CN110677858B - Transmission power and computing resource allocation method based on task migration period of Internet of things - Google Patents
Transmission power and computing resource allocation method based on task migration period of Internet of things Download PDFInfo
- Publication number
- CN110677858B CN110677858B CN201911026324.1A CN201911026324A CN110677858B CN 110677858 B CN110677858 B CN 110677858B CN 201911026324 A CN201911026324 A CN 201911026324A CN 110677858 B CN110677858 B CN 110677858B
- Authority
- CN
- China
- Prior art keywords
- terminal
- task
- migration
- transmission power
- internet
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000013508 migration Methods 0.000 title claims abstract description 117
- 230000005012 migration Effects 0.000 title claims abstract description 115
- 230000005540 biological transmission Effects 0.000 title claims abstract description 103
- 238000000034 method Methods 0.000 title claims abstract description 46
- 238000013468 resource allocation Methods 0.000 title claims abstract description 22
- 238000012545 processing Methods 0.000 claims abstract description 33
- 238000005265 energy consumption Methods 0.000 claims abstract description 21
- 238000004364 calculation method Methods 0.000 claims description 20
- 230000008569 process Effects 0.000 claims description 14
- 241000255581 Drosophila <fruit fly, genus> Species 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 9
- 230000001976 improved effect Effects 0.000 claims description 5
- 239000000126 substance Substances 0.000 claims description 5
- 238000004458 analytical method Methods 0.000 claims description 2
- 230000001939 inductive effect Effects 0.000 claims description 2
- 241000255588 Tephritidae Species 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 230000001965 increasing effect Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 3
- 238000005562 fading Methods 0.000 description 3
- 235000013305 food Nutrition 0.000 description 3
- 239000000654 additive Substances 0.000 description 2
- 230000000996 additive effect Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000004083 survival effect Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000019637 foraging behavior Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003595 spectral effect Effects 0.000 description 1
- 238000002945 steepest descent method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000008093 supporting effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W28/00—Network traffic management; Network resource management
- H04W28/16—Central resource management; Negotiation of resources or communication parameters, e.g. negotiating bandwidth or QoS [Quality of Service]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W16/00—Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
- H04W16/22—Traffic simulation tools or models
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W52/00—Power management, e.g. TPC [Transmission Power Control], power saving or power classes
- H04W52/04—TPC
- H04W52/30—TPC using constraints in the total amount of available transmission power
- H04W52/34—TPC management, i.e. sharing limited amount of power among users or channels or data types, e.g. cell loading
- H04W52/346—TPC management, i.e. sharing limited amount of power among users or channels or data types, e.g. cell loading distributing total power among users or channels
Abstract
The invention discloses a transmission power and computing resource allocation method based on an Internet of things task migration period, which comprises the following steps: s1, determining a migration period of a k migration task of a terminal
S2, determining energy consumption of k migration task of terminal
S3, migrating the task according to the terminal k
And energy consumption
Constructing a minimization processing model of a terminal computing task migration period
S4, minimizing processing model based on terminal computing task migration period
And reasonably distributing transmission power and computing resources. The transmission power and computing resource distribution method based on the task migration period of the Internet of things can efficiently and reasonably distribute the joint pilot frequency, the data transmission power and the computing resource.
Description
Technical Field
The invention relates to the field of Internet of things, in particular to a transmission power and computing resource allocation method based on a task migration period of the Internet of things.
Background
Edge Computing (EC) deploys an edge server in a radio access network closer to an application site, provides data processing capability, and creates a low-latency, highly reliable service environment. By utilizing the edge computing network, the terminal of the Internet of things can transfer the computation-intensive tasks to the edge server for processing, so that the processing capacity and the energy consumption of the terminal of the Internet of things are reduced; and data generated by massive Internet of things terminals can be accessed to a nearby edge server, so that processing time delay is reduced, and return load is lightened. The massive MIMO (multi-input multi-output) technology can improve spectral efficiency and multiplexing capability, and also improve energy efficiency using diversity gain and array gain by increasing the number of base station antennas and using spatial resources. The massive MIMO and the EC are fused to form a massive MIMO-EC Internet of things system, and powerful migration computing service can be provided for the Internet of things terminal.
In 2015, aiming at a single-cell MIMO-EC network, Spain and Teloneia university provide a combined optimal migration ratio and computing resource allocation method to realize energy consumption-time delay compromise; aiming at a multi-cellular MIMO-EC network, the university of Roman Sabinsa of Italy provides a joint communication resource and computing resource allocation method; the Arizona State university of America proposes an uplink resource allocation method aiming at minimizing energy consumption based on distributed fog computing; in 2016, the new jersey institute of technology, usa, proposed a method for joint antenna selection, communication resource, computing resource and backhaul resource allocation for multi-cell multi-user edge computing networks. In 2017, Beijing university of transportation designs a large-scale MIMO system architecture of amorphous cells (free-cells) supporting edge computing, a base station is provided with an edge cloud server, a central server is used as a cloud computing central server, and a method for joint task migration, edge cloud association and transmission power distribution is provided. In 2018, a migration decision and calculation and communication resource allocation method is provided by Kuebec university in Canada aiming at a large-scale MIMO-EC cellular network and aiming at minimizing the maximum weighted energy consumption sum of all terminal devices. However, the above resource allocation method does not effectively solve the problems of joint pilot, data transmission power allocation and computational resource allocation.
Therefore, in order to solve the above problems, a transmission power and computing resource allocation method based on the task migration period of the internet of things is needed, which can efficiently and reasonably allocate the joint pilot frequency, the data transmission power and the computing resource.
Disclosure of Invention
In view of this, the present invention provides a transmission power and calculation resource allocation method based on a task migration period of the internet of things, which can efficiently and reasonably allocate joint pilot, data transmission power and calculation resources.
The invention discloses a transmission power and computing resource allocation method based on an Internet of things task migration period, which comprises the following steps:
s1, determining a migration period of a k migration task of a terminal
Wherein k is a terminal number in the Internet of things, and the value of k is 1,2, … and N;
s2, determining energy consumption of k migration task of terminal
S3, migrating the task according to the terminal k
And energy consumption
Constructing a minimization processing model of a terminal computing task migration period
S4, minimum processing model based on terminal computing task migration period
And allocating the transmission power and the computing resources to obtain the minimum value of the migration period of the terminal computing task, and taking the transmission power and the computing resources corresponding to the obtained minimum value as optimal allocation.
Further, in step S1, the migration cycle of the terminal k migration task is determined according to the following formula
Wherein, tpIs the transmission time of the pilot sequence;
migrating the transmission time of task data for a terminal k;
the processing time of the task is migrated for terminal k.
Further, in step S2, the energy consumption of the terminal k migration task is determined according to the following formula
Wherein, tpIs the transmission time of the pilot sequence;
a pilot sequence transmission power for terminal k;
migrating the transmission time of task data for a terminal k;
the transmission power of the migration task data for terminal k.
Further, in step S3, a minimization process model of the terminal calculation task transition period is determined according to the following formula
Wherein, the first and the second end of the pipe are connected with each other,
a pilot sequence transmission power for terminal k;
the transmission power of the migration task data of the terminal k; f. ofkAssigning edge servers to terminalsThe CPU working frequency of the terminal k;
is based on
And fkSolving the minimum value of the migration period of the terminal calculation task; max is the maximum value in the task migration period calculated by all terminals; t is tpIs the transmission time of the pilot sequence;
migrating the transmission time of task data for a terminal k;
migrating the processing time of the task for the terminal k; k is the number of terminals of the Internet of things, and the value is 1,2, … and N; f. ofSThe maximum CPU working frequency of the edge server; p is a radical ofmaxIs the maximum transmission power of the terminal.
Further, a minimization processing model for the migration period of the terminal computing task
The function relation and the constraint condition of the auxiliary function F (p) are subjected to inductive analysis to obtain the auxiliary function F (p)p,puF, τ, μ) by means of an auxiliary function F (p)p,puF, tau, mu) minimization process model for terminal computing task migration period
Carrying out rapid solving; wherein the auxiliary function F (p) is determined according to the following formulap,pu,f,τ,μ)：
F(pp,pu,f,τ,μ)＝τ+μ(A+B+C)；
K is the number of terminals of the Internet of things, and the value of K is 1,2, … and N; p is a radical ofpIs the set of pilot sequence transmission powers of the terminal, taking the value as
puA set of transmission powers for migrating task data for the terminal, valued as
f is the set of CPU working frequency distributed to the terminal by the edge server, and the value is f ═ f1,f2,......,fK](ii) a τ is an auxiliary variable, and
mu is a penalty coefficient; A. b and C are and minimization process models
The penalty term associated with the constraint in (1).
Further, penalty terms A, B and C are determined according to the following formulas, respectively:
further, step S4 specifically includes:
s41: transmitting power p according to pilot sequence of terminalpMigration task data transmission power puAnd the CPU working frequency f distributed to the terminal by the edge server, and constructing an initialization point
Wherein N is 1, …, N;
s42: based on a penalty function method, according to an auxiliary function F (p)p,pu,f,τ,μ)＝τ+μ(A+B+ C), pair initialization point
Processing to obtain N suboptimal solutions;
s43: selecting the energy consumption constraint meeting the terminal in the N suboptimal solutions, and selecting the corresponding objective function value
The minimum suboptimal solution is used as the minimum suboptimal solution;
s44: and taking the transmission power of the terminal pilot sequence and the transmission power of the migration task data in the minimum suboptimal solution as the initialization positions of the drosophila groups, and executing an improved drosophila optimization algorithm to obtain the transmission power and the calculation resources corresponding to the minimum value of the terminal calculation migration task period.
