CN105611574B - A method of combining dynamic access and subcarrier distribution under the super-intensive network based on caching - Google Patents
A method of combining dynamic access and subcarrier distribution under the super-intensive network based on caching Download PDFInfo
- Publication number
- CN105611574B CN105611574B CN201510994459.2A CN201510994459A CN105611574B CN 105611574 B CN105611574 B CN 105611574B CN 201510994459 A CN201510994459 A CN 201510994459A CN 105611574 B CN105611574 B CN 105611574B
- Authority
- CN
- China
- Prior art keywords
- user
- access point
- subcarrier
- access
- value
- 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
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
- H04W28/08—Load balancing or load distribution
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W48/00—Access restriction; Network selection; Access point selection
- H04W48/08—Access restriction or access information delivery, e.g. discovery data delivery
- H04W48/10—Access restriction or access information delivery, e.g. discovery data delivery using broadcasted information
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W48/00—Access restriction; Network selection; Access point selection
- H04W48/20—Selecting an access point
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W72/00—Local resource management
- H04W72/04—Wireless resource allocation
- H04W72/044—Wireless resource allocation based on the type of the allocated resource
- H04W72/0453—Resources in frequency domain, e.g. a carrier in FDMA
Abstract
The method for combining dynamic access and subcarrier distribution under the invention discloses a kind of super-intensive network based on caching, the specific steps are as follows: users multiple first transmit request information to all access points, searching cache contents simultaneously；Then each access point judges whether there is the cache contents of active user K request, and all access points for meeting user K send respective attribute parameter to local control, and best access point is distributed to user K by local control；Otherwise, user K directly sends to remote server and requests, and obtains content；Remote server is analyzed according to user request information using popularity, and buffer update is completed；After last each user matches with respective access point, subcarrier distribution is carried out, makes to communicate between user and access point.Advantage is: comprehensive Multiple factors complete access selection, the promotion of resources management efficiency and the dynamic allocation of subcarrier are realized, so that the availability of frequency spectrum is obviously improved.
Description
Technical field
The invention belongs to networking and resource allocation techniques field, in particular to a kind of super-intensive network second lines of a couplet based on caching
The method for closing dynamic access and subcarrier distribution.
Background technique
Super-intensive network is the strong candidate technology of 5G, and super-intensive networking technology, can be real by increasing base station deployment density
The tremendous increase of existing channeling efficiency, greatlys improve power system capacity, meets thousand times of capacity increased requirements of 5G.However, more
Intensive network is disposed so that network topology is more complicated, and existing content distribution mechanism is realizing the seas such as picture, audio, video
While measuring information transmission, repeats to transmit there are a large amount of content, frequency spectrum resource etc. is caused and is greatly wasted.For this
Caching technology is introduced super-intensive network by problem, by carrying out content caching in access point or core net, it is possible to reduce redundancy
Backhaul link consumption and network delay is effectively reduced, to improve spectrum utilization efficiency and efficiency utilization rate in data transmission.
Under the dense network based on caching, physical layer document 1: is based in the intensive wireless network that backhaul link is limited
The throughput gain optimization method of caching proposes a kind of novel caching wireless network architecture being limited based on backhaul link, and
And propose physical layer buffering scheme under this architecture to improve throughput of system, but the program only consider backhaul link etc. because
Element, and show that result to node B cache amount of capacity is relevant.
Document 2: the method for joint route and content caching optimization in heterogeneous network is selected and is cached using joint route
The problem of scheme of distribution carries out resource allocation optimization, and the program is lower considered is single, does not account for the resources benefit such as frequency effect, efficiency
With rate problem.
Under the scene of super-intensive, the prior art does not consider that the load balancing of access point, carries out the dynamic choosing of access point
It selects.In addition, not accounting for the optimization of spectrum efficiency, the resources such as sub-carrier are allocated optimization.
Summary of the invention
The present invention, which is directed to, efficiently cannot provide service, frequency spectrum using the storage resource of access point in the prior art for user
It can not achieve maximization with efficiency utilization of resources rate, propose joint dynamic under a kind of super-intensive network based on caching and connect
Enter the method with subcarrier distribution,
Specific step is as follows:
Step 1: multiple users transmit request information to all access points with broadcast mode, cache contents are found；
Solicited message refers to cache contents；Number of users is O；
Step 2: each access point judges whether there is in the caching of active user's request using user K as active user
Hold, if some access point is idle and there are the cache contents, access point feedback 1 enters step three to user K；Otherwise anti-
Feedback 0；Enter step four；
1≤K≤O；
Step 3: all access points for meeting user K send respective attribute parameter to local control, local control will
Best access point distributes to user K；
All access points for meeting user K are m；The attribute parameter of each access point include buffer memory capacity, time delay and
Signal-to-noise ratio etc., total n attribute；
Specific step is as follows:
Step 301 is locally controlled to n attribute parameter according to the solicited message of user K for each candidates
Relative importance between middle every two attribute parameter is compared one by one, obtains decision matrix M:
Relative importance between each attribute parameter of each candidates is compared by local control, is somebody's turn to do
The decision matrix M of candidates:
Wherein aijRepresent the relative importance fiducial value of attribute parameter i and attribute parameter j in access point；
Step 302 is normalized decision matrix M, the decision matrix B after being standardized:
Wherein bijIt represents in access point to fiducial value aijValue after normalization；
Step 303 verifies the consistency of decision matrix B, judges whether decision matrix is effective, if effectively, into
Row step 304, otherwise return step 302；
Consistency ratio CR is defined as follows:
Wherein, CI indicates inconsistency index:λmaxIt is the maximum eigenvalue of decision matrix B, n is to determine
The number of attribute parameter in plan matrix B, RI are known Aver-age Random Consistency Index；
As CR < 0.1, it is believed that decision matrix B has acceptable consistency, otherwise reconfigures decision matrix B.
Step 304 obtains the comprehensive weight vectors ω generated of n attribute parameter in decision matrix B；
ω=(ω1,ω2,...ωj,...,ωn)
ωjFor the weight of j-th of attribute parameter；
Step 305 is directed to m candidates, generates the state matrix S of all properties parameter；
State matrix S is m row n column, and every a line represents n attribute parameter of each access point；
Wherein, smnIndicate the value of corresponding n-th of the attribute parameter of m-th of access point.
Weight vectors ω is multiplied to obtain weighted decision matrix Q with state matrix S by step 306:
Step 307, according to weighted decision matrix Q, determine best access scheme QbestWith worst access scheme Qworst,
Qbest=(ω1·s1best,ω2·s2best,...ωj·sjbest,...,ωn·snbest)
Qworst=(ω1·s1worst,ω2·s2worst,...ωj·sjworst,...,ωn·snworst)
sjbestIndicate optimum value in j-th of attribute parameter of all m access points；sjworstIndicate all m access points
Worst-case value in j-th of attribute parameter；
Step 308 is directed to some access point l, calculates separately candidate access scheme xljWith best access scheme QbestEurope
Family name's distance, and candidate access scheme xljWith worst access scheme QworstEuclidean distance；
Candidate access scheme xljWith best access scheme QbestEuclidean distance be Qlbest, in particular to candidate access scheme
xljEach attribute parameter and attribute optimal value sjbestEuclidean distance, it is as follows:
Candidate access scheme xljWith worst access scheme QworstEuclidean distance be Qlworst, in particular to candidate access side
Case xljEach attribute parameter and attribute worst-case value sjworstEuclidean distance, it is as follows:
Step 309 is directed to some access point l, calculates candidate scheme xljPreference value P between preferred planl。
Formula is as follows:
Wherein PlWhat is represented is preference value of the user K for first of access point.
Step 310 calculates all users respectively for the preference value of each access point, and arranged in sequence, generation user select
Select matrix R；
It is as follows:
Wherein, PmORepresent the preference value that the O user selects m access point.All O users selection that every a line represents
The preference value of each identical access point；The column of user's selection matrix R indicate that each user selects the preference of each diverse access point
Value；
Step 311 is directed to each user, and the corresponding access point of maximum preference value is distributed to the user by local control, and
The user and matched access point are deleted, is sequentially allocated, until all users complete network insertion.
Step 4: user K does not obtain the response of any access point, then user K directly sends to remote server and requests,
Obtain content；Remote server is analyzed according to user request information using popularity, and buffer update is completed；
Step 5: carrying out subcarrier distribution after each user matches with respective access point, making between user and access point
It is communicated.
Step 501, initialization t easet ofasubcarriers and user's set；
T easet ofasubcarriers are N={ n'| n'=1,2 ..., N }, user's set I={ i'| i'=1,2 ..., I }, distribution
To the sub-carrier indices X of user i'i'=φ.
Step 502, according to water-filling algorithm calculate each user corresponding to each subcarrier transmission power and channel capacity.
Transmission power pi',n':
pi',n'Indicate the power to user i' distribution subcarrier n', PtotIndicate maximum transmission power；γi',n'Indicate to
The signal-to-noise ratio of family i' distribution subcarrier n'.
Channel capacity Ci',n':
B represents the overall system bandwidth；
Step 503 is directed to subcarrier n', calculates separately the channel capacity of the subcarrier under each user, and carry out descending
Maximum value is selected in arrangement
Subcarrier n' initial value is 1；
It indicates are as follows:
Subcarrier n' is distributed to maximum channel capacity value by step 504Corresponding user, and by subcarrier n' from son
It is removed in carrier set N, return step 503 continues the next subcarrier of ordinal selection, distributes until by all subcarriers.
Step 505 fills the water the subcarrier after distribution, according to the transmission power p of subcarrieri',n'Computing system frequency spectrum
Utilization efficiency；
It is as follows that system spectrum utilization efficiency maximizes objective function:
The condition to be met are as follows:
{ai',n'Indicate subcarrier distribution set, being worth is 0 or 1,1 to indicate subcarrier n' distributing to user i', and 0 indicates son
Carrier wave n ' it is not allocated to user i'；
First constraint condition indicates the restrictive condition of general power, wherein PtotIndicate maximum transmission power；Second constraint
Condition indicates that the power of user i' distribution subcarrier n' is more than or equal to 0；The last one constraint condition indicates each subcarrier only
It can distribute primary.
The present invention has the advantages that
1) a kind of method for, combining dynamic access and subcarrier distribution under super-intensive network based on caching, according to emulation
As a result as can be seen that this method effectively improves spectrum utilization efficiency, this result demonstrates the mechanism under intensive scene
Meet the feasibility and applicability of multiple business demand.
2) a kind of method for, combining dynamic access and subcarrier distribution under super-intensive network based on caching, can integrate
Multiple factors complete access selection, realize the promotion of resources management efficiency.
3) a kind of method for, combining dynamic access and subcarrier distribution under super-intensive network based on caching, may be implemented
The dynamic allocation of subcarrier, so that spectrum utilization efficiency is obviously improved.
Detailed description of the invention
Fig. 1 is system model schematic diagram of the invention；
Fig. 2 is system model block architecture diagram of the invention；
Fig. 3 is the method flow for combining dynamic access and subcarrier distribution under the super-intensive network the present invention is based on caching
Figure；
Fig. 3 a is the flow chart that best access point is distributed to user by local control of the invention；
Fig. 3 b is the flow chart that each user of the present invention and respective access point carry out subcarrier distribution；
Fig. 4 is the weight factor emulation schematic diagram of lower 4 users of multiple attribute decision making (MADM) algorithm of the present invention；
Fig. 5 is inventive network access selection ordering of optimization preference emulation schematic diagram；
Fig. 6 is spectrum utilization efficiency of the present invention and user's number relational graph；
Fig. 7 is present system spectrum efficiency and Between Signal To Noise Ratio figure；
Specific embodiment
Below in conjunction with attached drawing, the present invention is described in further detail.
As shown in Figure 1, connecting multiple access points, access point as core net using remote server in super-intensive network
A part of content is cached according to content popularit analysis, each user is separately connected an access point, as shown in Fig. 2, access point
Send respective attribute parameter to local control, local control is according to user request information and access point the attribute ginseng received
Amount is user's dynamic select access point using multiple attribute decision making (MADM) matrix, in selection course, needs to comprehensively consider each of access point
Attribute, such as buffer memory capacity, time delay, signal-to-noise ratio carry out more attribute access selections.After completing dynamic access process, in greedy algorithm
On the basis of complete subcarrier distribution, AP and subcarrier are carried out based on power, spectral bandwidth and cache size limitation is sent
It chooses, using maximum spectral efficiency as target, completes subcarrier distribution.
The present invention finds cache contents using broadcast mode first according to user request information, and feedback result is indicated with S,
If some access point is idle and there are the cache contents, it is denoted as S=1, is otherwise denoted as S=0.Access point believes the attribute of oneself
Such as buffer memory capacity, time delay, signal-to-noise ratio are ceased, sends local control to.If user does not obtain the response of any access point, directly
It connects to send to server end and request, obtain content, server end is analyzed according to user request information using popularity, is completed slow
Deposit update；If user obtains access point response, requested according to the content of user and type of service, carries out hierarchy analysis, selection
Optimal access point provides service by the access point that local control notice is selected, completes access selection, update access choice set
Close Q.
Each user is calculated by water-filling algorithm under the premise of known channel state information according to access selection result
The channel capacity of corresponding subcarrier, it is excellent to the subcarrier with optimum channel capacity on the basis of calculated channel capacity
It first distributes, is then iterated operation, be finally completed subcarrier distribution.
A method of combining dynamic access and subcarrier distribution under the super-intensive network based on caching, as shown in figure 3, tool
Steps are as follows for body:
Step 1: multiple users transmit request information to all access points with broadcast mode simultaneously, cache contents are found；
Solicited message includes: cache contents；Number of users is O；
Step 2:, according to the solicited message of user K, each access point judges whether there is using user K as active user
The cache contents of active user's request, if some access point is idle and there are the cache contents, access point feedback 1 is to user
K enters step three；Otherwise 0 is fed back；Enter step four；
1≤K≤O；
Step 3: all access points for meeting user K send respective attribute parameter to local control, local control will
Best access point distributes to user K；
All access points for meeting user K are m；The attribute parameter of each access point include buffer memory capacity, time delay and
Signal-to-noise ratio etc., total n attribute；
Local control carries out hierarchy analysis, selects optimal access point according to the request content and type of service of user,
It notifies the access point selected to provide service, completes access selection, then update access selection set Q.
As shown in Figure 3a, the specific steps are as follows:
Step 301 is locally controlled according to the solicited message of user K for n attribute parameter of each candidates
Relative importance between every two attribute parameter is compared one by one, obtains decision matrix M:
Cache contents are different, different for the attribute specification of access point selection, and construction adjudicates matrix to indicate each candidate
Important sexual intercourse between each attribute parameter of access point, by comparing the importance between different attribute two-by-two, to determine certain
The significance level of a attribute usually carries out assignment to importance degree by 1~9 proportion quotiety, gives 1~9 scale in table 1
Meaning:
Table 1
Relative importance between each attribute parameter of each candidates is compared by local control, is somebody's turn to do
The decision matrix M of candidates:
Wherein aijRepresent the relative importance fiducial value of parameter i and parameter j in access point；
Step 302 is normalized decision matrix M, the decision matrix B after being standardized:
Wherein bijIt represents in access point to fiducial value aijValue after normalization；
Step 303 carries out consistency desired result to the decision matrix B after standardization；Judge whether decision matrix is effective, if
Effectively, step 304 is carried out, otherwise return step 302；
Before calculating each attribute weight, need to carry out consistency desired result to decision matrix.Because if decision matrix mistake
In deviateing consistency, the weight vectors being calculated will not have credibility, it is therefore necessary to carry out to the consistency of decision matrix B
Verification.It is defined as follows:
bik×bkj=bij, i, j, k=1,2 ..., n (3)
bikFor the relative importance fiducial value of parameter i and parameter k in access point after standardization；
If equation is set up, then it represents that judgment matrix is with uniformity.Indicate that policymaker is carrying out the two of attribute
Two when comparing, and thinking is with uniformity.But since the thinking of people has certain subjectivity, it is difficult to keep absolute consistent
Property, so introducing the concepts such as inconsistency index CI, consistency ratio CR to measure the consistency of matrix:
Wherein, λmaxIt is the maximum eigenvalue of decision matrix B, n is the number of attribute parameter in decision matrix B, and RI is known
Aver-age Random Consistency Index, it is as shown in the table:
Table 2
| n | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
| RI | 0 | 0 | 0.58 | 0.90 | 1.12 | 1.24 | 1.32 | 1.41 | 1.45 |
As CR < 0.1, it is believed that decision matrix B has acceptable consistency, otherwise reconfigures decision matrix B.
Step 304 calculates the comprehensive weight vectors ω generated of n attribute parameter in decision matrix B；
ω=(ω1,ω2,...ωj,...,ωn)
ωjFor the weight of j-th of attribute parameter；The present invention is selected using the feature vector after normalization as network insertion
Weight vectors.
The state matrix S of all properties parameter of m step 305, generation candidate's access node；
State matrix S is m row n column, and every a line represents n attribute parameter of each access point；
Wherein, smnIndicate the value of corresponding n-th of the attribute parameter of m-th of access point.
Weight vectors ω is multiplied to obtain weighted decision matrix Q with state matrix S by step 306:
Step 307, according to weighted decision matrix Q, determine best access scheme QbestWith worst access scheme Qworst。
Best access scheme is the best situation for selecting each parameter, such as buffer memory capacity, just
Maximum value is selected, for power consumption, just selects minimum value.And worst access scheme is then in contrast, calculation is as follows:
Qbest=(ω1·s1best,ω2·s2best,...ωj·sjbest,...,ωn·snbest) (8)
Qworst=(ω1·s1worst,ω2·s2worst,...ωj·sjworst,...,ωn·snworst) (9)
sjbestIndicate optimum value in j-th of attribute parameter of all m access points；sjworstIndicate all m access points
Worst-case value in j-th of attribute parameter；
Step 308 is directed to some access point l, calculates separately candidate access scheme xljWith best access scheme QbestEurope
Family name's distance, and candidate access scheme xljWith worst access scheme QworstEuclidean distance；
Candidate access scheme xljWith best access scheme QbestEuclidean distance be Qlbest, in particular to candidate access scheme
xljEach attribute parameter and attribute optimal value sjbestEuclidean distance, it is as follows:
Candidate access scheme xljWith worst access scheme QworstEuclidean distance be Qlworst, in particular to candidate access side
Case xljEach attribute parameter and attribute worst-case value sjworstEuclidean distance, it is as follows:
Step 309 is directed to some access point l, calculates candidate scheme xljPreference value P between preferred planl。
By calculating preference value, the ratio of distance between candidate network and optimal network and worst network is obtained.It calculates
Formula is as follows:
Wherein PlWhat is represented is preference value of the user K for first of access point.
Maximum preference value of the user K to each access point: Pk=(P1k,P2k,...Plk,...Pmk)；
Step 310 calculates all users respectively for the preference value of each access point, and arranged in sequence, generation user select
Select matrix R；It is as follows:
Wherein, PmORepresent the preference value that the O user selects m access point.The O user selection that every a line represents is each
The preference value of access point；
Step 311 is directed to each user, and the corresponding access point of maximum preference value is distributed to the user by local control, and
The user and matched access point are deleted, is sequentially allocated, until all users complete network insertion.
Local control selects maximum preference value, by the corresponding access point of the preference value according to the column of user's selection matrix R
Corresponding user is distributed to, matched user and access point are deleted after distribution, continues to be sequentially allocated, until all users complete net
Network access.
The each column expression of user's selection matrix R, preference value of each user to m access point；
The simulation parameter of 3 access point of table
| Signal-to-noise ratio (/dBm) | Time delay (/ms) | Covering radius (km) | Buffer memory capacity (TB) | Power consumption (W) | |
| Access point 1 | 70 | 25 | 50 | 100 | 1/100 |
| Access point 2 | 60 | 50 | 40 | 200 | 1/100 |
| Access point 3 | 73 | 80 | 70 | 400 | 1/50 |
| Access point 4 | 50 | 30 | 30 | 200 | 1/70 |
According to content popularit analytic approach, content hit rate is related with Zipf index.In this emulation, setting Zipf ginseng
Number is 0.8, cache hit rate 0.7, without loss of generality, set to obtain user's number that access point responds as 4, remaining 2 use
Family needs to obtain content from back-end server.This 4 users are in different regions, there is different business.During user 1 belongs to
Heart user requests web browsing business；User 2 belongs to central user, requests streaming media service；User 3 belongs to edge customer, asks
Seek streaming media service；User 4 belongs to edge customer, requests web browsing business.According to this section propose multiple attribute decision making (MADM) algorithm,
Simulation result is as shown in figure 4, abscissa illustrates 5 parameters: signal-to-noise ratio, time delay, covering radius, buffer memory capacity, power consumption.It is vertical to sit
Mark indicates the weighted value of corresponding each parameter, it can be seen that requests for different users, weight factor has very big difference.
For streaming media service, mainly one-way transmission does not need two-way real time communication, of less demanding in time delay, still
File is generally large, needs higher buffer memory capacity, and user 2 and user 3 are request streaming media service, so time delay factor is weighed
Weight is lower, and buffer memory capacity Factor Weight is higher.For web browsing business, buffer memory capacity is also low, but needs higher clothes
Be engaged in quality and lower time delay, so more demanding to signal-to-noise ratio, if user 1 and user 4 request web browsing business, time delay because
The weight of son is higher, and the weight of the buffer memory capacity factor is lower.Central user is lower to covering radius requirement, so user 1 and use
The weight of the covering radius factor at family 2 is lower；And the weight of the 4 covering radius factor of user 3 and user is higher.
According to the attribute information of weight factor and each access point, each user has been obtained for the selection preference of access point
Value, simulation result are as shown in Figure 5:
As seen from Figure 5, for the same access point, parameter information is constant, but different user meter
Obtained preference value is but different, and illustrates that method proposed in this paper has fully considered the demand information of user, for same
For user, network demand is also constant, but the preference value being calculated under different access points is also different,
Illustrate that algorithm has also fully considered the parameter information of access point.Thus, network insertion selection algorithm proposed in this paper combines
User demand and access point performance are effective bases for carrying out subcarrier distribution.The network insertion selection result of end user is such as
Shown in table 4:
4 user of table accesses selection result
| User | Access point back-end server |
| User 1 | Access point 3 |
| User 2 | Access point 2 |
| User 3 | Access point 1 |
| User 4 | Access point 4 |
| User 5 | Back-end server |
| User 6 | Back-end server |
Step 4: user K does not obtain the response of any access point, then user K directly sends to remote server and requests,
Obtain content；Remote server is analyzed according to user request information using popularity, and buffer update is completed；
Step 5: carrying out subcarrier distribution after each user matches with respective access point, making between user and access point
It is communicated.
Each user is calculated by water-filling algorithm under the premise of known channel state information according to access selection result
The channel capacity of corresponding subcarrier, it is excellent to the subcarrier with optimum channel capacity on the basis of calculated channel capacity
It first distributes, is then iterated operation, be finally completed subcarrier distribution.
It completes user to access after selection, on the basis of greedy algorithm, proposes one kind and maximized based on power system capacity
The preferential allocation algorithm of subcarrier.Under the premise of known channel state information, the corresponding son of each user is calculated by water-filling algorithm
The channel capacity of carrier wave preferentially divides the subcarrier with optimum channel capacity then on the basis of the channel capacity of calculating
Match.
As shown in Figure 3b, the specific steps are as follows:
Step 501, initialization t easet ofasubcarriers and user's set；
T easet ofasubcarriers are N={ n'| n'=1,2 ..., N }, user's set I={ i'| i'=1,2 ..., I }, distribution
To the sub-carrier indices X of user i'i'=φ.
Step 502, according to water-filling algorithm calculate each user corresponding to each subcarrier transmission power and channel capacity.
Transmission power pi',n':
pi',n'Indicate the power to user i' distribution subcarrier n', PtotIndicate maximum transmission power；γi',n'Indicate to
The signal-to-noise ratio of family i' distribution subcarrier n'.
Channel capacity CI', n':
B represents the bandwidth of total system framework model；
Step 503 is directed to subcarrier n', calculates separately the channel capacity of the subcarrier under each user, and carry out descending
Maximum value is selected in arrangement
Subcarrier n' initial value is 1；
It indicates are as follows:
As sub-carrier channels capacity arranges in descending order are as follows:
Subcarrier n' is distributed to maximum channel capacity value by step 504Corresponding user, and by subcarrier n' from son
It is removed in carrier set N, return step 503 continues the next subcarrier of ordinal selection, distributes until by all subcarriers.
Step 505 fills the water the subcarrier after distribution, according to the transmission power p of subcarrieri',n'Computing system frequency spectrum
Utilization efficiency；
It is as follows that system spectrum utilization efficiency maximizes objective function:
The condition to be met are as follows:
B represents the bandwidth of total system framework model, { ai',n'Indicate subcarrier distribution set, being worth is that 0 or 1,1 expression will be sub
Carrier wave n ' user i' is distributed to, 0 expression subcarrier n' is not allocated to user i'；
First constraint condition indicates the restrictive condition of general power, wherein PtotIndicate maximum transmission power；Second constraint
Condition indicates that the power of user i' distribution subcarrier n' is more than or equal to 0；The last one constraint condition indicates each subcarrier only
It can distribute primary.
Emulation uses rayleigh fading channel model in the present embodiment, and sub-carrier number is 64, and number of users is 16, general power
PtotFor 1W, noise power spectral density N0It is 1MHz for 10e-8W/Hz, total bandwidth B.And with minimum capacity maximum method (MAX-
MIN it) is compared with unfair portion method (FPS) both classical sub-carrier wave distribution methods, simulation result is as shown in Figure 6 and Figure 7:
As seen from Figure 6, in the identical situation of user's number, the obtained spectrum utilization of algorithm proposed by the present invention
Efficiency is much higher than proportional fair algorithm and minimum capacity maximizes method, because algorithm of the invention integrally considers power system capacity,
Subcarrier and power are distributed according to channel state information, to substantially increase the resource utilization of system.Ratio is public simultaneously
The availability of frequency spectrum of flat algorithm is slightly above minimum capacity and maximizes method, because while all considering fairness, but ratio justice is calculated
Method considers power system capacity maximization simultaneously, and minimum capacity maximization method only considered the distribution of subcarrier between users,
The self-adjusted block of power between subcarriers is not accounted for, so that spectrum utilization efficiency rate is lower.
Fig. 7 is the curve that spectrum utilization efficiency changes with signal-to-noise ratio, it can be seen that in the case where identical signal-to-noise ratio, herein
The obtained spectrum utilization efficiency of the algorithm of proposition is much higher than proportional fair algorithm and minimum capacity maximizes method, this is because
This algorithm distributes subcarrier by being constantly iterated water filling calculating channel capacity, can preferably avoid having poor letter
The subcarrier of road quality distributes to user, so as to improve system spectrum utilization efficiency.
Claims (3)
1. a kind of method for combining dynamic access and subcarrier distribution under super-intensive network based on caching, which is characterized in that packet
Include following steps:
Step 1: multiple users transmit request information to all access points with broadcast mode, cache contents are found；
Number of users is O；
Step 2: each access point judges whether there is the cache contents of active user's request using user K as active user,
If some access point is idle and there are the cache contents, access point feedback 1 enters step three to user K；Otherwise 0 is fed back；
Enter step four；
Step 3: all access points for meeting user K send respective attribute parameter to local control, local control will be best
Access point distributes to user K；
All access points for meeting user K are m；
Step 4: user K does not obtain the response of any access point, then user K directly sends to remote server and requests, and obtains
Content；Remote server is analyzed according to user request information using popularity, and buffer update is completed；
Step 5: carrying out subcarrier distribution after each user matches with respective access point, making to carry out between user and access point
Communication；
It specifically includes:
Step 501, initialization t easet ofasubcarriers and user's set；
T easet ofasubcarriers are N={ n'| n'=1,2 ..., N }, and user's set I={ i'| i'=1,2 ..., I } distributes to use
The sub-carrier indices X of family i'i'=φ；
Step 502, according to water-filling algorithm calculate each user corresponding to each subcarrier transmission power and channel capacity；
Transmission power pi',n':
pi',n'Indicate the power to user i' distribution subcarrier n', PtotIndicate maximum transmission power；γi',n'It indicates to user i'
Distribute the signal-to-noise ratio of subcarrier n'；
Channel capacity Ci',n':
B represents overall system bandwidth；
Step 503 is directed to subcarrier n', calculates separately the channel capacity of the subcarrier under each user, and carry out descending arrangement
Select maximum value
Subcarrier n' initial value is 1；
Subcarrier n' is distributed to maximum channel capacity value by step 504Corresponding user, and by subcarrier n' from subcarrier
It is removed in set N, return step 503 continues the next subcarrier of ordinal selection, distributes until by all subcarriers；
Step 505 fills the water the subcarrier after distribution, according to the transmission power p of subcarrieri',n'Computing system spectrum utilization
Efficiency；
It is as follows that system spectrum utilization efficiency maximizes objective function:
The condition to be met are as follows:
ai',n'Indicate subcarrier distribution set, being worth is 0 or 1,1 to indicate subcarrier n' distributing to user i', and 0 indicates subcarrier n'
It is not allocated to user i'；
First constraint condition indicates the restrictive condition of general power, wherein PtotIndicate maximum transmission power；Second constraint condition
Indicate that the power of user i' distribution subcarrier n' is more than or equal to 0；The last one constraint condition indicates that each subcarrier can only divide
With primary.
2. combining the side of dynamic access and subcarrier distribution under a kind of super-intensive network based on caching as described in claim 1
Method, which is characterized in that each access point includes n attribute parameter.
3. combining the side of dynamic access and subcarrier distribution under a kind of super-intensive network based on caching as described in claim 1
Method, which is characterized in that the step 3 specifically:
Step 301, according to the solicited message of user K, for each candidates, local control in n attribute parameter often
Relative importance between two attribute parameters is compared one by one, obtains decision matrix M:
aijRepresent the relative importance fiducial value of attribute parameter i and attribute parameter j in access point；aij> 0；aii=1；
Step 302 is normalized decision matrix M, the decision matrix B after being standardized:
Wherein bijIt represents in access point to fiducial value aijValue after normalization；
Step 303 verifies the consistency of decision matrix B, judges whether decision matrix is effective, if effectively, walked
Rapid 304, otherwise return step 302；
Consistency ratio CR is defined as follows:
Wherein, CI indicates inconsistency index:λmaxIt is the maximum eigenvalue of decision matrix B, n is decision matrix
The number of attribute parameter in B, RI are known Aver-age Random Consistency Index；
As CR < 0.1, it is believed that decision matrix B has acceptable consistency, otherwise reconfigures decision matrix B；
Step 304 obtains the comprehensive weight vectors ω generated of n attribute parameter in decision matrix B；
ω=(ω1,ω2,...ωj,...,ωn)
ωjFor the weight of j-th of attribute parameter；
Step 305 is directed to m candidates, generates the state matrix S of all properties parameter；
State matrix S is m row n column, and every a line represents n attribute parameter of each access point；
Wherein, smnIndicate the value of corresponding n-th of the attribute parameter of m-th of access point；
Weight vectors ω is multiplied to obtain weighted decision matrix Q with state matrix S by step 306:
Step 307, according to weighted decision matrix Q, determine best access scheme QbestWith worst access scheme Qworst,
Qbest=(ω1·s1best,ω2·s2best,...ωj·sjbest,...,ωn·snbest)
Qworst=(ω1·s1worst,ω2·s2worst,...ωj·sjworst,...,ωn·snworst)
sjbestIndicate optimum value in j-th of attribute parameter of all m access points；sjworstIndicate the jth of all m access points
Worst-case value in a attribute parameter；
Step 308 is directed to some access point l, calculates separately candidate access scheme xljWith best access scheme QbestEuclidean away from
From, and candidate access scheme xljWith worst access scheme QworstEuclidean distance；
Candidate access scheme xljWith best access scheme QbestEuclidean distance be Qlbest, in particular to candidate access scheme xlj's
Each attribute parameter and attribute optimal value sjbestEuclidean distance, it is as follows:
Candidate access scheme xljWith worst access scheme QworstEuclidean distance be Qlworst, in particular to candidate access scheme xlj
Each attribute parameter and attribute worst-case value sjworstEuclidean distance, it is as follows:
Step 309 is directed to some access point l, calculates candidate scheme xljPreference value P between preferred planl；
Formula is as follows:
Wherein PlWhat is represented is preference value of the user K for first of access point；
Step 310 calculates all users respectively for the preference value of each access point, and arranged in sequence, generation user select square
Battle array R；
Wherein, PmORepresent the preference value that the O user selects m access point；All O users selection that every a line represents is each
The preference value of identical access point；The column of user's selection matrix R indicate that each user selects the preference value of each diverse access point；
Step 311 is directed to each user, and the corresponding access point of maximum preference value is distributed to the user by local control, and is deleted
The user and matched access point, are sequentially allocated, until all users complete network insertion.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510994459.2A CN105611574B (en) | 2015-12-25 | 2015-12-25 | A method of combining dynamic access and subcarrier distribution under the super-intensive network based on caching |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510994459.2A CN105611574B (en) | 2015-12-25 | 2015-12-25 | A method of combining dynamic access and subcarrier distribution under the super-intensive network based on caching |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105611574A CN105611574A (en) | 2016-05-25 |
| CN105611574B true CN105611574B (en) | 2019-02-01 |
Family
ID=55991041
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510994459.2A Active CN105611574B (en) | 2015-12-25 | 2015-12-25 | A method of combining dynamic access and subcarrier distribution under the super-intensive network based on caching |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105611574B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106900019B (en) * | 2016-06-07 | 2019-04-30 | 中国移动通信有限公司研究院 | A kind of multi-access point selection method and device |
| CN105898807B (en) * | 2016-06-08 | 2019-04-23 | 北京邮电大学 | A kind of selection of joint access point and resource allocation under super-intensive network from cure method |
| CN107567053B (en) * | 2016-06-30 | 2020-06-26 | 华为技术有限公司 | Data transmission method, equipment and system |
| CN106304307B (en) * | 2016-08-16 | 2019-11-12 | 上海交通大学 | A kind of resource allocation methods under heterogeneous network converged |
| CN106912079B (en) * | 2017-02-20 | 2020-04-03 | 北京邮电大学 | Combined user access selection and resource allocation method in cache heterogeneous network |
| CN108738103B (en) * | 2017-04-13 | 2021-03-02 | 电信科学技术研究院 | Resource allocation method and device |
| CN107172667B (en) * | 2017-05-19 | 2019-10-15 | 北京邮电大学 | Management method, system, electronic equipment and the mobile terminal of access point cluster |
| CN107466099B (en) * | 2017-07-31 | 2020-01-10 | 北京邮电大学 | Interference management self-optimization method based on non-orthogonal multiple access |
| CN107613565B (en) * | 2017-08-30 | 2020-09-04 | 东莞理工学院 | Wireless resource management method in full-duplex ultra-dense network |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103096335A (en) * | 2012-12-26 | 2013-05-08 | 陈宏滨 | Optimization method of spectrum efficiency and energy efficiency of wireless communication system |
| CN103281786A (en) * | 2013-06-04 | 2013-09-04 | 北京邮电大学 | Method for optimizing resources of family base station double-layer network based on energy efficiency |
| CN104066096A (en) * | 2014-07-03 | 2014-09-24 | 东南大学 | Super-dense heterogeneous-network optimal power coordination method based on improved particle swarm |
| CN105188089A (en) * | 2015-08-05 | 2015-12-23 | 东南大学 | Load balancing method based on integral optimization of user connection and interference management in ultra-dense heterogeneous network |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070161150A1 (en) * | 2005-12-28 | 2007-07-12 | Intel Corporation | Forming ultra dense 3-D interconnect structures |
-
2015
- 2015-12-25 CN CN201510994459.2A patent/CN105611574B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103096335A (en) * | 2012-12-26 | 2013-05-08 | 陈宏滨 | Optimization method of spectrum efficiency and energy efficiency of wireless communication system |
| CN103281786A (en) * | 2013-06-04 | 2013-09-04 | 北京邮电大学 | Method for optimizing resources of family base station double-layer network based on energy efficiency |
| CN104066096A (en) * | 2014-07-03 | 2014-09-24 | 东南大学 | Super-dense heterogeneous-network optimal power coordination method based on improved particle swarm |
| CN105188089A (en) * | 2015-08-05 | 2015-12-23 | 东南大学 | Load balancing method based on integral optimization of user connection and interference management in ultra-dense heterogeneous network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105611574A (en) | 2016-05-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105611574B (en) | A method of combining dynamic access and subcarrier distribution under the super-intensive network based on caching | |
| Liu et al. | Economically optimal MS association for multimedia content delivery in cache-enabled heterogeneous cloud radio access networks | |
| CN110417847B (en) | Method and device for user access and content caching in unmanned aerial vehicle communication network | |
| CN107566194B (en) | Method for realizing cross-domain virtual network mapping | |
| CN105960024B (en) | User's discovery and resource allocation methods in a kind of D2D communication based on social activity perception | |
| CN109413724A (en) | A kind of task unloading and Resource Allocation Formula based on MEC | |
| CN108834080B (en) | Distributed cache and user association method based on multicast technology in heterogeneous network | |
| CN108156596B (en) | Method for supporting D2D-cellular heterogeneous network combined user association and content caching | |
| CN109905918A (en) | A kind of NOMA honeycomb car networking dynamic resource scheduling method based on efficiency | |
| CN112020103A (en) | Content cache deployment method in mobile edge cloud | |
| CN111182570A (en) | User association and edge computing unloading method for improving utility of operator | |
| CN103687023B (en) | Optimization Radio Resource method based on time delay differentiated service and proportionality rate constraint | |
| CN107302801B (en) | QoE-oriented double-layer matching game method in 5G mixed scene | |
| CN109194763A (en) | Caching method based on small base station self-organizing cooperative in a kind of super-intensive network | |
| He et al. | Resource allocation based on graph neural networks in vehicular communications | |
| CN108900355A (en) | A kind of multistage edge network resource allocation methods in star ground | |
| CN106792995B (en) | User access method for guaranteeing low-delay content transmission in 5G network | |
| Liu et al. | MEC-assisted flexible transcoding strategy for adaptive bitrate video streaming in small cell networks | |
| Zhang et al. | DMRA: A decentralized resource allocation scheme for multi-SP mobile edge computing | |
| CN108900617A (en) | A kind of three layers of cooperative caching method of mist wireless access network | |
| CN108271172A (en) | A kind of honeycomb D2D communication joint clusters and content disposition method | |
| CN104540203B (en) | Wireless body area network efficiency optimization method based on independent sets | |
| CN108924934A (en) | Heterogeneous network interference management method based on multi dimensional resource distribution | |
| Le et al. | Joint cache allocation with incentive and user association in cloud radio access networks using hierarchical game | |
| TW201228432A (en) | Joint subcarrier usage ratio and power allocation method, system using the same, base station and controller using the same |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |