CN108596746A - Mobile intelligent perception and its resource allocation based on two way auction and incentive mechanism method - Google Patents
Mobile intelligent perception and its resource allocation based on two way auction and incentive mechanism method Download PDFInfo
- Publication number
- CN108596746A CN108596746A CN201810360743.8A CN201810360743A CN108596746A CN 108596746 A CN108596746 A CN 108596746A CN 201810360743 A CN201810360743 A CN 201810360743A CN 108596746 A CN108596746 A CN 108596746A
- Authority
- CN
- China
- Prior art keywords
- buyer
- seller
- candidate
- benchmark
- algorithm
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/08—Auctions
Landscapes
- Business, Economics & Management (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Marketing (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Entrepreneurship & Innovation (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention is based on the mobile intelligent perceptions and its resource allocation of two way auction and incentive mechanism method suitable for mobile intelligent perception, some users are limited to the perception resource and data acquisition conditions of equipment of itself, high-speed can not be provided and there is multifarious quality data resource, the resource by other equipment is needed to participate in the scene of mobile intelligent perception.Under the scene, on the one hand, smart machine can provide more rich resource by the powerful calculating ability of thin cloud, on the other hand can be there is the smart machine of the high credit worthiness of superperformance state by means of other, establish the data contact network between equipment and equipment.This process is related to the two way auction between smart machine and thin cloud or smart machine and smart machine, can use Game Theory thus, study its resource allocation and exciting torque.
Description
Technical field
The present invention relates to the two-way resource allocation and excitation algorithm technical field between smart machine in mobile intelligent perception,
More particularly to mobile intelligent perception and its resource allocation and incentive mechanism method based on two way auction.
Background technology
In mobile intelligent perception, main participant is the smart machine that user carries.Although current smart machine is more
It is new regenerate it is very fast, but in transmitting high speed data and during some have the perception task that higher quality requires, general smart machine
The opposite low side because of its configuration and performance is still suffered from, thus the problem of the task of not competent certain high-end requirements, at this time
The poor smart machine of performance needs to seek the high performance resource supply promotion performance of itself, desired to preferably participate in
In the mobile gunz perception task participated in, high performance resource supply side can be active other high prestige smart machines,
It can also be the thin cloud during mobile gunz calculates.
When the performance of smart machine is restricted by itself configuration, the application that thin cloud carries out resource on the one hand can be sought
With distribution, the calculated performance for itself participating in mobile intelligent perception is improved using the powerful calculating ability of thin cloud.Whole process can be with
It regards application of the mobile cloud computing in intelligent perception as, therefore belongs to mobile intelligent perception and calculate (Mobile Crowd
Sensing and Computing, MCSC) scope.
On the other hand it can be distributed to the user's request for data for obtaining good prestige in perception task.Meanwhile for frequency
Numerous user that good prestige is obtained in mobile gunz perception task, the data acquisition ability and offer high quality number of its own
According to ability also have certain competitiveness, so when other low prestige users want fast lifting prestige, and be not intended to lead to
It crosses the higher thin cloud of expense and promotes self performance, other high prestige smart machines that expense can be selected slightly lower provide good number
According to resource.This is a reciprocal process, designs corresponding excitation distribution system, and the resource-sharing and distribution to both sides have
Prodigious excitation.
Invention content
In order to solve problem above, the present invention provides a kind of mobile intelligent perception and its resource allocation based on two way auction
And incentive mechanism method by the scene abstract modeling process in pairs to auction trade, and is transported for above-mentioned described scene
The design of resource allocation and incentive mechanism is carried out with the model of two way auction, for this purpose, the present invention is based on two way auctions
Mobile intelligent perception and its resource allocation and incentive mechanism method, be as follows:
(1) system modelling;
Considering, which has the m resource supply side i.e. sellers that can provide resource and n smart machine to carry user, buys
Family, resource allocation problem between the two can be modeled as a unidirectional multi-item double auction model, each buyer's secret
Offer their quotation, each seller's secret submits they charge to auctioner, namely each to take part in auction
The information of all other men is not known completely;
For each buyer bi∈ B, B={ b1,b2,...,bnFor, each buyer biTo the quotation vector of the seller
It can be expressed asHereIndicate buyer biTo seller sj∈ S, S={ s1,s2,...,smReport
The matrix of valence, quotation contains the quotation vector of all buyers, is denoted as R=(R1;R2;...;Rn), each in S is sold
For family, the vector of charging of seller is expressed as A=(A1,A2,...,Am), A herejIndicate seller sj∈ S's charges, buyer for
For different sellers, there is the preference of oneself to seller since buyer can be directed to different mission requirements, different are sold
There is a different quotations in family, and seller is due to only focusing on and sharing the resource of itself and therefrom collect remuneration, the rope of seller
Valence will not distinguish buyer, although the quotation of buyer is secrecy for seller, resource supply side is being auctioned
Before, it is desired nonetheless to some specific information are provided such as calculating capacity, data prestige quality and network bandwidth, in order to
User carries out valuation according to these characteristics to selected resource supply side, and the cost information of resource supply side is kept absolutely secret;
Given vector B, S, R, A, auctioner determine triumph buyerAnd winning sellersWSAnd WBBetween
Matching relationship γ:{j:sj∈WS}→{i:bi∈WB, select auctioner and triumph buyer b after the both parties of triumphi
∈WBBetween settlement price PBi, auctioner pay winning sellers sj∈WSRemuneration PSj, in order to emphasize triumph both parties
Between matching, in certain situations also use PBijAnd PSijIndicate the settlement price that buyer need to pay and the report that seller obtains
Reward;
Other than above-mentioned buyer's settlement price and seller's remuneration, the effectiveness of both parties additionally depends on buyer to seller
Other required services of service and the valuation of cost of offer, enable Vi jIndicate buyer biFrom seller sjThe valence serviced
Value, CjIndicate seller siThe cost of service, buyer b are providediValue vector can be expressed as Vi=(Vi 1,Vi 2,...,Vi m), it is right
Matching i=γ (j), buyer b are bought-sold in oneiWith seller sjEffectiveness indicate as follows:
Also UB is usedijAnd USijIndicate buyer biWith seller sjBetween successful match after effectiveness, it is clear that need to ensure
Effectiveness UBi> 0 namely smart machine carry user and are collected into the valence obtained after the resource from resource supply side as buyer
Value is higher than it and pays the settlement price of auctioner, so, UBiIt shows smart machine and carries user to the resource that is collected into
Satisfaction, for seller, the effectiveness US of the resource supply side as selleriWhat is represented is that the remuneration that it is obtained is more than
The degree of itself cost namely the effectiveness of resource supply side indicate that it shares the resource getable profit of institute later;
(2) algorithm designs;
It is the believable auctioner of third party to manipulate entirely auction in two way auction, which needs root in auction
The buyer's set W to win is determined according to the agreement mechanism of auctionB, winning seller's set WS, matching relationship between both parties,
Settlement price set PB between triumph buyerwWith the remuneration PS for paying the triumph sellerw, mathematic(al) representation be Ψ=(B, S,
R, A), it is proposed that a kind of resource allocation mechanism DAIM based on two way auction;
DAIM algorithms are made of two sub- algorithms, i.e., rough candidate's matching algorithm and one-to-one matching algorithm, rough
In candidate's matching algorithm, rough triumph candidate collection is obtained, and by these set as input, in one-to-one matching algorithm
In obtain final one-to-one matching winning results;
In rough candidate matches algorithm, auctioner is each participation seller s firstjDetermine the candidate of buyer, then
Determine the remuneration for needing the settlement price collected to buyer and paying seller, here, there are one very crucial benchmark, reports
Valence and needs of charging are compared with benchmark benchmark, to be selected, use A-jIndicate that all sellers charge middle removal
Seller sjVector of charging later.The benchmark compared each time, by vectorial A-jIt is obtained after middle removal maximum value and minimum value
AverageIt indicates;
For seller sjDetermine that the buyer to win is candidate, according to BjThere are two types of situations for the number of middle element;
Work as BjIn element only there are one when, namely only there are one buyer biQuotation be no less than Aj:IfR and k and AjWhen≤benchmark meets simultaneously, then buyer is added to the candidate collection W of buyerB, together
When the price is fixed at benchmark;In the case of other, buyer cannot reach successful transaction with seller;
Work as BjIn element there are two or it is more when, namely have more than a buyer biQuotation be no less than Aj:Such as
Fruit does not have buyer b if highest quotation is less than benchmark in thisiSeller s can be obtainedjService;In the case of other,
There is the buyer of highest quotation or if buyer's more than one of highest quotation randomly chooses one of them, can be added into and buy
Family candidate collection WB, while sjIt is added into seller's candidate collection WS, the settlement price that pays needed for the buyer of selection and corresponding
The remuneration of seller is the maximum value in the high quotations of benchmark and second;
Find benchmark algorithm be selection win dealing side important evidence, while be also to determine buyer's settlement price and to
The important evidence for giving seller's remuneration, the selection for benchmark, the influence in order to avoid extreme offer data to benchmark, simultaneously
Use for reference the information of charging of other sellers, selection removes seller and charges the data vector that current seller in vector charges after data,
The average after the maximin in the vector is rejected as benchmark benchmark;
In rough candidate matches algorithm, due to buyer's candidate collection WBIn buyer may match two or more
Seller's candidate collection WSIn seller, it is therefore desirable to execute 3.4 one-to-one matching algorithm of algorithm, determine only one most for buyer
The excellent seller, auctioner's selection can allow corresponding buyer to reach the seller of maximum utility, likewise, if there is multiple sellers can be with
Effectiveness is allowed to reach maximum, then one of them seller of random selection, after having executed one-to-one matching algorithm, each buyer bo(j)
WBHave and corresponds matched winning sellers s therewithj∈WS。
It is further improved as the present invention, DAIM algorithms are as follows in the step 2 algorithm design:
Input:Buyer set B, seller set S, buyer bid matrix R, seller charge vectorial A;
Output:Triumph buyer's set WB, winning sellers set WS, match γ, triumph buyer's settlement price set PBw, win
Seller's remuneration set;
First, the assignment from rough candidate matches algorithm (B, S, R, A)
Then, from one-to-one matching algorithmMiddle assignment (WB,WS,γ,PBw,PSw);
Finally obtain WB,WS,γ,PBw,PSwValue.
It is further improved as the present invention, rough candidate matches algorithm is as follows in the step 2 algorithm design:
Input:Buyer set B, seller set S, buyer bid matrix R, seller charge vectorial A;
Output:Buyer's candidate collection WB, seller's candidate collection WS, candidate matchesCandidate buyer's set of prices PBw, candidate sells
Family's remuneration collection;
First, by WB, WS, PBw, PSwAll it is assigned a value of empty set;
When meeting sjIt is recycled when ∈ S;
Benchmark benchmark=is obtained first finds benchmark algorithm (A),
If BjMiddle element only have 1 so;
IfA simultaneouslyj≤ benchmark is so;
PBij=PSj=benchmark;
PBw←PBw∪{PBij},PSw←PSw∪{PSj};
If BjMiddle element is more than 1;
According to marked price to BjMiddle element is ranked up, i.e.,Collection after note sequence is combined into
IfSo;
IfIn a quotation of preceding t (t >=2) be it is the same so;
At random in BjPreceding t buyer in select bi;
If other situations;
Middle first b for selecting to have highest quotationi;
PBw←PBw∪{PBij},PSw←PSw∪{PSj};
End loop.
It is further improved as the present invention, it is as follows to find benchmark algorithm in the step 2 algorithm design:
Input:Buyer set B, seller set S, buyer bid matrix R, seller charge vectorial A;
Output:Buyer's candidate collection WB, seller's candidate collection WS, candidate matchesCandidate buyer's set of prices PBw, candidate sells
Family's remuneration collection;
First s is being rejected in vectorial AjForm vector A-j;
Then according to the sequence charged from small to large to A-jIt is ranked up;
And form vector A after rejecting head and the tail element-j;
It finally returns to
It is further improved as the present invention, one-to-one matching algorithm is as follows in the step 2 algorithm design:
Input:Buyer's candidate collection WB, seller's candidate collection WS, candidate matchesCandidate buyer's set of prices PBw, candidate sells
Family remuneration collection PSw
Output:Triumph buyer's set WB, winning sellers set WS, match γ, triumph buyer's price set PBw, winning sellers
Remuneration set PSw
When meeting WSThe middle different seller s of any twoα,sβ∈WSCycle is executed when (α ≠ β) condition
If σ (α)=σ (β) is so
IfSo
J' ← the selection from { α, β } at random;
If other situations
WS←WS\{sj'};
The present invention provides a kind of mobile intelligent perception based on two way auction and its resource allocation and incentive mechanism method, strives
There is the urgent user for participating in mobile awareness task to be limited to the sensing capability of equipment of itself some, needs to ask to other equipment
Resource and data are asked, to make up itself sensing capability, and improves the scene of itself credit worthiness.For such scene, participate in using
Family needs to obtain required resource by modes such as transaction to other high-end smart machines, which can be described as both sides by inch of candle
It is traded.The buyer-seller relationship between equipment can be established by two way auction theory thus, to realize low-performance equipment and height
Resource allocation problem between performance equipment.
Description of the drawings
Fig. 1 is individual rationality experimental result picture of the present invention;
Fig. 2 is triumph buyer authenticity experimental result picture of the present invention;
Fig. 3 is buyer's authenticity experimental result picture of the invention of not winning;
Fig. 4 is winning sellers authenticity experimental result picture of the present invention;
Fig. 5 is the non-winning sellers authenticity experimental result picture of the present invention
Fig. 6 is present system effectiveness contrast and experiment figure.
Specific implementation mode
Present invention is further described in detail with specific implementation mode below in conjunction with the accompanying drawings:
The present invention provides a kind of mobile intelligent perception based on two way auction and its resource allocation and incentive mechanism method, this
Invention provides a kind of mobile intelligent perception and its resource allocation and incentive mechanism method based on two way auction, is retouched for above-mentioned
The scene stated carries out resource point by the scene abstract modeling process in pairs to auction trade, and with the model of two way auction
Match and the design of incentive mechanism.
Embodiment 1:In order to verify the calculating validity of DAIM algorithms, set according to the number of such as table 1, each group all random
100 examples are generated, and the operation time of this 100 examples are averaged to ensure the reliability of result.Have to calculating
The test experiments of effect property are carried out in Windows10 systems, and processor isCoreTMi7-6650UCPU@2.20GHz
2.21GHz, RAM 16.0GB.Shown in the setting of experiment and the following table of result.
The step of experiment carries out is that buyer's number n fixed first is 50, changes seller's m numbers from 50 to 400, is with 50
The area of a room is incremented by, and runs DAIM algorithms, wherein each group of record randomly generates 100 experiments, and records the operation of this 100 times experiments
Average time;Later, the number m of fixed seller is 50, changes buyer's number n from 50 to 400, is incremented by for the area of a room with 50, operation
DAIM algorithms, 100 experiments of same each group of carry out, record the average time of operation.It is as shown in the table for run time result.From
As can be seen that DAIM algorithms, which meet, calculates validity in table.
Embodiment 2:In order to verify the individual rationality of DAIM algorithms, in hardware configuration ring identical with above-mentioned calculating validity
DAIM algorithms are run under border, are specifically set as 100 sellers and 25 buyers, analyze successful matching between buyer and seller
Seller later charges, the relationship between buyer's settlement price and buyer's quotation.From attached drawing 1 as can be seen that in DAIM
Each buyer for, be all not above its quotation with the settlement price of auctioner, seller that each is won is from auction
That remuneration obtained of quotient will not be less than his charge.Therefore, DAIM algorithms meet individual rationality.For the knot of individual rationality
Fruit illustrates, when the resource of smart machine itself is not abundant enough, needs to resource supply side request higher calculated performance or more
When height completes quality, resource supply side can obtain enough compensation, therefore can have enough excitations to promote resource supply side
Share resource.On the other hand, smart machine user has collected required resource, and the expense paid thus is not above
The value of these resources itself.Therefore, mobile device user is similarly energized the more rich resource supply side request of whereabouts not
The resource of foot.
Due to being paired with PB for all triumphs in DAIM algorithmsij=PSj, budget equalization also may be used in DAIM algorithms
To ensure.For auctioner, while the resource allocation between resource supply side and smart machine end can be helped, also not
The phenomenon that suffering the deficit.
Embodiment 3:In order to verify the authenticity of DAIM algorithms, in the both parties generated at random randomly choose buyer and
The seller observes the variation of buyer/seller effectiveness caused by the variation of quotation/seller of buyer charged, such as attached drawing 2 to Fig. 5 institutes
Show, buyer and the seller meet the property of authenticity.Experiment is divided into 4 son experiments to carry out, respectively to triumph buyer, do not win
The effectiveness variation of buyer, winning sellers, non-winning sellers carry out experimental study analysis.The setting configuration of experiment such as above-mentioned individual is managed
Property experiment it is identical, be all the true Journal of Sex Research that 100 sellers and 25 buyers carry out DAIM algorithms.
Embodiment 4:The concept of system utility may be defined as the matching of Successful Transaction to quantity, in order to more preferably compare, define
Normalizing system utility is:
Obviously, normalization system utility can preferably embody the effectiveness of performance of two way auction.For actual scene,
If the quotation of buyer is far smaller than charging for seller, the possibility that trade can be successfully established between this both parties is micro-
Its is micro-, under normal circumstances, only the seller charge with buyer quotation difference be not king-sized in the case of, could successfully by
The two matches.
Attached drawing 6 describes DAIM algorithms and the system utility of TASC algorithms introduced in related work before is with Seller Number
The situation of change of variation and variation.Obviously, DAIM algorithms, can reach higher system utility, also, no matter TASC or
DAIM algorithms, system utility are all to keep relatively stable fluctuating change with the variation of Seller Number.The institute of TASC algorithms
To there is more low system utility to essentially consist in, one seller may be divided in the maximum matching algorithm that allocated phase uses
Dispensing one has the buyer of very low quotation, however such distribution is decided by that the stage of price is invalid in last triumph collection
, because such ineffective assignment algorithm, causes the effectiveness of TASC algorithms to be consistently lower than DAIM algorithms.Therefore, this patent proposes
Improvement project be effective.
The DAIM algorithms of simulating, verifying of the present invention design meet the desirable properties of Game Theory description, i.e.,:
(1) validity is calculated.
(2) individual rationality.
(3) budget equalization.
(4) system utility:In order to weigh design two way auction mechanism both parties' successful match situation, with successfully handing over
Easy quantity assesses the system utility of auction.Successful Transaction is defined as the matched number between the buyer and seller finally to win
Amount
(5) authenticity.
The above described is only a preferred embodiment of the present invention, being not the limit for making any other form to the present invention
System, and according to the technical essence of the invention made by any modification or equivalent variations, still fall within present invention model claimed
It encloses.
Claims (5)
1. the mobile intelligent perception and its resource allocation based on two way auction and incentive mechanism method, are as follows, special
Sign is:
(1) system modelling;
Consider there is the m resource supply side i.e. sellers that resource can be provided and n smart machine carrying user i.e. buyer, two
Resource allocation problem between person can be modeled as a unidirectional multi-item double auction model, the offer of each buyer's secret
Their quotation, each seller's secret submits them to charge to auctioner, namely each to take part in auction is not completely
Know the information of all other men;
For each buyer bi∈ B, B={ b1,b2,...,bnFor, each buyer biIt can be with to the quotation vector of the seller
It is expressed asHereIndicate buyer biTo seller sj∈ S, S={ s1,s2,...,smQuotation, report
The matrix of valence contains the quotation vector of all buyers, is denoted as R=(R1;R2;...;Rn), each seller in S is come
It says, the vector of charging of seller is expressed as A=(A1,A2,...,Am), A herejIndicate seller sj∈ S's charges, and buyer is for difference
Seller for, due to buyer can be directed to different mission requirements have the preference of oneself to seller, have to different sellers
Different quotations, and seller, due to only focusing on and sharing the resource of itself and therefrom collect remuneration, charging for seller is
Buyer will not be distinguished, although the quotation of buyer be for seller secrecy, resource supply side before being auctioned,
Still need to provide some specific information such as calculating capacity, data prestige quality and network bandwidth, in order to user
Valuation is carried out to selected resource supply side according to these characteristics, the cost information of resource supply side is kept absolutely secret;
Given vector B, S, R, A, auctioner determine triumph buyerAnd winning sellersWSAnd WBBetween
With relationship γ:{j:sj∈WS}→{i:bi∈WB, select auctioner and triumph buyer b after the both parties of triumphi∈WB
Between settlement price PBi, auctioner pay winning sellers sj∈WSRemuneration PSj, in order to emphasize between triumph both parties
Matching, in certain situations also use PBijAnd PSijIndicate the settlement price that buyer need to pay and the remuneration that seller obtains;
Other than above-mentioned buyer's settlement price and seller's remuneration, the effectiveness of both parties additionally depends on buyer and provides seller
Service it is required other service and cost valuation, enable Vi jIndicate buyer biFrom seller sjThe value serviced, CjTable
Show seller siThe cost of service, buyer b are providediValue vector can be expressed as Vi=(Vi 1,Vi 2,...,Vi m), for one
Buy-sell matching i=γ (j), buyer biWith seller sjEffectiveness indicate as follows:
Also UB is usedijAnd USijIndicate buyer biWith seller sjBetween successful match after effectiveness, it is clear that need ensure effectiveness
UBi> 0 namely smart machine carrying user are collected into the value obtained after the resource from resource supply side as buyer and want
The settlement price of auctioner is paid higher than it, so, UBiIt shows smart machine and carries user's expiring to the resource that is collected into
Meaning degree, for seller, the effectiveness US of the resource supply side as selleriWhat is represented is that the remuneration that it is obtained is more than it
The degree of body cost namely the effectiveness of resource supply side indicate that it shares the resource getable profit of institute later;
(2) algorithm designs;
It is the believable auctioner of third party to manipulate entirely auction in two way auction, which needs in auction according to bat
The agreement mechanism sold determines the buyer's set W to winB, winning seller's set WS, matching relationship between both parties, win
Settlement price set PB between buyerwWith the remuneration PS for paying the triumph sellerw, mathematic(al) representation be Ψ=(B, S, R,
A), it is proposed that a kind of resource allocation mechanism DAIM based on two way auction;
DAIM algorithms are made of two sub- algorithms, i.e., rough candidate's matching algorithm and one-to-one matching algorithm, rough candidate
In people's matching algorithm, rough triumph candidate collection is obtained, and by these set as input, in one-to-one matching algorithm
To final one-to-one matching winning results;
In rough candidate matches algorithm, auctioner is each participation seller s firstjIt determines the candidate of buyer, then determines to need
The settlement price to be collected to buyer and the remuneration for paying seller, here, there are one very crucial benchmark, quotation and rope
Valence needs are compared with benchmark benchmark, to be selected, use A-jIndicate that all sellers charge middle removal seller sj
Vector of charging later.The benchmark compared each time, by vectorial A-jThe average obtained after middle removal maximum value and minimum valueIt indicates;
For seller sjDetermine that the buyer to win is candidate, according to BjThere are two types of situations for the number of middle element;
Work as BjIn element only there are one when, namely only there are one buyer biQuotation be no less than Aj:IfR and k and AjWhen≤benchmark meets simultaneously, then buyer is added to the candidate collection W of buyerB, together
When the price is fixed at benchmark;In the case of other, buyer cannot reach successful transaction with seller;
Work as BjIn element there are two or it is more when, namely have more than a buyer biQuotation be no less than Aj:If this
If highest quotation is less than benchmark in the middle, there is no buyer biThe service of seller sj can be obtained;In the case of other, have
The buyer that highest quotation or if highest quotation the random selection of buyer's more than one one of them, buyer can be added into
Candidate collection WB, while sjIt is added into seller's candidate collection WS, the settlement price that pays needed for the buyer of selection and sell accordingly
The remuneration of family is the maximum value in the high quotations of benchmark and second;
Finding benchmark algorithm is the important evidence of selection triumph dealing side, while being also to determine buyer's settlement price and giving and selling
The important evidence of square remuneration, the selection for benchmark, the influence in order to avoid extreme offer data to benchmark are used for reference simultaneously
The information of charging of other sellers, selection remove seller and charge the data vector that current seller in vector charges after data, reject
The average after maximin in the vector is as benchmark benchmark;
In rough candidate matches algorithm, due to buyer's candidate collection WBIn buyer may match two or more sellers
Candidate collection WSIn seller, it is therefore desirable to execute 3.4 one-to-one matching algorithm of algorithm, determine that only one is optimal for buyer
The seller, auctioner's selection can allow corresponding buyer to reach the seller of maximum utility, likewise, if there is multiple sellers can allow effect
It is maximum with reaching, then randomly choose one of them seller, after having executed one-to-one matching algorithm, each buyer bo(j)∈WB
Have and corresponds matched winning sellers s therewithj∈WS。
2. mobile intelligent perception and its resource allocation and incentive mechanism side according to claim 1 based on two way auction
Method, it is characterised in that:DAIM algorithms are as follows in the step 2 algorithm design:
Input:Buyer set B, seller set S, buyer bid matrix R, seller charge vectorial A;
Output:Triumph buyer's set WB, winning sellers set WS, match γ, triumph buyer's settlement price set PBw, winning sellers
Remuneration set;
First, the assignment from rough candidate matches algorithm (B, S, R, A)
Then, from one-to-one matching algorithmMiddle assignment (WB,WS,γ,PBw,PSw);
Finally obtain WB,WS,γ,PBw,PSwValue.
3. mobile intelligent perception and its resource allocation and incentive mechanism side according to claim 1 based on two way auction
Method, it is characterised in that:Rough candidate matches algorithm is as follows in the step 2 algorithm design:
Input:Buyer set B, seller set S, buyer bid matrix R, seller charge vectorial A;
Output:Buyer's candidate collection WB, seller's candidate collection WS, candidate matchesCandidate buyer's set of prices PBw, candidate seller's report
Reward collection;
First, by WB, WS, PBw, PSwAll it is assigned a value of empty set;
When meeting sjIt is recycled when ∈ S;
Benchmark benchmark=is obtained first finds benchmark algorithm (A),
If BjMiddle element only have 1 so;
IfA simultaneouslyj≤ benchmark is so;
PBij=PSj=benchmark;
PBw←PBw∪{PBij},PSw←PSw∪{PSj};
If BjMiddle element is more than 1;
According to marked price to BjMiddle element is ranked up, i.e.,Collection after note sequence is combined into
IfSo;
IfIn a quotation of preceding t (t >=2) be it is the same so;
At random in BjPreceding t buyer in select bi;
If other situations;
Middle first b for selecting to have highest quotationi;
PBw←PBw∪{PBij},PSw←PSw∪{PSj};
End loop.
4. mobile intelligent perception and its resource allocation and incentive mechanism side according to claim 1 based on two way auction
Method, it is characterised in that:It is as follows that benchmark algorithm is found in the step 2 algorithm design:
Input:Buyer set B, seller set S, buyer bid matrix R, seller charge vectorial A;
Output:Buyer's candidate collection WB, seller's candidate collection WS, candidate matchesCandidate buyer's set of prices PBw, candidate seller's report
Reward collection;
First s is being rejected in vectorial AjForm vector A-j;
Then according to the sequence charged from small to large to A-jIt is ranked up;
And form vector A after rejecting head and the tail element-j;
It finally returns to
5. mobile intelligent perception and its resource allocation and incentive mechanism side according to claim 1 based on two way auction
Method, it is characterised in that:One-to-one matching algorithm is as follows in the step 2 algorithm design:
Input:Buyer's candidate collection WB, seller's candidate collection WS, candidate matchesCandidate buyer's set of prices PBw, candidate seller's report
Reward collection PSw
Output:Triumph buyer's set WB, winning sellers set WS, match γ, triumph buyer's price set PBw, winning sellers remuneration
Set PSw
WB←WB,WS←WS,PBw←PBw,PSw←PSw;
When meeting WSThe middle different seller s of any twoα,sβ∈WSCycle is executed when (α ≠ β) condition
If σ (α)=σ (β) is so
IfSo
J' ← the selection from { α, β } at random;
If other situations
WS←WS\{sj'};
PBw←PBw\{PBσ(j')j'},PSw←PSw\{PSj'},
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810360743.8A CN108596746A (en) | 2018-04-20 | 2018-04-20 | Mobile intelligent perception and its resource allocation based on two way auction and incentive mechanism method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810360743.8A CN108596746A (en) | 2018-04-20 | 2018-04-20 | Mobile intelligent perception and its resource allocation based on two way auction and incentive mechanism method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108596746A true CN108596746A (en) | 2018-09-28 |
Family
ID=63614318
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810360743.8A Pending CN108596746A (en) | 2018-04-20 | 2018-04-20 | Mobile intelligent perception and its resource allocation based on two way auction and incentive mechanism method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108596746A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112907340A (en) * | 2021-02-03 | 2021-06-04 | 西安电子科技大学 | Bidirectional auction model-based incentive method and system in crowd-sourcing perception |
CN114047971A (en) * | 2021-11-09 | 2022-02-15 | 北京中电飞华通信有限公司 | Edge computing resource allocation method and device |
-
2018
- 2018-04-20 CN CN201810360743.8A patent/CN108596746A/en active Pending
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112907340A (en) * | 2021-02-03 | 2021-06-04 | 西安电子科技大学 | Bidirectional auction model-based incentive method and system in crowd-sourcing perception |
CN112907340B (en) * | 2021-02-03 | 2023-12-01 | 西安电子科技大学 | Excitation method and system based on two-way auction model in crowd sensing |
CN114047971A (en) * | 2021-11-09 | 2022-02-15 | 北京中电飞华通信有限公司 | Edge computing resource allocation method and device |
CN114047971B (en) * | 2021-11-09 | 2023-12-08 | 北京中电飞华通信有限公司 | Edge computing resource allocation method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Ockenfels et al. | Online auctions | |
An et al. | Strategic agents for multi-resource negotiation | |
Zhang et al. | Auction-based resource allocation in cognitive radio systems | |
CN110544147A (en) | multitask cross-server resource allocation method based on two-way auction in MEC | |
Sharghivand et al. | A comprehensive survey on auction mechanism design for cloud/edge resource management and pricing | |
Zaman et al. | Combinatorial auction-based mechanisms for vm provisioning and allocation in clouds | |
Hassan et al. | Reputation and user requirement based price modeling for dynamic spectrum access | |
US20160267578A1 (en) | System and method for implementation of sharing economy in a web environment | |
CN105282246B (en) | The method of perception task distribution based on auction mechanism | |
Yue et al. | A double auction-based approach for multi-user resource allocation in mobile edge computing | |
CN108596746A (en) | Mobile intelligent perception and its resource allocation based on two way auction and incentive mechanism method | |
Chichin et al. | Truthful market-based trading of cloud services with reservation price | |
Alahdadi et al. | A truthful and budget-balanced double auction model for resource allocation in cloud computing | |
CN107484174A (en) | The double-deck auction research of cooperative cognitive radio network resources | |
CN111343595B (en) | D2D task unloading system and multi-type resource transaction method thereof | |
Wang et al. | eBay in the clouds: False-name-proof auctions for cloud resource allocation | |
Agmon et al. | Preventing collusion in cloud computing auctions | |
Farajian et al. | Market-driven continuous double auction method for service allocation in cloud computing | |
Li et al. | Resource allocation for mobile blockchain: A hierarchical combinatorial auction approach | |
Zhang et al. | Reverse Auction‐Based Services Optimization in Cloud Computing Environments | |
Wu et al. | Budget-feasible Procurement Mechanisms in Two-sided Markets. | |
Santhiya et al. | Survey on auction based scheduling in grid and cloud environment | |
Zhao et al. | Efficient grid task-bundle allocation using bargaining based self-adaptive auction | |
Chichin et al. | Adaptive market mechanism for efficient cloud services trading | |
Qiao et al. | Mechanism design for cross-market task crowdsourcing |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180928 |