CN107343303B - Based on the routing optimization method of Duality Decomposition in wireless Mesh netword - Google Patents
Based on the routing optimization method of Duality Decomposition in wireless Mesh netword Download PDFInfo
- Publication number
- CN107343303B CN107343303B CN201710557122.4A CN201710557122A CN107343303B CN 107343303 B CN107343303 B CN 107343303B CN 201710557122 A CN201710557122 A CN 201710557122A CN 107343303 B CN107343303 B CN 107343303B
- Authority
- CN
- China
- Prior art keywords
- channel
- probability
- user
- optimization
- present
- 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
- H04W40/00—Communication routing or communication path finding
- H04W40/24—Connectivity information management, e.g. connectivity discovery or connectivity update
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/14—Routing performance; Theoretical aspects
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The present invention provides the routing optimization method based on Duality Decomposition in a kind of wireless Mesh netword, include: the available capacity value of each link determined according to the constraint condition of network layer in wireless Mesh netword route optimization process and the constraint condition of physical layer resources distribution, the channel use information of each link, constructs Network Optimization Model；Optimized model is described using Lagrangian, network optimization subproblem is obtained and physical layer optimizes subproblem；The dual variable of Lagrangian is initialized, and convex optimization method processing network optimization subproblem is respectively adopted, obtains the first processing result；Subproblem is optimized using genetic algorithm processing physical layer, obtains second processing result；Optimal solution is obtained when the first processing result and second processing result restrain.The above method can effectively reduce the solution complexity of Optimized model, and the result solved can satisfy the requirement for requiring optimal solution.
Description
Technical field
A kind of cognitive radio networks of the present invention, and in particular to the road based on Duality Decomposition in wireless Mesh netword
By optimization method.
Background technique
Wireless Mesh netword is a kind of new distribution type radio bandwidth access network, has merged WLAN and Ad Hoc net
The advantage of network has the characteristics that self-organizing, selfreparing, multi-hop cascade, high-speed, wide coverage, improve the network coverage,
Increase all various aspects such as network capacity, reduction up-front investment and all show very big advantage, is expected to become " last one kilometer " broadband
The ideal solution of access problem is just causing the extensive concern of industry.
With the development of wireless technology, number of users is increasing, and the demand to service quality is constantly promoted, limited frequency
Spectrum resource has become the serious hindrance for obtaining high performance data services, and traditional fixed frequency spectrum allocation model undoubtedly asks this
Topic makes the matter worse.In recent years, the generation and development of cognitive radio technology (Cognitive Radio, CR) provide for people
New resolving ideas.CR enables to cognitive user to wait for an opportunity insertion authority frequency spectrum, is expected to solve by dynamically distributing idle frequency spectrum
Contradiction between growing wireless communication needs and limited frequency spectrum resource, cognitive radio Mesh network come into being.
Routing optimality is that cognitive radio Mesh network needs issues that need special attention.Since routing algorithm directly determines respectively
The flow demand of link.And in cognitive radio Mesh network, link capacity in addition to used power control, channel distribution machine
It makes closely related, is also influenced by channel detection method.However, there is no consider Channel Detection probability in calculation of capacity at present
Influence to link capacity.In view of the Channel Detection based on energy has become being widely used in present cognitive radio technology
How idle frequency spectrum detecting method is based on energy detection mechanism, effectively assessment link capacity, it has also become cognitive radio Mesh network
The major issue that must be taken into consideration in routing optimality.
Summary of the invention
To solve the problems of the prior art, the present invention provides the routing based on Duality Decomposition in a kind of wireless Mesh netword
Optimization method.
In a first aspect, the present invention provides the routing optimization method based on Duality Decomposition in a kind of wireless Mesh netword, comprising:
Step 1: being distributed according to the constraint condition of network layer in wireless Mesh netword route optimization process and physical layer resources
Constraint condition, determine based on the channel use information of link each in wireless Mesh netword the available capacity value of each link, structure
Build the Optimized model of wireless Mesh netword；
Step 2: the Optimized model being described using Lagrangian, and the optimization problem of Optimized model is divided
Solution obtains network optimization subproblem and physical layer optimization subproblem；
Step 3: initializing the dual variable of Lagrangian；
Step 4: based on the dual variable after initialization, the network optimization subproblem being handled using convex optimization method, is obtained
Obtain the first processing result；And based on the dual variable after initialization, physical layer optimization is handled using genetic algorithm and is asked
Topic obtains second processing result；
Step 5: judging whether the first processing result and second processing result restrain, if convergence, by the first processing result
It is exported with second processing result as optimal solution；
Step 6: if the first processing result and second processing result in step 5 do not restrain, according to the first processing result
With second processing as a result, being updated to dual variable, and re-execute the steps 4 and step 5, with based on update to mutation
Amount obtains the first processing result using convex optimization method；And the dual variable based on update, second is obtained using genetic algorithm
Processing result simultaneously judges whether convergent process.
Optionally, before the Optimized model for constructing wireless Mesh netword, the method also includes:
The effective of each link is determined based on cognitive user, the corresponding channel use information of primary user in wireless Mesh netword
Capability value.
Optionally, each chain is determined based on cognitive user, the corresponding channel use information of primary user in wireless Mesh netword
The available capacity value on road, comprising:
It can be correctly detecting and work as when it is idle for perceiving present channel for each channel in wireless Mesh netword
Preceding channel is idle probability as the first probability；
When perceiving present channel and being occupied by primary user, can error detection to the probability that present channel is the free time as the
Two probability；
The primary user that will acquire occupies the probability of present channel as third probability, and the primary user that will acquire releases
The probability of present channel is put as the 4th probability；
Obtain time user corresponding first when primary user occupies present channel, while secondary user uses current channel communications
SINR；
It obtains and works as the vacant present channel of primary user, and secondary user uses when current channel communications users corresponding second
SINR；
According to first probability, the second probability, third probability, the 4th probability and the first SINR, the 2nd SINR, chain is obtained
Road uses the available capacity value of present channel.
Optionally, when it is idle for perceiving present channel, can be correctly detecting present channel is idle probability conduct
First probabilityInclude:
Alternatively, if present channel is Gaussian channel,
Where it is assumed that the accumulative acquisition energy of channel m is Ym, smThe state for indicating channel m, works as smWhen=1, channel m is indicated
For the free time, otherwise sm=0, indicate that channel m is occupied；
F (γ) is the distribution probability for receiving signal, For mean receiving power, τ is decision threshold, and u is
Time-bandwidth product；
And/or
When perceiving present channel and being occupied by primary user, can error detection to the probability that present channel is the free time as the
Two probabilityInclude:
Alternatively, if present channel is Gaussian channel,
Wherein, Pf,AIt is idle probability for the error detection under Gaussian channel to present channel.
Optionally, the primary user that will acquire occupies the probability of present channel as third probability, and will acquire
Primary user discharges the probability of present channel as the 4th probability, comprising:
Defining third probability isThen the 4th probability is
Optionally, time user couple when primary user occupies present channel, while secondary user uses current channel communications is obtained
The first SINR answered, comprising:
Wherein, the power of communications of the primary user of busy channel m is Pm, when secondary user i and time user j is carried out using channel m
When communication,For the first SINR,Interference for other secondary users using channel m to secondary user j, N0For white noise
Power, HijFor the path loss of secondary user i and time user j；
And/or
It obtains and works as the vacant present channel of primary user, and secondary user uses when current channel communications users corresponding second
SINR, comprising:
For when primary user's free time, and secondary user's accurate judgement channel idle, secondary user i and time user j at this time
When being communicated using channel m, the 2nd SINR of acquisition.
Optionally, according to first probability, the second probability, third probability, the 4th probability and the first SINR, second
SINR obtains link eijUse the available capacity value of present channel, comprising:
Wherein, UijThe whole available capacity that present channel is used between time user i and secondary user j is indicated, to use when secondary
Available capacity when family i is communicated with time user j using channel m,
Variable is distributed for link circuit resource, ifIndicate that time user i is communicated on channel m with time user j, otherwise,
Optionally, according to the constraint item of the constraint condition and physical layer resources distribution that construct network layer in wireless Mesh netword
The available capacity value of part, the channel determined based on channel use information each in wireless Mesh netword constructs Wireless Mesh network
The Optimized model of network, comprising:
Assuming that in wireless Mesh netword include N number of node and Q route requests, the objective function of Optimized model are as follows: max:
∑q∈Qlog(1+λq)
s.t.
Wherein, < sq,dq,rq> it is input parameter,λ is to determine physical layer resources distribution and Route Selection
Optimized variable, objective function and UijIt include nonlinear restriction；
sq、dq、rqSource node, the destination node, flow demand of route requests q ∈ [1,2 ..., Q] are respectively indicated,
Indicate the flow that route requests q is transmitted from node i → j or j → i, i, j ∈ V, i ≠ j；
For any user i, j (i, j ∈ V), 1 letter can only be distributed between time user i and secondary user j every time by defining
Road, that is,
Wherein,Indicate time channel distribution situation of user i, ifThenI ≠ j, m ∈ the whole network
Channel set OC,OrOtherwise,To any time user i,
Optionally, network optimization subproblem
Constraint condition are as follows: the constraint condition six to constraint condition nine；
Physical layer optimizes subproblem
Constraint condition are as follows: the constraint condition one to constraint condition five；
Obtaining dual problem according to Lagrangian is,
Wherein,For Lagrange coefficient, that is, dual variable；
The update mode of dual variable are as follows:
For the t times iteration step length of link e.
Optionally, the network optimization subproblem is handled using convex optimization method, obtains the first processing result, comprising:
Network optimization subproblem is decomposed into Q routing subproblem, using cvx optimization tool to each routing subproblem into
Row solves,
The solution of all routing subproblems is synthesized, the solution as network layer optimization subproblem.
The device have the advantages that as follows:
Firstly, the effective appearance for the constraint condition and link distributed based on network layer in wireless Mesh netword and physical layer resources
Magnitude constructs the Route Optimization Model of cognitive radio Mesh network；Then, pass through the side of embodying of excavation optimization problem
Formula, channel distribution on the optimized throughput subproblem that entire optimization problem is decomposed into network layer and physics-link layer with
Power control subproblem；Simultaneously propose based on Lagrange antithesis Solving mechanism, by dual variable to two sub-problems into
Row is adjusted, and obtains the cross-layer optimizing method with maximum throughput.
In addition, the link available capacity value in the embodiment of the present invention is based on comprehensive channel distribution, power control and frequency spectrum
What detection probability determined, it is ensured that the accuracy of calculating.
Detailed description of the invention
Fig. 1 is the flow diagram of the method for the embodiment of the present invention；
Fig. 2 is the structural schematic diagram of the embodiment of the present invention；
Fig. 3 is the schematic diagram of individual UVR exposure scheme in the embodiment of the present invention；
Fig. 4 is the schematic diagram of the crossover mechanism based on substring in the embodiment of the present invention；
The convergent schematic diagram of algorithm when Fig. 5 is different step-lengths in the embodiment of the present invention.
Specific embodiment
In order to preferably explain the present invention, in order to understand, with reference to the accompanying drawing, by specific embodiment, to this hair
It is bright to be described in detail.
All technical and scientific terms used herein with to belong to those skilled in the art of the invention usual
The meaning of understanding is identical.Term as used herein in the specification of the present invention is intended merely to description specific embodiment
Purpose, it is not intended that in the limitation present invention.Term " and or " used herein includes one or more relevant listed items
Any and all combinations.
The embodiment of the present invention provides the routing optimization method based on Duality Decomposition in a kind of wireless Mesh netword, such as Fig. 1 institute
Show, this method can include:
Step 101: according to the constraint condition of network layer in wireless Mesh netword route optimization process and physical layer resources point
The constraint condition matched, the available capacity value that each link is determined based on the channel use information of link each in wireless Mesh netword,
Construct the Optimized model of wireless Mesh netword；
Step 102: the Optimized model being described using Lagrangian, and the optimization problem of Optimized model is divided
Solution obtains network optimization subproblem and physical layer optimization subproblem；
Step 103: initializing the dual variable of Lagrangian；
Step 104: based on the dual variable after initialization, the network optimization subproblem is handled using convex optimization method,
Obtain the first processing result；And based on the dual variable after initialization, the physical layer is handled using genetic algorithm and optimizes son
Problem obtains second processing result；
Step 105: judging whether the first processing result and second processing result restrain, if convergence, the first processing is tied
Fruit and second processing result are exported as optimal solution；
Step 106: if the first processing result and second processing result in step 105 do not restrain, according to the first processing
As a result the dual variable with second processing as a result, be updated to dual variable, and based on update repeats step 104 and step
105 process.
That is, in step 106, it, can be according to the first processing result and second processing result to first if not restraining
Dual variable after beginningization is updated, and then during repeating step 104, can be used based on the dual variable of update
The convex optimization method processing network optimization subproblem, obtains the first processing result；And based on the dual variable after initialization,
The physical layer is handled using genetic algorithm and optimizes subproblem, obtains second processing as a result, repeating pair in step 105 in turn
The process that first processing result and second processing result are judged.
In the present embodiment before step 101, it can also carry out following steps 100:
Step 100: each chain is determined based on cognitive user, the corresponding channel use information of primary user in wireless Mesh netword
The available capacity value on road.
For example, sub-step 01: for each channel in wireless Mesh netword, when it is idle for perceiving present channel, energy
Enough being correctly detecting present channel is idle probability as the first probability；
Sub-step 02: when perceive present channel occupied by primary user when, can error detection to present channel be it is idle
Probability is as the second probability；
Sub-step 03: the primary user that will acquire occupies the probability of present channel as third probability, and will acquire
Primary user discharge present channel probability as the 4th probability；
Sub-step 04: time user when primary user occupies present channel, while secondary user uses current channel communications is obtained
Corresponding first SINR；
Sub-step 05: obtaining and work as the vacant present channel of primary user, and secondary user uses when current channel communications users
Corresponding 2nd SINR；
Sub-step 06: according to first probability, the second probability, third probability, the 4th probability and the first SINR, second
SINR obtains the available capacity value that link uses present channel.
The constraint condition distributed in the method for the present embodiment based on network layer in wireless Mesh netword and physical layer resources and
The available capacity value of link, constructs the Route Optimization Model of cognitive radio Mesh network；Then, by excavating optimization problem
Mode is embodied, on the optimized throughput subproblem that entire optimization problem is decomposed into network layer and physics-link layer
Channel distribution and power control subproblem；The antithesis Solving mechanism based on Lagrange is proposed simultaneously, by dual variable to two
A subproblem is adjusted, and obtains the cross-layer optimizing method with maximum throughput.
In addition, the link available capacity value in the embodiment of the present invention is based on comprehensive channel distribution, power control and frequency spectrum
What detection probability determined, it is ensured that the accuracy of calculating.
For the scheme for being better understood from the embodiment of the present invention, each step is described in detail below in conjunction with Fig. 2.
First: constructing the basic model of cross-layer optimizing
(1) detection probability calculator, for present channel can be correctly detecting when it is idle for perceiving present channel
It is idle probability as the first probability；
It can be used whether existing channel-aware algorithm perception present channel is idle in the present embodiment.
Specifically, defining binary variable smThe state for indicating channel m, works as smWhen=1, indicate that channel is the free time, otherwise sm
=0, indicate that channel is occupied.F (γ) is the distribution probability for receiving signal, and meets following expression
Wherein,For mean receiving power.When using energy detection method, it is assumed that the accumulative acquisition energy of channel m is
Ym then correctly judges channel for idle probability are as follows:
Further, when channel is Gaussian channel, the first probabilityIt is represented by
Wherein τ is decision threshold, and u is time-bandwidth product.
(2) probability of false detection calculator, for when perceive present channel occupied by primary user when, can error detection to currently
Channel is idle probability as the second probability
The idle state or occupied state of existing channel-aware algorithm channel perception are used in the present embodiment.
Specifically, the second probability of erroneous detection is represented by
If present channel is Gaussian channel,
(3) idle probability getter, the primary user for will acquire occupy the probability of present channel as third probability,
And the primary user that will acquire discharges the probability of present channel as the 4th probability.
That is, definitionFor the third probability of channel m free time, thenIt is general for channel m the occupied 4th
Rate.
(4) operating condition SINR calculator is used to obtain when primary user's occupancy present channel, while secondary user uses currently
Time user corresponding first SINR when channel communication.
That is, for calculating when primary user's work, and secondary user's false judgment channel idle, secondary user makes at this time
When being communicated with channel m, the SINR value of acquisition.
Specifically, the power of communications for defining the primary user of busy channel m is Pm, when secondary user i and time user j use letter
When road m is communicated,It is represented by
WhereinInterference for other secondary users using channel m to secondary user j, N0For white noise acoustical power, HijFor node
The path loss of i and node j.
(5) idle condition SINR calculator, for obtaining when primary user's occupancy present channel, and secondary user uses current letter
Time user corresponding 2nd SINR when road communicates.
That is, for calculating when primary user's free time, and secondary user's accurate judgement channel idle, secondary user makes at this time
When being communicated with channel m, the SINR value of acquisition.
Specifically, when secondary user i is communicated with time user j using channel m,It is represented by
(6) link capacity calculator, for according to first probability, the second probability, third probability, the 4th probability and the
One SINR, the 2nd SINR obtain link eijUse the available capacity value of present channel.
That is: for calculating considering power control, channel distribution and detection probability and probability of false detection after, acquisition it is effective
Capability value.
Specifically, when secondary user i is communicated with time user j using channel m, available capacityIt is represented by
And the whole available capacity between secondary user i and secondary user j are as follows:
WhereinVariable is distributed for link circuit resource.IfIndicate that time user i is logical with time user j on channel m
Letter.Otherwise,
(7) network layer constrains composer, the constraint condition obeyed for obtaining routing algorithm.
Specifically, definition < sq,dq,rq> indicate q-th of route need end to end in network, wherein sq、dq、rqPoint
It Biao Shi not the source node of route requests q ∈ [1,2 ..., Q], destination node, flow demand.DefinitionIndicate that routing is asked
The flow for asking q to transmit from node i → j (j → i), wherein i, j ∈ V, i ≠ j.
Then according to flow Conservation Relationship, have,
Wherein, the flow that formula (9) requires route requests q to flow through any chain road is non-negative；Formula (10) indicate except source node and
Except destination node, the flow conservation of each node, i.e. inflow flow are equal to outflow flow；Formula (11) is indicated for arbitrarily routing
Request, the flow flowed out from source node are at least the λ of its request amountqTimes；Formula (12) indicates to be flowed into any route requests
The flow of destination node is at least the λ of its request amountqTimes；Formula (13) indicates that the sum of all flows for flowing through a link cannot surpass
Cross the available capacity of the link.
(8) physical layer resources assignment constraints composer: the constraint item obeyed for obtaining physical layer resources allocation algorithm
Part.
Specifically, for any user i, j (i, j ∈ V), it is contemplated that the parallel transmission of data is easily led between secondary user
The arrival of cause data packet is out-of-order, in order to guarantee the arrival consistency of data packet, defines between time user i and secondary user j each
1 channel can be distributed, that is,
DefinitionIndicate time channel distribution situation of user i.IfThenI ≠ j, m ∈ OC,OrOtherwise,For any time user i, the number of channel which can use simultaneously by
The limitation of network interface number, that is,
(9) Optimized model composer, for according to the constraint condition and physical layer for constructing network layer in wireless Mesh netword
The available capacity of the constraint condition of resource allocation, the channel determined based on channel use information each in wireless Mesh netword
Value, constructs the Optimized model of wireless Mesh netword.
That is, comprehensively considering frequency spectrum perception probability, probability of false detection, resource allocation in cognitive radio Mesh network about for constructing
The Network Optimization Model of beam, route restriction.
It is known that including N number of node and Q route requests in cognitive radio Mesh network, meet ratio justice
Property routing optimality problem be search for optimal channel distribution and power control scheme to maximize the whole network ratio justice effectiveness
Function.In form, which can be expressed as,
max:∑q∈Qlog(1+λq) (16)
s.t.
Wherein, < sq,dq,rq> it is input parameter.Pay attention toλ is to determine Internet resources distribution and routing choosing
The optimized variable selected, objective function and UijIt include nonlinear restriction.It is non-thread that the optimization problem belongs to MIXED INTEGER in form
The convex optimization problem of property, and belong to np hard problem, it can not be solved by optimization routine tool such as CPLEX etc..
It should be noted that optimization problem is variable related with routing layer, wherein constraint and net by physical layer a~e
The constraint of network difference 16f~16i and interconnection constraint 16j.Constraint 16j is carried out Lagrangian rewriting, optimization problem can be divided
Solution.After decomposition, that is, it can be divided into individual physical layer optimization problem and network layer optimization problem.
(10) it Duality Decomposition optimizer: is based primarily upon Duality Decomposition method and above-mentioned model is optimized, and solve optimal
Routed path and resource allocation vector.
For example, Lagrangian, which can be used, describes the Optimized model, and the optimization problem of Optimized model is divided
Solution obtains network optimization subproblem and physical layer optimization subproblem；
The dual variable for initializing Lagrangian, based on the dual variable after initialization, at convex optimization method
The network optimization subproblem is managed, the first processing result is obtained；And based on the dual variable after initialization, using genetic algorithm
The physical layer optimization subproblem is handled, second processing result is obtained；
Judge whether restrain according to the first processing result and second processing result, if convergence, by the first processing result and
Second processing result is exported as optimal solution, otherwise, based on the first processing result and second processing result to pair after initialization
Mutation amount is updated, and the dual variable based on update reacquires the first processing result and second processing result.
Second: model decomposing method used in Duality Decomposition optimizer
It is found that the cross-layer optimizing problem in cognitive radio Mesh network based on SINR is same from above-mentioned optimization problem description
When the constraint condition comprising network layer and physical layer, and contacting for upper layer and lower layer is to flow through the total flow of link no more than
The intrinsic capacity of the link, i.e. inequality constraints (16j).For this purpose, the present embodiment is it is further proposed that the model based on dual variable
Decomposition method.
Step 1: Lagrangian corresponding to solving optimization problem is,
Wherein,For Lagrange coefficient.
Step 2: above-mentioned function is deformed, following warping function is obtained:
Step 3: network optimization subproblem is solved from warping functionOptimize subproblem with physical layer
That is,
Step 4: obtaining dual problem according to Lagrangian is,
Third: model optimization method used in Duality Decomposition optimizer
The cross-layer optimizing based on Duality Decomposition is used in the present embodiment.
Step1: setting t=0 initializes Lagrange coefficient, that is, dual variable
Step2: based on the dual variable after initialization, network layer is solved using convex optimization method and optimizes subproblem
Step3: based on the dual variable after initialization, physical layer problem is solved using genetic algorithm
Step4: according to the solving result of Step2 and Step3, judging whether majorized function i.e. dual problem has restrained,
If convergence is exported the result of Step 2 and Step 3 as optimal solution；
If not restraining, the projection subgradient of each of the links e in dual problem (22) is updated according to formula (23), that is, is updated and drawn
Ge Lang coefficient (dual variable),
Wherein,For the t times iteration step length of link e；
At this point, the number of iterations t=t+1, the knot for obtaining Step 2 and Step3 is repeated based on updated Lagrange coefficient
Fruit, and the deterministic process of Step 4 is executed again.
In conjunction with Fig. 2 analytic explanation, in Fig. 2 after solution, corresponding step is that dual variable updates, and the present embodiment can not
Limiting the step is before the step of judging convergence or after the step of judging convergence, before the step of judging convergence
Or it is ok later.As long as after judgement does not restrain in the present embodiment, obtaining above-mentioned son according to the dual variable based on update
As a result.
4th: network layer subproblem, which solves, to be illustrated
Network layer optimization subproblem in Step2 is solved, specific steps include:
Step2-1: optimize subproblem indicates again,
S.t. (16f)~(16i)
Step2-2: bundle PROBLEM DECOMPOSITION is Q routing subproblem.Since each routing subproblem is stringent convex
Optimization problem, therefore each subproblem is solved using cvx optimization tool herein,
Step2-3: synthesizing the solution of all subproblems, and in this, as the solution of network layer optimization subproblem.
5th: physical layer problem solving illustrates
Physical layer optimization subproblem in Step3 is solved, the present embodiment is using genetic algorithm to physical layer problem
It is solved.Its key step includes:
Step3-1: the coding mode using network node as chromosome is used, and based on integer and real number combined coding side
Method generates initialization population；
It specifically includes: to the channel selection of arbitrary node i, using substringIt indicates, the substring is by integer vectorsWith real vectorComposition, whereinIndicate the channel that node i is currently distributed
Set, C are number of available channels in network, and I is the network interface number of node,
Indicate the power ratio of each network interface distribution in node i.
Fig. 3 illustrates the integer coding effect based on network node, wherein number of available channels 4, usable interface number are 2.
Indicate that the 1st interface of node 1 uses channel 2, the 2nd interface uses
Channel 3.
Indicate that the 1st interface uses power2nd interface uses power grade
Since any two interface cannot use same channel in same node, thenOrDeng
It will be regarded as unreasonable encoding scheme, it repaired during initialization of population, to accelerate convergence process.
Step3-2: the encoding scheme based on Step3-1 generates M coding individual and is used as initial population.
Step3-3: after individual UVR exposure is completed, resource allocation and perception probability can comprehensively considered by formula (8) determination
Each capacity value；
Step3-4: individual fitness is evaluated, and evaluates the adaptation value function of each individual in initialization population.This
The individual fitness that embodiment defines is based on current capacities priceThe whole network maximum available capacity that can be obtained
Step3-5: with genetic algorithm to Evolution of Population.Wherein main operation mainly includes the probability choosing based on roulette
The basic evolutional operation such as select, intersect, making a variation.
Particularly, for the crossover operation in genetic algorithm, in order to guarantee the gene of parent parents, this hair as far as possible
It is bright to use the cross method based on substring, using substring as basic operation unit, and crossover location is randomly choosed, can guarantee to hand over
New individual after fork meets the constraint condition of coding, and specific crossover process is as shown in Figure 4.
For the mutation operation in genetic algorithm, the present invention executes corresponding variation behaviour for according to the different type of substring
Make.For channel substring, variation executes change from maximum available channel range；For power substring, make a variation from most
High-power range executes change.In addition, will also be repaired to unreasonable individual after executing mutation operation, to accelerate to restrain
Process.
Above-described embodiment has fully considered influence of the activity of primary user in cognitive radio system to secondary user, thus energy
More accurately measure the available capacity between link.
The present embodiment constructs the optimization problem based on dual variable and decomposes mechanism, and the solution that can effectively reduce former problem is multiple
Polygamy, and make the result solved very close to optimal solution.
In addition, above-mentioned model optimization method has preferable convergence, the search of optimal solution can be faster completed.
In the region 1000*600, the cognitive radio Mesh comprising 2 primary users and 20 users is randomly generated
Network.Each node is equipped with 3 intelligent radio network interface cards, 4 channels, maximum transmission power 16.The maximum transmission distance of node
For 250m.The transmission coverage area of primary user is 150m, and path-loss factor γ is 4, the bandwidth H of each channelmFor 54Mbps,
Reception threshold value is 10dbm.
The convergent of the algorithm under asynchronous elongate member is primarily looked at, 2 route needs, step-length are randomly generated in network
Step is respectively set to 0.004,0.006 and 0.008.Experimental result is as shown in Figure 5.As seen from Figure 5, using the equal energy of different step-lengths
Optimal solution is enough converged to, and step-length is bigger, convergence rate is faster, but respective waveforms oscillation is bigger.
Finally, it should be noted that above-described embodiments are merely to illustrate the technical scheme, rather than to it
Limitation；Although the present invention is described in detail referring to the foregoing embodiments, those skilled in the art should understand that:
It can still modify to technical solution documented by previous embodiment, or to part of or all technical features into
Row equivalent replacement；And these modifications or substitutions, it does not separate the essence of the corresponding technical solution various embodiments of the present invention technical side
The range of case.
Claims (8)
1. based on the routing optimization method of Duality Decomposition in a kind of wireless Mesh netword characterized by comprising
Step 1: the pact distributed according to the constraint condition of network layer in wireless Mesh netword route optimization process and physical layer resources
Beam condition, the available capacity value that each link is determined based on the channel use information of link each in wireless Mesh netword construct nothing
The Optimized model of line Mesh network；Step 2: the Optimized model being described using Lagrangian, and to the excellent of Optimized model
Change problem is decomposed, and network optimization subproblem and physical layer optimization subproblem are obtained；Step 3: initialization Lagrangian
Dual variable；
Step 4: based on the dual variable after initialization, the network optimization subproblem being handled using convex optimization method, obtains the
One processing result；And based on the dual variable after initialization, the physical layer is handled using genetic algorithm and optimizes subproblem, is obtained
Obtain second processing result；
Step 5: judging whether the first processing result and second processing result restrain, if convergence, by the first processing result and the
Two processing results are exported as optimal solution；
Step 6: if the first processing result and second processing result in step 5 do not restrain, according to the first processing result and
Two processing results, are updated dual variable, and the dual variable based on update repeats the process of step 4 and step 5, until
Convergence；
Wherein, the Optimized model that wireless Mesh netword is constructed in the step 1 specifically includes:
Assuming that including N number of node and Q route requests, the objective function of Optimized model in wireless Mesh netword are as follows:
s.t.
Wherein, < sq, dq, rq> it is input parameter,λ is the optimization for determining physical layer resources distribution and Route Selection
Variable, objective function and UijIt include nonlinear restriction；
Sq, dq, rq respectively indicate route requestsSource node, destination node, flow demand,Indicate that routing is asked
The flow for asking q to transmit from node i → j or j → i, i, j ∈ V, i ≠ j；
For any user i, j (i, j ∈ V), 1 channel can only be distributed between time user i and secondary user j every time by defining, that is,
Wherein,Indicate time channel distribution situation of user i, ifThenThe channel of i ≠ j, m ∈ the whole network
Collect OC,OrOtherwise,To any time user i,
Wherein,It indicates, when perceiving currently empty idle, to be correctly detecting for each channel in wireless Mesh netword
Present channel is idle probability；
pf mIt indicates for each channel in wireless Mesh netword, it, being capable of mistake when perception present channel is occupied by primary user
Detect that present channel is idle probability；
Indicate that the primary user that will acquire for each channel in wireless Mesh netword occupies the probability of present channel；
Indicate that the primary user that will acquire for each channel in wireless Mesh netword discharges the probability of present channel；
It is corresponding to indicate to obtain time user when primary user occupies present channel, while secondary user uses current channel communications
SINR；
It indicates to obtain and works as the vacant present channel of primary user, and is secondary with corresponding using when current channel communications users
SINR；
UijIndicate the whole available capacity that present channel is used between time user i and secondary user j；
For the available capacity when secondary user i is communicated with time user j using channel m；
Variable is distributed for link circuit resource；
λqThe transmission ratio data of data is routed for q-th；
E is the digraph that all nodes are formed according to channel distribution and power control commands in whole network；
For the objective function of Optimized model；
Wherein, the Optimized model is described using Lagrangian in the step 2, and to the optimization problem of Optimized model into
Row decompose specific steps include:
Step 2-1: Lagrangian rewriting is carried out to the formula of constraint condition ten in the step 1, obtains corresponding Lagrange
Function:
Wherein,For Lagrange coefficient；
Step 2-2: deforming the Lagrangian in the step 2-1, obtains warping function:
Step 2-3: network optimization subproblem is solved from the warping function that the step 2-2 is obtainedWith physical layer
Optimize subproblem
Constraint condition are as follows: the constraint condition six to constraint condition nine；
Constraint condition are as follows: the constraint condition one to constraint condition five；
That is,
Step 2-4: obtaining dual problem according to Lagrangian is,
Wherein, in step 6 dual variable update mode are as follows:
For the t times iteration step length of link e.
2. the method according to claim 1, wherein building wireless Mesh netword Optimized model before, it is described
Method further include:
The available capacity of each link is determined based on cognitive user, the corresponding channel use information of primary user in wireless Mesh netword
Value.
3. according to the method described in claim 2, it is characterized in that, based on cognitive user, primary user couple in wireless Mesh netword
The channel use information answered determines the available capacity value of each link, comprising:
Current letter can be correctly detecting when it is idle for perceiving present channel for each channel in wireless Mesh netword
Road is idle probability as the first probability；
When perceiving present channel and being occupied by primary user, can error detection to present channel be idle probability as second generally
Rate；
The primary user that will acquire occupies the probability of present channel as third probability, and the primary user's release that will acquire is worked as
The probability of preceding channel is as the 4th probability；
Obtain time user corresponding first when primary user occupies present channel, while secondary user uses current channel communications
SINR；
It obtains and works as the vacant present channel of primary user, and secondary user uses when current channel communications users corresponding second
SINR；
According to first probability, the second probability, third probability, the 4th probability and the first SINR, the 2nd SINR, obtaining link makes
With the available capacity value of present channel.
4. according to the method described in claim 3, it is characterized in that, can correctly be detected when it is idle for perceiving present channel
It is idle probability as the first probability to present channelInclude:
Alternatively, if present channel is Gaussian channel,
Where it is assumed that the accumulative acquisition energy of channel m is Ym, smThe state for indicating channel m, works as smWhen=1, indicate that channel m is sky
Spare time, otherwise sm=0, indicate that channel m is occupied；
F (γ) is the distribution probability for receiving signal, For mean receiving power, τ is decision threshold, and u is the time
Bandwidth product；
And/or
When perceiving present channel and being occupied by primary user, can error detection to present channel be idle probability as second generally
Rate pf m, comprising:
Alternatively, if present channel is Gaussian channel,
Wherein, PF, AIt is idle probability for the error detection under Gaussian channel to present channel；
Pr(|) is the expression formula of conditional probability, wherein Pr(Ym> τ | sm=1) it indicates to tire out on channel m when channel m is idle
Meter obtains energy YmMore than the probability of threshold value τ；
Wherein, the PD, AIt is idle probability to be correctly detecting present channel under Gaussian channel；
Qu() is broad sense Marcum Q function.
5. according to the method described in claim 4, it is characterized in that, the probability that the primary user that will acquire occupies present channel is made
The probability of present channel is discharged as the 4th probability for third probability, and the primary user that will acquire, comprising:
Defining third probability isThen the 4th probability is
6. according to the method described in claim 5, it is characterized in that, obtaining when primary user occupies present channel, while secondary user
Use when current channel communications corresponding first SINR of user, comprising:
Wherein, the power of communications of the primary user of busy channel m is Pm, when secondary user i is communicated with time user j using channel m
When,For the first SINR,Interference for other secondary users using channel m to secondary user j, N0For white noise acoustical power,
HijFor the path loss of secondary user i and time user j；
And/or
It obtains and works as the vacant present channel of primary user, and secondary user uses when current channel communications users corresponding second
SINR, comprising:
For when primary user's free time, and secondary user's accurate judgement channel idle, secondary user i and time user j are used at this time
When channel m is communicated, the 2nd SINR of acquisition；
Wherein,The power of communications with node j when channel m is communicated is selected for node i.
7. according to the method described in claim 6, it is characterized in that, according to first probability, the second probability, third probability,
4th probability and the first SINR, the 2nd SINR obtain link eijUse the available capacity value of present channel, comprising:
Wherein, UijIndicate the whole available capacity that present channel is used between time user i and secondary user j,
For the available capacity when secondary user i is communicated with time user j using channel m,
Variable is distributed for link circuit resource, ifIndicate that time user i is communicated on channel m with time user j, otherwise,
Wherein, V is the set of any user node in whole network；
WmFor the bandwidth value of each channel m.
8. being asked the method according to the description of claim 7 is characterized in that handling the network optimization using convex optimization method
Topic obtains the first processing result, comprising:
Network optimization subproblem is decomposed into Q routing subproblem, each routing subproblem is asked using cvx optimization tool
Solution,
The solution of all routing subproblems is synthesized, the solution as network layer optimization subproblem.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710557122.4A CN107343303B (en) | 2017-07-10 | 2017-07-10 | Based on the routing optimization method of Duality Decomposition in wireless Mesh netword |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710557122.4A CN107343303B (en) | 2017-07-10 | 2017-07-10 | Based on the routing optimization method of Duality Decomposition in wireless Mesh netword |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107343303A CN107343303A (en) | 2017-11-10 |
| CN107343303B true CN107343303B (en) | 2019-11-01 |
Family
ID=60219775
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710557122.4A Active CN107343303B (en) | 2017-07-10 | 2017-07-10 | Based on the routing optimization method of Duality Decomposition in wireless Mesh netword |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107343303B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108449267B (en) * | 2018-03-15 | 2020-08-07 | 东北大学 | Reliable routing algorithm based on link quality estimation |
| CN108495339B (en) * | 2018-03-21 | 2021-07-20 | 浙江理工大学 | Wireless network traffic load sharing method based on momentum improved dual decomposition |
| CN111404815B (en) * | 2020-03-06 | 2021-03-16 | 武汉大学 | Constrained routing method based on deep learning |
| CN113490179A (en) * | 2021-07-19 | 2021-10-08 | 北京信息科技大学 | Unmanned aerial vehicle coverage optimization method based on signal-to-interference-and-noise ratio probability perception |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008091725A1 (en) * | 2007-01-25 | 2008-07-31 | Cisco Technology, Inc. | Path optimization for mesh access points in a wireless mesh network |
| CN101420321A (en) * | 2008-11-27 | 2009-04-29 | 电子科技大学 | SDH network planning design method for multi-modularized optical fiber |
| CN103096415A (en) * | 2013-01-15 | 2013-05-08 | 东北大学 | Route optimizing device and method catering to perceive wireless mesh network |
| CN103200689A (en) * | 2013-04-01 | 2013-07-10 | 东南大学 | Link distribution method used for multichannel wireless Mesh network |
| CN104053205A (en) * | 2013-03-13 | 2014-09-17 | 中国科学院大学 | Wireless sensor network routing method based on quasi-optimization principle |
-
2017
- 2017-07-10 CN CN201710557122.4A patent/CN107343303B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008091725A1 (en) * | 2007-01-25 | 2008-07-31 | Cisco Technology, Inc. | Path optimization for mesh access points in a wireless mesh network |
| CN101420321A (en) * | 2008-11-27 | 2009-04-29 | 电子科技大学 | SDH network planning design method for multi-modularized optical fiber |
| CN103096415A (en) * | 2013-01-15 | 2013-05-08 | 东北大学 | Route optimizing device and method catering to perceive wireless mesh network |
| CN104053205A (en) * | 2013-03-13 | 2014-09-17 | 中国科学院大学 | Wireless sensor network routing method based on quasi-optimization principle |
| CN103200689A (en) * | 2013-04-01 | 2013-07-10 | 东南大学 | Link distribution method used for multichannel wireless Mesh network |
Non-Patent Citations (1)
| Title |
|---|
| Lagrange对偶结合粒子群优化的无线Mesh网络信道分配方案;郑金芳;《计算机应用研究》;20170131;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107343303A (en) | 2017-11-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Pandey et al. | Low-latency and energy-balanced data transmission over cognitive small world WSN | |
| Li et al. | Multi-dimensional conflict graph based computing for optimal capacity in MR-MC wireless networks | |
| CN107343303B (en) | Based on the routing optimization method of Duality Decomposition in wireless Mesh netword | |
| US20110188409A1 (en) | Communication Network Topology Determination | |
| Tam et al. | A decomposition-based multi-objective optimization approach for balancing the energy consumption of wireless sensor networks | |
| CN103945482A (en) | Energy optimization routing algorithm in consideration of energy balance | |
| Jia et al. | A genetic approach on cross-layer optimization for cognitive radio wireless mesh network under SINR model | |
| CN103596182A (en) | Method for managing spectrum resources in D2D communications | |
| CN105554888A (en) | Link multi-rate-based multi-radio frequency multi-channel wireless Mesh network channel allocation algorithm | |
| Chuang et al. | Dynamic multiobjective approach for power and spectrum allocation in cognitive radio networks | |
| CN102387585B (en) | Communication method and communication nodes for wireless communication network | |
| Wang et al. | Joint multicast routing and channel assignment for multi-radio multi-channel wireless mesh networks with hybrid traffic | |
| CN115866787A (en) | Network resource allocation method integrating terminal direct transmission communication and multi-access edge calculation | |
| Robert et al. | Genetic algorithm optimized fuzzy decision system for efficient data transmission with deafness avoidance in multihop cognitive radio networks | |
| CN105493602A (en) | Dynamic clustering for radio coordination in a virtual network | |
| CN115134928B (en) | Wireless Mesh network congestion control method with optimized frequency band route | |
| Németh et al. | Throughput of ideally routed wireless ad hoc networks | |
| Wang et al. | Joint offloading decision and resource allocation in vehicular edge computing networks | |
| CN107371166B (en) | Cross-layer optimization device and method based on spectrum aggregation in wireless network | |
| Flushing et al. | Relay node placement for performance enhancement with uncertain demand: A robust optimization approach | |
| Amar et al. | Online learning for shortest path and backpressure routing in wireless networks | |
| Ren et al. | Joint spectrum allocation and power control in vehicular communications based on dueling double DQN | |
| CN113285832B (en) | NSGA-II-based power multi-mode network resource optimization allocation method | |
| Yousefvand et al. | Interference-constraint spectrum allocation model for cognitive radio networks | |
| Dai et al. | Effective Channel Assignment Based on Dynamic Source Routing in Cognitive Radio Networks. |
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 |