CN111866181A

CN111866181A - Block chain-based task unloading optimization method in fog network

Info

Publication number: CN111866181A
Application number: CN202010796536.4A
Authority: CN
Inventors: 黄晓舸; 刘鑫; 王永生; 陈前斌
Original assignee: Chongqing University of Post and Telecommunications
Current assignee: Nanjing Qingke Liangu Technology Service Co ltd
Priority date: 2020-08-10
Filing date: 2020-08-10
Publication date: 2020-10-30
Anticipated expiration: 2040-08-10
Also published as: CN111866181B

Abstract

The invention relates to a task unloading optimization method in a fog network based on a block chain, and belongs to the technical field of mobile communication. There are two unloading modes in this method: off-load to the fog server FS and off-load to the D2D group of devices to achieve optimization of compute-intensive task processing latency and energy consumption. The FS and D2D device groups are incentivized to provide task offload services for SMEs through incentives such as resource coin awards. A shareholder delegation certification mechanism DPoS based on SMEs voting is provided, wherein a block chain conducts shareholder voting on an FS through SMEs in the mechanism, and an FS verification set is selected. The manager FS is responsible for packaging transaction transactions to generate new blocks, and the new blocks are added into the block chain after verification and audit of the verifier FS. The method can realize the effectiveness of resource allocation by jointly optimizing the task execution cost of the system and the consensus cost of the block chain system on the premise of ensuring the safety of SMEs data.

Description

Block chain-based task unloading optimization method in fog network

Technical Field

The invention belongs to the technical field of mobile communication, and relates to a task unloading optimization method in a fog network based on a block chain.

Background

The internet of things market is experiencing an unprecedented period of rapid development. Gartner forecasts that the number of terminals of the internet of things reaches 208 hundred million by 2020, and the composite growth rate is 34%. In accordance with a report issued in the united states on the trend of 2016 + 2045 emerging technology, over 1000 billion devices were connected to the Internet by 2045. Although there are many application scenarios for the internet of things, the security and privacy issues have not been adequately addressed. The internet of things will penetrate the aspects of people's lives and enter various industries, thereby creating a variety of application scenarios, such as smart cities, smart homes, wearable devices and automobile internet. The internet has turned to a cloud-based architecture, with cloud computing providing outsourced computing and storage functions for end users. However, with the rapid increase of traffic in recent years, the data with ultra-large capacity is transmitted to the cloud server, which not only brings heavy burden to communication bandwidth, but also causes intolerable transmission delay, reduces the satisfaction degree of the terminal user on the service quality, and a large number of internet of things terminals bring great challenges to network capacity and centralized platform performance. However, attractive numbers are hidden in huge challenges and crisis, while at the same time more demands are focused on large internet terminal devices with access information security risks, such as terminal-to-counterfeit, node control, data manipulation and DDoS attacks. According to statistics, the number of network attacks on internet equipment is increased by 280% in the last half of 2017. By 2021, the business' expenditure in information security would increase from 835 billion dollars today to 1199 billion dollars. Because the computing power and the storage capacity of the terminal equipment of the internet of things are insufficient, the traditional network security protection technology cannot meet the security requirement of the internet of things, and therefore a new internet of things security protection technology is urgently needed to solve the increasingly serious problems of the security and the privacy of the internet of things. The unique technical characteristics of the block chain can effectively solve the problems of safety and privacy faced by the development of the Internet of things, and provide support such as trust, transparency and distributed storage for the Internet of things, so that a high-efficiency, reliable and safe distributed Internet of things network is constructed, and the safety and privacy of the Internet of things user are effectively guaranteed.

In conclusion, the invention designs a task unloading optimization model in the fog network based on the block chain. Under this model, SMEs can offload compute-intensive tasks to other nodes with idle resources. The nodes with idle resources may be FSs or a D2D group of devices nearby. In addition, according to the scheme, a corresponding optimization algorithm is provided according to an application scene of the unloading model, and the improvement of the system performance is realized through the joint optimization of the task execution cost and the block chain system consensus cost on the premise of guaranteeing the system security and the user data privacy.

Disclosure of Invention

In view of the above, the present invention provides a method for optimizing task offloading in a fog network based on a block chain.

In order to achieve the purpose, the invention provides the following technical scheme:

a task unloading optimization method in a fog network based on a block chain comprises the following steps:

s1: two offload modes based on blockchain techniques;

s2: D2D device group decision scheme;

s3: a DPoS consensus mechanism based on SMEs voting;

s4: a block chain based fog network performance optimization algorithm;

in step S1, the proposed model of the present scheme includes three layers: the first layer is an equipment layer (front end) which comprises a smart phone, smart wearable equipment, a smart vehicle and the like; the second layer is a fog service layer (near end) comprising FSs and various APs; the third layer is a core cloud service layer (far end), and powerful central cloud computing services are provided at the far end.

SMEs at the front end provide better interactivity and responsiveness for the user. However, due to limited SME computing power and capacity, the compute-intensive tasks must be offloaded to FSs and idle D2D devices that can provide computing resources nearby. According to the signal-to-noise ratio threshold, the D2D devices which are idle nearby the SMEs to be unloaded form a D2D device group, and the SMEs can be selectively unloaded to a nearby FSs or D2D device group.

The near end FSs may support most traffic in the network as well as resource requests such as real-time data processing, data caching and computation forking. FSs thus provides better communication quality and performance for data delay time sensitive users. Since FSs has sufficient computational and storage resources, FSs can perform computational offload and blockchain consensus tasks simultaneously, and thus can act as blockchain nodes.

The remote core cloud server provides more powerful computing power (e.g., big data processing). The goal of this model is to execute the compute-intensive and delay-sensitive part of the application in the edge network and communicate with the core cloud for data synchronization.

In step S2, the computation-intensive task a of SMEk in cell n_n,kUsing parameter tuples < D_n,k,τ_n,kX > is represented by D_n,kThe size of the task is represented, and the unit is bit; tau is_n,kRepresenting the maximum tolerated delay for task completion; x represents the CPU cycle consumed to complete 1bit of data, and the unit is (CPU cycles/bit). When S is_n,kRequest to offload to D2D device group because SMEs in D2D device group have their respective computing capabilities f_n,k'Different and there is also a gap in the channel conditions between them and the SMEs to be offloaded, i.e. S_n,kTransmission rate with SMEs in D2D device group

And also different. Defining a vector α ═ α_n,1,α_n,2,...,α_n,k',...,α_n,K'},k'＝1,2,...,K',α_n,k'∈[0,1]To characterize the fraction of computation offload tasks offloaded by SMEs in the D2D device group. By assuming S_n,kEqual to the sum of the transmission phase delay and the execution phase delay of SMEs in the D2D device group, that is:

simultaneous alpha_n,1+α_n,2+...+α _n,k'1, α ═ α can be obtained_n,k'The solution of.

Further, in step S3, the present invention proposes a DPoS consensus mechanism based on SMEs voting. The block uplink can be divided into three stages, namely a block production stage, a block sharing stage and a block verification stage.

In the block production phase, FSs is responsible for collecting transaction transactions for SMEs in each cell, and then broadcasting to the whole network, manager

Responsible for packing transaction and producing blocks, the input data size of transaction is D_pExpressed in units of bit; manager

Is calculated by f_w'Means that the CPU power consumption of the manager is P_w'＝x_w'(f_w')³Wherein x is_w'Presentation manager

The calculated energy efficiency coefficient. The latency and energy consumption of the block production phase can be expressed as:

E^p＝P_w'T^p＝x_w'(f_w')²D_pX (3)

in the block sharing stage, the manager

The new block is signed and then broadcasted to all verifiers in the block chain system

Block size S_bExpressed in units of bit; manager

And verifier

With R between_w'，wExpressed in units of (bit/s); manager

A transmission power of

The latency and energy consumption of the block sharing phase can be expressed as:

in the block verification stage, a verifier receiving a new block

Comparing the hash value calculated by the hash algorithm with the digital signature to verify the accuracy of the new block, and if the hash value is equal to the digital signature, indicating that the new block is not tampered and returning a feedback message for block verification; otherwise, no response is made. Finally, if all the audit results are correct, the new block is added to the block chain according to the timestamp. Assume that the input data size of the verification phase is D_vThe unit is bit; verifying device

Is calculated by f_wPresentation, verification device

CPU power consumption of P_w＝x_w(f_w)³Wherein x is_wPresentation verifier

The calculated energy efficiency coefficient. The latency and energy consumption of the block verification stage can be expressed as:

in summary, the total delay and total energy consumption of the blockchain system can be expressed as:

further, in step S4, the present solution considers a task unloading model in a block chain-based fog network, and the system task execution cost can be expressed as

Wherein λ₀，λ₁Respectively representing scalar weights, characterizing S_n,kI.e., the relative importance of latency and energy consumption to the system task execution. The consensus cost of the blockchain system can be expressed as

Wherein λ₂Indicating the relative preference of the blockchain system for the consensus process latency and energy consumption. Assume that the weighting factor between the system task execution cost and the blockchain system consensus cost is ζ₁ζ and₂means that the objective function is guaranteed to be in the same order of magnitudeThe optimized objective function model may be expressed as U ═ ζ₁U₁(_n,k,α_n,k',β_k,s)+(1-ζ₁)ζ₂U₂(f_n,f_w,f_w')。

The invention has the beneficial effects that: the present invention offloads compute-intensive tasks to other nodes with idle resources. In particular, the present invention contemplates two offloading modes, one to nearby FSs and the other to the D2D group of devices. The system performance is improved by the joint optimization of the system task execution cost and the block chain system consensus cost.

Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention may be realized and attained by the means of the instrumentalities and combinations particularly pointed out hereinafter.

Drawings

For the purposes of promoting a better understanding of the objects, aspects and advantages of the invention, reference will now be made to the following detailed description taken in conjunction with the accompanying drawings in which:

FIG. 1 is a diagram of a model for task offloading in a block chain based fog network;

FIG. 2 is a diagram of a DPoS consensus mechanism model based on SMEs voting;

Detailed Description

The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention in a schematic way, and the features in the following embodiments and examples may be combined with each other without conflict.

Wherein the showings are for the purpose of illustrating the invention only and not for the purpose of limiting the same, and in which there is shown by way of illustration only and not in the drawings in which there is no intention to limit the invention thereto; to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product; it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.

The same or similar reference numerals in the drawings of the embodiments of the present invention correspond to the same or similar components; in the description of the present invention, it should be understood that if there is an orientation or positional relationship indicated by terms such as "upper", "lower", "left", "right", "front", "rear", etc., based on the orientation or positional relationship shown in the drawings, it is only for convenience of description and simplification of description, but it is not an indication or suggestion that the referred device or element must have a specific orientation, be constructed in a specific orientation, and be operated, and therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes, and are not to be construed as limiting the present invention, and the specific meaning of the terms may be understood by those skilled in the art according to specific situations.

Referring to fig. 1, the model in fig. 1 combines two offloading manners, FC offloading and D2D offloading, and uses a blockchain technique to maintain a continuous tamper-resistant ledger database, so as to prevent malicious nodes from tampering transaction information, and ensure security of the entire network and privacy of user data. To meet the latency requirements of SMEs for performing compute-intensive tasks, SMEs choose to offload compute tasks to nearby FSs or groups of D2D devices. The proposal adopts a DPoS consensus mechanism and a manager based on SMEs voting

Responsible for packaging transactions to generate new blocks and broadcast them to W-1 validators

Then passes through a verifier

If the verification result is correct, the new block is managed by the manager

Added to the blockchain.

1. Network communication model

Consider an OFDMA system with S uplink subcarriers, a total system bandwidth of B, and a bandwidth of S subcarriers

The present scheme assumes that each subcarrier can only be allocated to one SME to avoid interference between SMEs.

Respectively, S on a subcarrier S_n,kChannel gain with SME k' in the D2D device group, and S on subcarrier S_n,kChannel gain with FS n. Using vector β ═ β_k,1,β_k,2,...,β_k,s,...,β_k,SDenotes S_n,kTransmission power allocated on subcarrier S, S_n,kP for total transmission power_n,kIs expressed as σ²And a_k,s E 0,1 represents the noise power and subcarrier allocation indication parameters, respectively. I.e. when subcarrier S is allocated to S_n,kThen a is_k,s1 is ═ 1; otherwise a_k,s＝0。

1) Offloading to D2D device group (offloading mode 0):

assume that the number of SMEs K' in the D2D device group is less than or equal to the number of subcarriers S. S_n,kThe transmission rate offloaded to SME k' over subcarrier s can be expressed as:

S_n,kthe uplink transmission rate offloaded to SME k' can be expressed as:

2) offloading to fog server (offloading mode 1):

S_n,koffloading to F over subcarrier s_nThe transmission rate of (d) may be expressed as:

S_n,kthe uplink transmission rate offloaded to FS n can be expressed as:

2. computational offload model

Suppose S_n,kBoth the nearby FSs and D2D groups of devices have computing power to provide task execution services. Compute intensive task A_n,kMay be offloaded to FSs or a group of D2D devices.

1) Offloading to D2D device group (offloading mode 0):

the time delay of the unloading mode 0 comprises two parts of time delay of a transmission phase and time delay of an execution phase, and the time of a result acquisition phase is omitted. Since for some typical fog calculation services the results sent back in the downlink are small, approximately equal to 1/30 where the results were sent in the uplink, the downlink delay is negligible.

Transmission phase delay: in the unloading mode, S_n,kSelecting to request an offload task from a nearby D2D device group, S_n,kThe transmission power with SME k' is expressed as

The transmission delay and power consumption at this stage can be expressed as:

and (3) executing stage time delay: in this offload model, SMEs in the D2D device group have their respective computational capabilities f_n,k'In a different sense, with x_n,k'Representing the calculated energy efficiency coefficients of SMEs in the D2D device group. In this offload mode, compute task A_n,kThe execution phase latency and energy consumption in SMEs in the D2D device group can be expressed as:

in summary, the total time delay under the unloading mode 0 can be obtained

And total energy consumption

Can be expressed as:

2) offloading to fog server (offloading mode 1):

the delay of SME selection unloading to the fog server comprises three parts of transmission phase delay, execution phase delay and queue phase delay.

Transmission phase delay: in the unloading mode, S_n,kSelecting F to nearby_nAn offload task is requested. S_n,kThe transmission power with FS n is expressed as

The transmission delay and power consumption at this stage can be expressed as:

and (3) executing stage time delay: modeling computational power consumption as P_n＝x_n(f_n)³Wherein x is_nRepresenting the calculated energy efficiency coefficient of the fog server, f_nIs represented by F_nThe unit of the computing power of (1) is (CPU cycles/s). f. of_nCan be flexibly adjusted by DVS technology to meet the requirements of users, and uses a vector F ═ F₁,f₂,...,f_n,...,f_NN, denotes F, 1,2_nIs S_n,kThe allocated computing resources. Under this offload mode, compute intensive task A_n,kThe execution latency and energy consumption in SMEs in the D2D device group can be expressed as:

delay of queuing stage: when decisions are made taking into account SMEs, the task buffer waits for queuing delays introduced by the tasks. Suppose in F_nIs Q of the total number of CPU cycles processed in the task buffer_n. By using

Is represented by F_nThe static circuit power of (1). Then compute task A_n,kThe queuing phase delay and energy consumption of (a) can be expressed as:

in summary, the total time delay T in the unloading mode 1 can be obtained₁ ^tolAnd total energy consumption

Expressed as:

3. optimization problem modeling

The optimization method aims to jointly optimize the task execution cost and the block chain system consensus cost under the condition of meeting the requirements of Quality of service (QoS) of system task execution and a block chain system so as to improve the system performance. Thus, the optimization problem can be modeled as:

constraint C1 ensures that the task offload mode decision is valid, and C2 indicates that the sum of the task allocation proportions in offload mode 0 is equal to 1. C3 denotes S_n,kThe sum of the transmission power allocation ratios over S subcarriers is equal to 1. C4 indicates that the total delay for task completion in either offload mode 0 or offload mode 1 does not exceed the maximum delay limit τ_n,kC5 shows that the sum of the total energy consumption of the system task execution and the total energy consumption of the block chain system consensus process is not more than the maximumEnergy consumption E^maxC6 indicates that the computational power allocated to FS by the system cannot exceed the total computational power

4. System flow

Fig. 2 is a flow chart of task offloading in a block chain-based fog network, specifically including the following steps:

step 201: initializing a system;

step 202: SMEs register as legal entity at LBS, LBS grants identity address I_n,kAnd certificate C_n,kThey may uniquely represent SME k. Then, the public key PK_n,kAnd a private key SK_n,k，＜PK_n,k,SK_n,k>. and wallet address { WAL) for implementing resource currency transactions_n,kSending the data to SMEs;

step 203: SMEs apply for LBS to join a block chain network;

step 204: LBS verifies whether the SMEs node is effective, if the SMEs node passes the verification, the SMEs node agrees to join the network, otherwise, no response is made;

step 205: after the authentication of the LBS is verified, the SMEs send the resource requirement to the LBS, the FSs send the available resource and the QoS constraint to the LBS, and the LBS writes the available resource, the resource requirement and the QoS constraint into an intelligent contract of resource transaction;

step 206: automatically triggering intelligent contracts, and selecting by SMEs to offload tasks to FSs or D2D device groups nearby;

step 207: s_n,kSelecting to unload to the D2D equipment group, namely unloading mode 0;

step 208: s_n,kSelecting a nearby D2D device group to offload request message R_n,kDigitally signing R_n,kAnd S for digital signature_n,kThe private key is encrypted and sent to the D2D device group;

step 209: D2D equipment group uses S_n,kDecrypting and verifying the digital signature and the transaction by the public key, and executing the computing task A if the verification is passed_n,kOtherwise, not responding to the uninstall request message;

step 210: the D2D equipment group executes the calculation task and returns the calculation result to the S_n,k；

Step 211: deriving resource prices, S, from intelligent contracts for resource transactions_n,kPaying a number of resource coins to the D2D device group by price;

step 212: s_n,kFSs for selecting unloading to the vicinity, namely unloading mode 1;

step 213: s_n,kSelect fog Server F_nFor the offload request message R_n,kDigitally signing R_n,kAnd S for digital signature_n,kIs encrypted and sent to F_n；

Step 214: f_nWith S_n,kDecrypting and verifying the digital signature and the transaction by the public key, and executing an uninstalling task A if the verification is passed_n,kOtherwise, not responding to the uninstall request message;

step 215: f_nExecuting the calculation task and returning the calculation result to S_n,k；

Step 216: resource price, S, derived from intelligent contracts for resource transactions_n,kPaying F by price_nA quantity of resource coins from S_n,kWallet address transfer to F_nThe wallet address of;

step 217: SMEs vote FSs in their covered area, one person for one vote, the weight of SME voting is independent of the share weight value;

step 218: FSs of top M names of ticket number as verification set, which includes a manager

And M-1 verifiers

Step 219: m FSs in the authentication set act as managers in turn

Step 220: s_n,kSending the transaction record to the associated F_n；

Step 221: f_nBroadcasting the transaction record;

step 222: manager

Constructing a new block consisting of all transactions, signing and broadcasting to the verifier in the blockchain system

Step 223: verifier for receiving new block

The correctness of the hash value and the digital signature are verified;

step 224: manager

From more than two-thirds of verifiers

Receiving a submit message;

step 225: manager

Putting the new block adding timestamp into a block chain;

step 226: verifying device

Only returning the block header content including the last block hash value, the current block hash value, the timestamp and the Merkel root value to the SMEs so as to save the storage space of the SMEs;

step 227: the algorithm ends.

Finally, the above embodiments are only intended to illustrate the technical solutions of the present invention and not to limit the present invention, and although the present invention has been described in detail with reference to the preferred embodiments, it will be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions, and all of them should be covered by the claims of the present invention.

Claims

1. A task unloading optimization method in a fog network based on a block chain is characterized in that: the method comprises the following steps:

s1: two offload modes based on blockchain techniques;

s2: device-to-device communication, i.e., D2D device group decision scheme;

s3: a right of stock entrustment certification mechanism DPoS consensus mechanism based on SMEs voting of the intelligent mobile equipment;

s4: and (3) a block chain-based fog network performance optimization algorithm.

2. The method for optimizing task offloading in a block chain based fog network as claimed in claim 1, wherein: in step S1, a block chain-based fog network is established, which is composed of an equipment layer, a fog service layer, and a cloud service layer;

the device layer consists of SMEs and comprises a smart phone, smart wearable devices and smart vehicles;

the fog service layer comprises N FSs units with set omega_a＝{F₁,F₂,...,F_n,...,F_NN is 1,2, ·, N denotes; there are K SMEs in the coverage of FSn, with the set Ψ_a＝{S_n,1,S_n,2,...,S_n,k,...,S_n,KK is 1,2, ·, K denotes; s_n,kRepresenting SME k within FSn coverage;

two unloading modes exist in the fog network, namely unloading to an FS or D2D equipment group; using vector as ready-to-use_n,1,_n,2,...,_n,k,...,_n,KK, K SMEs within the coverage of the FSn,_n,ke 0,1 represents the unloading mode,_n,k0 represents S_n,kOffloading to D2D device group, i.e., offloading mode 0;_n,k1 represents S_n,kUnload to FS, unload mode 1.

3. The method for optimizing task offloading in a block chain based fog network as claimed in claim 2, wherein: in step S2, assuming that each SME can access a nearby FS, S is set_n,kThe D2D device group has K 'SMEs, K' < K, and the set psi is used_bIt is shown that,

S_n,kD2D device group set Ψ_bIs Ψ_aMiddle S_n,kIs not less than a threshold S_thSME of (i)

4. The method for optimizing task offloading in a block chain based fog network as claimed in claim 1, wherein: in step S3, the SME votes for its preferred nodes to form a verification set according to the equity value of the FS; each SME only has one ticket, and the weight of the ticket is not in direct proportion to the stock weight, so that the influence of the stock weight on the election of the verification set is reduced; the top W FS with the most votes form a verification set, the verification set comprises 1 manager FS and W-1 verifier FS, and the verification set is represented as

It is shown that the manager FS is,

which represents the verifier FS, is shown,

FS in the validation set acts as a block manager in turn, with the rest W-1The individual verifiers FS perform block verification.

5. The method for optimizing task offloading in a block chain based fog network as claimed in claim 4, wherein: the DPoS consensus mechanism is as follows: manager

Packaging the transaction records to generate a new block and broadcasting the new block to W-1 verifiers

Pass through the verifier

Verification and audit of, manager

Adding the block with correct verification result into the block chain;

the block chain system is divided into three working stages, namely a block production stage, a block sharing stage and a block verification stage, wherein the time delay of the three stages is respectively T_p，T_s，T_vRepresents; corresponding energy consumption respectively using E_p，E_s，E_vRepresents; the total delay of the blockchain system is used

Means, total energy consumption

And (4) showing.

6. The method for optimizing task offloading in a block chain based fog network as claimed in claim 1, wherein: in step S4, when the SMEs adopts the offload mode 1, the method includes three stages of delays, i.e., a transmission stage delay T₁ ^upExecution phase delay T₁ ^exeAnd queuing stageSegment delay T₁ ^que；

Energy consumption is determined by the energy consumption of the transmission stage

Execution phase energy consumption

And energy consumption in queuing stage

Composition is carried out;

the total time delay and the total energy consumption of the task completion of the unloading mode 1 are respectively T₁ ^tol，

Represents;

when SMEs adopts the unloading mode 0, the SMEs have limited computing capacity, only one task can be processed in the same stage, and the time delay comprises the time delay of a transmission stage

And execution phase delay

And execution phase energy consumption

Composition is carried out;

the total time delay and the total energy consumption of the task execution of the unloading mode 0 are respectively used

Represents; weighting factor ζ by optimizing task execution cost and block chain system consensus cost₁，1-ζ₁So as to improve the system performance.