CN109831759B - Three-dimensional D2D matching algorithm based on software defined wireless network - Google Patents
Three-dimensional D2D matching algorithm based on software defined wireless network Download PDFInfo
- Publication number
- CN109831759B CN109831759B CN201910171397.3A CN201910171397A CN109831759B CN 109831759 B CN109831759 B CN 109831759B CN 201910171397 A CN201910171397 A CN 201910171397A CN 109831759 B CN109831759 B CN 109831759B
- Authority
- CN
- China
- Prior art keywords
- matching
- user
- algorithm
- network
- full
- 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
Images
Landscapes
- Mobile Radio Communication Systems (AREA)
Abstract
The invention provides a three-dimensional D2D matching algorithm based on a Software Defined Wireless Network (SDWN). under the framework that the software defined network and the wireless network are fused, a controller senses the device cache information in the global network and performs centralized control on users, so that D2D full-duplex communication is realized. The invention realizes the maximization of the user transmission rate of the D2D in the whole network by establishing a mixed integer programming problem by utilizing the DAC (D2D-ware caching) caching strategy and the characteristics of the full-duplex D2D. Aiming at the problem of transmission rate reduction caused by sharing CUE user spectrum resources by full-duplex D2D users, the algorithm applies a power control algorithm with low complexity and high operation speed; and based on a graph theory model, a three-dimensional matching algorithm is constructed to perform D2D full-duplex matching and multiplexing channel allocation, so that one-to-one matching is realized. The algorithm is different from the existing three-dimensional matching algorithm based on price, the algorithm complexity is low, and the realization effect is slightly good.
Description
Technical Field
The invention belongs to a three-dimensional D2D matching algorithm for joint power control under a software-defined wireless network architecture, and relates to an SDWN architecture, a DAC (digital-to-analog converter) cache strategy and a mixed integer programming solving problem, which are particularly represented in a power control algorithm and a three-dimensional matching algorithm based on a hypergraph model.
Background
Software Defined Networking (SDN) is a software-based networking technology that calls for separation of control and data. The software defined network architecture divides the network into an application layer, a control layer and a data layer. The application layer provides an open programming interface and a network view to the control layer through a northbound interface, the control layer comprises a controller and a network operating system, and controls the data layer to process, forward and collect data through a southbound interface. The decoupling, open programmable interface and centralized control of the control plane and the data plane enable the software-defined network technology to have unique advantages in the aspects of simplifying network infrastructure and improving network management efficiency. Therefore, the academic community tries to merge two network architectures, namely a content-centric network and a software-defined network, in order to create a network architecture which can not only be efficiently managed, but also satisfy mass data traffic distribution.
The design idea of the software defined network is introduced into a wireless network, according to the basic characteristics of the SDN, a controller can extract all information of the whole network by using a global view of the network to improve the effectiveness of network resource allocation, and then the whole network optimization can be realized through module operation in the controller.
On the other hand, although there is a large amount of content stored in the core network, in practice only a small amount of more popular content can be requested by most mobile data users, and the content follows a certain content popularity distribution, which is shown by a lot of research to be a Zipf distribution. Since most of the data traffic is due to repeated downloads of some popular content, such as popular video and music, researchers have focused on content caching to prevent redundant traffic generation. Caching generally has two ways: first, edge caching, where content is stored on an auxiliary node, such as a cell base station, to reduce backhaul consumption of a cell by reducing transmission of a core network; secondly, the device cache, that is, the content is stored on the user equipment, and the load of the base station is reduced by reducing the cellular transmission, thereby increasing the rate of the cellular cell users and reducing the dynamic energy consumption of the base station. Since the buffered content is provided instantaneously and communicated via the D2D buffered at the local device, the delay for the user to accept the signal is also lower.
With the increase of mobile data traffic, attention is paid to wireless co-frequency simultaneous full duplex communication. Compared with half-duplex communication, full-duplex can potentially improve the spectrum efficiency by one time, realize more flexible spectrum use and reduce end-to-end time delay. Full-duplex D2D communication also helps to achieve the high spectral efficiency and high energy requirements of 5G systems. The channel gain, hop gain, multiplexing gain and duplex gain can improve the overall spectrum efficiency of the system and reduce the energy consumption of users, and the efficient spectrum utilization can obviously increase the number of access users, thereby reducing the deployment density of base stations in a future ultra-dense network and reducing the complexity and cost of the network.
Full-duplex communication based on simultaneous co-frequency transceiving can further improve the performance of D2D communication, but also brings more complicated inter-user interference and self-interference to the existing network, and the interference affects the quality of received signals, and if the interference cannot be effectively controlled, the overall performance and user service experience of the network can be seriously damaged. Therefore, D2D resource optimization configuration and interference management mechanism have been the focus of research in the industry.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention aims to provide a three-dimensional D2D matching algorithm based on a software-defined wireless network, which is used for carrying out centralized control on users based on DAC (digital-to-analog converter) equipment cache under the framework that the software-defined network and the wireless network are fused so as to realize D2D full-duplex communication.
In order to achieve the purpose, the invention adopts the technical scheme that:
a three-dimensional D2D matching algorithm based on a software-Defined Wireless network is characterized in that under an SDWN (software Defined Wireless network) framework with a software-Defined network and a Wireless network fused, a controller senses device cache information in a global network and performs centralized control on users so as to realize D2D full-duplex communication, and by utilizing a DAC (D2D-ware caching) cache strategy and the characteristics of full-duplex D2D, the maximization of user transmission rate of the full-network D2D is realized by establishing a mixed integer programming problem:
the problem of transmission rate reduction caused by sharing CUE user spectrum resources by full-duplex D2D users is solved by using a power control algorithm with low complexity and high operation speed;
and constructing a three-dimensional matching algorithm based on a graph theory model, performing D2D full-duplex matching and multiplexing channel allocation, realizing one-to-one matching and achieving full-network optimization.
The SDWN architecture is divided into two parts: the SDN controller is set in a control plane layer, the controller can extract all information of the whole network by utilizing a global view of the network to improve the effectiveness of network resource allocation, and then the whole network optimization is realized through module operation in the controller; at the data plane layer, a single cell is used as a model, and according to the content cached by the user and the requested content, the user can be matched for carrying out D2D full-duplex communication.
The DAC caching strategy specifically comprises the following steps:
3.1) assume that the user requests content from a library of L files, each user can only store up to K files, K < L, each user can only request one file at a time, and their request obeys Ziff's Law:
where ξ is the popularity index of the ziff distribution, describing one bias of the popularity distribution, and depending on the content type, then q isiThe probability that the user requests the ith file is expressed, and j is a quantity series;
3.2) divide the 2K most popular contents of the L files in the library into two sets of equal numbers of complementary overlaps, defined as set CAAnd set CBI.e. each set has K contents, and sets CAProbability of being requested hAAnd set CBIs requested for content hBAre approximately equal, is recorded as
The two sets are randomly distributed to each user with equal probability, namely 50%, the caching user is defined as CU, and the storage set C is definedAThe user of (A) is UE A, the storage set CBThe user of (1) is UE B;
3.3) when a UE A requests set CBAnd there is a UE B request set C aroundAThen the two can be matched for D2D full duplex communication;
3.4) if there are no CU users around that satisfy the condition or with a probability of 1-hA-hBRequesting file, then receiving signal from base station, namely becoming down user;
3.5) if the content of the self buffer is requested, the receiving signal is not needed, namely, the self-linking is carried out.
The mixed integer programming problem specifically comprises the following steps:
4.1) in order to maximize the weighted sum of all D2D pairs, the buffered users are paired by matching method and the appropriate spectrum is allocated to the corresponding transmission link, i.e. this is a problem of D2D pair matching joint channel allocation, in order to avoid excessive interference to the cellular network, the D2D transmission is power controlled, and the system performance is optimized by combining the binary decision variable and D2D pair and the Cellular User (CUE) power variable:
4.2) firstly, aiming at the problem of transmission rate reduction caused by sharing CUE user spectrum resources by D2D users, decomposing a mixed integer programming problem into an optimized power distribution problem by using a power control algorithm, converting a non-convex problem into a convex problem which is easy to solve, and calculating an approximate optimal solution of the problem by using an iterative algorithm;
4.3) secondly, carrying out D2D full duplex matching and multiplexing channel allocation by using a three-dimensional matching method based on a hypergraph, realizing one-to-one matching and improving the total transmission rate of D2D to the maximum extent.
Defining a hypergraph HG (v,), wherein v is a vertex set and is a hyperedge set, and a weighted hypergraph, namely each edge e ∈ has a weight w (e), in the invention, a pair of matched cache users and CUE users are represented as three vertexes at the hyperedge, each hyperedge has a weight value to display a D2D transmission rate, and transforming the hypergraph HG (v,) into a common graph G (v ', '), wherein each vertex of the graph G is a set of three vertexes of the hypergraph and has a weight value to represent a D2D transmission rate, and the edge ' represents a non-empty set intersection between two vertexes of the graph G, namely two vertexes are combined with one or more weightsThe same element, graph G, is equivalent to the intersection of the sub-graphs, i.e. G ═ G1∪G2∪...∪Gn；
Defining the maximum weighted matching set as omega, and finally aiming at searching the maximum weighted value omega (omega), the algorithm starts from the first subgraph, searches the vertex with the maximum weight to the set omega, and deletes the vertex in the rest subgraphs at the same time, and so on until all the subgraphs are traversed circularly, and the final matching result is the set omega.
The algorithm flow under the SDWN framework specifically comprises the following steps:
5.1) a Base Station (BS) in a data plane under the SDWN architecture collects all user information and channel information in a cell and sends the user information and the channel information to a power control module in a D2D matching module of a control plane for power control distribution calculation;
5.2) the power control module in the SDWN control plane feeds back all power distribution information to the D2D matching module which is also the control plane for D2D matching and channel distribution calculation;
5.3) the D2D matching module of the control plane sends the calculated matching information to the base station in the data plane, and then the base station carries out the final user D2D full duplex communication and channel allocation.
The algorithm of the invention is different from the existing three-dimensional matching algorithm based on price, and the like, and has low algorithm complexity and slightly good realization effect.
Drawings
FIG. 1 is a diagram of a software defined wireless network architecture;
FIG. 2 is a hypergraph-based weighted three-dimensional matching graph;
FIG. 3 is a flow chart of an algorithm under the SDWN architecture;
FIG. 4 is a simulation diagram of the relationship between popularity index ξ and cache hit rate;
fig. 5 is a simulation diagram of the relationship between the number of the users caching different files K, the number of the caching users N and the total rate of D2D.
Detailed Description
The embodiments of the present invention will be described in detail below with reference to the drawings and examples.
The invention provides a three-dimensional D2D matching algorithm based on a Software Defined Wireless Network (SDWN). under an SDWN framework with a software defined network and a wireless network fused, a controller senses device cache information in a global network and performs centralized control on users so as to realize D2D full-duplex communication. The invention realizes the maximization of the user transmission rate of the D2D in the whole network by establishing a mixed integer programming problem by utilizing the DAC (D2D-ware caching) caching strategy and the characteristics of the full-duplex D2D. Aiming at the problem of transmission rate reduction caused by sharing CUE user spectrum resources by full-duplex D2D users, a power control algorithm with low complexity and high operation speed is applied; based on a graph theory model, a three-dimensional matching algorithm is constructed to perform D2D full-duplex matching and multiplexing channel allocation, so that one-to-one matching is realized, specifically:
1. system model
As shown in fig. 1, the SDWN architecture is divided into two parts: a control plane and a data plane.
1) Control plane layer
The control plane layer establishes an SDN controller that includes a series of modules, such as a resource allocation module and a D2D matching module, which can be used in the present invention. According to basic SDN characteristics, the controller can extract all information of the whole network by utilizing a global view of the network to improve the effectiveness of network resource allocation, and then the whole network optimization can be realized through module operation in the controller.
A feedback control loop is designed and proposed in a control plane, and a controller firstly collects full-network information through a base station, sends the information to a power control module, and executes a power control algorithm. The power control module is contained in the D2D matching module, the power optimization information is sent to the D2D matching module for matching and optimization, and finally the result is sent back to the base station for deployment.
2) Data plane layer
The OpenFlow protocol is deployed in a Base Station (Base Station-BS) of a data plane and is used for supporting information interaction between a controller and the BS.
In the data plane, inter-cell disregards for simplicityThe interference is modeled as a single cell. M uplink Cellular Users (CUE) are randomly distributed in a cell, and are defined as CUE ═ CUE1,...,CUEk...,CUEMN Cache users-CU, defined as CU ═ CU }1,...,CUn...,CUN}. The CU users may be D2D matched for full duplex communication based on the content cached by the CU users and the requested content. The pair of D2D needs to multiplex the channel resource of CUE users, each CUE occupies one channel without interference, and each pair of D2D occupies at most one channel, and each channel is occupied by at most one pair of D2D.
2. DAC caching strategy
Assuming that the user requests content from a library of L files, each user can only store up to K (K < L) files, each user can only request one file at a time, and their requests obey the law of Zigh:
where ξ is the popularity index of the ziff distribution, describing one bias in the popularity distribution, and depending on the content type, generally, ξ is between 0.5 and 1.5.
The 2K most popular contents of the L files in the library are divided into two sets of equal numbers of complementary overlaps, defined as set C, according to the DAC caching policyAAnd set CBI.e. each set has K contents and the probability (sum) that the contents in the two sets are requested is roughly equal, noted as
The two sets are randomly distributed with equal probability (1/2) to the respective users, i.e. CU users. Defining a storage set CAThe user of (A) is UE A, the storage set CBWhen a CU user requests a file, the user of (1) is UE B, and there are three cases:
1) when a UE A requestsSet C is solvedBAnd there is a UE B request set C aroundAD2D full duplex communication may be performed;
2) if no CU users satisfying the condition around the CU users exist or the probability is 1-hA-hBRequesting file, then receiving signal from base station, namely becoming down user;
3) if the self-cached content is requested, no signal is required to be received, i.e. autonomous linking.
According to the DAC buffer strategy, defining a request set C in the UE ABIs a user set of
Requesting set C among UE BAIs a user set of
When certain communication distance conditions are met, RB and RA may be matched for full-duplex D2D matching.
3. Data layer plane model
In the data plane, CUE ═ { CUE is defined1,...,CUEk...,CUEMAnd the CU users capable of matching and forming full-duplex D2D communication request set C in the UE A according to the DAC caching strategyBTo a user
And request set C among UE BATo a user
Assuming that the channel allocation of the CUEs has been predetermined in advance, each CUE occupies one channel, i.e., channel CHkIs assigned to CUEkThere is no interference with each other. Each D2D pair occupies at most one channel and each channel is occupied by at most one D2D pair. The following describes an optimized power allocation algorithm under full-duplex D2D communication.
In each pair of D2D, the two full-duplex D2D users transmit and receive signals on the same frequency at the same time, and thus self-interfere with each other. In addition, in the same channel, co-channel interference exists between the D2D pair and the CUE, as shown in fig. 1. The interference model for the full-duplex D2D system is more complex than for the half-duplex D2D cellular system. The base station will receive the joint channel interference of the two users of D2D pair. At the same time, each D2D user experiences channel interference from the CUE and self-interference from its transmit antenna.
Suppose RBiAnd RAjCan form D2D pairs, and is combined with CUEkShared channel CHk. D2D user RBi(RAj) Accept from RAj(RBi) Signals, self-interference due to full duplex transmission and interference from CUEkThe interference of (2). Thus RBiAnd RAjIn the channel CHkThe signal to interference plus noise ratio (SINR) of (c) is:
wherein the content of the first and second substances,
represents RBi、RAjThe power of the transmission is transmitted,
represents CUEkβ of1And β2In order to be a self-interference coefficient,
represents RBiAnd RAjThe gain of the channel in between is increased,
represents CUEkTo RBiThe interference channel of (a) is,
represents CUEkTo RAjInterference channel of, N0Representing gaussian white noise.
According to Shannon theory, D2D is paired in channel CHkThe total transmission rate achievable is:
the base station will accept the signal from CUEkAnd from D2D pairs RBi、RAjThe interference of (2). So that the base station is on channel CHkThe SINR of (1) is as follows:
wherein the content of the first and second substances,
represents CUEkThe channel gain with the BS is determined,
represents RBiThe interference channel to the BS is transmitted,
represents RAjAn interfering channel to the BS.
Likewise, the achievable transmission rates are:
4. problem model construction
It is an object of the present invention to meet user requirements through D2D technology while maximizing the transmission rate of D2D links. Taking the D2D pair transmission rate as the objective function:
where ρ isijFor binary parameters, when RBiAnd RAjAt a distance d betweenmaxWithin the range, it is 1, otherwise it is 0.
In order to maximize the sum of all D2D pairs, the present invention applies an efficient matching method to pair RB and RA while allocating the appropriate spectrum to the transmission link, i.e., it is a problem of D2D pair matching joint channel allocation. Meanwhile, to avoid excessive interference to the cellular network, D2D transmissions should be power controlled.
Define binary variable X ═ Xi,j,kThe D2D pair matching and channel matching conditions are adopted when x isi,j,k1 denotes multiplex CHkRB ofiAnd RAjThe D2D pair was successfully established. Joint binary decision variable xi,j,kAnd power variation
To optimize system performance. A mixed integer programming model is constructed:
P1:
s.t.:C1:
C2:
C3:
C4:
condition C1 represents the transmission power range of the D2D pair and guarantees that the maximum value is not exceeded
Condition C2 ensures that the CUE transmission power does not exceed the maximum value
Condition C3 guarantees QoS requirements for cellular links, and the three inequalities of condition C4 represent each RBiCan only match up to one RAjAnd vice versa, while each channel can be allocated to only one pair of D2D at most.
5. D2D pair matching and channel allocation algorithm
1) Power control algorithm
To solve the P1 mixed integer programming problem, P1 is decomposed, first solving the problem of optimizing power allocation. Multiplexing CUEs for each pairkRB ofi、RAjIs provided with
P2:
s.t.:C1:
C2:
C3:
P2 is a non-convex problem, thus converting the non-convex problem into a convex problem that is easy to solve.