The invention has the beneficial effects that: according to the transmission power and calculation resource distribution method based on the task migration period of the Internet of things, the minimum value of the task migration period of the terminal of the Internet of things is calculated, and the combined transmission power and calculation resources are efficiently and reasonably distributed.
Drawings
The invention is further described below with reference to the following figures and examples:
FIG. 1 is a schematic flow diagram of the process of the present invention;
FIG. 2 is a schematic diagram of a massive MIMO-EC network system according to the present invention;
FIG. 3 is a schematic diagram of time slot allocation of a compute task migration cycle according to the present invention;
FIG. 4 is a maximum computation task migration period of the present invention
Changing a curve along with the number of terminals of the Internet of things;
FIG. 5 is a maximum computation task migration period of the present invention
Changing a curve along with the task migration bit number of the terminal of the Internet of things;
FIG. 6 is a maximum computation task migration period of the present invention
The curve varies with the number of base station antennas.
Detailed Description
The invention is further described with reference to the accompanying drawings, in which:
the invention discloses a transmission power and computing resource allocation method based on an Internet of things task migration period, which comprises the following steps:
s1, determining a migration period of a k migration task of a terminal
Wherein k is a terminal number in the Internet of things, and the value of k is 1,2, … and N;
s2, determining energy consumption of k migration task of terminal
S3, migrating the task according to the terminal k
And energy consumption
Constructing a minimization processing model of a terminal computing task migration period
S4, minimum processing model based on terminal computing task migration period
And allocating the transmission power and the computing resources to obtain the minimum value of the migration period of the terminal computing task, and taking the transmission power and the computing resources corresponding to the obtained minimum value as optimal allocation.
Massive MIMO-EAs shown in fig. 2, the internet of things system C includes K single-antenna internet of things terminals and a base station, and the base station is configured with M antennas and an edge server, and has communication and calculation capabilities. The K single-antenna Internet of things terminals have concurrent task data migration requirements, and the Internet of things terminal K belongs to the quadruplet for task migration of K
There is described a method of, wherein,
is the number of bits to be migrated for the task,
is the number of bits of the processing result,
is the task survival time, XkIs task calculation density (CPU instruction cycle number/bit), namely the terminal K of the internet of things belongs to K
The bit task is transferred to the edge server for processing through wireless transmission, and the required CPU instruction cycle number is
After which the task is calculated
The bit processing result is fed back to the terminal of the Internet of things to require the survival time of the task
And (4) completing the operation.
K internet of things terminals are accessed to the base station based on Space Division Multiple Access (SDMA), and the edge server distributes independent Virtual Machines (VMs) for each internet of things terminal to process migration tasks in parallel. Within the channel coherence time, K pilot sequences with the length of K
And allocating each terminal of the Internet of things to perform pilot-assisted channel estimation, and performing signal detection by the base station based on the channel estimation result, wherein all pilot sequences are orthogonal to each other.
In this embodiment, in step S1, the calculation task migration period is divided into three stages, i.e., pilot sequence transmission, task data transmission, and migration task processing, where the time is tp,tu,tvmAs shown in fig. 3.
In the pilot frequency sequence transmission stage, the terminal K of the Internet of things belongs to K and transmits power
Transmitting a pilot sequence phikTo the base station, the base station adopts Minimum Mean Square Error (MMSE) criterion to execute channel estimation, and obtains a channel estimation vector g of the terminal to the base stationk＝[gk1,…,gkM]T：
Wherein, betakRepresenting a large-scale fading factor from a terminal K to a base station channel, including path loss and shadow effect;
an additive white gaussian noise vector representing zero mean, unit variance,
and representing additive white Gaussian noise from the terminal K of the Internet of things to the antenna M of the base station, wherein the antenna M belongs to the channel M.
In the task data transmission stage, the terminal K of the Internet of things belongs to K to transmit power
Sending migration task data bits
Received signal vector y to base station, base stationk(i)＝[y1(i),…,yM(i)]T：
The base station uses the channel estimation result gkDetection of the received signal, i.e. the received signal vector y, is performed based on a Maximum Ratio Combining (MRC) criterionk(i) Multiplied by GH＝[g1,...,gK]HDecoding all migration task data bits sent by all terminals of the Internet of things
The transmission rate of the terminal K of the Internet of things belonging to K is as follows:
Rk＝log2(1+SINRk) (3)
wherein the content of the first and second substances,
in the migration task processing stage, the edge server allocates independent Virtual Machines (VMs) for different Internet of things terminals to provide parallel processing of the migration tasks, and the processing performance depends on the CPU working frequency allocated to the edge server. By fSThe maximum CPU working frequency of the edge server is represented, and the CPU working frequency f allocated to the terminal K belonging to the Internet of things belongs to Kk≤fSProcessing the terminal migration task
The time of (a) is:
in conclusion, the terminal K of the internet of things belongs to the K migration task
The calculation task migration period is as follows:
wherein the content of the first and second substances,
in this embodiment, in step S2, the internet of things terminal K belongs to K to process the migration task
The required energy consumption comprises two parts of pilot sequence transmission energy consumption and migration task data transmission energy consumption, and the energy consumption mainly depends on transmission power and transmission time.
Determining the energy consumption of the k migration task of the terminal according to the following formula
Wherein, tpIs the transmission time of the pilot sequence;
a pilot sequence transmission power for terminal k;
migrating the transmission time of task data for a terminal k;
the transmission power of the migration task data for terminal k.
In this embodiment, in step S3, the pilot sequence transmission power, the migration task data transmission power, and the CPU operating frequency allocated by the edge server are used as decision variables to minimize the load on all terminals of the internet of thingsEstablishing a joint transmission power allocation and calculation resource allocation model in a large-scale MIMO-EC internet of things system by taking the maximum calculation task migration period as an optimization target
Wherein the content of the first and second substances,
a pilot sequence transmission power for terminal k;
the transmission power of the migration task data of the terminal k; f. ofkAllocating the CPU working frequency of the terminal k to the edge server;
is based on
And fkSolving the minimum value of the migration period of the terminal calculation task; max is the maximum value in the task migration period calculated by all terminals; t is tpIs the transmission time of the pilot sequence;
migrating the transmission time of the task data for the terminal k;
migrating the processing time of the task for the terminal k; k is the number of terminals of the Internet of things, and the value is 1,2, … and N; f. ofSThe maximum CPU working frequency of the edge server; p is a radical ofmaxIs the maximum transmission power of the terminal; c1 is the energy consumption constraint for terminal k; c2 is a computational resource constraint of the edge server; c3 is the transmit power constraint for the terminal.
In this embodiment, in step S4, the optimization model is analyzed
The following conclusion can be obtained, where the optimal solution is obtained, the migration periods of the computing tasks of all the terminals of the internet of things are equal, that is, the computing resource allocation satisfies the equation set:
in conjunction with the boundary conditions of the computational resource constraint C2, the above equation is abbreviated as:
solving the above equation can yield f1Then, computing the computing resource f ═ f { f) distributed by the edge server to all internet of things terminals according to equation (8)1,…,fKAnd the allocation result depends on the transmission power of the pilot sequence and the transmission power of the migration task data.
To achieve optimal power allocation, an auxiliary variable τ is introduced
Conversion to equivalent optimization model
Wherein, C0 indicates that the computation task migration period of each internet of things terminal is less than a constant.
To pair
Constructing an auxiliary function:
F(pp,pu,f,τ,μ)＝τ+μ(A+B+C) (11)
wherein the content of the first and second substances,
mu is a penalty coefficient, and the value is a positive number which is sufficiently large; penalty A, B and C are:
the model will be further optimized
Conversion into an optimization model
If it is
Decision variable (p)p,puF, τ, μ) is in the feasible region, then there is F (p)p,puF, τ, μ) ═ τ; if it is
Decision variable (p)p,puF, τ, μ) are not in the feasible region, the helper function F (p) is setp,puThe penalty term coefficient μ in f, τ, μ) is an increasing and approaching infinite series of numbers { μtAnd t is 1,2, …, and is the iteration number, which aims to punish the decision variables which are not in the feasible domain, and force the iteration process to approach the feasible domain continuously.
Computing resources according to equations (8) and (9)Assignment result f ═ f1,…,fKWhich resource allocation depends on the pilot sequence transmission power and the migration task data transmission power. The implementation flow of the joint transmission power allocation and computing resource allocation method (IFOA-PFSA) in the massive MIMO-EC IOT system is described in the following algorithm 1:
1: randomly initializing N power allocation initiation points
Obtaining a computing resource allocation f according to equations (8) and (9)(n)N is 1, …, N, and sets the auxiliary variable target value τ;
2: constructing an auxiliary function F (p) according to equation (11)p,puF, τ, μ) ═ τ + μ (a + B + C), based on the current initial point
Executing the algorithm 2 to obtain N suboptimal solutions;
3: among the N sub-optimal solutions, the selection satisfies the energy constraint
And corresponding objective function value
Minimum sub-optimal solution
4: to be provided with
As a fitness function, the sub-optimal solution obtained from step 3
Extracting pilot frequency transmission power and migration task data transmission power of all internet of things terminals
As a group of fruit fliesInitial position of body, execution of algorithm 4, output of optimal solution
Wherein the algorithm 2 is based on a penalty function method for each initial point determined in the algorithm 1
Processing to obtain a suboptimal solution corresponding to the initial point
For the nth initial point
The flow of executing algorithm 2 is as follows:
1: is provided with
The power control sub-optimal solution precision is epsilon and the punishment coefficient series is mut1,2, …; wherein, the iteration number t ← 1;
2: constructing an auxiliary function F ((p)p,pu,f)(t-1),τ,μt)；
3: solving for minF ((p) using Algorithm 3p,pu,f)(t-1),τ,μt)；
4: if (p)p,pu,f)(t)-(pp,pu,f)(t-1)||<Epsilon, stopping iteration and outputting a suboptimal solution:
otherwise, t ← t +1, return to step 2, and execute downward in turn.
The algorithm 3 adopts the steepest descent method to solve the minF ((p)p,pu,f)(t-1),τ,μt). The flow of algorithm 3 is as follows:
1: for (p)p,pu,f)(0)And the accuracy delta of the steepest descent gradient is given,number of iterations m ← 0;
2: calculating gradients
If it is
Stopping iteration and outputting a suboptimal solution:
otherwise, turning to the step 3 and sequentially executing downwards;
3: according to
Search iteration step size lambdamLet us order
4: if | | | F (p)p,pu,f,μt)(m+1)-F(pp,pu,f,μt)(m)||<δ, output suboptimal solution:
stopping iteration; otherwise, m ← m +1, go to step 2, and execute downwards in sequence.
The algorithm 4 is an improved fruit fly optimization algorithm, and is an intelligent algorithm based on fruit fly foraging behavior deduction, namely, the fruit flies firstly use good smell to collect smell to search for food sources, and then use sharp vision to find the food and the fellows after flying to the vicinity of the food, and fly to the fellows gathering place.
The algorithm 4 is divided into four stages, which are sequentially: initialization, an olfactory search process, a visual search process and iterative processing, wherein the implementation process comprises the following steps:
stage 1: initialization
1: the fruit fly population scale is pilot frequency transmission power and migration task data transmission power (p) of the terminal of the Internet of thingsp,pu) Maximum number of iterations T, single fly of DrosophilaLine distance (Δ p)p,Δpu) Initial position of Drosophila population
Namely the initial position of each individual K1, …, K in the fruit fly population
t←1；
And (2) stage: olfactory search process
2: each individual K is 1, …, K uses smell search, in the t smell search, endows it with fixed flight distance
Direction of flight at random
New location information:
3: calculating the odor concentration
4: selecting the group of Drosophila K having the greatest concentration of odor
Fruit fly individual k*Recording its maximum odor concentration value and corresponding position as
And (3) stage: visual search process
5: maintaining maximum odor concentration value and corresponding fruit fly position
Other individual fruit flies fly to this location using vision, i.e.
And (4) stage: t ← T +1, repeating stage 2 and stage 3 until the number of iterations reaches T
6: output of
Corresponding position Xk(t),Yk(t) is the optimal position of the individual, and the optimal positions of all the individuals are the optimal solution
The implementation effect of the method is explained by taking an internet of things system applied to an intelligent substation as an example. As shown in FIG. 2, the single-cell massive MIMO-EC IOT system has a base station deployed in an intelligent substation, and a configured edge server with a maximum CPU working frequency fSThe number of configured antennas M is 32-256 at 100GHz, and the number K of terminals of the Internet of things with task migration requirements is [10,20 ]]Uniformly distributed, and for the internet of things terminal K1max0.2W, energy constraint
Number of task migration bits
The distance d between the terminal of the Internet of things and the base station is [10,60 ]]And m are uniformly distributed.
Setting a pilot sequence transmission time tpThe power control suboptimal solution precision epsilon is 0.005W and the steepest descent gradient precision delta is 10ms-4Penalty term coefficient series [ mu ]tIn an improved drosophila optimization algorithm, the maximum iteration time T is 10, and the iteration step length of the pilot transmission power and the data transmission power of the internet of things terminal, namely the single flying distance Δ p of the drosophila, is 10p＝Δpu＝0.001W。
In channel estimation based on pilot frequency assistance, a large-scale fading factor from a terminal K of the Internet of things to a base station channelβk＝zk/(dk)α，zkRepresenting a shadow fade with a logarithm obeying a mean of 0 and a variance of
Normal distribution of (d)kAnd the communication distance between the terminal K of the Internet of things and the base station is represented, and alpha is a path loss index. In this embodiment, α is 3.7, σshThe base station has obtained a communication distance d using a wireless positioning technique at 8dBkI.e. the large-scale fading factor beta from the terminal K of the internet of things to the base station channelkAre known.
The performance of the classical drosophila optimization algorithm (TFOA) was analyzed in comparison to the performance of the present method (IFOA-PFSA) as follows:
FIG. 4 illustrates a maximum computation task migration period
According to the variation curve of the number of the terminals of the Internet of things, the number M of the antennas configured in the base station is 32, and the number of the task migration bits of each terminal of the Internet of things
Maximum CPU operating frequency f of edge serverS100 GHz. As can be seen from the figure, the maximum computation task migration period of the IFOA-PFSA and TFOA increases with the number of terminals in the internet of things, which is due to the following reasons: on one hand, in a large-scale MIMO system with a given number of antennas, as the number of terminals of the Internet of things increases, mutual interference increases, the transmission rate decreases, and the data transmission time of a migration task increases; on the other hand, because the computing resources are limited, the computing resources allocated to each internet of things terminal are reduced as the number of the internet of things terminals increases, which also causes the maximum computing task migration period to increase. But the performance of the IFOA-PFSA is better than that of the TFOA, and the performance is more and more remarkable as the number of terminals of the Internet of things increases.
FIG. 5 illustrates a maximum computation task migration period
Bit migration along with internet of things terminal taskThe number change curve, the base station configuration antenna number M is 32, the internet of things terminal number K is 10, and the maximum CPU working frequency f of the edge serverS100 GHz. As can be seen from the figure, the maximum computation task migration periods of the IFOA-PFSA and the TFOA both increase with the increase of the number of the task migration bits of the terminal of the internet of things, and the reasons are as follows: on one hand, under the condition that the transmission rate is unchanged, the transmission time is increased due to the increase of the task migration data volume; on the other hand, an increase in the amount of data that needs to be processed by the same computing resource increases the processing time. But the performance of the IFOA-PFSA is better than that of the TFOA, and the number of the bit of the terminal task migration of the Internet of things is increased more and more remarkably.
FIG. 6 illustrates a maximum computation task migration cycle
According to the variation curve of the number of the base station antennas, the number K of the terminals of the internet of things is 10, and the number of the task migration bits of each terminal of the internet of things
Maximum CPU operating frequency f of edge serverS100 GHz. As can be seen, the maximum task migration period of both IFOA-PFSA and TFOA decreases with the number of antennas, because: the number of the base station antennas is increased, so that larger space diversity gain can be provided for each Internet of things terminal, SINR is increased, the transmission rate is improved, and the transmission time of task data is reduced.
Finally, the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all of them should be covered in the claims of the present invention.
Claims (6)
1. A transmission power and computing resource allocation method based on an Internet of things task migration period is characterized in that: the method comprises the following steps:
s1, determining a migration period of a k migration task of a terminal
Wherein k is a terminal number in the Internet of things, and the value of k is 1,2, … and N;
s2, determining energy consumption of k migration task of terminal
S3, migrating the task according to the terminal k
And energy consumption
Constructing a minimization processing model of a terminal computing task migration period
Determining a minimization processing model of a terminal computing task migration period according to the following formula
Wherein the content of the first and second substances,
a pilot sequence transmission power for terminal k;
the transmission power of the migration task data of the terminal k; f. ofkAllocating the CPU working frequency of the terminal k to the edge server;
is based on
And fkSolving the minimum value of the migration period of the terminal calculation task; max is the maximum value in the task migration period calculated by all terminals; t is tpIs the transmission time of the pilot sequence;
migrating the transmission time of task data for a terminal k;
migrating the processing time of the task for the terminal k; k is the number of terminals of the Internet of things, and the value is 1,2, … and N; f. ofSThe maximum CPU working frequency of the edge server; p is a radical ofmaxIs the maximum transmission power of the terminal;
s4, minimum processing model based on terminal computing task migration period
And allocating the transmission power and the computing resources to obtain the minimum value of the migration period of the terminal computing task, and taking the transmission power and the computing resources corresponding to the obtained minimum value as optimal allocation.
2. The method for allocating transmission power and computing resources based on the task migration period of the internet of things according to claim 1, wherein: in step S1, the migration cycle of the terminal k migration task is determined according to the following formula
Wherein, tpIs the transmission time of the pilot sequence;
migrating the transmission time of task data for a terminal k;
the processing time of the task is migrated for terminal k.
3. The method for allocating transmission power and computing resources based on the task migration period of the internet of things according to claim 1, wherein: in step S2, the energy consumption of the terminal k migration task is determined according to the following formula
Wherein, tpIs the transmission time of the pilot sequence;
a pilot sequence transmission power for terminal k;
migrating the transmission time of task data for a terminal k;
the transmission power of the migration task data for terminal k.
4. The internet of things-based notebook of claim 1The transmission power and calculation resource allocation method of the service migration period is characterized in that: minimization processing model for terminal computing task migration period
The function relation and the constraint condition of the auxiliary function F (p) are subjected to inductive analysis to obtain the auxiliary function F (p)p,puF, τ, μ) by means of an auxiliary function F (p)p,puF, tau, mu) minimization process model for terminal computing task migration period
Carrying out rapid solving; wherein the auxiliary function F (p) is determined according to the following formulap,pu,f,τ,μ)：
F(pp,pu,f,τ,μ)＝τ+μ(A+B+C)；
K is the number of terminals of the Internet of things, and the value of K is 1,2, … and N; p is a radical ofpIs the set of pilot sequence transmission powers of the terminal, taking the value as
puA set of transmission powers for migrating task data for the terminal, valued as
f is the set of CPU working frequency distributed to the terminal by the edge server, and the value is f ═ f1,f2,......,fK](ii) a τ is an auxiliary variable, and
mu is a penalty coefficient; A. b and C are and minimization process models
The penalty term associated with the constraint in (1).
5. The method for allocating transmission power and computing resources based on the task migration period of the internet of things according to claim 4, wherein: penalty terms A, B and C are determined according to the following formulas, respectively:
6. the method for allocating transmission power and computing resources based on the task migration period of the internet of things according to claim 1, wherein: step S4 specifically includes:
s41: transmitting power p according to pilot sequence of terminalpMigration task data transmission power puAnd the CPU working frequency f distributed to the terminal by the edge server, and constructing an initialization point
Wherein N is 1, …, N;
s42: based on penalty function method, according to auxiliary function F (p)p,puF, τ, μ) ═ τ + μ (a + B + C), for the initialization point
Processing to obtain N suboptimal solutions;
s43: selecting the energy consumption constraint meeting the terminal in the N suboptimal solutions, and selecting the corresponding objective function value
Is the minimum sub-optimal solution as the minimumSuboptimal solution;
s44: and taking the transmission power of the terminal pilot sequence and the transmission power of the migration task data in the minimum suboptimal solution as the initialization positions of the drosophila groups, and executing an improved drosophila optimization algorithm to obtain the transmission power and the calculation resources corresponding to the minimum value of the terminal calculation migration task period.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911026324.1A CN110677858B (en) | 2019-10-25 | 2019-10-25 | Transmission power and computing resource allocation method based on task migration period of Internet of things |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911026324.1A CN110677858B (en) | 2019-10-25 | 2019-10-25 | Transmission power and computing resource allocation method based on task migration period of Internet of things |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110677858A CN110677858A (en) | 2020-01-10 |
| CN110677858B true CN110677858B (en) | 2022-05-17 |
Family
ID=69084095
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911026324.1A Active CN110677858B (en) | 2019-10-25 | 2019-10-25 | Transmission power and computing resource allocation method based on task migration period of Internet of things |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110677858B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111372260B (en) * | 2020-03-09 | 2021-09-07 | 西安交通大学 | Network load balancing method based on flow prediction and drosophila optimization algorithm |
| CN111586146B (en) * | 2020-04-30 | 2022-04-22 | 贵州电网有限责任公司 | Wireless internet of things resource allocation method based on probability transfer deep reinforcement learning |
| CN112214301B (en) * | 2020-10-29 | 2023-06-02 | 华侨大学 | Smart city-oriented dynamic calculation migration method and device based on user preference |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109151864A (en) * | 2018-09-18 | 2019-01-04 | 贵州电网有限责任公司 | A kind of migration decision and resource optimal distribution method towards mobile edge calculations super-intensive network |
| CN109286664A (en) * | 2018-09-14 | 2019-01-29 | 嘉兴学院 | A kind of computation migration terminal energy consumption optimization method based on Lagrange |
| CN109413676A (en) * | 2018-12-11 | 2019-03-01 | 西北大学 | Combine the edge calculations moving method of lower uplink in a kind of ultra dense heterogeneous network |
| CN109819046A (en) * | 2019-02-26 | 2019-05-28 | 重庆邮电大学 | A kind of Internet of Things virtual computing resource dispatching method based on edge cooperation |
| CN109905470A (en) * | 2019-02-18 | 2019-06-18 | 南京邮电大学 | A kind of expense optimization method for scheduling task based on Border Gateway system |
| CN109992419A (en) * | 2019-03-29 | 2019-07-09 | 长沙理工大学 | A kind of collaboration edge calculations low latency task distribution discharging method of optimization |
| CN110187964A (en) * | 2019-05-07 | 2019-08-30 | 南京邮电大学 | The deadline minimizes mist computation migration method in scenes of internet of things |
-
2019
- 2019-10-25 CN CN201911026324.1A patent/CN110677858B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109286664A (en) * | 2018-09-14 | 2019-01-29 | 嘉兴学院 | A kind of computation migration terminal energy consumption optimization method based on Lagrange |
| CN109151864A (en) * | 2018-09-18 | 2019-01-04 | 贵州电网有限责任公司 | A kind of migration decision and resource optimal distribution method towards mobile edge calculations super-intensive network |
| CN109413676A (en) * | 2018-12-11 | 2019-03-01 | 西北大学 | Combine the edge calculations moving method of lower uplink in a kind of ultra dense heterogeneous network |
| CN109905470A (en) * | 2019-02-18 | 2019-06-18 | 南京邮电大学 | A kind of expense optimization method for scheduling task based on Border Gateway system |
| CN109819046A (en) * | 2019-02-26 | 2019-05-28 | 重庆邮电大学 | A kind of Internet of Things virtual computing resource dispatching method based on edge cooperation |
| CN109992419A (en) * | 2019-03-29 | 2019-07-09 | 长沙理工大学 | A kind of collaboration edge calculations low latency task distribution discharging method of optimization |
| CN110187964A (en) * | 2019-05-07 | 2019-08-30 | 南京邮电大学 | The deadline minimizes mist computation migration method in scenes of internet of things |
Non-Patent Citations (2)
| Title |
|---|
| Fog Computing:Towards Minimizing Delay in the Internet of Things;Ashkan Yousefpour 等;《2017 IEEE 1st international conference on edge computing》;20171231;17-24 * |
| 一种能效优先的物联网任务协同迁移策略;周龙雨等;《物联网学报》;20190630;第3卷(第2期);64-71 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110677858A (en) | 2020-01-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110677858B (en) | Transmission power and computing resource allocation method based on task migration period of Internet of things | |
| CN110166090B (en) | Large-scale MIMO downlink unicast beam domain power distribution method with optimal energy efficiency | |
| KR101087873B1 (en) | Method and apparatus to support sdma transmission in a ofdma based network | |
| CN110289895B (en) | Large-scale MIMO downlink power distribution method based on energy efficiency and spectrum efficiency joint optimization | |
| CN105162507B (en) | Two benches method for precoding based on letter leakage noise ratio in extensive MIMO FDD systems | |
| CN112468196B (en) | Power distribution method in de-cellular large-scale MIMO system based on PZF precoding | |
| CN109104225A (en) | A kind of optimal extensive MIMO Beam Domain multicast transmission method of efficiency | |
| CN113708804B (en) | Whale algorithm-based user scheduling and simulated beam selection optimization method | |
| CN110311715B (en) | Large-scale MIMO non-orthogonal unicast and multicast transmission power distribution method with optimal energy efficiency | |
| CN111970033B (en) | Large-scale MIMO multicast power distribution method based on energy efficiency and spectrum efficiency joint optimization | |
| CN106160806B (en) | Method and apparatus for performing interference coordination in wireless communication system | |
| Nimmagadda | Enhancement of efficiency and performance gain of massive MIMO system using trial-based rider optimization algorithm | |
| CN109951219B (en) | Low-cost large-scale non-orthogonal multi-access method | |
| CN109787672B (en) | Large-scale MIMO lattice point offset channel estimation method based on parameter learning | |
| CN114302487B (en) | Energy efficiency optimization method, device and equipment based on self-adaptive particle swarm power distribution | |
| CN107872255B (en) | Pilot frequency scheduling method suitable for large-scale MIMO cellular mobile communication network | |
| Alharbi et al. | A time-and energy-efficient massive mimo-noma mec offloading technique: A distributed admm approach | |
| Shurman et al. | Pilot contamination mitigation in massive MIMO-based 5G wireless communication networks | |
| CN113922849B (en) | User grouping and power distribution method under millimeter wave MIMO-NOMA system | |
| CN113258985B (en) | Energy efficiency optimization method for single-station multi-satellite MIMO (multiple input multiple output) upper injection system | |
| CN110545204B (en) | Resource allocation method and server based on external penalty function and fruit fly optimization | |
| CN115133969A (en) | Performance improving method of millimeter wave large-scale MIMO-NOMA system | |
| CN108063656A (en) | A kind of new pilot distribution method suitable for extensive MIMO cellular networks | |
| CN115604824A (en) | User scheduling method and system | |
| CN109004962B (en) | Stratospheric large-scale MIMO user side beam forming method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |