5G network slice access method for power service terminalTechnical Field
The invention belongs to the technical field of 5G network communication, and particularly relates to a 5G network slice access method of a power service terminal.
Background
With the development of 5G networks, the differentiated demands of users on network services are increasingly significant. Traditional network deployment approaches meet different requirements through independent infrastructure, resulting in high equipment cost and low energy efficiency. The network slicing technique divides a physical network into a plurality of logical networks to carry traffic with different quality of service (QoS) requirements;
The current network slice access is realized by an Analytic Hierarchy Process (AHP), but the analytic hierarchy process depends on expert experience, decision results are easily affected by subjective factors, and the prior scheme does not consider user satisfaction and network operation cost.
Disclosure of Invention
The invention aims to solve the technical problems of high subjectivity, poor dynamic adaptability and unbalanced interests of users and networks in the prior art, and provides a power service terminal 5G network slice access method capable of realizing network slice selection with low cost, high service quality and user satisfaction, aiming at the defects in the prior art, and the invention adopts the following technical scheme:
The method comprises the steps of constructing a network slice QoS parameter matrix, obtaining index weight of the network slice QoS parameter matrix, obtaining random consistency ratio of the parameter matrix, judging and adjusting the matrix according to the random consistency ratio until the parameter matrix meets consistency requirements, obtaining a reference sequence, carrying out gray correlation analysis on the reference sequence and a comparison sequence, combining the index weight with correlation coefficients obtained by a gray correlation analysis method to obtain comprehensive correlation coefficients, and determining an optimal network slice according to the comprehensive correlation coefficients of the network slice;
The method for constructing the network slice dynamic selection game model comprises the steps of constructing a user priority preference sorting list, constructing a network priority preference sorting list based on comprehensive relevance coefficients, and carrying out multi-round matching game according to the priority level based on the preference sorting list so as to dynamically bind a user with the network slice.
Further, the method for constructing the network slice QoS parameter matrix comprises the following steps:
The user determines and evaluates QoS parameters of the network according to the preference of the selected network slices, and m network slices are arranged in the network environment, and n QoS parameters selected by the user are arranged;
Where aij represents the value of the jth QoS parameter of the ith network slice.
Further, the method for obtaining the index weight of the network slice QoS parameter matrix comprises the following steps:
normalizing the column vectors of the multi-target parameter matrix A to obtain a matrix B, wherein Bij is an element in the matrix B;
summing the rows of matrix B to obtain column vectorIs a column vectorElements of (a) and (b);
Alignment vectorNormalizing to obtain a feature vector W, wherein Wi is an element in the feature vector W, namely the index weight of QoS;
W=[w1,w2,...,wn]T。
further, the method for obtaining the random consistency ratio of the parameter matrix and judging and adjusting the matrix to meet the consistency requirement according to the random consistency ratio comprises the following steps:
The maximum feature root lambdamax of the judgment matrix A is calculated, and the formula of lambdamax is calculated as follows:
calculating a deviation consistency index CI of a judgment matrix:
CI=(λmax-n)/(n-1)
to ensure the reliability of the parameter matrix, the consistency check is performed on the matrix, and the random consistency ratio CR of the parameter matrix is calculated as follows:
CR=CI/RI
where n is the order of the target parameter matrix and RI is a random consistency index.
When CR is less than or equal to 0.1, the matrix meets the consistency requirement, and when CR is more than 0.1, the matrix does not meet the consistency requirement, and the matrix is adjusted and consistency test is carried out again.
Further, the method for acquiring the reference sequence and carrying out gray correlation analysis on the reference sequence and the comparison sequence comprises the following steps:
The reference sequence is represented as X0＝(X0(1),X0(2),···,X0 (n)), and gray correlation analysis is performed on the reference sequence and the comparison sequence, wherein gray correlation coefficients are represented as follows:
Where ζ is a resolution coefficient, vi (k) is a relationship coefficient of the kth QoS parameter in the ith network slice, and xi (k) represents a value of the kth QoS parameter in the ith network slice.
Further, the method for determining the optimal network slice according to the comprehensive association coefficient of the network slice comprises the following steps:
the gray correlation coefficient vi (k) obtained by a CR calculation formula of the comparison sequence of the network slice and the reference sequence is formed into a matrix C, and the evaluation values of the evaluation indexes of different schemes are obtained;
combining the index weight obtained by the analytic hierarchy process with the association coefficient obtained by the gray association analysis process to obtain a comprehensive association coefficient V:
V=WC
Wherein W is a weight set of QoS parameters, C is a gray correlation coefficient matrix, and V is a comprehensive correlation coefficient.
Further, the method for constructing the user priority preference ordered list based on the comprehensive relevance coefficient comprises the following steps:
according to the comprehensive relevance coefficient, carrying out preference sequencing on the network slices, and generating a priority sequencing U (N) of the network slices by a user;
and according to parameters required by the network slice, the preference ordering is carried out on the users, and the priority ordering N (U) of the network slice to the users is generated, wherein the required parameters comprise time delay required by the users when using a certain service, data rate required by the users, packet loss rate required by the users, communication cost required by the users and bandwidth required by the users.
Further, the method for dynamically binding the user and the network slice based on the preference ordering list and carrying out multiple rounds of matched games according to the priority level comprises the following steps:
First round matching, wherein a user selects a highest priority network based on U (N), and each network screens the user according to the size of a current matching window to establish a preliminary matching pair;
window dynamic adjustment, namely updating a matching window of a network end for unmatched users, deleting matched users and supplementing new candidate users;
Iterative matching, namely performing multiple times of matching according to the updated priority order, and dynamically adjusting a window after each time of matching until a termination condition is met;
downgrade matching-enabling secondary priority network selection for consecutive non-matching users.
Compared with the prior art, the invention has the following beneficial effects:
1. through fusion Analytic Hierarchy Process (AHP) and gray correlation analysis, user preference and network slice QoS parameter quantization are combined, subjectivity of the traditional method depending on expert experience is reduced, and scientificity and objectivity of network slice selection are improved.
2. Based on a multi-round dynamic game matching mechanism, the user and network double-end priority ordering is combined, the dynamic allocation of resources according to needs is realized, the resource redundancy or deficiency caused by static allocation is avoided, the network resource utilization rate is improved, the dynamic adjustment and secondary priority degradation matching is performed through windows under the network resource shortage scene, the service interruption is avoided, the basic communication requirements of the user are ensured, and the fault-tolerant capability and the continuity of the system are obviously improved.
3. The optimal network slice is precisely matched through the comprehensive relevance coefficient, and the optimal slice is preferentially accessed to high-priority service (such as power emergency communication and real-time monitoring) by combining with a degradation matching strategy, so that the service standard reaching rate of the delay sensitive service is improved, the QoS requirement of a user and the network operation cost (such as communication cost and bandwidth occupation) are considered, and the redundant resource allocation is reduced and the network operation cost is reduced through dynamic gaming on the premise of meeting the service performance.
Drawings
The present invention will be described in further detail with reference to the accompanying drawings.
FIG. 1 is a schematic flow chart of the present invention.
Detailed Description
For a better understanding of the present invention, the content of the present invention will be further clarified below with reference to the examples and the accompanying drawings, but the scope of the present invention is not limited to the following examples only. In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the present invention. It will be apparent, however, to one skilled in the art that the invention may be practiced without one or more of these details.
Embodiment 1 referring to FIG. 1-, the method for accessing the 5G network slice of the power service terminal of the embodiment comprises the following steps of
Constructing a static decision analysis model:
Constructing a network slice QoS parameter matrix, wherein the 5G power network is more focused on network quality and user experience. The user will choose the optimal network slice access according to the two indexes. Many factors affect network quality and user experience, such as network bandwidth, network throughput, transmission power, transmission delay, network security, and the like. These factors will all have an impact on the user's selection of the optimal network slice. First, the user determines the QoS parameters of the evaluation network according to the preference of selecting the network slice by himself. Assuming that there are m network slices in the network environment, there are n QoS parameters selected by the user. The invention adopts a 1-9 scale method commonly used in analytic hierarchy process to construct a multi-target parameter matrix A as follows.
Where aij represents the value of the jth QoS parameter of the ith network slice.
And (3) obtaining index weight of the network slice QoS parameter matrix, namely normalizing the column vector of the target parameter matrix A to obtain a matrix B, wherein Bij is an element in the matrix B.
Summing the rows of matrix B to obtain column vectorIs a column vectorIs a component of the group.
Alignment vectorNormalization results in feature vector W, Wi being the element in feature vector W, i.e. the weight of QoS.
Then
W=[w1,w2,...,wn]T
The random consistency ratio of the parameter matrix is obtained by calculating the maximum characteristic root lambdamax of the judgment matrix A and calculating lambdamax as follows:
calculating a deviation consistency index CI of a judgment matrix:
CI=(λmax-n)/(n-1)
To ensure the reliability of the parameter matrix, it is necessary to perform a consistency check on the matrix, and a random consistency ratio CR of the parameter matrix is calculated as follows:
CR=CI/RI
where n is the order of the target parameter matrix and RI is a random consistency index.
When CR is less than or equal to 0.1, the matrix is considered to meet the consistency requirement. Otherwise, the matrix must be adjusted and the consistency check re-performed until the matrix meets the requirements. RI is given in the following table, where n is the matrix order.
Random consistency index RI value
| n | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
| RI | 0.00 | 0.00 | 0.58 | 0.90 | 1.12 | 1.24 | 1.32 | 1.41 | 1.45 |
Gray correlation analysis:
First, a reference sequence and a comparison sequence are determined. Then the obtained sequence value of the optimal value of each network slice target QoS parameter
As a reference sequence. Represented by X0＝(X0(1),X0(2),···,X0 (n)). And then carrying out grey correlation analysis on the reference sequence and the comparison sequence. The gray correlation coefficient is expressed as follows:
Where ζ is the resolution coefficient, vi (k) is the relationship coefficient of the kth QoS parameter in the ith network slice. xi (k) represents the value of the kth QoS parameter (QoS index factor value) in the ith network slice.
And (3) comparing the comparison sequence of each network slice with the reference sequence, and forming a matrix C by using a gray correlation coefficient vi (k) obtained by a CR calculation formula to obtain the evaluation values of the evaluation indexes of different schemes.
Combining the index weight obtained by the analytic hierarchy process with the association coefficient obtained by the gray association analysis process to obtain a comprehensive association coefficient V:
V=WC
Wherein W is a weight set of QoS parameters, C is a gray correlation coefficient matrix, and V is a comprehensive correlation coefficient.
And finally, judging which is the optimal network slice according to the magnitude of the comprehensive association coefficient of each network slice.
Network slice selection algorithm based on matching game:
the algorithm is based on the idea of matching game theory, and by combining GAHP algorithm, not only the speed and accuracy of network slice selection are ensured, but also the satisfaction degree of both users and networks can be considered.
The matching game can be divided into one-to-one, many-to-one and many-to-many matching relationships according to the matching relationship.
Assume that the network set is
Net={net1,net2,…,netm}
User set as
User={user1,user2,…,usern}
The matching game adopts a many-to-one matching game, namely, in the 1 st matching game, all users simultaneously carry out the matching game. In each round of matching process, the participants arranged in front of the network ordering U (N) and the user ordering N (U) complete matching first, and users without matching can complete matching after multiple matching. The specific steps of the matched game are as follows.
1. And (5) determining and collecting parameters. The first part is the parameters needed by the user to order the network slices, and this part of parameters can be used to construct the priority of the network slices according to the comprehensive relevance coefficient mentioned above. The second part is the parameter needed by the network slice to order the users, and the part of the parameter is determined as the time delay, the data rate, the packet loss rate, the communication cost and the bandwidth needed by the users when the users use a certain service according to the operation requirement of the network operators.
2. Determination of preference ordering. In the present model, the preference ordering of the network slices by the user and the preference ordering of the network slices by the user are classified.
3. The 1 st match. It is assumed that the users that have completed matching are not in the user set. And each user in the user side sorts U (N) according to own network priority, and selects the network with the highest priority. The network side needs to set the size of the matching window, i.e. how many users are considered in one-time matching of the network side. If the network end user is located in the matching window, the matching is successful.
4. Translation of the matching window. If the network j is the highest priority network in the user i, if the user i exists in the user priority window of the network j, the matching is completed. If the matching between the user i and the network j is unsuccessful, namely, no user i exists in the network window. After the 1 st matching is finished, the matching window needs to be translated. The window of the network j deletes the users successfully matched in the 1 st matching of the network j and the users successfully matched in the 1 st matching of other networks, and adds the users which do not participate in the matching in the priority sorting.
5. The 2 nd match. Network j is still the highest priority network of user i, and the matching window of network j has shifted. If user i is in the translated matching window, the matching is completed. The 2 nd match may be performed multiple times.
6. The user matching fails. If the user i does not complete the matching in the 1 st and 2 nd matches, the user i selects a network with a progressive priority. The above process is repeated until all user matches are completed.
The beneficial effects are that:
1. through fusion Analytic Hierarchy Process (AHP) and gray correlation analysis, user preference and network slice QoS parameter quantization are combined, subjectivity of the traditional method depending on expert experience is reduced, and scientificity and objectivity of network slice selection are improved.
2. Based on a multi-round dynamic game matching mechanism, the user and network double-end priority ordering is combined, the dynamic allocation of resources according to needs is realized, the resource redundancy or deficiency caused by static allocation is avoided, the network resource utilization rate is improved, the dynamic adjustment and secondary priority degradation matching is performed through windows under the network resource shortage scene, the service interruption is avoided, the basic communication requirements of the user are ensured, and the fault-tolerant capability and the continuity of the system are obviously improved.
3. The optimal network slice is precisely matched through the comprehensive relevance coefficient, and the optimal slice is preferentially accessed to high-priority service (such as power emergency communication and real-time monitoring) by combining with a degradation matching strategy, so that the service standard reaching rate of the delay sensitive service is improved, the QoS requirement of a user and the network operation cost (such as communication cost and bandwidth occupation) are considered, and the redundant resource allocation is reduced and the network operation cost is reduced through dynamic gaming on the premise of meeting the service performance.
Finally, it is noted that the above-mentioned embodiments are merely for illustrating the technical solution of the present invention, and that other modifications and equivalents thereof by those skilled in the art should be included in the scope of the claims of the present invention without departing from the spirit and scope of the technical solution of the present invention.