First, condition C3 was changed to
Wherein the content of the first and second substances,
in order to meet the maximum D2D rate, the power of the cellular user is preferably smaller, since the interference generated is relatively smaller, but the QoS of the CUE is also met, and thus
The lower bound can be taken off directly, i.e.
Will be provided with
The following forms were converted:
wherein
In addition
The D2D transmission power will transition to
Question P2 may be reduced to the form of question P3
P3:
s.t.:C1：
It can be seen that conditions C1 and C2 of problem P3 are both linear, but the utility function is still non-concave, so problem P3 is still non-convex. The P3 problem was solved according to lemma 1, the proof of lemma 1 is in appendix a.
Introduction 1: definition R (z) ═ log2(1+ z), z ≧ 0, a given positive value of z0∈ [0, ∞) is the lower bound of R (z)
R(z,z0)＝alog2z+b
Wherein
Based on the theorem 1, the condition C1 that the problem P3 is satisfied is defined, and the power of C2 is
Will be provided with
To be converted into
Wherein, a1、b1And a2、b2Are respectively as
The problem P3 may thus be converted into:
P4:
s.t.:C1：
C2：
using logarithmic transformation
The problem P4 is transformed into the following form:
P5:
s.t.:C1：
C2：
wherein
Problem P5 is the equation transformation of P4, and solving P5 is equivalent to solving P4. Therefore, we have the following two theorems.
Theorem 1: problem P5 is a concave function and conditions C1 and C2 are convex constraints, so problem P5 is a strict convex optimization problem and can be solved by standard algorithms. The proof is in appendix B.
Theorem 2: it can be shown that the transmission rate obtained by the problem P4 is better than that obtained by the problem
The incoming transmission rate. Definition of
The optimal solution for problem 4 would be
The demonstration is in appendix C.
The specific algorithm is shown in fig. 3.
Appendix A:
proof of lemma 1:
is provided with
When in use
With z ═ z0，
Minimum value of f (z)
So that f (z) ═ R (z) -R(z,z0)＝log2(1+z)-(alog2z + b) is equal to or greater than 0, therefore, R (z) ≧ 0R(z,z0) And theory 1 proves completion.
Appendix B:
Appendix C:
based on the introduction of 1, can obtain
The first inequality shows
Is that
The second inequality indicates
Is the optimal solution to the problem P4.
2) Three-dimensional matching algorithm based on hypergraph
With the goal of maximizing the transmission rate of the D2D network, the buffer-based RA and RB users are formulated to perform D2D matching and reusable channel allocation. Defining all possible potential combinations of pairable RA, RB and CUE users as the set Φ { (i, j, k) }, meaning RBiAnd RAjMatching into D2D pairs and multiplexing CUEkAnd channel resources, a weighted three-dimensional matching graph can be constructed, and a three-dimensional matching method based on a hypergraph is provided based on equipment cache so as to improve the total transmission rate of the D2D to the maximum extent. As shown in fig. 2.
The hypergraph is defined as n-uniform.rb users, RA users and CUE users can be represented as three vertices at the hyper-edge, each hyper-edge has a weight n, which is derived by a power control algorithm, and therefore, device cache based D2D matching and channel allocation can be expressed as a weighted 3-uniform hypergraph.
The matching result of the hypergraph HG is a subset of the hyperedges, and no hyperedge contains the same vertex, i.e. the one-to-one matching condition is satisfied, and the matching base number is the number of matching edges. The multidimensional maximum weighted match of a weighted hypergraph HG is a subset of the matching hyperedges of the graph HG. In the present invention, the goal is to solve the 3-dimensional maximum weighted matching.
First, the hypergraph HG ═ v, (v,) is transformed into G ═ v ', where each vertex of graph G is a set of three vertices of the hypergraph, and the edge ' is all intersections between vertices v '. In the weighted 3-uniform hypergraph, the hyper-edges of HG are equivalent to the vertices of graph G, and thus each vertex of graph G has a weight value representing the matching D2D transmission rate. Defining the maximum weight matching set as Ω and Ω as a subset of all combinations (i, j, k), the present invention seeks an independent set with the maximum weight ω (Ω). As shown in fig. 2, the vertices of the weighted graph G are all potential combinations Φ { (i, j, k) }, meaning RBiAnd RAjMatching into D2D pairs and multiplexing CUEkChannel resources, then each vertex has a weight. The edge of G represents a non-empty set intersection between two vertices, i.e., two combined with one or more identical elements. In this case, graph G is equivalent to the intersection of the subgraphs, i.e. G ═ G1∪G2∪...∪Gn. The purpose of the power allocation algorithm is to find the maximum weight ω (Ω), as follows:
1. initialization, setting l to 0 and setting the initial value of the problem to
The error is greater than 0;
2. iteration
3.l＝l+1；
4.
5. Calculating corresponding coefficients
6. Solving a convex optimization problem P5 and obtaining an optimal solution
7. Will be provided with
Performing exponential transformation to obtain the optimal solution of the problem P4
8. Up to
9. End of iteration
10. Return to
The algorithm starts from the first subgraph, finds the vertex with the largest weight to the set omega, and deletes the vertex in the rest subgraphs. And repeating the steps until all the sub-graphs are circularly traversed, wherein the final matching result is the set omega.
3) Complexity analysis
And analyzing and comparing the hypergraph-based three-dimensional matching algorithm with the price-based three-dimensional matching algorithm. Since the main difference between the two algorithms is the final matching method, only the temporal complexity (number of iterations) of this part is considered.
The number of RB users is known to be N1The number of RA users is N2The CUE user number is M. In the price-based three-dimensional matching algorithm, the complexity of each RB user for matching the RB and CUE cell blocks which are most matched is as follows
Wherein
The final iteration number after price is increased based on the step length s, so the complexity of the algorithm matching process is
(N2Not less than M) or
(M≥N2). In the three-dimensional matching algorithm based on the hypergraph, the matching algorithm is mainly concentrated on step2 part in the algorithm 2, and the algorithm can know that min (N) can be generated at most to realize one-to-one matching result1,N2M) subgraphs, the algorithm loops through each subgraph to find the highest weight vertex, so the complexity of the algorithm matching process is O (min (N)1,N2M)). It can be seen that the computation amount of the hypergraph-based three-dimensional matching algorithm is small.
6. Algorithm flow under SDWN
The algorithm flow is shown in fig. 3, and the specific operation process is as follows:
1) the BS collects the intra-cell user information and the channel information and sends the intra-cell user information and the channel information to a power control module of a D2D matching module of the control plane to perform power allocation calculation of algorithm 1.
2) The power control module feeds back information to the D2D matching module for D2D matching of algorithm 2 and channel allocation calculation.
3) The D2D matching module sends control information to the base station for final user D2D matching and channel assignment.
7. Simulation experiment
The performance of the hypergraph-based three-dimensional D2D matching and channel allocation algorithm is researched through simulation, and compared with the price-based three-dimensional matching algorithm and the random matching algorithm, the simulation result shows that the algorithm constructed by the method can achieve better effect compared with other two algorithms. Some of the simulation parameters are set forth in the table below.
The cache hit rate is defined as the percentage of requests in the device that result in a cache hit FIG. 4 is a correlation of the popularity index ξ with the cache hit rate.
Matching and channel assignment algorithm by D2D:
it can be known that, when the number K of the device cache files is larger, the cache hit rate is larger, and as ξ is increased, the cache hit rate is also larger, but not more than 0.5, and zigh's law is satisfied.
Fig. 5 simulates a relationship between the number N of the cached users and the total rate of D2D when the number of the cached files of the user is different, i.e., a cellular user M is 8, and a popularity index ξ is 1.0, when the number of the cached files of the device is 100, the total rate is obviously greater than the total rate D2D when K is 50, but when K is 200, the total rate D2D is smaller than the total rate K100 and is greater than the total rate K50, as can be seen from fig. 4, when K is 200, the self-caching hit rate is greater, the cached users can perform self-linking without the D2D link to meet the requirement, so the number of D2D matches is less, and the total rate is also greater than the total rateIs small. The conclusion is that the content requested by the user belongs to the set C as the number of files cached by the device is largerAAnd set CBIs greater, so the greater the likelihood that D2D matches the communication, the greater the overall rate of D2D; however, when the number of the user cache files is large enough, the larger the content requested by the user belongs to the content cached by the user, the less the D2D is needed to match the communication, and therefore, the total D2D rate is slightly smaller.
In a word, the invention provides a three-dimensional D2D matching algorithm based on Software Defined Wireless Network (SDWN), under the framework that the software defined network and the wireless network are fused, the DAC equipment cache is based on, and the centralized control is carried out on the users, so that the D2D full-duplex communication is realized.
Claims (5)
1. Under an SDWN (software Defined Wireless network) framework with a software Defined network and a Wireless network fused, centralized control is performed on users through perception of device cache information in a global network by a controller, so that D2D full-duplex communication is achieved, and by utilizing a DAC (D2D-aware) cache strategy and the characteristics of full-duplex D2D, maximization of a D2D user transmission rate in the full network is achieved by establishing a mixed integer programming problem:
the problem of transmission rate reduction caused by sharing cellular user spectrum resources by full-duplex D2D users is solved by using a power control algorithm;
constructing a three-dimensional matching algorithm based on a graph theory model, and performing D2D full-duplex matching and multiplexing channel allocation to realize one-to-one matching;
the DAC caching strategy is characterized by comprising the following steps:
3.1) assume that the user requests content from a library of L files, each user can only store up to K files, K < L, each user can only request one file at a time, and their request obeys Ziff's Law:
where ξ is the popularity index of the ziff distribution, describing one bias of the popularity distribution, and depending on the content type, then q isiThe probability that the user requests the ith file is expressed, and j is the quantity series;
3.2) divide the 2K most popular contents of the L files in the library into two sets of equal numbers of complementary overlaps, defined as set CAAnd set CBI.e. each set has K contents, and sets CAProbability of being requested hAAnd set CBIs requested for content hBAre approximately equal, is recorded as
The two sets are randomly distributed to each user with equal probability, namely 50%, the caching user is defined as CU, and the storage set C is definedAThe user of (A) is UE A, the storage set CBThe user of (1) is UE B;
3.3) when one UE A requests set CBAnd there is a UE B request set C aroundAThen the two can be matched for D2D full duplex communication;
3.4) if there are no CU users around that satisfy the condition or with a probability of 1-hA-hBRequesting a file, receiving a signal from a base station, namely, becoming a downlink user;
3.5) if the self-cached content is requested, the signal does not need to be received, namely, the self-linking is carried out.
2. The three-dimensional D2D matching algorithm based on software defined wireless network according to claim 1, wherein: the SDWN architecture is divided into two parts: the SDN controller is set in a control plane layer, the controller extracts all information of the whole network by utilizing a global view of the network to improve the effectiveness of network resource allocation, and then the whole network optimization is realized through module operation in the controller; and at the data plane layer, a single cell is used as a model, and the users are matched according to the content cached by the users and the requested content, so that the D2D full-duplex communication is carried out.
3. The three-dimensional D2D matching algorithm based on software defined wireless network according to claim 1, wherein: the mixed integer programming problem specifically comprises the following steps:
4.1) in order to maximize the weighted sum of all D2D pairs, the buffered users are paired by matching method and the appropriate spectrum is allocated to the corresponding transmission link, i.e. this is a problem of D2D pair matching joint channel allocation, in order to avoid excessive interference to the cellular network, the D2D transmission is power controlled, and the system performance is optimized by combining the binary decision variable and D2D pair and the Cellular User (CUE) power variable:
4.2) firstly, decomposing the mixed integer programming problem into an optimized power distribution problem by using a power control algorithm, converting a non-convex problem into a convex problem which is easy to solve, and calculating an approximate optimal solution of the problem by using an iterative algorithm;
4.3) secondly, carrying out D2D full duplex matching and multiplexing channel allocation by using a three-dimensional matching method based on a hypergraph, realizing one-to-one matching and improving the total transmission rate of D2D to the maximum extent.
4. The three-dimensional D2D matching algorithm based on software defined wireless network as claimed in claim 3, wherein in said 4.3), we define a hypergraph HG ═ (v,), where v is a set of vertices and a set of hyper-edges, and a weighted hypergraph, i.e. a weight w (e) is provided for each edge e ∈, and a matchable pair of cache users and CUE users are represented as three vertices on the hyper-edges, and each hyper-edge has a weight to show D2D transmission rate, and we transform the hypergraph HG ═ v,) into a general graph G ═ v ',', where each vertex v 'of graph G is a set of three vertices of the hypergraph and has a weight to show D2D transmission rate, and edge' represents the non-empty set intersection between two vertices of graph G, i.e. two combined with one or more same elements, graph G is equivalent to the intersection of the graph G ═ G subgraph1∪G2∪…∪Gn(ii) a Definition ofThe maximum weighted matching set is omega, the final purpose is to find the maximum weighted value omega (omega), the algorithm starts from the first subgraph, the vertex with the maximum weight is found to the set omega, the vertex in the rest subgraphs is deleted, and the like until all the subgraphs are traversed circularly, and the final matching result is the set omega.
5. The three-dimensional D2D matching algorithm based on software defined wireless network as claimed in claim 1, wherein: the algorithm flow under the SDWN framework specifically comprises the following steps:
5.1) a Base Station (BS) in a data plane under the SDWN architecture collects all user information and channel information in a cell and sends the user information and the channel information to a power control module in a D2D matching module of a control plane for power control distribution calculation;
5.2) the power control module in the SDWN control plane feeds back all power distribution information to the D2D matching module which is also the control plane for D2D matching and channel distribution calculation;
5.3) the D2D matching module of the control plane sends the calculated matching information to the base station in the data plane, and then the base station carries out the final user D2D full duplex communication and channel allocation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910171397.3A CN109831759B (en) | 2019-03-07 | 2019-03-07 | Three-dimensional D2D matching algorithm based on software defined wireless network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910171397.3A CN109831759B (en) | 2019-03-07 | 2019-03-07 | Three-dimensional D2D matching algorithm based on software defined wireless network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109831759A CN109831759A (en) | 2019-05-31 |
| CN109831759B true CN109831759B (en) | 2020-08-04 |
Family
ID=66865611
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910171397.3A Active CN109831759B (en) | 2019-03-07 | 2019-03-07 | Three-dimensional D2D matching algorithm based on software defined wireless network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109831759B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112261628B (en) * | 2020-10-26 | 2022-07-08 | 杭州梦视网络科技有限公司 | Content edge cache architecture method applied to D2D equipment |
| CN113163408B (en) * | 2021-02-05 | 2023-02-07 | 北京中电飞华通信有限公司 | Wireless resource allocation method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3026963A1 (en) * | 2014-11-25 | 2016-06-01 | Nokia Technologies Oy | Method and apparatus for optimizing transmissions in wireless networks |
| CN105722236A (en) * | 2016-02-23 | 2016-06-29 | 重庆邮电大学 | Resource distribution method for supporting full-duplex D2D communication in cellular network |
| CN107959982A (en) * | 2017-12-31 | 2018-04-24 | 西安科技大学 | A kind of software definition cellular network EPC frameworks and D2D communication partner method |
-
2019
- 2019-03-07 CN CN201910171397.3A patent/CN109831759B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3026963A1 (en) * | 2014-11-25 | 2016-06-01 | Nokia Technologies Oy | Method and apparatus for optimizing transmissions in wireless networks |
| CN105722236A (en) * | 2016-02-23 | 2016-06-29 | 重庆邮电大学 | Resource distribution method for supporting full-duplex D2D communication in cellular network |
| CN107959982A (en) * | 2017-12-31 | 2018-04-24 | 西安科技大学 | A kind of software definition cellular network EPC frameworks and D2D communication partner method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109831759A (en) | 2019-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109413724B (en) | MEC-based task unloading and resource allocation scheme | |
| WO2022121985A1 (en) | Static and dynamic combined millimeter wave beam resource allocation and optimization method | |
| CN111132077A (en) | Multi-access edge computing task unloading method based on D2D in Internet of vehicles environment | |
| CN108495340B (en) | Network resource allocation method and device based on heterogeneous hybrid cache | |
| CN106912079B (en) | Combined user access selection and resource allocation method in cache heterogeneous network | |
| CN109194763B (en) | Caching method based on small base station self-organizing cooperation in ultra-dense network | |
| CN107682935B (en) | Wireless self-return resource scheduling method based on system stability | |
| WO2023179010A1 (en) | User packet and resource allocation method and apparatus in noma-mec system | |
| CN110602722B (en) | Design method for joint content pushing and transmission based on NOMA | |
| CN111556572A (en) | Spectrum resource and computing resource joint allocation method based on reinforcement learning | |
| CN111010684A (en) | Internet of vehicles resource allocation method based on MEC cache service | |
| Yu et al. | Millimeter-wave backhaul traffic minimization for CoMP over 5G cellular networks | |
| Guo et al. | On the asynchrony of computation offloading in multi-user MEC systems | |
| Li et al. | A delay-aware caching algorithm for wireless D2D caching networks | |
| CN105873219A (en) | GASE based TDMA wireless Mesh network resource allocation method | |
| CN109831759B (en) | Three-dimensional D2D matching algorithm based on software defined wireless network | |
| CN106792451A (en) | A kind of D2D communication resource optimization methods based on Multiple-population Genetic Algorithm | |
| CN109788540B (en) | Power control and channel allocation method based on energy collection in D2D system | |
| CN111615129A (en) | Resource allocation method in NOMA-based multi-user mobile edge computing system | |
| CN114880046B (en) | Low-orbit satellite edge computing and unloading method combining unloading decision and bandwidth allocation | |
| CN104254130A (en) | Method, device and system for allocating D2D (Device-to-Device) user link and cellular user shared resources | |
| CN112203308A (en) | Satellite ground fusion network data transmission method | |
| CN109981340B (en) | Method for optimizing joint resources in fog computing network system | |
| CN112654058A (en) | Mobile edge computing offload and resource allocation algorithm in D2D multicast network | |
| CN115866787A (en) | Network resource allocation method integrating terminal direct transmission communication and multi-access edge calculation |
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 |