge computing and unloading method under the scene of the Internet of things, a computing and unloading model under the scene of the Internet of things is firstly constructed, then the solving of the computing and unloading strategy is converted into an integer nonlinear optimization problem, and the optimal computing and unloading strategy is obtained by continuously and iteratively solving by combining a sparrow searching algorithm. The sine and cosine algorithm thought is introduced in the process of algorithm, so that the capability of the algorithm to jump out of the local optimal solution is improved. (2) By introducing the Levy flight strategy and the random walk strategy, the search can be comprehensively performed in the search space, and meanwhile, the search of the local optimal solution is kept, so that the optimization capacity of an algorithm is improved, each user device in the system model can reasonably unload the task to the cloud or edge server, the system benefit is improved, and the task discarding rate is reduced.
Drawings
FIG. 1 is a schematic flow chart of an edge computing and unloading method in an Internet of things scene;
FIG. 2 is a schematic diagram of an edge computing system model in an Internet of things scenario;
FIG. 3 is a graph of network benefit versus task offloading under four algorithms.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application more clear, the technical solutions of the present application will be clearly and completely described below with reference to specific embodiments of the present application and corresponding drawings. It will be apparent that the described embodiments are some, but not all, embodiments of the application.
S1, constructing a cloud edge collaborative edge computing system model of the Internet of things: the model comprises a group of mobile devices U, a group of edge servers S and a cloud server M with ultra-large computing power.
S2, initializing relevant parameters of the edge computing system model.
The set of S21, mobile users and edge servers is denoted by u= {1,2, … U } and s= {1,2, … S } respectively. Each mobile device U e U generates one computing task L at a time u =<c L ,d L ,T a ,T b>, wherein cL Representing the workload of completing a computational task, d L Representing the amount of input data required to transfer program execution from the mobile device to the server, T a and Tb Respectively representing ideal time delay and maximum tolerable time delay.
S22, defining unloading strategy according to system modelWherein Q represents a mobile device generated set of computing tasks, wherein +.>Is a binary variable +.>Indicating that task L is offloaded to s.epsilon.S, +.>Indicating that task L is offloaded to cloud M.
S23, the position of the mobile device is determined byIndicating that the location of the edge server is defined by +.>Representation, where x u ,y u Respectively representing the abscissa and the ordinate of the mobile device, and assuming that the height of the mobile device is 0; x is x s ,y s Respectively representing the abscissa and ordinate of the edge server, and H represents the height of the edge server.
And S3, establishing a time delay model and an energy consumption model aiming at the cloud server and the edge server.
S31, the channel gain from the mobile U e U to the MEC server S e S in the uplink can be expressed as:
in the formula h0 The channel power gain at a transmission power of 1W and a transmission distance of 1m is shown.
S32, assuming that the locations of the mobile device and the edge server do not change during a time interval, the transmission rate from the mobile device to the edge server may be expressed as:
in the formula, u' noteq, B and sigma 2 Power, p= { p, representing system bandwidth and noise of uplink, respectively us U e U, S e S represents the transmit power level from the mobile U e U to the edge server S e S.
S33, respectively obtaining time delay and energy consumption of the edge server and the cloud through the system model established in the S1.
The time delay of task offloading to the edge server can be expressed as:
wherein Rus (t) represents an uplink transmission rate; c (C) us Representing the computing power of the edge server.
S332, the time delay of task offloading to the cloud server may be expressed as:
in the formula dL Representing an amount of input data required to transfer program execution from the mobile device to the server; considering that the position of the cloud server is always unchanged and is far away from the mobile device, the relative distance between the cloud server and the mobile device can be approximately constant, because we can assume that the mobile device is unloaded to the cloud end at the rate R um Is a fixed value. C (C) um Representing the computing power of the cloud.
S333, the energy consumption of task offloading to the cloud server or the edge server may be expressed as:
the energy consumption of task unloading comprises the energy consumption of task transmission and the energy consumption of task processing, and the task transmission energy consumption of the U E U task of the mobile equipment is defined as
The energy consumption of processing tasks by an edge server or cloud server may be expressed as
wherein ps Representing average power consumption of edge servers, p m Representing the average power consumption of the cloud server.
S4, when the edge server of a certain area is idle, the mobile equipment of the area directly unloads the task to the edge server; if the edge server of a certain area is not idle, the mobile equipment can select to unload the task to the cloud server M, and a system benefit function is constructed according to the task;
s41, when the mobile devices in the same area compete for the edge server in the area to finish task unloading within ideal time delay, defining the efficiency eta of the edge system L The method comprises the following steps:
wherein :T a and Tb Respectively represents ideal time delay and maximum tolerable time delay, T avg Representing the average time delay;
s42, when a mobile user has a lot of tasks waiting for unloading in a certain area, the waiting time of the tasks increases. In some extreme cases, the computing task may either be discarded, at which point the mobile user may choose to offload the computing task to the cloud server, but at a cost defined as:
o L =θ·η L +(1-θ)E pro
where θ represents the weighting factor, here taken to be 0.5.
S43, constructing a calculation unloading system model network benefit in the scene of the Internet of things according to S3 and S4 as follows
wherein eL Representing the normalized energy consumption of the device,
and S5, using an improved sparrow searching algorithm to continuously and iteratively update the sparrow position and obtain the optimal unloading strategy and system benefit value.
S51, setting a sparrow population in an algorithm and initializing related parameters;
and S52, continuously updating the sparrow position in each optimizing through multiple iterations, finally finding out the sparrow position with the optimal fitness value, taking the sparrow position into the spectrum allocation model built in the step S1, and obtaining the system benefit according to the step S43.
In the original sparrow search algorithm, as the iteration number increases, the dimension of the sparrow individual gradually decreases, so that the search space also decreases, and the sparrow individual is easy to fall into a local optimal value. To improve this problem, the sparrow finder's location update strategy employs the Sine and Cosine (SCA) algorithm concept and introduces a nonlinear sine learning factor. In the early stage of searching, the learning factor has a larger value, which is beneficial to global exploration, and in the later stage of searching, the learning factor has a smaller value, which is beneficial to improving the local development capability and accuracy, in particular, the learning factor formula and the improved finder position formula are as follows:
ω=ω min +(ω max -ω min )·sin(tπ/iter max )
s522, in the optimization problem, in order to improve the search efficiency, different search strategies may be adopted. The Levy flight strategy can enable individuals to be widely distributed in the search space, and is favorable for finding out a global optimal solution; the random walk strategy can enable the individual to search in a relatively dense area, which is beneficial to finding out a local optimal solution. The two strategies are combined, so that the search can be comprehensively performed in the search space, and meanwhile, the search of the local optimal solution is kept, so that the optimization capacity of the algorithm is improved.
(1) Levy flight strategy
In the population optimizing process, a larger searching step length is needed in the early stage to improve the searching capability of the global optimal solution, and the searching capability of the local optimal solution is improved by reducing the step length in the later stage. The Levy flight strategy is a non-Gaussian random gait, and can be alternatively explored by a high-frequency short distance and a low-frequency long distance in the searching process, so that the Levy flight is introduced into a sparrow adder update formula to improve the optimizing effect. The improved formula is as follows:
where d is the vector dimension, and the calculation formula of levy strategy is as follows:
where Γ (x) = (x-1) +.! Q is a random number obeying normal distribution, r 3 ,r 4 ∈[0,1]Random numbers in the range, and xi takes a value of 1.5.
(2) Random walk strategy
The random walk strategy enhances the diversity of the population by introducing two modes of mixed variation and intersection, can greatly improve the capability of local optimization of the algorithm, and can accelerate the searching of the optimal solution by introducing the random walk strategy into the intelligent optimization algorithm. The position movement formula of the sparrow jointer after the random walk strategy is introduced is as follows:
wherein , and />Two random solutions at iteration number t are respectively represented, τ is a scaling factor, and is subject to uniform distribution of (0, 1).
S523, obtaining the current maximum network benefit value through the sparrow position of the current wheel, and recording the corresponding sparrow position.
S524, comparing the maximum network benefit value obtained in the round with the global optimal network benefit value, if the network benefit value in the round is better than the global optimal benefit value, updating the sparrow position, and updating the global optimal network benefit value into the current network benefit value.
