CN103796211B - Joint Power and method for channel allocation in a kind of cognition wireless network - Google Patents

Abstract

The invention discloses joint Power and method for channel allocation in a kind of cognition wireless network, it is comprised the following steps：（1）Interference between primary user on each channel in multichannel cognition wireless network and cognitive user, suffered between each cognitive user is realized quantifying and limiting；（2）The utility function for defining the access frequency spectrum behavior of each cognitive user is handling capacity, and the distributed utility function optimal solution for solving cognitive user is carried out to reach the optimal of overall network performance by building non-cooperative game framework；（3）Introduce resource occupation behavior of the corresponding cost function to cognitive user to fix a price, limit the selfishness of cognitive user；（4）A latent betting model is built into using non-cooperative game, it is ensured that its convergence.It is an advantage of the invention that：Joint Power and Channel Assignment Problems are solved using cost function and latent betting model are distributed, Internet resources distributional equity and spectrum utilization efficiency is improved.

Description

Joint Power and method for channel allocation in a kind of cognition wireless network

Technical field

The invention belongs to wireless communication technology field, joint Power and letter in more particularly to a kind of cognitive radio system Channel allocation method.

Background technology

Cognitive radio (CR:Cognitive Radio) it is a kind of new technology for improving frequency spectrum resource utilization rate.It is made It is a kind of intelligent communication system that can perceive extraneous communication environment, it is allowed to the frequency of multidimensional is carried out on time domain, frequency domain and spatial domain Spectrum is multiplexed and shared.The core concept of CR is：By regulating networks running parameter, quantifying and limiting to primary user (PUs: Primary Users) interference on the premise of, it is allowed to the secondary user (SU based on cognitive radio function: Secondary User), also referred to as cognitive user, collectively referred to herein as cognitive user, i.e. CR) opportunistic dynamic access mandate frequency range, so as to improve The availability of frequency spectrum.Mitola is the scholar for proposing cognitive radio concept earliest, and he emphasizes that cognitive radio is based on model theory Demonstrate,prove and obtain a kind of radio of specified function in radio association area.In other words, cognitive radio is received based on frequency spectrum Limit network with correspondence primary user, it is necessary to reciprocally negotiate and select a good opportunity and access a kind of technology that corresponding channel realizes resource-sharing.To mesh Before untill, some government departments and radio standard tissue have launched respectively corresponding policy and standard.As the United States Federal leads to Federal communications commission (FCC:Federal Communications Commission) proposed in succession in 2002,2003 it is corresponding Unauthorized device needs to have the ability and the frequency spectrum share mode based on interference temperature for perceiving vacant frequency range；Britain communicates Office (Ofcom:Office of Commnications) issue Britain TVWS frequency ranges using emphasizing to set up complete in rule English TVWS frequency range databases, it is contemplated that equipment cost, do not require that equipment possesses frequency spectrum perception ability.

Game theory is a kind of mathematical tool of research strategy form, for being determined under prediction and optimization imperfect competition The decision-making of plan main body and equalization problem.Due to that can analyze and predict behavior and the income of rationality user in competitive environment, game It is used in wireless network resource configuration and Power Control by increasing.On the one hand, in many cognition wireless networks of isomery In coexisted environment, the finiteness of usable spectrum resource causes that user's competitive behavior becomes selfish.User does not take into account that other users Or the interests of whole network, and one-side number one of pursuing is maximized.On the other hand, cognitive radio is substantially autonomous , by academic environment, itself running parameter is adjusted, so as to improve, optimize network performance.For single cognition wireless network scene The interaction of middle multi-user and influence, can be studied by game modeling.In theory of games framework, defined cognitive without The line network user is participant, and corresponding new configured transmission (frequency, power etc.) selection is behavior collection, the obtained net of correspondence Network performance is the effectiveness collection of policymaker, determines the corresponding betting model of special scenes, the row of many-side research cognition wireless network For：First, by building the shared betting model of the dynamic spectrum between the network user, user is analyzed under the game framework of specification Behavior, and determine running parameter adjustable strategies；Secondly, the various optimized criterions of game theory energy design frequency spectrum share problem, In general, it is a multi-objective optimization question that frequency spectrum optimization is utilized, it is difficult to its optimal solution is obtained, and game theory is by seeking not With the balanced criterion under game strategies, optimal solution of game can be obtained；Furthermore, non-cooperative game can just be obtained merely with local information The shared optimal resource allocation scheme of dynamic spectrum, is particularly well-suited to the dynamic BTS configuration of distributed cognition wireless network.

Non-cooperative game is the important branch of theory of games, is widely used in cognitive radio networks resource allocation.Such as pin To ad-hoc network application scenes, N.Nie et al. proposes the distributed self-adaption channel point based on latent game and irrepentant study With algorithm, set up latent betting model and solve adaptive channel allocation convergence of algorithm sex chromosome mosaicism well；And in order that must recognize Know that the network user will not produce the skew problem of utility function, N.Nie et al. to also proposed based on irrepentant study because of selfishness Distributed self-adaption channel allocation algorithm.Although this algorithm has taken into full account fairness of the cognitive user in competitive resource Problem, but disturbing of not accounting for that primary user is subject to；In addition, the optimization selection problem of power is not also solved well Certainly.A.Mostaani and M.F.Sabahi proposes a kind of new utility function design on the Research foundation of N.Nie et al., as a result Show, the gambling process under price mechanism is a kind of latent game, can converge on Nash Equilibrium.Additionally, Goodman et al. is carried first The power control algorithm based on non-cooperative game is gone out, the algorithm points out NPG (Non-cooperative Power control Game) one and only one Nash Equilibrium point, but research shows that the Nash Equilibrium point is pareto ineffective.Therefore, Goodman et al. introduces cost mechanism, establishes a kind of non-cooperative power control betting model (NPGP based on linear price: Non-cooperative Power control Game of Pricing), the design of its cost function is competing in order to suppress user The selfishness striven, makes resource of each user to using pay a price.Result shows that NPGP can improve the handkerchief of Nash Equilibrium point Tired support efficiency, but be not optimal, its deficiency is that utility function produces degenerate solution in zero point power degeneration.Therefore, C.W.Sung The non-linear price mechanism for being produced to user and disturbing proportion to be directly proportional is introduced with W.S.Wong, and use is constructed from information theory view The utility function at family.What above researcher generally considered is to carry out a metamessage (power or letter using non-cooperative game model Road) iterative, it is disadvantageous in that the control of no joint Power and Channel assignment.

Channel distribution and power control techniques are the primary study contents of resource allocation.Song Wu China et al. proposes that one kind is based on The combined channel of non-cooperative game and the PIWF algorithms of power distribution, Pareto efficiency is improved by pricing mechanism.Although price Mechanism can select cost function, but best price function seek need central entity control, be not suitable for distributed network. Hao He, Jie Chen et al. constructs the latent betting model of a combined channel selection and power distribution, before interference-limited Put, abstract is a nonlinear optimal problem that can improve cognitive radio networks handling capacity and consideration user fairness.This Outward, the special object function of each transmission node, and the latent betting model of structure is defined to seek the optimal solution of problem step by step.Research Result shows that the game of substep orderliness can converge to Nash Equilibrium point, and meet interference constraints limitation.Wherein, Hao He, Jie The nonlinear optimal problem that Chen et al. is proposed is as follows：

Will be from nodeTo nodeTransmitting-receiving node to being defined as.If channel,RepresentOn simultaneously All nodes pair of transmission, then receiverReception signal with interference the ratio between () can be expressed as：

Wherein,It is nodeArriveTransmission gain,It is receiverThe thermal noise at place, usually constant,It is channelInThe transmission power of node.

Define transmitting-receiving node pairHandling capacity be：

FromIf expression formula can be seen thatLess than 1, then the numerical value for obtaining is negative.This means now believing Make an uproar than () too it is low can not continue transmitting in addition the total throughout of whole network can be had a negative impact because additionally Interference always influence whether other links.

Based on above-mentioned analysis, selected by combined channel and power distribution is to optimize above mentioned problem：

Maximize

Subject to

WhereinIt is the set of transmitting-receiving node pair.It is to be operated in channelOn primary user.It is that a two-value becomes Amount, if channelOn linkIt is active, then it is 1；Otherwise it is 0.ConstraintsMean Have and can only be in channel groupIn a channel.

The nonlinear optimal problem is intended to maximize total throughout.All transmitting nodes are intended to maximize Object function, thus, not only need consider object function in, it is also contemplated that the interference that primary users are born.Thus, it is fixed Justice is operated in channelOn primary userIt is as follows：

WhereinIt is that the cognition wireless network that primary user y is subject to is disturbed,It isThe maximum dry that middle primary user is subject to Disturb,It is then the summation of all channel maximum interferences.Threshold value is exceeded for the interference on channel, further definition：

WhereinRepresent the total interference on these channels.Then, using the definition for providing before, target letter can be defined Number is：

Above method presence is disadvantageous in that：First, the connection of distributed AC servo system power and channel is not realized really Close distribution；Second, do not take into full account interfering between cognitive user.However, to realize distributed associating power and letter Road is distributed, and needs to interact bulk information between each cognitive user, this adds increased the complexity of each cognitive user, is being realized On be also a technical barrier.

The content of the invention

The technical problems to be solved by the invention are just to provide joint Power and channel distribution in a kind of cognition wireless network Method, it can be fully taken into account between primary user and cognitive user between (CR-to-PU), each cognitive user (CR-to-CR) AF panel, joint Power and Channel Assignment Problems are solved using cost function and latent betting model are distributed, improve network money Source distributional equity and spectrum utilization efficiency.

In order to solve the above-mentioned technical problem, technical scheme is comprised the following steps：

1st, between the primary user on each channel in multichannel cognition wireless network and cognitive user, each cognitive user Between suffered interference realize quantifying and limiting；

2nd, the utility function for defining the access frequency spectrum behavior of each cognitive user is handling capacity, by building non-cooperative game Framework carries out the distributed utility function optimal solution for solving cognitive user to reach the optimal of overall network performance；

3rd, resource occupation behavior of the corresponding cost function of introducing to cognitive user is fixed a price, and limits cognitive user oneself Private；

4th, it is built into a latent betting model using non-cooperative game, it is ensured that its convergence.

Step 1 ensure that cognitive user to the interference of primary user not over the interference range that primary user can bear, and The co-channel interference problem between cognitive user is taken into full account；Step 3 lead-in sexual valence lattice mechanism limits the selfishness of cognitive user Sexual behaviour, it is ensured that the fairness of the competitive resource between user；Step 4 using non-cooperative game theory be built into one it is latent rich Model is played chess, its convergence is ensured in theory；Problem cumbersome in centralized Control is distributed to each cognitive user by step 2, respectively Individual cognitive user maximize itself effectiveness using distributed orderly behavioral strategy, selection to oneself optimal channel and Corresponding transmission power, so as to reduce the expense of frequency spectrum distribution, improves network throughput and its performance.

Brief description of the drawings

Brief description of the drawings of the invention is as follows：

Fig. 1 is the network (WSN) emulation system of one embodiment of the invention；

Fig. 2 is resource allocation flow chart of the invention；

Fig. 3 is the flow chart of single cognitive user joint Power of the invention and channel iterations；

Fig. 4 is latent function convergence schematic diagram in the present invention.

In Fig. 1,1. primary user base station；2. primary user；3. cognitive user.

Specific embodiment

The invention will be further described with reference to the accompanying drawings and examples：

The present invention is comprised the following steps：

1st, to (CR-to-PU) between the primary user on each channel in multichannel cognition wireless network and cognitive user, Interference between each cognitive user suffered by (CR-to-CR) is realized quantifying and limiting.

It is assumed that the ad-hoc cognitive radio networks being made up of multiple cognitive users, around with the presence of primary user,In the range of random distributionIndividual cognitive user pair, a cognitive user is to including a transmitting node and Individual receiving node；Frequency spectrum byKIndividual orthogonal channel composition,K＜N, while havingIndividual primary user's node, each primary user can be with work Make on one or more of the channels；Defined cognitive userTransmission power vector,It is cognition UserIn channelOn transmission power；Channel, then receiverReceptionCan be expressed as：

Wherein,It is cognitive userChannelTransmission existsLocate receive signal interference ratio (),It is nodeArriveTransmission gain；It is receiverThe thermal noise at place, usually constant；It is channelInThe transmitting work(of node Rate；It is disturbance equation,Be exceptOther cognitive users in addition：

SettingIt is the maximum interference power of primary user (PU).So, channelIn all cognitive users (CR) to each lead Total interference of user (PU) must is fulfilled for：

WhereinIt is channelMiddle transmitting nodeTo primary userObstacle gain,It is similarIt is dry Disturb equation.Setting thresholdingThe patient maximum interference of institute actual far below primary users, then the protection for primary user is asked Topic can just be realized very well.

