CN114980198B - Resource allocation method for combining MIMO mobile edge computing network IA and IN - Google Patents
Resource allocation method for combining MIMO mobile edge computing network IA and IN Download PDFInfo
- Publication number
- CN114980198B CN114980198B CN202210571856.9A CN202210571856A CN114980198B CN 114980198 B CN114980198 B CN 114980198B CN 202210571856 A CN202210571856 A CN 202210571856A CN 114980198 B CN114980198 B CN 114980198B
- Authority
- CN
- China
- Prior art keywords
- user
- edge
- edge node
- representing
- computing network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 43
- 238000013468 resource allocation Methods 0.000 title claims abstract description 16
- 239000011159 matrix material Substances 0.000 claims abstract description 78
- 238000004364 calculation method Methods 0.000 claims abstract description 67
- 230000005540 biological transmission Effects 0.000 claims description 9
- 230000010076 replication Effects 0.000 claims description 3
- 238000009825 accumulation Methods 0.000 claims description 2
- 230000014509 gene expression Effects 0.000 claims description 2
- 230000026676 system process Effects 0.000 claims description 2
- 238000004891 communication Methods 0.000 abstract description 27
- 238000013461 design Methods 0.000 abstract description 5
- 238000005516 engineering process Methods 0.000 description 14
- 238000004088 simulation Methods 0.000 description 13
- 238000006386 neutralization reaction Methods 0.000 description 9
- 230000000750 progressive effect Effects 0.000 description 6
- 230000007547 defect Effects 0.000 description 5
- 238000005265 energy consumption Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000001934 delay Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 230000001747 exhibiting effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Classifications
-
- 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
- H04W28/0231—Traffic management, e.g. flow control or congestion control based on communication conditions
- H04W28/0236—Traffic management, e.g. flow control or congestion control based on communication conditions radio quality, e.g. interference, losses or delay
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
- H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
- H04B7/0413—MIMO systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
- H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
- H04B7/0413—MIMO systems
- H04B7/0456—Selection of precoding matrices or codebooks, e.g. using matrices antenna weighting
-
- 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
- H04W28/08—Load balancing or load distribution
- H04W28/0827—Triggering entity
- H04W28/0838—User device
-
- 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
- H04W28/08—Load balancing or load distribution
- H04W28/09—Management thereof
- H04W28/0958—Management thereof based on metrics or performance parameters
- H04W28/0967—Quality of Service [QoS] parameters
- H04W28/0975—Quality of Service [QoS] parameters for reducing delays
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Quality & Reliability (AREA)
- Mobile Radio Communication Systems (AREA)
- Radio Transmission System (AREA)
Abstract
The invention discloses a resource allocation method for combining MIMO mobile edge computing networks IA and IN, which solves the problems that the prior art does not multiplex communication resources, eliminates interference, needs infinite time slot expansion and cannot be suitable for MIMO scenes IN engineering practice. The implementation steps are that the calculation tasks are processed according to the calculation load; each user uploads the processed sub-calculation tasks to an edge node, designs a precoding matrix and a decoding matrix in an uploading stage based on the IA to eliminate interference, and calculates uplink time delay; and the edge node downloads the expected value to the user, designs a precoding matrix and a decoding matrix of the downloading stage based on the combination of the IA and the IN to eliminate interference, and calculates the downlink delay. The invention multiplexes communication resources, does not need to expand time slot in the process of eliminating interference, and is applicable to any system configured mobile edge computing network.
Description
Technical Field
The invention belongs to the technical field of communication, and further relates to a resource allocation method combining a multi-input multi-output MIMO (multiple input multiple output) mobile edge computing network interference alignment IA (interference alignment) and interference neutralization IN (interference neutralization) technology in the technical field of wireless communication. The method and the device can be used for a mobile edge computing network, and the degree of freedom of the network is improved by eliminating interference in the network, so that the communication time delay is reduced.
Background
Mobile edge computing is widely used in emerging services, such as smart navigation, online gaming, etc., which may be implemented to provide computing resources for mobile users. The mobile edge computing network is mainly divided into four stages, namely a task allocation stage, a task uploading stage, a task computing stage and a result downloading stage. In the task uploading stage and the result downloading stage, if the simultaneous same-frequency transmission data is not considered, the utilization rate of communication resources is lower. If the multiplexing communication resource transmits data, a large amount of interference is caused in the network due to the need of transmitting a large amount of data, thereby affecting the performance of the network. Thus, in order to better accomplish the task, research into a resource allocation method of the mobile edge computing network is required.