S525, judging whether the iteration number of the round of judgment reaches the maximum iteration number, if not, carrying out S52 again, otherwise, outputting the optimal network benefit value and the corresponding sparrow position.
The method for calculating and unloading the edges in the scene of the Internet of things is subjected to experimental simulation:
the MATLAB 2019b is adopted for carrying out algorithm simulation, and the simulation scene mainly comprises: the cloud server comprises a cloud server, a plurality of edge servers and a plurality of user devices. And (3) randomly generating the positions of the user equipment and the edge equipment without considering the performance difference of the edge equipment, wherein the parameters in the experiment are respectively set as follows: the number of user equipments u=32, the number of edge servers s=8, the uplink channel bandwidth b=20×10 6 Hz, maximum transmission power P of user equipment us The computing power of the edge server and cloud is C =5 us =60GHz,C um =120GHz。
The simulation results are shown in fig. 3, and the network total benefits of the Improved Sparrow Search Algorithm (ISSA), the Genetic Algorithm (GA), the Particle Swarm Optimization (PSO) and the basic Sparrow Search Algorithm (SSA) are given. As can be seen from the simulation verification results, the overall network benefit is increased continuously and is stabilized at a constant gradually as the iteration number is increased. As the improved sparrow searching algorithm is easier to jump out of the local extremum, the improved sparrow searching algorithm has the maximum network total benefit and has faster convergence speed than other algorithms.
The above detailed description of the preferred embodiments of the present application does not provide any limitation to the present application. Any person skilled in the art will make any equivalent substitution or modification to the technical solution and technical content disclosed in the application without departing from the scope of the technical solution of the application, and the technical solution of the application is not departing from the scope of the application.
Claims (6)
1. The edge computing and unloading method in the scene of the Internet of things is characterized by comprising the following steps of:
s1, constructing a cloud edge collaborative internet of things edge computing and unloading system model: the model comprises a group of mobile devices U, a group of edge servers S and a cloud server M with computing power;
s2, initializing related parameters of an edge computing system model;
s3, establishing a time delay model and an energy consumption model aiming at the cloud and the edge server;
s4, when the edge server of a certain area is idle, the mobile equipment of the area directly unloads the task to the edge server; if the edge server of a certain area is not idle, the mobile equipment can select to unload the task to the cloud server M, and a system benefit function is constructed according to the task;
s5, respectively introducing sine and cosine SCA algorithm ideas, levy flight strategies and random walk strategies into a sparrow position updating strategy to improve the sparrow position updating strategy, and continuously and iteratively updating the sparrow position by using an improved sparrow searching algorithm to obtain an optimal unloading strategy and a system benefit value, wherein the method specifically comprises the following steps of:
s51, setting a sparrow population in an algorithm and initializing related parameters;
and S52, continuously updating the sparrow position in each optimizing through multiple iterations, and finally finding out the sparrow position with the optimal fitness value, and taking the sparrow position into the calculated unloading system model built in the step S1 to obtain network benefits.
2. The method for edge computing and offloading in an internet of things scenario of claim 1, wherein S2 specifically comprises:
s21, the set of mobile devices and edge servers are denoted by u= {1,2, … U } and s= {1,2, … S } respectively, each mobile device U e U generating one computing task L at a time u =<c L ,d L ,T a ,T b>, wherein cL Representing the workload of completing a computational task, d L Representing the amount of input data required to transfer program execution from the mobile device to the server, T a Respectively represent ideal time delay, T b Representing the maximum tolerable delay;
s22, defining unloading strategy according to system modelWherein Q represents a mobile device generated set of computing tasks, wherein +.>Is a binary variable +.>Indicating that task L is offloaded to s.epsilon.S, +.>Indicating that task L is offloaded to cloud M;
s23, the position of the mobile device is determined byIndicating that the location of the edge server is defined by +.>Representation, where x u ,y u Respectively representing the abscissa and the ordinate of the mobile device, and assuming that the height of the mobile device is 0; x is x s ,y s Respectively representing the abscissa and ordinate of the edge server, and H represents the height of the edge server.
3. The method for edge computing and unloading in the scene of the internet of things according to claim 1, wherein the step S3 specifically comprises:
s31, the channel gain from the mobile U e U to the MEC server S e S in the uplink can be expressed as:
in the formula h0 The channel power gain when the transmission power is 1W and the transmission distance is 1m is shown;
s32, assuming that the locations of the mobile device and the edge server do not change during a time interval, the transmission rate from the mobile device to the edge server may be expressed as:
in the formula, u' noteq, B and sigma 2 Power, p= { p, representing system bandwidth and noise of uplink, respectively us U e U, S e S, represents the transmit power level from the mobile U e U to the edge server S e S;
s33, respectively obtaining time delay and energy consumption of the edge server and the cloud server M through the system model established in the S1.
4. The method for offloading edge computation in an internet of things scenario according to claim 3, wherein the time delay and energy consumption specific contents of the edge server and the cloud server M in S33 include:
the time delay of task offloading to the edge server can be expressed as:
wherein Rus (t) represents an uplink transmission rate; c (C) us Representing computing power of the edge server;
s332, the time delay of task offloading to the cloud server may be expressed as:
in the formula dL Representing an amount of input data required to transfer program execution from the mobile device to the server; the relative distance between the cloud server and the mobile device is approximately constant, and the rate R of unloading the mobile device to the cloud is equal to the rate R um For a fixed value, C um Representing computing power of the cloud server;
s333, the energy consumption of task offloading to the cloud server or the edge server may be expressed as:
the energy consumption of task unloading comprises the energy consumption of task transmission and the energy consumption of task processing, and the task transmission energy consumption of the mobile equipment U epsilon U is defined as:
the energy consumption of the edge server or cloud server processing task is expressed as:
wherein ps Representing average power consumption of edge servers, p m Representing the average power consumption of the cloud server.
5. The method for edge computing and unloading in the internet of things scenario of claim 1, wherein S4 specifically comprises:
s41, when the mobile devices in the same area compete for the edge server in the area to finish task unloading within ideal time delay, defining the efficiency eta of the edge system L The method comprises the following steps:
wherein :T a and Tb Respectively represents ideal time delay and maximum tolerable time delay, T avg Representing the average time delay;
s42, when a mobile device in a certain area is processing a task, if the mobile device in the area selects to offload the task to the cloud server M, the offload cost is defined as:
o L =θ·η L +(1-θ)E pro
wherein θ represents a weighting coefficient, 0.5;
s43, constructing a calculation unloading system model network benefit under the scene of the Internet of things according to S3 and S4, wherein the calculation unloading system model network benefit is as follows:
wherein eL Representing the normalized energy consumption of the device,
6. the method for edge computing and unloading in the internet of things scenario of claim 1, wherein updating the sparrow position in each optimization in S52 specifically comprises:
s521, a sine and cosine SCA algorithm idea is adopted in a position updating strategy of the sparrow finder, and a nonlinear sine learning factor is introduced; the learning factor formula and the improved finder position formula are as follows:
ω=ω min +(ω max -ω min )·sin(tπ/iter max )
in the formula ,r1 Is [0,2 pi ]]Random number in r 2 Is [0,2 ]]The random number in the memory, omega is a learning factor,when the iteration number is t, the position of the ith sparrow in the jth dimension is represented by R 2 ∈[0,1]And ST e [0.5,1 ]]Respectively representing an early warning value and a safety value;
s522, combining the Levy flight strategy and the random walk strategy, comprehensively exploring in a search space, and simultaneously keeping searching for a local optimal solution;
levy flight is introduced into the sparrow adder update formula, and the improved formula is as follows:
where d is the dimension of the vector,
the calculation formula of the levy flight strategy is as follows:
where Γ (x) = (x-1) +.! Q is a random number obeying normal distribution, r 3 ,r 4 ∈[0,1]Random numbers in the range, and value of xi is 1.5;
the sparrow jointer position updating formula after the random walk strategy is introduced is as follows:
wherein , and />Respectively representing two random solutions when the iteration times are t, wherein tau is a scaling factor and is subject to uniform distribution of (0, 1);
s523, obtaining the current maximum network benefit value through the sparrow position of the wheel, and recording the corresponding sparrow position;
s524, comparing the maximum network benefit value obtained in the round with the global optimal network benefit value, if the network benefit value in the round is better than the global optimal benefit value, updating the sparrow position, and updating the global optimal network benefit value into the current network benefit value;
s525, judging whether the iteration times of the round reach the maximum iteration times, if not, carrying out S52 again, otherwise, outputting the optimal network benefit value and the corresponding sparrow position.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310792578.4A CN116827992A (en) | 2023-06-30 | 2023-06-30 | Edge computing and unloading method based on Internet of things scene |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310792578.4A CN116827992A (en) | 2023-06-30 | 2023-06-30 | Edge computing and unloading method based on Internet of things scene |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116827992A true CN116827992A (en) | 2023-09-29 |
Family
ID=88140545
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310792578.4A Pending CN116827992A (en) | 2023-06-30 | 2023-06-30 | Edge computing and unloading method based on Internet of things scene |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116827992A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117768464A (en) * | 2023-11-13 | 2024-03-26 | 重庆理工大学 | Block chain calculation task unloading method based on group intelligent reinforcement learning |
-
2023
- 2023-06-30 CN CN202310792578.4A patent/CN116827992A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117768464A (en) * | 2023-11-13 | 2024-03-26 | 重庆理工大学 | Block chain calculation task unloading method based on group intelligent reinforcement learning |
CN117768464B (en) * | 2023-11-13 | 2024-07-09 | 重庆理工大学 | Block chain calculation task unloading method based on group intelligent reinforcement learning |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112512056B (en) | Multi-objective optimization calculation unloading method in mobile edge calculation network | |
CN111726826A (en) | Online task unloading method in base station intensive edge computing network | |
CN112105062B (en) | Mobile edge computing network energy consumption minimization strategy method under time-sensitive condition | |
CN116260871A (en) | Independent task unloading method based on local and edge collaborative caching | |
CN116827992A (en) | Edge computing and unloading method based on Internet of things scene | |
CN114564304A (en) | Task unloading method for edge calculation | |
Chen et al. | A distributed game theoretical approach for credibility-guaranteed multimedia data offloading in MEC | |
CN115659803A (en) | Intelligent unloading method for computing tasks under unmanned aerial vehicle twin network mapping error condition | |
CN114626298A (en) | State updating method for efficient caching and task unloading in unmanned aerial vehicle-assisted Internet of vehicles | |
CN112667406A (en) | Task unloading and data caching method in cloud edge fusion heterogeneous network | |
CN116112981A (en) | Unmanned aerial vehicle task unloading method based on edge calculation | |
CN114116061B (en) | Workflow task unloading method and system in mobile edge computing environment | |
CN115696452A (en) | Game method for joint optimization of unloading decision and resource allocation in cloud-edge cooperative computing | |
CN117580105B (en) | Unmanned aerial vehicle task unloading optimization method for power grid inspection | |
CN117156493A (en) | QoE-based vehicle networking task unloading and resource allocation method | |
CN111930435A (en) | Task unloading decision method based on PD-BPSO technology | |
CN116880923A (en) | Dynamic task unloading method based on multi-agent reinforcement learning | |
CN113950059B (en) | Unmanned aerial vehicle relay auxiliary user task unloading method and system | |
CN116137724A (en) | Task unloading and resource allocation method based on mobile edge calculation | |
Zhuang et al. | Task Offloading Strategy for Unmanned Aerial Vehicle Power Inspection Based on Deep Reinforcement Learning | |
CN114513814A (en) | Edge network computing resource dynamic optimization method based on unmanned aerial vehicle auxiliary node | |
CN110933687A (en) | User uplink and downlink access method and system based on decoupling | |
Tao et al. | UAV-assisted resource allocation strategy in energy harvesting edge computing system | |
CN114756294B (en) | Mobile edge computing and unloading method based on deep reinforcement learning | |
CN117687797B (en) | Vehicle networking calculation unloading method based on VIFOX optimization algorithm |
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 |