For certain cognitive user in network, on certain particular channel except receive people having a common goal's neighbor node interference it Outward, itself also produces interference to people having a common goal's neighbor node.Quantify the co-channel interference that certain cognitive user is bornIt is as follows：

Wherein,Represent other cognitive users to present nodeTransmission power,It is other cognitive users to current NodeLink gain,It is similarDisturbance equation.

Similarly, the interference that the cognitive user is produced to other people having a common goal's neighbor nodesIt is as follows：

Then cognitive userThe interference from other cognitive users born and the interference to the generation of other cognitive users Summation can be expressed as：

2nd, the utility function for defining the access frequency spectrum behavior of each cognitive user is handling capacity, by building non-cooperative game Framework carry out the distributed utility function optimal solution for solving cognitive user received to reach the optimal of overall network performance, that is, it is assorted Weighing apparatus.

First, defining a game is.In betting model,It is cognitive net The set of all transmitting nodes of network；It is on transmitting nodeBehavioral strategy；It is object function, transmitting node is expected Maximize object function.

Secondly, each cognitive user nodes utility function is definedIt is the handling capacity of transmitting-receiving node, handling capacity Formula is as follows：

In view of the co-channel interference between unauthorized user, should ensure that the interference between each cognitive user will not cause reception Signal interference ratio at machineLess than the minimum signal interference ratio that can be demodulated, in corresponding to the design of cognitive user utility function, we Have been found thatLess than 1, then the numerical value for obtaining is negative.This means nowValue is too low to cause the transmitting node can not to send out Penetrate, and to the handling capacity contribution negative growth of whole network.

Again, the interference of CR-to-PU, CR-to-CR is quantified, it is as follows as interference constraints condition：

Wherein,It is to be operated in channelOn primary user；It is disturbance equation.ConstraintsMeaning Taste transmitting nodeTo corresponding receiving nodeHave and can only be in channel groupIn a channel, but it is multiple Cognitive user can share same channel；This constraints is embodied in signal interference ratioComputing formula in.

3rd, the utility function behavior of fleeing from of lead-in sexual valence lattice function pair cognitive user punished, suppresses cognitive user Selfish sexual behaviour；

In view of the interference suffered by primary user, definition is operated in channelOn primary userIt is as follows：

WhereinIt is primary userThe cognition wireless network being subject to is disturbed,It isThe maximum dry that middle primary user is subject to Disturb.

For certain cognitive user in network, on certain particular channel except the interference by people having a common goal's neighbor node it Outward, itself also produces interference to people having a common goal's neighbor node.Defined cognitive userCo-channel interference sum be：

For ad-hoc networks, distributed network architecture enables that the cognitive user in network freely determines the communication of oneself Parameter simultaneously correspondingly improves the communication link (message transmission rate higher, lower bit error rate etc.) of itself, but available The restricted gender of frequency spectrum resource and maximum transmitted general power can cause resource of vying each other between cognitive user, one of its negative effect It is exactly that Pareto efficiency is low in a balanced way for network.In the environment reached without mandatory constraints, cognitive user even meeting Benefit according to itself flees from original utility function.Therefore, considering both the above interference, used as limitation is cognitive The foundation of family selfishness sexual behaviour, user is by the corresponding network throughput of acquisition of bidding.Lead-in sexual valence lattice function pair is cognitive to be used The utility function behavior of fleeing from family punished, suppresses the selfish sexual behaviour of cognitive user.Defined cognitive userCost function It is as follows with total utility function：

Wherein、It is constant, as the coefficient of linear cost function.

4th, it is built into a latent betting model using non-cooperative game, it is ensured that its convergence.

The purpose of game is to reach equilibrium, that is, the strategy of all participants is attained by a kind of stable state.Receive assorted Equilibrium is that one kind of static game of complete information is balanced, is defined as follows：

GameIn, wherein strategy combinationBe one receive it is assorted Weighing apparatus, if relative to participant,It is in the selection of given remaining participant's optimal policySituation Lower participantOptimal policy, i.e.,：

From definition it can be seen that Nash Equilibrium be game a kind of stable state, participant reach Nash Equilibrium point when Wait, none of participant can unilaterally deviate from existing state.

Solving Nash Equilibrium is relatively difficult thing, and distribution is solved and is more the increase in difficulty.Generally do not let on person with The expansion of game, iteration carries out accumulation observation, and wishes that process can converge to equilibrium point.Although such case is not most Limits simulate a gambling process, but when scene is modeled as latent game, can guarantee that the convergence of gambling process Property, that is, ensure that it converges to Nash Equilibrium.

Latent game is a kind of betting model that utility function meets specific condition.

Definition：GameIn, there is latent functionMeet following some condition, The then game is latent game.

i)

ii)

WhereinIt is sign function.It is accurately latent game to meet condition i)；Meet condition ii) it is the latent game of order.

The characteristics of latent game is：When all participants make decisions in order, latent game ensure that gambling process is having Nash Equilibrium point is converged in limit step.The searching of latent function is the precondition of latent game modeling, represents game in convergence process Middle overall interests lifting, in other words, any one participant makes more preferable decision-making in convergence process and can cause latent function Value further lifting, this point is it will be apparent that individual interest is consistent with system overall interests in game of diving from definition 's.

Used as latent function, the utility function of each cognitive user is：

If each user fromStrategy change overWhen,Change be more thanIt is corresponding to become Change, then we just by network modelling into the latent game of order, and then can ensure the existence of Nash Equilibrium point.

Prove as follows：

IfIt is latent function, then there is a definitely latent functionMeet following formula：

Work as presenceWhen, sign function meets：

If right, hypothesis establishment below：

Then there is following expression：

Defined from latent game, as long as meeting above-mentioned inequality assumed condition,It isLatent letter Number, the latent game definition of satisfaction order, network can be modeled as latent betting model.

Latent gambling process is the process that cognitive user constantly updates own channel selection and transmission power information.Each is cognitive Channel selection strategies of the user first according to current other cognitive users calculate effectiveness when it uses other channels；Then it is every Utility function value when individual cognitive user is by relatively currently used other channels selects current optimum utility channel in order； And then each cognitive user selects suitable power by calculating the effectiveness under different transmission power；By cycle calculations, obtain A kind of equilibrium state.

Be achieved in that withAs the utility function expression formula of latent function, game is illustrated Utility function value or latent functional value are constantly liftings with the accumulation of iterative process, in other words, each ginseng in game The Channel assignment and power distribution strategies of itself are constantly updated and optimize by comparison procedure repeatedly with person's (cognitive user), point The contribution of cloth causes that overall network effectiveness is improved, and tends towards stability (Nash Equilibrium).

Temperature interference model sharing frequency spectrum resource is based between cognition network and primary user's network；Total utility function is letter RoadMiddle cognitive userThe handling capacity for being obtained is subtracted to disturb the cost function as cost.For the protection to primary user, Cognitive network resource is far from enough.If in the absence of the condition for suppressing transmitting node, then each cognitive user is accessing channel Shi Huicong number ones set out and unilaterally increase effectiveness, some inappropriate transmitting nodes can unilaterally select maximum transmitting Power emission will certainly so cause other suitable transmitting nodes because of the total interference to primary user without considering other situations Can not launch close to threshold value.Therefore introduceAs to cognitive user limitation (、It is constant, as The coefficient of linear cost function).Cognitive user is selecting most suitable power and channel to maximize the handling capacity of itself acquisition When, it is necessary to consider the various interference produced by corresponding transmission power, and therefrom obtain the generation after frequency spectrum resource as it with this Valency.

Introduce, correspondingly cognitive user can tend to select most suitable channel, i.e., the cognitive user is at this During multi-channel operationIts value in other multi-channel operations can be less than.Obviously, the handling capacity of cognition wireless network and fairness meeting Thus it is improved.The selected of value will be a practical problems；It is too small, then may to cause interference with and suppress too small and cause Network performance reduction；It is excessive, then may cause interference with suppression it is too strong and cause lifted network throughput become very difficult；Numerical value may influence convergence rate.From point of theory, as the factor embodied to primary user's AF panel,Preferably take larger Value, to ensure rapidly reduce interference during the excessive even more than primary user's thresholding of interference, protection primary user.

Introduce.IfIt is a kind of macro adjustments and controls to cognitive user nodes hardness, thenCan be a kind of Microcosmic regulation and control softly to cognitive user.Cognitive user weighs effectiveness and cost, and the messaging parameter to oneself is suitably adjusted It is whole.If the Internet resources itself asked for are more, then the cost that it is paid is higher；If resource occupying lacks, that Corresponding paying is also lower.Value also will be a practical problems,Value should with cognitive user obtain gulp down The amount of telling, to adjust effectiveness of each cognitive user in Nash Equilibrium, improves Pareto effect with an order of magnitude or smaller Rate.

Each cognitive user continuously increases the object function of itself so as to reach Nash Equilibrium point in accurate game herein It is a kind of practical method.In order to realize the Continuous behavior of each cognitive user, we introduce a kind of simple Stochastic accessing machine System：Each cognitive user withProbability updating itself strategy.In particular, each cognitive user is in each time slot Incipient stage withWhether probability decides whether to carry out data transmission in CCCH, i.e., according to current effectiveness letter Several value of utilities makes new decision-making.If it succeeds, corresponding cognitive user access network, carries out data transmission；Otherwise after It is continuous to intercept channel.Assuming that cognitive user can be perceived correctly to all channels, the resource request and signaling of other cognitive users are handed over Mutual information is all broadcasted in the reachable CCCH of the whole network.Cognitive user is able to know that respective channel each other Selection information.The program ensure that average time point only one of which cognitive user is taken action, and at the same time surpass It is then non-zero to cross the probability that a cognitive user takes action.When two simultaneously or multiple cognitive user is taken action, dive Function may temporarily decline, but this will not destroy the trend of potential function overall monotone increasing.

The distribution of selection joint Power and the two uncertain factors of channel, are effectively divided resource by distributed algorithm Match somebody with somebody, as shown in Figure 2 and Figure 3.Wherein Fig. 2 is overall flow figure, and Fig. 3 is the supplement of Fig. 2, to determine the channel of optimum utility to cut Access point, provides the concrete operations flow of each cognitive user.Algorithm is as follows：

Step1：Setting ad-hoc network topologies, randomly generate cognitive user pair, calculate communication node link gain, initialize cognitive user effectiveness, Channel assignment, power, coefficient of bidding、, give cognition User node target throughput；Initialization random access procedure；

Step2：Each cognitive user calculates current power selection successively respectively, Channel assignmentUnder effectiveness And handling capacity；

Step3：Judge all cognitive user nodes, if so, terminating algorithm；Otherwise, step4 is carried out；

Step4：Judge whether that all cognitive user nodes all reach maximum transmission power, if so, terminating algorithm；Otherwise, Carry out step5；

Step5：Judge whether to reach maximum iteration, if so, terminating algorithm；Otherwise, step6 is carried out；

Step6：Present cognitive user node calculates current power successivelyUnder, use the effectiveness of other available channels, and handling capacity；

Step7：Present cognitive user node compares the effectiveness under different channels selection, select with maximum utility Channel simultaneously updates, while calculating corresponding handling capacity；Otherwise, step3 is returned；

Step8：Judge the handling capacity that present cognitive user node is obtainedIf then reducing node transmitting work( Rate, and enter step2；Otherwise, increase node transmitting power, and enter step2.

Compared with the scheme for combining of original power and channel distribution, there is following advantage in this programme：

1st, distributed cognition network association power and Channel Assignment Problems are effectively solved using non-cooperative game model, will be total The Nonlinear Optimization Problem of object function (aggregate network throughput) is changed into the object function of distributed cognition user.

2nd, the convergence that corresponding latent betting model ensures network is built, that is, ensures the existence of Nash Equilibrium.

3rd, using AF panel as cognitive user cost function, user (gulped down by the corresponding Internet resources of acquisition of bidding The amount of telling).

4th, the interference of CR-to-PU, CR-to-CR is taken into full account, the quantization of AF panel is realized and is limited, and as price Function.

Embodiment

The present invention is directed to ad-hoc network models, and distribution solves ascending resource in single multichannel cognition wireless network (power and channel) assignment problem.The type of cognitive user can be mobile phone, vehicle radio station, PDA or other.As shown in Figure 1 One network (WSN) emulation system is：Cognitive userIt is randomly dispersed in the region of 100m*100m, available orthogonal number of channel K= 4, maximum iteration is 1000 times.All cognitive user random selection transmission powers and available channel are used as init state. It is assumed that all cognitive users can be perceived correctly to all channels, the resource request and Signalling exchange information of other cognitive users are all Broadcasted in the reachable CCCH of the whole network.Cognitive user be able to know that each other respective transmission power and Corresponding channel selection information.Controlled between in the absence of center, the implementation of this method places one's entire reliance upon between each cognitive user Information exchange, each cognitive user makes a policy in strict accordance with algorithm.Flow of the specific steps as shown in above-mentioned Fig. 2, Fig. 3.

Fig. 4 is latent function (cost function) convergence schematic diagram, dive function convergence Process is consistent with whole network convergence process, because latent betting model causes that any one participant makees in convergence process When going out more preferable decision-making all overall network performance can further lifted.The utility function design for introducing price mechanism is filled The AF panel between the AF panel and cognitive user to primary user point is embodied, the fairness and frequency of resource allocation is improved The utilization rate of spectrum.

Preferable embodiment of the invention is the foregoing is only, is not intended to limit the invention, belonging to any present invention Technical staff, do not depart from disclosed herein spirit and scope on the premise of, can implement formal and details on Make any modification and change, but scope of patent protection of the invention, still must be with the scope of which is defined in the appended claims It is defined.

Claims (4)

1. joint Power and method for channel allocation in a kind of cognition wireless network, it is characterised in that comprise the following steps：

(1) between the primary user on each channel in multichannel cognition wireless network and cognitive user, each cognitive user it Between suffered interference realize quantifying and limiting；

(2) utility function for defining the access frequency spectrum behavior of each cognitive user is handling capacity T_ij, by building non-cooperative game frame Frame carries out the distributed utility function optimal solution for solving cognitive user to reach the optimal of overall network performance；

(3) introduce resource occupation behavior of the corresponding cost function to cognitive user to fix a price, limit the selfishness of cognitive user Property；

(4) it is built into a latent betting model using non-cooperative game, it is ensured that its convergence；

In step (1)：

It is assumed that the ad-hoc cognitive radio networks being made up of multiple cognitive users, around with the presence of primary user, D*D's In the range of random distribution N (1,2 ... n) individual cognitive user pair, a cognitive user including a transmitting node and one to receiving Node；Frequency spectrum is made up of K orthogonal channel, K ＜ N, while there is M primary user's node, each primary user can be operated in one Or on multiple channels；The transmission power vector P of defined cognitive user i_i=[p_i(1), p_i(2)... p_i(c)], p_i(c)It is cognitive user i Transmission power on channel c；

Channel c ∈ K, then the reception SINR of receiver j can be expressed as：

${\mathrm{\γ}}_{ijc}=\frac{P_{ijc}{G}_{ij}}{\underset{s\≠i,s=1}{\overset{N}{\Σ}}{P}_{src}{G}_{sj}f(s,j)+N_0}$

Wherein, γ_ijcIt is the signal interference ratio (SINR) of cognitive user i channels c transmission receptions at j, G_ijIt is channel of the node i to j Transmission gain；N₀It is the thermal noise at receiver j, usually constant；P_ijcIt is the transmission power of i-node in channel c；F (s, j) It is disturbance equation, s is other cognitive users in addition to i：

Set τ as the maximum interference power of primary user, then, total interference of all cognitive users to each primary user in channel c It must is fulfilled for：

$\underset{i=1}{\overset{N}{\&Sigma;}}{P}_{ijc}{G}_{iyc}f(i,y)<\mathrm{\&tau;}$

Wherein G_iycIt is to launch section i points to the obstacle gain of primary user y in channel c, f (i, y) is the disturber for being similar to f (s, j) Journey；

For certain cognitive user in network, on certain particular channel in addition to receiving the interference of people having a common goal's neighbor node, Itself also produces interference to people having a common goal's neighbor node；Quantify the co-channel interference I that certain cognitive user is born_ijcIt is as follows：

$I_{ijc}=\underset{j=1,j=i}{\overset{N}{\&Sigma;}}{P}_{ijc}{G}_{ji}f(j,i)$

Wherein, P_jicRepresent transmission power of other cognitive user nodes to present node i, G_jiIt is that other cognitive user nodes are arrived The link gain of present node i, f (j, i) is the disturbance equation of similar f (s, j)；

The interference I ' that the cognitive user is produced to other people having a common goal's neighbor nodes_ijcIt is as follows：

$I_{ijc}^{\&prime;}=\underset{j=1,j=i}{\overset{N}{\&Sigma;}}{P}_{ijc}{G}_{ij}f(i,j)$

The interference from other cognitive users and the interference summation to the generation of other cognitive users that then cognitive user i is born Can be expressed as：

I_ic=I_ijc+I′_ijc；

In step (2)：

First, it is G={ N, { S to define a game_i}_i∈N, { u_i}_i∈N}；In betting model, N is the institute of cognition wireless network There are the set of transmitting node, S_iIt is the behavioral strategy on transmitting node i；u_iIt is object function, transmitting node expectation maximization mesh Scalar functions；

Secondly, each cognitive user nodes utility function u is defined_i(S_i, S_-i) it is the handling capacity T of transmitting-receiving node_ijc, throughput equation It is as follows：

u_i(s_i, s_-i)=T_ijc=10log (γ_ijc)

Furthermore, quantifying the interference of CR-to-PU, CR-to-CR, it is as follows as interference constraints condition：

γ_ijc＞ γ_min

$\underset{c\&Element;K}{\&Sigma;}{f}_{ijc}(i,j)=1$

f_ijc(i, j)≤P_ijc≤P_maxf_ijc(i, j) c ∈ K

$\underset{i=1}{\overset{N}{\&Sigma;}}{P}_{ijc}{G}_{iyc}f(i,y)<\mathrm{\&tau;},y\&Element;M_c,c\&Element;K$

Wherein, M_cIt is the primary user being operated on channel c, f_ijc(i, j) is disturbance equation, constraintsMeaning Taste transmitting node i to be had and can only be in channel group K a channel to corresponding receiving node j, but multiple is recognized Know that user can share same channel；

In step (3)：

Definition is operated in the primary user M on channel c_cIt is as follows：

$I_{yc}=\underset{i=1}{\overset{N}{\&Sigma;}}{P}_{ijc}{G}_{iyc}f(i,y)$

c∈K y∈M_c

Wherein I_ycIt is that the cognition wireless network that primary user y is subject to is disturbed, I^c _maxIt is M_cThe maximum interference that middle primary user is subject to；

Cost function P (the s of defined cognitive user i_i, s_-i) and total utility function u_ti(s_i, s_-i) as follows：

$P(s_i,s_{-i})=(\mathrm{\&alpha;}\&times;I_{\max }^c+\mathrm{\&beta;}\&times;I_{ic})$

$u_{ij}(s_i,s_{-i})=T_{ijc}-{\mathrm{\&alpha;I}}_{\max }^c-{\mathrm{\&beta;I}}_{ic}$

Wherein α, β are constants, used as the coefficient of linear cost function.

2. joint Power and method for channel allocation in a kind of cognition wireless network according to claim 1, it is characterised in that In step (4)：

P (s_i, s_-i) used as cost function, then the total utility function of each user is：

u_ti(s_i, s_-i)=- P (s_i, s_-i)+u_i(s_i, s_-i)

If each user is from (s_i, s_-i) strategy (s when changing over_i ^*, s_-i), the change of P () is more than u_i() changes accordingly, Then can be by network modelling into the latent game of order.

3. joint Power and method for channel allocation in a kind of cognition wireless network according to claim 2, it is characterised in that： WithUsed as latent function, α, β are coefficient undetermined；Network struction is turned into the latent game mould of order Type, can converge to Nash Equilibrium point and meet the limitation of interference constraints rapidly by distributed orderly game, it is ensured that The convergence of network.

4. joint Power and method for channel allocation in a kind of cognition wireless network according to claim 3, it is characterised in that The distribution of selection joint Power and the two uncertain factors of channel, by distributed algorithm, are effectively distributed resource；It is calculated Method is as follows：

Step1：Setting ad-hoc network topologies, randomly generate cognitive user pair, calculate communication node link gain gain (ii, Jj), initialization cognitive user effectiveness u_ti, Channel assignment ch, power power, factor alpha of bidding, β give cognitive user nodes mesh Mark handling capacity T_tar；Initialization random access procedure；

Step2：Each cognitive user calculates current power selection power, the effectiveness u under Channel assignment ch successively respectively_tiWith handle up Amount T_ij；

Step3：Judge all cognitive user nodes T_ij≥T_tar, if so, terminating algorithm；Otherwise, step4 is carried out；

Step4：Judge whether that all cognitive user nodes all reach maximum transmission power, if so, terminating algorithm；Otherwise, carry out step5；

Step5：Judge whether to reach maximum iteration, if so, terminating algorithm；Otherwise, step6 is carried out；

Step6：Present cognitive user node is calculated under current power power successively, uses the effectiveness u of other available channels_ti, with And handling capacity T_ij；

Step7：Present cognitive user node compares the effectiveness u under different channels selection_ti, select the channel with maximum utility And update, while calculating corresponding handling capacity T_ij；Otherwise, step3 is returned；

Step8：Judge the handling capacity T that present cognitive user node is obtained_ij≥T_tarIf then reducing node transmitting power, go forward side by side Enter step2；Otherwise, increase node transmitting power, and enter step2.