CN111083675A - Resource allocation method based on social perception in industrial Internet of things - Google Patents

Resource allocation method based on social perception in industrial Internet of things Download PDF

Info

Publication number
CN111083675A
CN111083675A CN201911221730.3A CN201911221730A CN111083675A CN 111083675 A CN111083675 A CN 111083675A CN 201911221730 A CN201911221730 A CN 201911221730A CN 111083675 A CN111083675 A CN 111083675A
Authority
CN
China
Prior art keywords
spj
social
sri
resource allocation
buyer
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
Application number
CN201911221730.3A
Other languages
Chinese (zh)
Inventor
孙文
刘家佳
郭鸿志
张海宾
岳燕林
宋强
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Northwestern Polytechnical University
Original Assignee
Northwestern Polytechnical University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Northwestern Polytechnical University filed Critical Northwestern Polytechnical University
Priority to CN201911221730.3A priority Critical patent/CN111083675A/en
Publication of CN111083675A publication Critical patent/CN111083675A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/70—Services for machine-to-machine communication [M2M] or machine type communication [MTC]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/02—Traffic management, e.g. flow control or congestion control
    • H04W28/0215—Traffic management, e.g. flow control or congestion control based on user or device properties, e.g. MTC-capable devices
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/02—Traffic management, e.g. flow control or congestion control
    • H04W28/0226—Traffic management, e.g. flow control or congestion control based on location or mobility
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W40/00—Communication routing or communication path finding
    • H04W40/02—Communication route or path selection, e.g. power-based or shortest path routing
    • H04W40/22—Communication route or path selection, e.g. power-based or shortest path routing using selective relaying for reaching a BTS [Base Transceiver Station] or an access point
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W40/00—Communication routing or communication path finding
    • H04W40/24—Connectivity information management, e.g. connectivity discovery or connectivity update
    • H04W40/32—Connectivity information management, e.g. connectivity discovery or connectivity update for defining a routing cluster membership

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention provides a resource allocation method based on social perception in an industrial Internet of things, which comprises the following steps: step 1, dividing IIoT MD into different physical clusters according to the proximity degree of geographic positions; step 2, utilizing an OHSSIM excitation mechanism or an RSIM excitation mechanism to carry out resource allocation on each physical cluster formed in the step 1; the invention considers that the tasks are unloaded through social perception and distance perception simultaneously, the participation of SR and SP is stimulated, the system efficiency can be improved to the maximum extent, and the resource sharing is carried out by combining network economy, distance perception and social perception under the D2D mode, so that the social benefit is maximized, and the defect of low efficiency of the resource sharing between the existing IIoTMD is overcome.

Description

Resource allocation method based on social perception in industrial Internet of things
Technical Field
The invention belongs to the field of industrial Internet of things, and particularly relates to a resource allocation method based on social perception in the industrial Internet of things.
Background
Industrial internet of things (IIoT) has undergone tremendous changes in the past few years as an application of the internet of things in industrial scenarios. People develop a plurality of industrial Internet of things projects in the fields of agriculture, food processing industry, environmental monitoring, safety supervision and the like. With the continuous emergence of a large number of IIoT Mobile Devices (MD), such as industrial monitors, industrial robots, industrial sensors, the large amount of traffic generated by these MDs puts a tremendous strain on the Radio Access Network (RAN). Furthermore, in the field of industrial automation control, closed-loop control and interlocking can only tolerate delays in the order of milliseconds (10-100 milliseconds) and transmission reliability higher than 99.99%, which presents challenges to existing network architectures.
Resource sharing between iiomds provides a new perspective to solve the above problems. IIoT MDs may be interconnected by D2D technology for data transfer and computation. Resource-constrained MDs can achieve high quality of service, e.g., high data rates, low latency, by offloading tasks directly to nearby resource-rich MDs.
While resource sharing between MDs has significant application prospects in IIoT, it also faces formidable challenges, one of which is network economy for task offloading. For task offloading, the service demander (SR, i.e. IIoT MD demanding the computing resource) and the service provider (SP, i.e. IIoT MD contributing the resource) belong to two different camps and both are profit driven. Given the limitations of IIoT MD in terms of energy and resources, in unprofitable situations, SPs with free resources are reluctant to provide their resources and SRs do not offload tasks to SPs. On the other hand, a significant portion of iiomds are carried manually or mounted on manned machines, which inevitably exhibit social relationships (such as friendship, consanguineous relationships, and coworkers) and regularity in the process of communicating with each other over D2D links. When an IIoT MD is close to other IIoT MDs and can satisfy the D2D link connection condition, it always prefers to communicate with IIoT MDs that are close in distance or socially close.
Existing work, while providing some insight into resource sharing and task offloading, does not consider network economics, distance awareness, and social relationships between IIoT MDs collectively.
Disclosure of Invention
The invention aims to provide a resource allocation method based on social sensing in an industrial Internet of things, which considers that tasks are unloaded through social sensing or distance sensing at the same time, stimulates the participation of SR and SP, can improve the system efficiency to the maximum extent, and utilizes the combination of network economy, distance sensing and social sensing to share resources under the D2D mode so as to maximize social benefits, thereby overcoming the defect of low efficiency of resource sharing among the existing IIoTMD.
In order to achieve the purpose, the invention adopts the technical scheme that:
the invention provides a resource allocation method based on social perception in an industrial Internet of things, which comprises the following steps:
step 1, dividing IIoT MD into different physical clusters according to the proximity degree of geographic positions;
and 2, performing resource allocation on each physical cluster formed in the step 1 by utilizing an OHSSIM excitation mechanism.
Preferably, in step 1, the IIoT MD is divided into different physical clusters according to the proximity of the geographic location, and the specific method is as follows:
each SP and the SR directly connected with the SP form a physical cluster.
Preferably, in step 2, resource allocation is performed on each physical cluster formed in step 1 by using an OHSIM excitation mechanism, and the specific method is as follows:
step 1, winner selection phase
Carrying out auction on the SPs in sequence according to the principle that the resource capacity is from large to small; collect all SRi, i ∈ NjBid b for SPj of the current auctioni,jSelecting a buyer winner using a knapsack algorithm, resulting in SPj of the best combination of buyer winners WjSocial benefits corresponding to SPj of the current auction
Figure BDA0002301045290000021
Wherein N isjSPj is the set of SRs of the physical cluster to which it belongs;
step 2, price decision stage
Calculating an optimal combination of buyer winners WjThe actual payment price p of SPj for each SRi pair ini。
Preferably, an optimal combination W of buyer winners is calculatedjBid b for SPj for each SRi ini,jThe specific method comprises the following steps:
the bid b for SRi pair SPj is calculated byi,j：
bi,j＝λ1ri,j+λ2ωi,j, (6)
Wherein λ is1，λ2Is a parameter and λ1+λ2＝1；ri,jIs the maximum data rate between different IIoT MDs; omegai,jSocial strength among different IIoT MDs;
calculating an optimal combination of buyer winners WjThe actual payment price p of SPj for each SRi pair iniThe specific method comprises the following steps:
Figure BDA0002301045290000031
a resource allocation method based on social perception in an industrial Internet of things comprises the following steps:
step 1, dividing IIoT MD into different physical clusters according to the proximity degree of geographic positions;
and 2, carrying out resource allocation on each physical cluster formed in the step 1 by utilizing an RSIM excitation mechanism.
Preferably, in step 1, the IIoT MD is divided into different physical clusters according to the proximity of the geographic location, and the specific method is as follows:
each SP and the SR directly connected with the SP form a physical cluster.
Preferably, in step 2, resource allocation is performed on each physical cluster formed in step 1 by using an RSIM incentive mechanism, and the specific method is as follows:
step 1, relay path selection phase
When the relay is needed from the SRi to SPj, the selection rule of the relay path is to select the relay node t and make the SRi bid the maximum through the node pair SPj;
step 2, winner selection phase
Carrying out auction on the SPs in sequence according to the principle that the resource capacity is from large to small; collect all SRi, i ∈ NjBid b for SPj of the current auctioni,jSelecting a buyer winner using a knapsack algorithm, resulting in SPj of the best combination of buyer winners WjSocial benefits corresponding to SPj of the current auction
Figure BDA0002301045290000032
Wherein N isjSPj is the set of SRs of the physical cluster to which it belongs;
step 3, price decision stage
Calculating an optimal combination of buyer winners WjThe actual payment price p of SPj for each SRi pair ini。
Preferably, an optimal combination W of buyer winners is calculatedjBid b for SPj for each SRi ini,jThe specific method comprises the following steps:
the bid b for SRi pair SPj is calculated byi,j：
bi,j＝max{λ1(ωi,t·ωt,j)+λ2(min(ri,t,rt,j)
Wherein λ is1，λ2Is a parameter and λ1+λ2＝1。
Preferably, an optimal combination W of buyer winners is calculatedjThe actual payment price p of SPj for each SRi pair iniThe specific method comprises the following steps:
Figure BDA0002301045290000041
compared with the prior art, the invention has the following beneficial effects:
according to the resource allocation method based on social sensing in the industrial Internet of things, the MD resource allocation scene in the industrial Internet of things under the constraints of distance sensing and social sensing is considered, in the existing D2D network, the SP can allocate the computing resources to the IIoT MDs, and due to the fact that data are directly transmitted between terminals, base station pressure is relieved, data transmission rate is improved, and meanwhile end-to-end transmission delay is reduced. When the wireless communication infrastructure is damaged or in a coverage blind area of a wireless network, the terminal can realize end-to-end communication by means of D2D and even access the cellular network, so that the link flexibility and the network reliability are improved;
the present invention takes into account that most IIoT MDs are carried or manipulated by humans and that they inevitably have human-like social attributes, taking social awareness will take into account the social relationships between MDs, which can not only improve the quality of service (e.g., success rate of task offloading) but also reduce the security risk.
In addition, according to the resource allocation method based on social perception in the industrial Internet of things, RSIM is utilized, SR can be forwarded through a relay MD to be connected to a remote SP demand resource, the SR of the D2D link directly connected with the SP can bid on the SP resource, and the SR connected with the SP through a relay node can also participate in bidding. Therefore, for a specific SP, the corresponding number of buyers will increase, and the SR also increases the corresponding selection space to improve the resource sharing efficiency.
Drawings
FIG. 1 is a IIoT mobile device resource allocation system architecture in an embodiment of the present invention;
FIG. 2 is a diagram of a physical cluster of IIoT mobile devices in an embodiment of the present invention;
FIG. 3 is a graph of bid price versus utility for the OHISIM and RSIM methods of the present invention;
FIG. 4 is a graph comparing SP resource utilization for 10 experiments of OHSIM and RSIM methods of the present invention;
FIG. 5 is a line graph showing the social benefit of the present invention with increasing number of SRs under different mechanisms;
FIG. 6 is a line graph showing the social benefit of the present invention with increasing number of SPs under different mechanisms;
figure 7 is a line graph of the task offload rate as a function of increased social strength for the OHSIM and RSIM methods of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings.
In order to solve the problem that resources are shared by MD (machine direction) based on social perception in the industrial Internet of things and solve one or more technical problems, the invention provides two task unloading modes, namely a single-hop social perception incentive mechanism (OHSI) and a one-step relay social perception incentive mechanism (RSIM), wherein the OHSI unloads a task to the MD with rich resources nearby under the single-hop limit by using a Vickrey-Clarke-Groves (VCG) auction; RSIM relaxes the transmit node constraints to two hops to achieve higher resource utilization.
The RSIM further improves the resource utilization rate, expands the range of task unloading, and unloads the tasks to other indirectly connected SPs through the relay equipment. Offloading tasks to indirectly connected SPs through relay nodes not only consumes channel resources and energy of the relay device, but also selection of relay paths is an NP-hard problem. Thus, the RSIM only considers the two-hop case, i.e. only one relay device is selected during task offloading. It is easy to prove that in the two-hop case, the time complexity of relay path selection is O (m + k).
For convenience of describing the specific implementation of the implementation example, the description of the model background related to the present application is first made:
as long as the distance between the two devices is less than the threshold value dmaxD2D links can be established between them, and the channel allocated to each physical link is orthogonal to the channels of the other D2D links under the control of the base station, so that interference between any two D2D links can be eliminated.
As shown in fig. 1, a base station is provided, which is a resource sharing scenario composed of several SPs and SRs, wherein a physical domain and a social domain are considered together.
As shown in fig. 1, in the physical domain, the physical link (i.e., D2D link) connects MDs that are physically located very close together, and the tasks of the SR are offloaded directly to the SP through the D2D link underneath the cellular network.
Social relations of social strength exist between IIoT MD in the social domain, and strong social relations can reduce safety risks brought by task unloading and improve QoS (quality of service); thus, when the SR offloads tasks to SPs over physical links, SPs with strong social relationships to the SR will be preferentially selected, e.g.:
assume that N IIoT MDs are involved, including M SRs and K SPs, and the IIoT MD set, SR set, and SP set are denoted as N ═ {1, 2.., N }, M ═ 1, 2.., M }, and K ═ 1, 2.., K }.
For each D2D link li,j(i, j. epsilon. N), B is seti,jIndicates assignment to li,jBandwidth of Ci,jIs represented byi,jA set of transmitting devices having the same cellular link signal frequency; then, D2D link li,jAnd the in-band interference between the cellular links can be expressed as:
Figure BDA0002301045290000061
wherein, PcIs the transmit power of the transmitting device c; dc,jIs when h isc,jIs the distance between c and j when the channel from c to the receiving device j responds; h isc,jIs the channel response from c to the receiving device j and α is a path loss exponent greater than 2.
In view of the above interference, the maximum data rate between different IIoT MDs is:
Figure BDA0002301045290000062
wherein, PiIs the transmit power, σ, of SRijIs SPj additive white gaussian noise.
Note that r is r if there is no direct D2D link between i and ji,jIt cannot be directly calculated by (2).
For example, in fig. 1, there is no direct D2D link between SR1 and SP1, and if SR1 wishes to offload tasks to SP1, then other nodes are required for relaying. Calculating between SR1 and SP1 if SR2 is selected as the relay nodeA more reasonable formula for the data rate would be rSR1，SP1＝min{rSR1，SR2，rSR2，SP1}。
Using Pi，jRepresenting the relay route from i to j, with its corresponding data rate r (P)i，j) Comprises the following steps:
r(Pi，j)＝min{ra，b|(a，b)∈Pi，j} (3)
wherein (a, b) is contained in Pi，jD2D link in (1), if i to j have no relay path, ri，j＝0。
The present application models social relationships as a weighted undirected graph G (V, E, W), where V ═ N is the IIoT device set and E ═ Ei，jI, j belongs to V and represents a social edge set; e indicates whether there is a social relationship between the devices, i.e. if there is a social relationship between two devices, they are connected with one edge. To quantitatively analyze the closeness of social relationships, each edge is assigned a range of (0, 1)]The scalar value of (a), i.e., the strength of the social relationship.
Since G is an undirected graph, W is a symmetric matrix. For the
Figure BDA0002301045290000071
If there is no social relationship between IIoT MDi and IIoT MD j, then ω isi，j＝ωj，i＝0。
In addition, setting the social strength of any IIoT device with itself to 0, i.e., ωi，i＝0，i∈V。
The method for calculating the social strength uses the Jaccard similarity coefficient as a method for calculating the social strength if an edge e exists between i and ji，jThen i and j are neighbors. We denote Γ (i) and Γ (j) as the neighbor sets of devices i and j, respectively, and then define the Jaccard similarity coefficient as the ratio of the number of common neighbors of i and j to their total number of neighbors, i.e.
Figure BDA0002301045290000072
If it is not
Figure BDA0002301045290000073
ωi，jThe Jaccard similarity coefficient is a visual representation of the strength of social interaction between the two IIoT MDs, 0.
However, if there is no direct D2D link between i and j, ωi，jIt cannot be directly calculated with the Jaccard similarity coefficient.
For example, in fig. 1, there is no direct D2D link between SR1 and SP1, assuming SR1 wishes to offload tasks to SP1, then other nodes are needed for relaying, if SR2 is selected as the relay node, then the more reasonable formula for calculating the social strength between SR1 and SP1 would be:
ωSR1，SR2×ωSR2，SP2
by Pi,jTo represent the relay path from i to j, the corresponding social strength ω (P)i,j) Is the product of all intensities on the path, i.e.
Figure BDA0002301045290000081
Wherein (a, b) is contained in Pi,jD2D link in (e), if i to j have no relay path, then ω (P)i,j)＝0。
The Vickrey-Clarke-Groves (VCG) auction is a multi-item auction mechanism in which each buyer submits a quantity of a desired item and a corresponding bid price willing to be paid for the item. The bidding information of all buyers is sealed and can be disclosed only after the auction is finished, and the auctioneer provides the buyers with commodities according to the principle of maximizing social benefit, and the price paid by each buyer winner is not higher than the corresponding bid.
W is expressed as a set of buyer winners, WjA set of buyer winners representing SPj resources are won, then WjThe following limitations need to be met:
1) the resources required for the SR are less than or equal to the resource capacity of the SP, i.e.
Figure BDA0002301045290000082
Wherein r isiIs the amount of resource, R, required by the SRijIs a resource capacity of SPj.
2) When there is a D2D path between SRi and SPj, bi,jRepresenting the bid of SRi for SPj. The bids by SRi for the seller's resources depend on social strength and channel status.
Generally, the higher the social strength, the faster the data transfer rate, and the more efficient the task offloading.
Thus, if there is a direct D2D link between i and j, the bid price can be represented by the following formula
bi,j＝λ1ri,j+λ2ωi,j, (6)
Wherein λ is1，λ2Is a parameter and λ1+λ2＝1。
The data rates used herein are standardized in order to correspond to social relationships. Unless otherwise specified, the data rate is defaulted to normalized.
If relaying is required from i to j, then a parameter λ is given1And λ2We select λ1r(Pi,j)+λ2ω(Pi,j) The path with the maximum value is taken as the unloading path, and the corresponding bids are:
bi,j＝max{λ1r(Pi,j)+λ2ω(Pi,j)|Pi,j∈Si,j}, (7)
wherein S isi,jIs the set of all paths from i to j.
When SRi gets SPj resource, it needs to pay SPj the corresponding price, and the effect of SRi is
ui＝bi,j-pi(8)
Wherein p isiIndicates the price that SRi needs to pay, if
Figure BDA0002301045290000091
uiWill be 0, meaning that the SRi bid failed.
The goal of resource sharing among IIoT MDs is to generate buyer winners to maximize social benefit, i.e.
Figure BDA0002301045290000092
Where M is a set of SRs, uiIs the utility of SRi, K is the SP set, riIs the amount of resource, R, required by the SRijIs a resource capacity of SPj.
The application provides two social awareness incentive task offloading methods, namely a single-hop social awareness incentive mechanism (OHSIM) and a one-step relay social awareness incentive mechanism (RSIM). Prior to the auction, we divided IIoT MDs into different physical clusters based on the proximity of geographic locations.
As shown in fig. 2, each SP and its directly connected SR form a physical cluster, the SP being the center of the cluster. The auction process will be conducted separately in each cluster.
The ohsim excitation mechanism steps are as follows:
1) winner selection stage
In the winner selection phase, we auction according to the resource capacity, and the SP with the largest resource capacity is auctioned first. When an SP's resource auction begins, all D2D buyers directly linked to the SP submit their bids to compete for the SP's resources.
The auctioneer selects a winner based on each buyer's bid and corresponding resource requirements to maximize social benefit.
The winner selection process may be equivalent to a knapsack problem, where the seller's resource capacity represents the maximum weight that the knapsack can withstand, the buyer's resource demand represents the weight of the item, and the buyer's bid represents the corresponding value of the item. Therefore, a dynamic programming based knapsack algorithm can be used for solving. Under the constraint of limited resources, the knapsack algorithm can calculate the optimal combination of winners of each physical cluster, thereby obtaining the maximum social benefit.
It should be noted that if an SP's SR neighbor has offloaded its own tasks at another seller, the SR can no longer participate in the auction.
When the SR has offloaded the task to the SP, there is no need to re-participate in other auctions, and if re-participate in an auction it needs to pay a premium without revenue, so no other agreement is needed to constrain it.
For example, in fig. 2, the resource capacity of SP3 is greater than the resource capacity of SP2, so the SP3 resource is auctioned first.
SR3 is a common neighbor of SP3 and SP2, and if the tasks of SR3 have been offloaded to SP3, SR3 will no longer participate in the resource auction of SP 2.
For SPj, after the first stage, the best combination W of buyer winners is obtainedjAnd corresponding social benefits
Figure BDA0002301045290000101
Wherein N isjIs a collection of SRs belonging to SPj physical cluster ranges.
2) Price decision phase
In the price decision phase, the price paid by each winner is marginal damage to other buyers, and therefore the price that SRi eventually needs to pay is calculated using marginal damage, specifically:
assume that there are no SRi in the resource auction at SPj (i ∈ N)j) Then, in the re-operation winner selection phase and calculating the operation result corresponding to SP social benefit, SPj is obtained as social benefit
Figure BDA0002301045290000102
Thus, the price ultimately paid by SRi may be expressed as:
Figure BDA0002301045290000103
because of WjIs an optimum combination, and is obtained from equation (10) assuming that i does not participate in the auction
Figure BDA0002301045290000104
If it is not
Figure BDA0002301045290000105
Then p isi＝bi,j. Otherwise, pi<bi,j(ii) a In other words, ui≧ 0, so OHSIM is individual rationality.
RSIM incentive mechanism steps are as follows:
1) relay path selection phase
In the relay path selection stage, the path selection problem under the two-hop constraint is only considered, so the path selection problem is reduced to the selection problem of a single relay node. Therefore, if relaying is required from SRi to SPj, the selection rule for the relay path is to select the relay node t to maximize the bid of SRi through this node pair SPj. From equations (3), (5), and (7), a final expression for the bid of SRi to SPj can be derived
bi,j＝max{λ1(ωi,t·ωt,j)+λ2(min(ri,t,rt,j) (11)
2) Winner selection stage
In the winner selection phase, the auctioneer needs to consider how to allocate the resources of the SPs according to the bid price and resource requirement information provided by the SR to maximize social benefits.
It should be noted that at this point, the physical cluster of each SP will extend to a two-hop range; for example, in fig. 2, SR1 may connect to SP2 through relay node SR2, and thus SR2 belongs to NSP2。
In RSIM, the knapsack algorithm is still used to obtain the optimal winner group that maximizes social benefit. The seller's resource capacity represents the maximum weight that the backpack can bear, the buyer's resource demand represents the weight of the item, and the buyer's bid represents the corresponding value of the item. As previously described, if the SR has offloaded its own tasks at another SP, the SR does not participate in the auction. During the auction process, the seller with the largest resource capacity is still in line with the principle of selling the resource preferentially. After the end of this phase, we can get the optimal winner combination WjAnd corresponding social benefits
Figure BDA0002301045290000111
3) Price decision phase
In the price decision stage, each winnerThe price paid is a marginal detriment to other buyers. If SRi wins SPj, the price p that SRi must pay can be derived according to equation 10i. Because of WjIs the optimal combination, i, assuming it is not participating in the auction, must have
Figure BDA0002301045290000112
If it is not
Figure BDA0002301045290000113
Then p isi＝bi,j. Otherwise, pi<bi,j. In other words, ui> 0, so RSIM is idiosyncratic.
The above-described method of the present invention is further illustrated by the following specific example.
Assume that 50 MDs are uniformly distributed in a space of 50m × 50m and that the BS is located at the center of the space. The maximum distance between devices capable of directly establishing the D2D link is Dmax30 m. The transmission power for all MD was 100 mW. The carrier bandwidth is set to 1MHz and the noise variance is set to 6X 10-10mW. Path loss exponent α is 4 the task size of each SR and the resource capacity of each SP are evenly distributed [1,4 ] respectively]And [4,10]The above. Lambda [ alpha ]1And λ2Set to 0.5. Social relationships of 50 MDs were constructed using kumpla model. To visualize the impact of social relationships on the performance of the proposed mechanism, we devised two corresponding mechanisms: single hop socieless information incentive mechanisms (OIWSI) and two hop socieless information incentive mechanisms (TIWSI). These two mechanisms were compared to OHSIM and RSIM, respectively.
To verify the authenticity of the OHSIM and RSIM, we randomly choose an SR to check how its utility varies at different bid prices.
As shown in fig. 3, in OHSIM, the real valuation of SR to SP service is 0.86, while in RSIM mechanism, the real valuation of SR to SP service is 0.72. When an SR submits a bid that is below its true value, its utility is zero, meaning it is a loser in the auction and cannot win the SP's service. When the SR submits a bid that is higher than its true value, the utility is a constant and does not exceed the utility received for its bid price as a true valuation. SR cannot gain additional revenue by submitting unreal prices, and therefore, it has no incentive to deliberately submit unreal bid prices.
FIG. 4 depicts SP resource utilization over 10 experiments. We randomly regenerated new simulation parameters before each experiment. Resource utilization was recorded after each experiment. Resource utilization refers to the ratio of the number of resources utilized by the SR to the original free resource capacity. In the block diagram of fig. 4, each block indicates that the resource utilization of the RSIM is always greater than that of the OHSIM in each experiment. After statistics on the resource utilization of the 10 experiments, the average utilization of OHSIM is about 0.64, and the average utilization of RSIM is about 0.75. Since we have established the D2D link according to the distance between IIoT MDs, if one MD is too far away from other MDs, a physical isolation point is formed, that is, the idle resources of the MD cannot be utilized, which results in a decrease in resource utilization.
To evaluate the impact of social relationships on task offloading success rates, we randomly selected SRi for verification. We iterate the proposed mechanism 100 times, fixing first the social strength between SRi and SPj in the same cluster as SRi (i.e., fixing ωi,j) Then at (0, 1)]Other elements in the W matrix are randomly generated within range (it is worth emphasizing that the element with value 0 is not altered in order to maintain social topology). In these 100 experiments, we recorded the number of times (in s) that the SRi was successfully servicediExpressed), the task offload success rate for SRi is defined as: si/100. By modifying ω a number of timesi,jThen, by performing the simulation in the above-described manner, we can plot a curve as shown in fig. 7. As can be seen from FIG. 7, as social relationships strengthen, the success rate of task offloading is also increasing. When the social relationship strength is equal to 1, the success rate of task offloading is less than 100%. This is because in our solution the proportion of social relations is only 0.5(λ)20.5). If the corresponding data rate ri,jToo small, it may not be able to offload tasks to SP j.
Fig. 5 depicts how the social benefit of the system varies with the number of SRs when the number of SPs is fixed (10 SPs). In the transverse direction, as the number of SR increases, the social benefit increases, but the growth rate decreases. This is because the resource capacity of the SP is gradually exhausted. In the vertical view, for the TIWSI mechanism and the OIWSI mechanism, since social information is not considered, their respective bid prices will be reduced according to equations 6 and 7. Therefore, TIWSI and OIWSI have a lower level of social benefit than RSIM and OHSIM. However, it can be observed that a higher social benefit can be obtained in the two-hop case, whether or not social information is considered, which indicates that the resources of the SP can be more fully utilized and more SRs can be served by the relay. In fig. 6, when the number of buyers remains the same (30 SRs), the corresponding social benefit increases as the number of sellers increases. With the increase of resources, the situation of over supply and over demand finally occurs, and the growth speed of social benefits is gradually slowed down. Similar to fig. 5, in vertical comparison, considering that social information leads to more social benefits, the performance of the two-hop mechanism is always better than that of the single-hop mechanism.
Table 1 is a symbol mapping table in the embodiment of the present invention, and please refer to Table 1 for symbols appearing in the embodiment of the present invention
Figure BDA0002301045290000131

Claims (9)

1. A resource allocation method based on social perception in an industrial Internet of things is characterized by comprising the following steps:
step 1, dividing IIoT MD into different physical clusters according to the proximity degree of geographic positions;
and 2, performing resource allocation on each physical cluster formed in the step 1 by utilizing an OHSSIM excitation mechanism.
2. The resource allocation method based on social perception in the industrial internet of things as claimed in claim 1, wherein in step 1, IIoT MD is divided into different physical clusters according to the proximity of geographic location, and the specific method is as follows:
each SP and the SR directly connected with the SP form a physical cluster.
3. The resource allocation method based on social perception in the industrial internet of things as claimed in claim 1, wherein in step 2, resource allocation is performed on each physical cluster formed in step 1 by using an OHSIM incentive mechanism, and the specific method is as follows:
step 1, winner selection phase
Carrying out auction on the SPs in sequence according to the principle that the resource capacity is from large to small; collect all SRi, i ∈ NjBid b for SPj of the current auctioni，jSelecting a buyer winner using a knapsack algorithm, resulting in SPj of the best combination of buyer winners WjSocial benefits corresponding to SPj of the current auction
Figure FDA0002301045280000011
Wherein N isjSPj is the set of SRs of the physical cluster to which it belongs;
step 2, price decision stage
Calculating an optimal combination of buyer winners WjThe actual payment price p of SPj for each SRi pair ini。
4. The method for resource allocation based on social perception in the industrial internet of things as claimed in claim 3, wherein the optimal combination W of the buyer winner is calculatedjBid b for SPj for each SRi ini，jThe specific method comprises the following steps:
the bid b for SRi pair SPj is calculated byi，j：
bi，j＝λ1ri，j+λ2ωi，j， (6)
Wherein λ is1，λ2Is a parameter and λ1+λ2＝1；ri，jIs the maximum data rate between different IIoT MDs; omegai，jSocial strength among different IIoTMD;
calculating optimal set of buyer winnersAnd WjThe actual payment price p of SPj for each SRi pair iniThe specific method comprises the following steps:
Figure FDA0002301045280000012
5. a resource allocation method based on social perception in an industrial Internet of things is characterized by comprising the following steps:
step 1, dividing IIoT MD into different physical clusters according to the proximity degree of geographic positions;
and 2, carrying out resource allocation on each physical cluster formed in the step 1 by utilizing an RSIM excitation mechanism.
6. The resource allocation method based on social perception in the industrial internet of things as claimed in claim 5, wherein in the step 1, the IIoT MD is divided into different physical clusters according to the proximity of the geographic location, and the specific method is as follows:
each SP and the SR directly connected with the SP form a physical cluster.
7. The resource allocation method based on social perception in the industrial internet of things according to claim 5, wherein in the step 2, resource allocation is performed on each physical cluster formed in the step 1 by using a RSIM incentive mechanism, and the specific method is as follows:
step 1, relay path selection phase
When the relay is needed from the SRi to SPj, the selection rule of the relay path is to select the relay node t and make the SRi bid the maximum through the node pair SPj;
step 2, winner selection phase
Carrying out auction on the SPs in sequence according to the principle that the resource capacity is from large to small; collect all SRi, i ∈ NjBid b for SPj of the current auctioni，jSelecting a buyer winner using a knapsack algorithm, resulting in SPj of the best combination of buyer winners WjSocial benefits corresponding to SPj of the current auction
Figure FDA0002301045280000021
Wherein N isjSPj is the set of SRs of the physical cluster to which it belongs;
step 3, price decision stage
Calculating an optimal combination of buyer winners WjThe actual payment price p of SPj for each SRi pair ini。
8. The method of claim 7, wherein the optimal combination W of buyer winners is calculatedjBid b for SPj for each SRi ini，jThe specific method comprises the following steps:
the bid b for SRi pair SPj is calculated byi，j：
bi，j＝max{λ1(ωi，t·ωt，j)+λ2(min(ri，t，rt，j))}
Wherein λ is1，λ2Is a parameter and λ1+λ2＝1。
9. The method of claim 7, wherein the optimal combination W of buyer winners is calculatedjThe actual payment price p of SPj for each SRi pair iniThe specific method comprises the following steps:
Figure FDA0002301045280000031
CN201911221730.3A 2019-12-03 2019-12-03 Resource allocation method based on social perception in industrial Internet of things Pending CN111083675A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201911221730.3A CN111083675A (en) 2019-12-03 2019-12-03 Resource allocation method based on social perception in industrial Internet of things

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201911221730.3A CN111083675A (en) 2019-12-03 2019-12-03 Resource allocation method based on social perception in industrial Internet of things

Publications (1)

Publication Number Publication Date
CN111083675A true CN111083675A (en) 2020-04-28

Family

ID=70312605

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201911221730.3A Pending CN111083675A (en) 2019-12-03 2019-12-03 Resource allocation method based on social perception in industrial Internet of things

Country Status (1)

Country Link
CN (1) CN111083675A (en)

Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2000057323A1 (en) * 1999-03-23 2000-09-28 The Trustees Of Columbia University In The City Of New York System and method for performing a progressive second price auction technique
US20050154639A1 (en) * 2004-01-09 2005-07-14 Zetmeir Karl D. Business method and model for integrating social networking into electronic auctions and ecommerce venues.
CN105282246A (en) * 2015-10-23 2016-01-27 上海交通大学 Method for distributing perception tasks based on auction mechanism
CN106485415A (en) * 2016-10-11 2017-03-08 安徽慧达通信网络科技股份有限公司 A kind of based on the mobile intelligent perception motivational techniques with budget for the relation between supply and demand
US20180034758A1 (en) * 2008-08-19 2018-02-01 Facebook, Inc. Resource management of social network applications
CN108200610A (en) * 2018-02-26 2018-06-22 重庆邮电大学 Take the intelligent perception resource allocation methods of distributed game
CN108269129A (en) * 2018-01-23 2018-07-10 重庆邮电大学 User's motivational techniques in a kind of mobile gunz sensing network based on reverse auction
CN108364198A (en) * 2018-02-12 2018-08-03 南京邮电大学 A kind of online motivational techniques of mobile crowdsourcing based on social networks
CN109492887A (en) * 2018-10-25 2019-03-19 浙江工商大学 Mobile intelligent perception motivational techniques based on gesture theory of games
US20210264451A1 (en) * 2020-02-20 2021-08-26 International Business Machines Corporation Cognitive evaluation of sensor environments and resource allocation
US20220022172A1 (en) * 2019-01-02 2022-01-20 Nokia Technologies Oy Industrial automation system communication

Patent Citations (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2000057323A1 (en) * 1999-03-23 2000-09-28 The Trustees Of Columbia University In The City Of New York System and method for performing a progressive second price auction technique
US20050154639A1 (en) * 2004-01-09 2005-07-14 Zetmeir Karl D. Business method and model for integrating social networking into electronic auctions and ecommerce venues.
US20180034758A1 (en) * 2008-08-19 2018-02-01 Facebook, Inc. Resource management of social network applications
CN105282246A (en) * 2015-10-23 2016-01-27 上海交通大学 Method for distributing perception tasks based on auction mechanism
CN106485415A (en) * 2016-10-11 2017-03-08 安徽慧达通信网络科技股份有限公司 A kind of based on the mobile intelligent perception motivational techniques with budget for the relation between supply and demand
CN108269129A (en) * 2018-01-23 2018-07-10 重庆邮电大学 User's motivational techniques in a kind of mobile gunz sensing network based on reverse auction
CN108364198A (en) * 2018-02-12 2018-08-03 南京邮电大学 A kind of online motivational techniques of mobile crowdsourcing based on social networks
CN108200610A (en) * 2018-02-26 2018-06-22 重庆邮电大学 Take the intelligent perception resource allocation methods of distributed game
CN109492887A (en) * 2018-10-25 2019-03-19 浙江工商大学 Mobile intelligent perception motivational techniques based on gesture theory of games
US20220022172A1 (en) * 2019-01-02 2022-01-20 Nokia Technologies Oy Industrial automation system communication
US20210264451A1 (en) * 2020-02-20 2021-08-26 International Business Machines Corporation Cognitive evaluation of sensor environments and resource allocation

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
WEN SUN: ""Social-Aware Incentive Mechanisms for D2D"", 《IEEE》 *
刘媛妮等: ""移动群智感知激励机制研究综述"", 《重庆邮电大学学报》 *

Similar Documents

Publication Publication Date Title
Iosifidis et al. An iterative double auction for mobile data offloading
Li et al. Incentive mechanisms for device-to-device communications
Xu et al. Resource allocation using a reverse iterative combinatorial auction for device-to-device underlay cellular networks
Wang et al. Distributed relay selection and power control for multiuser cooperative communication networks using Stackelberg game
CN107249217A (en) The Joint Task unloading of ad hoc mobile cloud network and resource allocation methods
Zhang et al. Double auction based multi-flow transmission in software-defined and virtualized wireless networks
CN110311723B (en) Pricing strategy-based computing resource allocation method for lunar space station communication system
Zhang et al. A double auction mechanism for virtual resource allocation in SDN-based cellular network
Abegaz et al. Blockchain-based resource trading in multi-UAV-assisted industrial IoT networks: A multi-agent DRL approach
Li et al. Economic-robust transmission opportunity auction in multi-hop wireless networks
CN103220757B (en) A kind of optimum relay selection method based on two-way auction model
CN112654058A (en) Mobile edge computing offload and resource allocation algorithm in D2D multicast network
Xu et al. Subcarrier and power optimization for device-to-device underlay communication using auction games
CN101854634A (en) Spectrum assignment method based on market in clustering self-organizing network
Wu et al. Collusion-resistant multi-winner spectrum auction for cognitive radio networks
Ng et al. Communication-efficient federated learning in UAV-enabled IoV: A joint auction-coalition approach
Zheng et al. Joint optimization of the deployment and resource allocation of UAVs in vehicular edge computing and networks
Wang et al. Auction pricing-based task offloading strategy for cooperative edge computing
CN103841622A (en) Relay self-adaptive selection method based on social attribute
CN111083675A (en) Resource allocation method based on social perception in industrial Internet of things
Lim et al. Dynamic resource allocation for hierarchical federated learning
CN108391257B (en) Resource allocation method based on auction theory in social network D2D scene
Jiang et al. Rally: Device-to-device content sharing in lte networks as a game
Hoang et al. Dynamic spectrum allocation with second-price auctions: When time is money
Maghsudi et al. Distributed downlink user association in small cell networks with energy harvesting

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

Application publication date: 20200428

RJ01 Rejection of invention patent application after publication