A resource allocation method for a mobile edge computing network is disclosed in patent literature "a task offloading and resource allocation method for a compromise between mobile edge computing network energy consumption and latency" (application number: 202111672596.6, application publication number: CN 114302457A) applied by university of hangzhou electronics and technology. The method comprises the following implementation steps: the method comprises the steps of firstly, obtaining network configuration information; secondly, taking the common minimization of the energy consumption and average calculation time delay as a target, and generating two sub-targets of an unloading decision and a resource allocation decision; thirdly, fixing the resource allocation decision, and determining an unloading decision with minimum energy consumption and average calculation time delay; fourthly, fixing the unloading decision, and calculating a resource allocation strategy with minimum energy consumption and average calculation time delay; fifth, the third step to the fourth step are iterated circularly to obtain an unloading decision and a resource allocation decision with minimum energy consumption and time delay total cost; and sixthly, determining task unloading and resource allocation of the mobile edge computing network. The method has the defects that the method does not consider the transmission data of the same time and the same frequency, does not consider multiplexing communication resources, influences the time delay of completing the task, and further influences the efficiency of completing the task.
Tao Meixia in its published paper "Exploiting Computation Replication for Mobile Edge Computing:A Fundamental Computation-Communication Tradeoff Study"(IEEE Transactions on Wireless Communications,vol.19,no.7,pp.4563-4578,July 2020.) discloses a method of resource allocation for a mobile edge computing network. The method comprises the following implementation steps: the first step, distributing task distribution and obtaining calculation load; uploading the task, and eliminating interference in the transmission process by using a progressive interference alignment technology; thirdly, completing calculation of the task on the server; and fourthly, downloading the result back to the user, and eliminating the interference in the transmission process by using a progressive interference alignment technology and an interference neutralization technology. The method has the defect that when the method adopts a progressive interference alignment technology to eliminate interference, infinite time slot expansion is needed, so that the method is a theoretical method and cannot be realized in engineering practice. In addition, the paper analyzes the mobile edge computing network with single input and single output SISO (single input single output), that is, the edge node and the user equipment are configured with only a single antenna, but the multi-antenna scene is not studied, and in actual life, the edge node and the user equipment are often configured with a plurality of antennas, so that the paper has a limitation.
Disclosure of Invention
The invention aims to solve the problems that the prior art does not consider multiplexing communication resources, interference elimination requires infinite time slot expansion and the prior art cannot be suitable for MIMO scenes in engineering practice.
The technical idea for realizing the purpose of the invention is as follows: the invention sets MIMO mobile edge computing network system parameters, and obtains computing load by computing the ratio of the total task bit number distributed to all edge nodes to the total input bit number of all user equipment. In the task uploading stage, according to the system configuration and the calculation load, a scheme of an interference alignment technology is designed, and according to the relation between the number of precoding matrixes and the number of interference at a receiving end, partial interference is selected to be aligned so as to compress the space occupied by the interference. The invention improves the degree of freedom of the network by eliminating the interference of the task uploading stage, thereby reducing the communication time delay and obtaining the relation between the communication time delay of the task uploading stage and the number of edge nodes, the number of users, the number of configured antennas and the calculation load. In the result downloading stage, according to the system configuration and the calculation load, a scheme of the result downloading stage combining the interference neutralization and interference alignment technology is designed, according to the size relation of the number of precoding matrixes and the number of interference at a receiving end, the space occupied by partial interference for neutralization or alignment to compress interference is selected, the degree of freedom of a network is improved by eliminating the interference of the result downloading stage, so that the communication time delay is reduced, and the relation between the communication time delay and the number of edge nodes, the number of users, the number of configured antennas and the calculation load in the result downloading stage is obtained. The invention multiplexes communication resources, does not need to expand time slot in the process of eliminating interference, and is applicable to any system configured mobile edge computing network.
In order to achieve the above object, the main steps of the present invention are as follows:
Step1, distributing computing tasks:
the mobile edge computing system processes the computing tasks according to each computing load, and equally divides the received input bit number of each user computing task to obtain a plurality of equally divided sub computing tasks of the user;
Step 2, uploading the calculation task to an edge node in the mobile edge calculation network:
Step 2.1, each user sends all sub-calculation tasks after equipartition to the edge node;
Step 2.2, each edge node receives all sub-calculation tasks sent by each user;
Step 3, designing a precoding matrix and a decoding matrix in an uploading stage:
Step 3.1, the user selects elements meeting the conditions from the precoding matrix of the complex domain to form a precoding matrix with the size of N r×ds in the uploading stage Obtained by the same methodA plurality of precoding matrices; the condition means that the values of each element in the tense space of the product matrix corresponding to the selected element are equal to obtain the value received by the edge nodeThe term-jammer computing tasks are all aligned to the same space; where N r denotes the number of user-configured antennas in the mobile edge computing network, d s denotes the length of the sub-computing tasks,Indicating that the ith user is sent toThe precoding matrix of the sub-computation task of the p-th multicast group,Represents the edge node set formed by the edge nodes except for the jth edge node in the edge node set, M represents the total number of users in the mobile edge computing network, K represents the total number of edge nodes in the mobile edge computing network, r represents the computing load in the mobile edge computing network,Representing the number of combinations of r edge nodes selected from the K edge nodes;
Step 3.2, the edge node selects elements meeting the conditions from the decoding matrix in the complex domain, and the composition size is Decoding matrix of uploading stage of (a)The condition means that the value of each element in the product matrix corresponding to the selected element is 0, so that the total number of the interference sub-calculation tasks eliminated by the edge node, the relation between the degree of freedom of each edge node and the number of the edge node, the number of users, the calculation load, the number of antennas configured by the edge node and the number of antennas configured by the users are obtained, and the total number of the interference sub-calculation tasks eliminated by each edge node is the same; where N t denotes the number of antennas configured by the edge node in the mobile edge computing network,A decoding matrix representing a j-th edge node;
and 4, calculating the uplink time delay of each user for transmitting the calculation subtask to the edge node according to the following formula:
wherein, Representing the uplink time delay of the ith user for transmitting the calculation subtask to the edge node, wherein an upper corner mark u represents an uplink symbol, and min { A, B } represents the minimum value of A and B;
Where N t represents the number of antennas configured by the edge node in the mobile edge computing network, and N r represents the number of antennas configured by the user in the mobile edge computing network;
step 5, calculating and downloading a user expected value:
Step 5.1, calculating a transmission expected value of each edge node, and transmitting the expected value of the node to a user thereof;
Step 5.2, each user receives the expected value sent by the edge node;
Step 6, designing a precoding matrix and a decoding matrix in a downloading stage:
step 6.1, the edge node selects elements meeting the conditions from the precoding matrix of the complex domain to form a precoding matrix of the downloading stage with the size of N t×ds Obtained by the same methodA plurality of precoding matrices; the condition means that the values of each element in the tensed space of the product matrix corresponding to the selected element are equal to obtain the received value of the userThe term interference expectations are all aligned to the same space, wherein,Representing a j-th edge node transmit set in a mobile edge computing networkThe kth edge node cooperation group is sent to a precoding matrix which is the interference expected value of the kth user;
Step 6.2, the user selects the elements meeting the conditions from the decoding matrix in the complex domain, the composition size is Decoding matrix of the download phase of (a)The condition means that the value of each element in the product matrix corresponding to the selected element is 0, so that the total number of interference expected values eliminated by the user, the relation between the degree of freedom of each user and the number of edge nodes, the number of users, the calculation load, the number of antennas configured by the edge nodes and the number of antennas configured by the users are obtained, and the total number of interference expected values which can be eliminated by each user is the same; wherein,A decoding matrix representing an ith edge node;
Step 7, calculating the downlink time delay for each edge node to transmit the expected value to the user according to the following formula:
wherein, The j-th edge node transmits the expected value to the downlink time delay of the user, the upper corner mark D represents a downlink symbol, and min { C, D } represents the minimum value in C and D;
Where N r represents the number of user configured antennas in the mobile edge computing network and N t represents the number of edge node configured antennas in the mobile edge computing network.
Compared with the prior art, the invention has the following advantages:
Firstly, the invention obtains the calculation load by calculating the ratio of the total task bit number distributed to all edge nodes to the total input bit number of all user equipment, designs the scheme of the interference alignment technology according to the system configuration and the calculation load to eliminate the interference in the task uploading stage, designs the scheme of the interference neutralization and the interference alignment technology according to the system configuration and the calculation load to eliminate the interference in the result downloading stage, and overcomes the defect that infinite time slot expansion is needed for eliminating the interference in the prior art, so that the invention improves the degree of freedom of the MIMO mobile edge calculation network and further reduces the communication delay.
Secondly, according to the scheme of the interference alignment technology in the task uploading stage and the scheme of the interference neutralization and interference alignment technology in the result downloading stage, the method and the device further obtain the expressions of communication time delay, the number of edge nodes, the number of users, the number of any configured antennas and the calculation load in the task uploading stage and the result downloading stage, overcome the defect that the prior art only aims at a single antenna scene, enable the method and the device to be suitable for the single antenna scene and also suitable for a multi-antenna scene in engineering practice, and enlarge the application range.
Drawings
FIG. 1 is a flow chart of the present invention;
fig. 2 is a diagram of simulation results of the present invention.
Detailed Description
The invention is further described below with reference to the drawings and examples.
The specific implementation steps of the implementation of the present invention will be further described with reference to fig. 1 and the embodiment.
And step 1, distributing tasks.
The calculation tasks are divided and uploaded for calculation, so that the completion of the calculation tasks can be quickened. However, in order to ensure that the computing tasks of each user are all computed, it is required that all the equally divided sub-computing tasks of each user are computed by the edge node and that each sub-computing task is independent from each other.
And the mobile edge computing system equally divides the received input bit number of each user computing task according to the computing load to obtain a plurality of equally divided sub computing tasks of the user. And each user uploads each sub-calculation task after the sub-calculation tasks are evenly divided to different edge node sets. In the embodiment of the invention, the number of input bits of the calculation task of the mth user is 8, and the calculation task of the user is divided into two irrelevant sub-calculation tasks according to the calculation load of 2, wherein the number of input bits of each sub-calculation task is 4. The mth user uploads each sub-computation task to two different edge nodes respectively. The edge node fairly treats the sub-computation tasks sent from all users.
Step 2, calculating the calculation load of each task according to the following formula:
Where r p denotes the computational load of the p-th task, Representing the equivalence to the sign, Σ representing the accumulation operation, j representing the jth edge node in the mobile edge computing network, e representing the belonging to the sign,Representing a set of edge nodes in a mobile edge computing network, i representing an ith user in the mobile edge computing network,Representing a set of users in a mobile edge computing network, phi representing the number of computing replications, ranging in value from 1 < phi < K,The representation is included in the symbol,Representing belonging symbols, W i,Φ represents a set of sub-computation tasks assigned to be computed at a j-th edge node in the mobile edge computing network, |·| represents an absolute value taking operation, M represents the number of users in the mobile edge computing network, and L represents the number of input bits per task in the mobile edge computing network.
In general, the computational load r is the average number of each input bit of the computational task at the edge node, thus exhibiting repeatability in computing the sub-computational task.
For a given feasible task allocation scheme, the overall task offloading process also includes two communication phases, one being a calculation task upload phase and the other being a user expectation download phase.
And step 3, uploading the calculation task to an edge node in the mobile edge calculation network.
Step 3.1, each user sends all sub-calculation tasks after equipartition to the edge node according to the following steps:
wherein, Representing the signal sent by the ith user in the mobile edge computing network to the edge node,Representing a set of multicast groups in a mobile edge computing network consisting of K edge nodes selected from r edge nodes,Represents the number of combinations of r elements from K elements, K represents the number of edge nodes in the mobile edge computing network, r represents the computational load of the computing task, j represents the jth multicast group in the mobile edge computing network,Denoted as the ith user in a mobile edge computing network sent to a collectionData of the j-th multicast group in (c),Is thatPrecoding matrix of data of the multicast group, N r is an antenna configured for a user, d s represents a length of a data signal,AndAll of the elements in (a) belong to the complex domain.
And 3.2, each edge node obtains a receiving signal according to the signal sent by the user.
And 4, designing a precoding matrix and a decoding matrix in an uploading stage.
In order to eliminate the interference signals in the received signals in step 3.2 and to reserve space for the desired signals in the received signals, an interference alignment technique is used to design the precoding matrix and decoding matrix of the uplink stage to achieve this.
Step 4.1, the user selects elements meeting the conditions from the precoding matrix of the complex domain to form a precoding matrix with the size of N r×ds in the uploading stageObtained by the same methodAnd precoding matrices. The condition means that the values of each element in the tense space of the product matrix corresponding to the selected element are equal to obtain the value received by the edge nodeThe term jammer calculation tasks are all aligned to the same space, where,Indicating that the ith user is sent toThe precoding matrix of the sub-computation task of the p-th multicast group,Representing the edge node set formed by the edge nodes except for the jth edge node in the edge node set.
Step 4.2, the edge node selects elements meeting the conditions from the decoding matrix in the complex domain, and the composition size isDecoding matrix of uploading stage of (a)The condition means that the value of each element in the product matrix corresponding to the selected element is 0, and the number of the interference sub-calculation tasks eliminated by the edge node is obtainedThe number of jammer calculation tasks that each edge node can eliminate is the same, wherein,Representing the decoding matrix of the jth edge node.
Step 5, calculating the uplink time delay of each user for transmitting the calculation subtask to the edge node according to the following formula:
wherein, The uplink time delay of the ith user transmitting the calculation subtask to the edge node is represented, the upper corner mark u represents an uplink symbol, and min { A, B } represents the minimum value in A and B.
Where N t represents the number of edge node configured antennas in the mobile edge computing network and N r represents the number of user configured antennas in the mobile edge computing network.
And 6, calculating and downloading a user expected value.
Step 6.1, calculating a transmission expected value of each edge node according to the following formula, and transmitting the expected value of the node to a user thereof:
wherein, Representing the expected value sent by the jth edge node in the mobile edge computing network to its corresponding user,Representing a set of edge node cooperation groups in the mobile edge computing, each cooperation group consisting of a jth edge node and r-1 edge nodes selected from the remaining K-1 edge nodes in the mobile edge computing network, the number of cooperation groups in the cooperation group set beingThe number of the two-dimensional space-saving type,Representing a collectionThe kth edge node cooperation group of (a) transmits the expected value to the corresponding ith user, and the expected value is collected in the networkThe calculation matrix of the edge node in the k-th edge node cooperation group is obtained by multiplying the calculation matrix with the calculation subtasks uploaded by the calculation matrix,Representing the jth edge node to send data to the userIs used for the pre-coding matrix of the (c),AndAll belong to the complex domain.
Step 6.2, each user receives the expected value sent by the edge node.
And 7, designing a precoding matrix and a decoding matrix in the downloading stage.
In order to eliminate the interference signals in the received signals in step 6.2 and reserve space for the expected signals in the received signals, a precoding matrix and a decoding matrix of a downlink stage are designed by adopting a combined technology of interference alignment and interference neutralization so as to achieve the purpose.
Step 7.1, the edge node selects elements meeting the conditions from the precoding matrix of the complex domain to form a precoding matrix of a downloading stage with the size of N t×ds Obtained by the same methodA plurality of precoding matrices; the condition means that the values of each element in the tensed space of the product matrix corresponding to the selected element are equal to obtain the received value of the userThe term interference expectations are all aligned to the same space, wherein,Representing a j-th edge node transmit set in a mobile edge computing networkThe kth edge node cooperation group is sent to a precoding matrix which is the interference expected value of the kth user.
Step 7.2, the user selects elements meeting the conditions from the decoding matrix in the complex domain, the composition size isDecoding matrix of the download phase of (a)The condition means that the value of each element in the product matrix corresponding to the selected element is 0, so that the total number of interference expected values eliminated by the user, the relation between the degree of freedom of each user and the number of edge nodes, the number of users, the calculation load, the number of antennas configured by the edge nodes and the number of antennas configured by the users are obtained, and the total number of interference expected values which can be eliminated by each user is the same; wherein,Representing the decoding matrix of the i-th edge node.
Step 8, calculating the downlink time delay for each edge node to transmit the expected value to the user according to the following formula:
wherein, The j-th edge node is represented by the downlink delay for transmitting the expected value to the user, the upper corner mark D is represented by the downlink symbol, and min { C, D } is represented by taking the minimum value of C and D.
Where N r represents the number of user configured antennas in the mobile edge computing network and N t represents the number of edge node configured antennas in the mobile edge computing network.
The effects of the present invention are further described below in conjunction with simulation experiments:
1. simulation experiment conditions:
The hardware platform of the simulation experiment of the invention is: the main frequency is 3.5GHz, and the Intel i75930kCPU processor with 16GB memory.
The software platform of the simulation experiment of the invention is: windows10 operating system and MatlabR2018b emulation software.
2. Simulation content and result analysis:
The simulation experiment of the invention adopts the method of the invention and a prior art (progressive interference alignment technology), in the range of time slot parameters n epsilon [1,10], the communication time delay in the mobile edge computing network is simulated by taking each time slot parameter as a unit, and the simulation result is recorded and the point connection is drawn, as shown in figure 2.
The mobile edge computing network in the simulation experiment comprises 4 edge nodes, 4 users and the computing load is 2, and the edge nodes and the user equipment in the network are only configured with a single antenna.
The prior art progressive interference alignment method refers to:
tao Meixia a progressive interference alignment method employed in a resource allocation method of a mobile edge computing network as set forth in its published paper "Exploiting ComputationReplication forMobile Edge Computing:A Fundamental Computation-Communication Tradeoff Study"(IEEE Transactions onWireless Communications,vol.19,no.7,pp.4563-4578,July2020.).
The effects of the present invention are further described below in conjunction with the simulation diagram of fig. 2.
Fig. 2 is a comparison graph of communication delays obtained by the method of the present invention and the method of the prior art respectively under a mobile edge computing network in a simulation experiment of the present invention. The abscissa of fig. 2 represents the time slot related parameters and the ordinate represents the normalized downlink delay NDLT (normalized downloading time). The curve marked with a fork in fig. 2 represents a simulated communication delay line using the prior art method, and the curve marked with a star represents a simulated communication delay line using the method of the present invention.
As can be seen from the two simulation curves in fig. 2, when n is smaller than 9, the communication delay obtained by the present invention is lower than that obtained by the prior art.
The simulation experiment results show that the resource allocation method combining interference alignment and interference neutralization is utilized, the defect that infinite time slot expansion is required for eliminating interference in the prior art is overcome, and communication time delay is further reduced.
Claims (3)
1. The method is characterized IN that a precoding matrix and a decoding matrix of an uploading stage and a downloading stage are respectively designed to obtain expressions of uplink time delay and downlink time delay, the number of edge nodes, the number of users, the calculation load, the number of antennas configured by the edge nodes and the number of antennas configured by the users; the method comprises the following steps:
Step1, distributing computing tasks:
the mobile edge computing system processes the computing tasks according to each computing load, and equally divides the received input bit number of each user computing task to obtain a plurality of equally divided sub computing tasks of the user;
Each of the computational loads is derived from the following equation:
Where r q denotes the computational load of the q-th task, Representing the equivalence to the sign, Σ representing the summation operation, j representing the jth edge node in the mobile edge computing network, e representing the belonging to the sign,Representing a set of edge nodes in a mobile edge computing network, i representing an ith user in the mobile edge computing network,Representing a set of users in a mobile edge computing network, phi representing the number of computing replications, ranging in value from 1 < phi < K,The representation is included in the symbol,Representing belonging to a symbol, W i,Φ represents a set of sub-computation tasks allocated to be computed at a jth edge node in the mobile edge computing network, |·| represents an absolute value taking operation, M represents the number of users in the mobile edge computing network, and L represents the number of input bits for each computation task in the mobile edge computing network;
Step 2, uploading the calculation task to an edge node in the mobile edge calculation network:
Step 2.1, each user sends all sub-calculation tasks after equipartition to the edge node;
Step 2.2, each edge node receives all sub-calculation tasks sent by each user;
Step 3, designing a precoding matrix and a decoding matrix in an uploading stage:
Step 3.1, the user selects elements meeting the conditions from the precoding matrix of the complex domain to form a precoding matrix with the size of N r×ds in the uploading stage Obtained by the same methodA plurality of precoding matrices; the condition means that the values of each element in the tense space of the product matrix corresponding to the selected element are equal to obtain the value received by the edge nodeThe term-jammer computing tasks are all aligned to the same space; where N r denotes the number of user-configured antennas in the mobile edge computing network, d s denotes the length of the sub-computing tasks,Indicating that the ith user is sent toThe precoding matrix of the sub-computation task of the p-th multicast group,Represents the edge node set formed by the edge nodes except for the jth edge node in the edge node set, M represents the total number of users in the mobile edge computing network, K represents the total number of edge nodes in the mobile edge computing network, r represents the computing load in the mobile edge computing network,Representing the number of combinations of r edge nodes selected from the K edge nodes;
Step 3.2, the edge node selects elements meeting the conditions from the decoding matrix in the complex domain, and the composition size is Decoding matrix of uploading stage of (a)The condition means that the value of each element in the product matrix corresponding to the selected element is 0, so that the total number of the interference sub-calculation tasks eliminated by the edge node, the relation between the degree of freedom of each edge node and the number of the edge node, the number of users, the calculation load, the number of antennas configured by the edge node and the number of antennas configured by the users are obtained, and the total number of the interference sub-calculation tasks eliminated by each edge node is the same; where N t denotes the number of antennas configured by the edge node in the mobile edge computing network,A decoding matrix representing a j-th edge node;
and 4, calculating the uplink time delay of each user for transmitting the calculation subtask to the edge node according to the following formula:
wherein, Representing the uplink time delay of the ith user for transmitting the calculation subtask to the edge node, wherein an upper corner mark u represents an uplink symbol, and min { A, B } represents the minimum value of A and B;
Where N t represents the number of antennas configured by the edge node in the mobile edge computing network, and N r represents the number of antennas configured by the user in the mobile edge computing network;
step 5, calculating and downloading a user expected value:
Step 5.1, calculating a transmission expected value of each edge node, and transmitting the expected value of the node to a user thereof;
Step 5.2, each user receives the expected value sent by the edge node;
Step 6, designing a precoding matrix and a decoding matrix in a downloading stage:
step 6.1, the edge node selects elements meeting the conditions from the precoding matrix of the complex domain to form a precoding matrix of the downloading stage with the size of N t×ds Obtained by the same methodA plurality of precoding matrices; the condition means that the values of each element in the tensed space of the product matrix corresponding to the selected element are equal to obtain the received value of the userThe term interference expectations are all aligned to the same space, wherein,Representing a j-th edge node transmit set in a mobile edge computing networkThe kth edge node cooperation group is sent to a precoding matrix which is the interference expected value of the kth user;
Step 6.2, the user selects the elements meeting the conditions from the decoding matrix in the complex domain, the composition size is Decoding matrix of the download phase of (a)The condition means that the value of each element in the product matrix corresponding to the selected element is 0, so that the total number of interference expected values eliminated by the user, the relation between the degree of freedom of each user and the number of edge nodes, the number of users, the calculation load, the number of antennas configured by the edge nodes and the number of antennas configured by the users are obtained, and the total number of interference expected values which can be eliminated by each user is the same; wherein,A decoding matrix representing an ith edge node;
Step 7, calculating the downlink time delay for each edge node to transmit the expected value to the user according to the following formula:
wherein, The j-th edge node transmits the expected value to the downlink time delay of the user, the upper corner mark D represents a downlink symbol, and min { C, D } represents the minimum value in C and D;
Where N r represents the number of user configured antennas in the mobile edge computing network and N t represents the number of edge node configured antennas in the mobile edge computing network.
2. The method for allocating resources by combining IA and IN a MIMO mobile edge computing network according to claim 1, wherein each user sends all sub-computing tasks after equipartition to its edge node IN step 2.1 is:
wherein, Representing all sub-computation tasks after equipartition sent by the ith user in the mobile edge computing network to its edge nodes, Σ representing the accumulation operation, j representing the sequence number of the multicast group in the mobile edge computing network edge node set,Representing a set of edge nodes for an i-th user in a mobile edge computing network, each element in the set being a multicast group, each multicast group being a combination of computing load individual edge nodes to which the user uploads its divided sub-computing tasks,Representing the transmission of an ith user to a set of edge nodes in a mobile edge computing networkA sub-computation task of a j-th multicast group, the sub-computation task belonging to a complex domain,Representing send sub-computing tasksWhich belongs to the complex domain.
3. The method for resource allocation IN association with IA and IN a MIMO mobile edge computing network according to claim 1, wherein the expected transmission value of each edge node IN step 5.1 is obtained by:
wherein, Representing the expected value sent by the jth edge node in the mobile edge computing network to its corresponding user, k representing the sequence number of the cooperating group set of edge nodes,Representing a set of edge node cooperation groups in a mobile edge computing network,Representing a collectionThe kth edge node cooperation group of (a) transmits the expected value to the corresponding ith user, and the expected value is collected in the networkThe calculation matrix of the edge node in the k-th edge node cooperation group is obtained by multiplying the calculation matrix with the calculation subtasks uploaded by the calculation matrix, the expected value belongs to a complex domain,Representing the jth edge node to send data to the userWhich belongs to the complex domain.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210571856.9A CN114980198B (en) | 2022-05-24 | 2022-05-24 | Resource allocation method for combining MIMO mobile edge computing network IA and IN |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210571856.9A CN114980198B (en) | 2022-05-24 | 2022-05-24 | Resource allocation method for combining MIMO mobile edge computing network IA and IN |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114980198A CN114980198A (en) | 2022-08-30 |
| CN114980198B true CN114980198B (en) | 2024-08-02 |
Family
ID=82956336
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210571856.9A Active CN114980198B (en) | 2022-05-24 | 2022-05-24 | Resource allocation method for combining MIMO mobile edge computing network IA and IN |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114980198B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103986509A (en) * | 2014-05-30 | 2014-08-13 | 西安电子科技大学 | Cooperative multi-point transmission method based on interference alignment and interference neutralization |
| CN107147468A (en) * | 2017-04-11 | 2017-09-08 | 西安电子科技大学 | Joint adaptive disturbance management method based on multi dimensional resource |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101587542B1 (en) * | 2009-05-15 | 2016-01-21 | 삼성전자주식회사 | Method of allocating resource for hierachical celluar system and transport frame for performing the method |
| CN107579763B (en) * | 2017-09-07 | 2020-07-10 | 西安电子科技大学 | Inter-cluster interference elimination method based on spatial fine high-resolution beams |
-
2022
- 2022-05-24 CN CN202210571856.9A patent/CN114980198B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103986509A (en) * | 2014-05-30 | 2014-08-13 | 西安电子科技大学 | Cooperative multi-point transmission method based on interference alignment and interference neutralization |
| CN107147468A (en) * | 2017-04-11 | 2017-09-08 | 西安电子科技大学 | Joint adaptive disturbance management method based on multi dimensional resource |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114980198A (en) | 2022-08-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109005551B (en) | Multi-user NOMA downlink power distribution method of non-ideal channel state information | |
| CN106160809B (en) | Mixed precoding method and device for multi-user multi-antenna system | |
| RU2689316C2 (en) | Wireless data transmission device, wireless data transmission system, wireless data transmission method and program | |
| WO2018212258A1 (en) | Receiver, communication system, and method implemented by computer for enabling both analog and digital beamforming in communication system | |
| CN105453465A (en) | Radio base station apparatus and scheduling method | |
| KR20080014213A (en) | Apparatus and method for low-complex scheduling in multiuser mimo system | |
| CN109104225A (en) | A kind of optimal extensive MIMO Beam Domain multicast transmission method of efficiency | |
| CN107408967A (en) | Wireless communication control method and wireless communication system | |
| CN111970033B (en) | Large-scale MIMO multicast power distribution method based on energy efficiency and spectrum efficiency joint optimization | |
| JP7170863B2 (en) | Method for enabling analog precoding and analog combining | |
| JPWO2006098387A1 (en) | Transmission method and transmission apparatus used in multi-antenna wireless communication system | |
| CN101675601A (en) | A method for communicating in a mimo context | |
| US11323164B2 (en) | Communication method and apparatus in cloud radio access network | |
| CN113824478A (en) | Broadband millimeter wave multi-user large-scale MIMO uplink spectrum efficiency optimization method assisted by discrete lens antenna array | |
| CN107094038B (en) | Method, device and system for adjusting power of antenna system | |
| Badrudeen et al. | Sub-connected structure hybrid precoding for millimeter-wave NOMA communications | |
| CN114980198B (en) | Resource allocation method for combining MIMO mobile edge computing network IA and IN | |
| CN110677858A (en) | Transmission power and computing resource allocation method based on task migration period of Internet of things | |
| JP4376805B2 (en) | Spatial multiplexing transmission method and transmitter | |
| TWI719914B (en) | Precoding method, base station and computing circuit | |
| Dong et al. | Cell-free ISAC massive MIMO systems with capacity-constrained fronthaul links | |
| CN111917444B (en) | Resource allocation method suitable for millimeter wave MIMO-NOMA system | |
| JP4593489B2 (en) | Wireless communication method and wireless communication device | |
| Mercy Sheeba et al. | Beamforming Techniques for Millimeter Wave Communications-A Survey | |
| Kaushik et al. | Energy efficiency maximization in millimeter wave hybrid MIMO systems for 5G and beyond |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |