CN102858019A - Down link space-time scheduling method of cognitive cellular network - Google Patents

Abstract

The invention discloses a down link space-time scheduling method of a cognitive cellular network. The method comprises the following steps of: step (1), collecting a channel state information vector between a multi-antenna cognitive base station and a single-antenna cognitive mobile terminal, and a channel state information vector between the cognitive base station and a single-antenna master user; step (2), initiating pheromones distributed on each side of a construction graph and heuristic information; step (3), generating an artificial ant; step (4), by virtue of the cognitive base station, calculating a double orthogonal wave beam shaping vector corresponding to the cognitive mobile terminal; step (5), allocating sending power to a data stream sent by the cognitive base station to the cognitive mobile terminal; step (6), determining whether a preset maximal number of iteration times is reached, and if so, stopping and outputting a optimal access set, wherein the maximal number of iteration times is a natural number; and step (7), updating pheromones distributed on all sides in the construction graph and returning to the step (3); and thus completing the down link space-time scheduling.

Description

Dispatching method when a kind of down link of cognitive Cellular Networks is empty

Technical field

The present invention relates to radio communication and signal of communication processing technology field in a kind of cognitive radio technology, dispatching method when particularly a kind of down link of the cognitive Cellular Networks based on cognitive radio is empty.

Background technology

Along with the continuous growth of wireless traffic and application, distributable radio spectrum resources is more and more nervous.At present, can distribute the scarcity of frequency spectrum to hinder greatly the sustainable development of radio communication.On the other hand, actual measured results shows that the most frequency spectrum that has distributed but is in the poor efficiency state.Under this background, cognitive radio (Cognitive Radio, CR) technology is arisen at the historic moment, the CR network by with hold primary user (the Primary User that authorizes frequency spectrum, PU) network coexisted, share PU in time and space and authorize frequency spectrum, thereby can improve existing utilance of authorizing frequency spectrum, for new wireless traffic and application provide bandwidth.

By utilizing the mandate frequency spectrum of PU, be expected to solve the deficient problem of frequency spectrum that current commercial mobile communication faces based on the cellular network (cognitive cellular network) of CR.In the down link of typical cognitive cellular network, the cognitive base station of many antennas (Cognitive Base Station, CBS) support multiplex data stream to transfer to the cognitive portable terminal (Cognitive Mobile Station, CMS) of different single antenna by space multiplexing technique.

When frequency spectrum share, for not affecting the operation of existing PU network, realize with it " transparent " coexistence, cognitive cellular network need to possess the ability of effectively avoiding the PU interference.Thereby different from the legacy cellular net is that in cognitive Cellular Networks, CBS transfers to the data-signal of cognitive portable terminal CMS and should avoid PU is brought any harmful interference.

For realizing effective utilization of frequency spectrum resource, maximized system capacity need to be dispatched between CMS.CBS by to self with PU and self and CMS between the time change fading channel estimation, and utilize the channel condition information of collecting to realize that the user dispatches, scheduling has user or user's subset realization transfer of data of optimum channel situation within preset time, thereby maximized system capacity is optimized the utilization to frequency spectrum resource.

Summary of the invention

Goal of the invention: technical problem to be solved by this invention is for the deficiencies in the prior art, dispatching method when providing a kind of down link of cognitive Cellular Networks empty.

In order to solve the problems of the technologies described above, the invention discloses a kind of down link of cognitive Cellular Networks dispatching method when empty, may further comprise the steps:

Step (1) gathers the vectorial h of channel condition information between the cognitive base station of many antennas and N the single antenna cognition portable terminal _i, i=1,2 ..., the channel condition information vector g between N and cognitive base station and M the single antenna primary user _j, j=1,2 ..., M;

Step (2), every limit e in the initialization design of graphics _{S, s+1}(j) the upper pheromones τ that distributes _{S, s+1}(j)=(τ _Max+ τ _Min)/2, and heuristic information η _{S, s+1}(j)=| h _j| ², s=0,1 ..., n _T-M, j=1,2 ..., N, wherein τ _MaxWith τ _MinBe respectively the upper bound and the lower bound of pheromones content, τ _MaxRange of set value be 10 ~ 20, τ _MinRange of set value be 0 ~ 10, | h _j| channel condition information vector h is asked in expression _jAmplitude, n _TNumber of antennas for cognitive base station;

Step (3) generates the only artificial ant of m, and wherein the range of set value of m is 5 ~ 30, places ant vertex v in design of graphics ₀The place, every ant is by probability P r (e _{S, s+1}(j)) select limit e _{S, s+1}(j) from vertex v _sMove to vertex v _S+1, s=0,1..., n _T-M; Record ant n _TThe path that moves through for-M time, the corresponding cognitive portable terminal in the limit in the selecting paths forms access set

Step (4), cognitive base station are calculated the dual orthogonal beams figuration vector corresponding to k cognitive portable terminal

Step (5), allocation of transmit power are sent to the data flow of cognitive portable terminal for cognitive base station; Calculate the access set that every ant is selected

Power system capacity

As follows:

Wherein,

For giving access set

In the transmitted power of distribution of flows of k CMS, The expression access set Channel condition information vector between k cognitive portable terminal and the cognitive base station,

Expression is to vector

Carry out matrix transpose operation,

Noise power for the receiver of cognitive portable terminal; Selecting access set corresponding to maximum system capacity is optimum access set

Step (6) judges whether the maximum iteration time that reaches default, then stops and exporting optimum access set if reach

Maximum iteration time is got natural number;

Step (7) is upgraded the pheromones that distributes on all limits in the design of graphics and is returned step (3); Scheduling when finishing thus the down link sky.

Ant is selected limit e in the step of the present invention (3) _{S, s+1}(j) probability P r (e _{S, s+1}(j)) calculate by the following method:

Wherein

Expression connect Vertex v _sWith vertex v _S+1The set on all limits, α and β are the weight coefficient corresponding to pheromones and heuristic information, the range of set value of α and β is 0 ~ 10.

Calculate the dual orthogonal beams figuration vector of k cognitive base station in the step of the present invention (4) according to following methods

To k cognitive base station, make up following matrix

Compute matrix The corresponding characteristic vector of 0 characteristic value

Normalization Obtain

The following water filling computing of foundation is to access set in the step of the present invention (5)

In user k allocation of transmit power

Wherein λ is the water level of water filling computing, λ〉0, λ satisfies

P _TotBe the total transmitted power in base station, P _MaxBe the maximum transmit power that distribute for single cognitive mobile terminal data stream, symbol

Be defined as:

Wherein

Corresponding x, the corresponding P of b _Max, a=0.

Limit e in the step of the present invention (7) _{S, s+1}(j) the pheromones τ on _{S, s+1}(j) upgrade by the following method:

${\mathrm{\τ}}_{s,s+1}\left(j\right)\←[(1-\mathrm{\ρ})\·{\mathrm{\τ}}_{s,s+1}\left(j\right)+\mathrm{\Δ}{\mathrm{\τ}}_{s,s+1}^{\mathrm{best}}{]}_{{\mathrm{\τ}}_{\min }}^{{\mathrm{\τ}}_{\max }},$

Wherein

The maximum system capacity of finding for ant up to now,

For

Corresponding path, ρ is the pheromones volatility coefficient, and its range of set value is that 0 ~ 1, Q is the constant of adjusting pheromones increment size, and its range of set value is 0.01 ~ 0.1.

Beneficial effect: the spatial reuse vector of the dual orthogonal transmission space multiplexing technique that proposes among the present invention by distributing quadrature to different user can satisfy cognitive Cellular Networks of future generation is disturbed control to PU requirement so that be zero to the interference of PU.Show by experiment, with respect to second best measure commonly used at present, the low complex degree user scheduling method based on ACO that proposes among the present invention can obtain the performance near best practice on the basis that increases certain amount of calculation.

Description of drawings

Below in conjunction with the drawings and specific embodiments the present invention is done further to specify, above-mentioned and/or otherwise advantage of the present invention will become apparent.

Fig. 1 is system of the present invention scene schematic diagram.

Fig. 2 is the corresponding design of graphics of user's scheduling problem.

Fig. 3 a ~ Fig. 3 c be under the different PU numbers the real-time system Capacity Ratio.

Fig. 4 is that the average system capacity under the different PU numbers compares.

Fig. 5 is the inventive method flow chart.

Embodiment

The downlink scenario of the cognitive Cellular Networks of future generation that the present invention considers, as shown in Figure 1, wherein cognitive cellular network is by 1 equipment n _TThe cognitive base station CBS of root antenna and N single antenna CMS to be accessed forms, and also has simultaneously M single antenna primary user PU.For ease of expression, use

With

Represent respectively the index set of CBS and PU.Channel coefficients between CBS and i the CMS is h _i, the channel coefficients between CBS and j the PU is g _jCBS supports that by space multiplexing technique multiplex data stream transfers to different CMS.To i circuit-switched data stream, be sent to the information symbol s of CMSi _iBe carried out transmitted power p _i, pass through subsequently wave beam forming vector w _iProcess.At last, at the transmitting antenna end of CBS, N circuit-switched data stream is sent to different CMS after superposeing.

The dispatching method when low complex degree based on ant group optimization (Ant Colony Optimization, ACO) of the present invention is empty maximizes cognitive Cellular Networks system information speed.The present invention has comprised in fact dual orthogonal transmission method for spacial multiplex that PU zero disturbs and low complex degree user choosing method two partial contents based on ACO.

(1) dual orthogonal transmission spatial reuse

Among the present invention, CBS comes signal transmission to CMS by the wave beam forming technology.Receive the CBS signal for ease of CMS, the selection of wave beam forming vector needs so that the CBS signal that transmits at different CMS links is orthogonal; Simultaneously for PU zero is disturbed, also need so that CBS signal and PU signal in orthogonal, thereby coexist with PU.

Order

The set of the CMS that expression is selected, then

In the beam shape-endowing weight value of k CMS

Selection need satisfy

Wherein

The expression access set

Channel condition information vector between i CMS and the CBS, subscript T represents transposition, subscript H represents conjugate transpose.

For realizing CMS access as much as possible, cognitive Cellular Networks adopts and supports the maximum number of user strategy among the present invention.Because be subjected to the constraint in the formula (1), CBS can support n at most _T-M CMS accesses simultaneously.

Dual orthogonal beams structure: the beam shape-endowing weight value of k CMS can calculate by the following method: CMSk makes up matrix

As follows:

Calculate by feature decomposition

The corresponding characteristic vector of 0 characteristic value

Pass through again normalization

Get final product to such an extent that the beam shape-endowing weight value of k CMS is

Wherein || || the mould of vector is asked in expression.Obviously

Unique condition that satisfies in the formula (1).

Optimum transmit power is distributed: because the CBS transmitted power is limited, CBS need allocation of transmit power to

In different CMS with maximization system information speed, namely find the solution:

Wherein

For giving access set

In the transmitted power of distribution of flows of k CMS,

Be the noise power of receiving terminal, P _MaxFor giving the maximum transmit power of single CMS distribution of flows, P _TotTotal transmitted power for CBS.By KKT(Karush Kuhn Tucker) condition, CBS can come allocation of transmit power by following water filling computing:

Wherein λ〉0 be the water level of water filling computing, the selection of its value need so that

(2) low complex degree user choosing method

Because dual orthogonal transmission spatial reuse can be supported n at most _T-M CMS accesses simultaneously, and in the mobile communication application scene of reality the number of users N of CMS usually greater than n _T-M's, thereby need to from N CMS, select optimum n _T-M user.Order

Represent given access set

The maximum system information rate that Shi Shuanchong orthogonal transmission space multiplexing technique is obtained, i.e. a formula mistake! Do not find Reference source.The optimal value of middle optimization problem.It is following discrete optimization problems of device that optimal user is selected:

Propose among the present invention to solve top problem based on the low complex degree user dispatching algorithm of ACO.In ACO, artificial ant is by moving construction solution at design of graphics (Construction Graph).In each iteration, every ant moves to another summit by the limit of design of graphics from a summit, constantly constructs partial solution.After solution was configured out fully, ant can stay on the limit of process a certain amount of pheromones.The amount of pheromones is relevant with the quality of solution, and the quality of solution is better, and then the quantity of pheromones is larger.Ant in the next iteration is further searched for promising zone and the lastest imformation element of solution space by the guide of pheromones.Make up and being described in further detail below of pheromones renewal process for the solution carried out in each iteration:

● separate structure

The corresponding design of graphics of the problem in the formula (5) as shown in Figure 2.Every limit e _{S, s+1}(j) corresponding to the CMS of an available access, s represents current vertex index, and s+1 represents next vertex index.Each ant is from vertex v ₀Set out, arrive next summit by selecting the limit.For the present invention, in vertex v ₀The time, can supply the number on the limit of ant selection is N, every movement once can supply the number on the limit of ant selection to subtract 1, passes through like this n _TAfter-M time movement, ant has been selected n _T-M bar limit arrives final summit

CMS corresponding to these limits is the solution of problem.In making up the process of separating, ant is selected CMS by a kind of chance mechanism.In vertex v _sThe time, ant selects a limit to arrive vertex v by a kind of mode of probability _S+1, it selects the probability of limit j to be:

Wherein

Expression connect Vertex v _sWith v _S+1The set on all limits, τ _{S, s+1}(j) expression limit e _{S, s+1}(j) the pheromones content on, η _{S, s+1}(j) expression limit e _{S, s+1}(j) the heuristic information value on, α and β are respectively the weight coefficient corresponding to pheromones and heuristic information, and its range of set value is all 0 ~ 10.The larger then algorithm of α value is subjected to the impact of pheromones larger in seeking the process of separating, otherwise the impact of the larger formula information that then is inspired of β is larger.

● pheromones is upgraded

The purpose that pheromones is upgraded is the pheromones content of increase and high-quality solution or potential high-quality decorrelation, reduces simultaneously the pheromones content with decorrelation inferior.The pheromones update rule of method is as follows:

${\mathrm{\τ}}_{\mathrm{ij}}\←[(1-\mathrm{\ρ})\·{\mathrm{\τ}}_{\mathrm{ij}}+\mathrm{\Δ}{\mathrm{\τ}}_{\mathrm{ij}}^{\mathrm{best}}{]}_{{\mathrm{\τ}}_{\min }}^{{\mathrm{\τ}}_{\max }},---\left(7\right)$

Wherein ρ is the pheromones volatility coefficient, and its range of set value is 0 ~ 1, in this scope the then larger so that pheromones in the last iteration of ρ value on this iteration to leave over impact less.τ _MaxWith τ _MinBe respectively the upper bound and the lower bound of pheromones content, its range of set value is respectively 10 ~ 20 and 0 ~ 10,

Be defined as follows:

Wherein Q is for regulating the constant of pheromones increment size, its range of set value is 0.01 ~ 0.1, the Q value is then larger so that the pheromones increase on the limit that the optimal solution that produces in this iteration comprises is faster in this scope, thereby so that algorithm can restrain faster, but separate second-rate; Otherwise algorithmic statement is slower, but the quality of separating is better.

Particularly, as shown in Figure 5, the invention discloses following steps:

Step (1), CBS gathers the channel condition information h between itself and N the CMS _i, i=1,2 ..., N and and M PU between channel condition information g _j, j=1,2 ..., M.

Step (2) arranges τ _Max=10, τ _Min=5, every limit e in the initialization design of graphics _{S, s+1}(j) the upper pheromones τ that distributes _{S, s+1}(j)=(τ _Max+ τ _Min)/2 and heuristic information η _{S, s+1}(j)=| h _j| ², s=0,1 ..., n _T-M, j=1,2 ..., N.

Step (3) generates the only artificial ant of m, and every ant is by following probability selection limit e _{S, s+1}(j):

Record ant n _TThe path that moves through for-M time, the corresponding CMS in the limit in the selecting paths forms access set

Ant is counted m and is set as between the 5-30, weight α, and β is set to α=2, β=1.

Step (4) is calculated CBS as follows corresponding to the dual orthogonal beams figuration vector of different CMS To k CMS, make up matrix

As follows:

Calculate

The corresponding characteristic vector of 0 characteristic value

Again by normalization

Step (5) is pressed the following formula allocation of transmit power Be sent to

In the data flow of CMSk:

λ wherein〉0 selection need so that

Calculate the access set that every ant is selected

Power system capacity as follows:

Selecting access set corresponding to maximum system capacity is optimum access set

Step (6) judges whether to reach maximum iteration time, and maximum iteration time is the natural number greater than 0, is traditionally arranged to be between 5 ~ 50, then stops and exporting if satisfy

Step (7) arranges ρ=0.04, and Q=0.02 upgrades the pheromones on all limits and is back to step (3) by following formula.

${\mathrm{\τ}}_{s,s+1}\left(j\right)\←[(1-\mathrm{\ρ})\·{\mathrm{\τ}}_{s,s+1}\left(j\right)+\mathrm{\Δ}{\mathrm{\τ}}_{s,s+1}^{\mathrm{best}}{]}_{{\mathrm{\τ}}_{\min }}^{{\mathrm{\τ}}_{\max }},$

Wherein

The maximum system capacity of finding for ant up to now,

For

Corresponding path.

Embodiment

Be the performance based on the user choosing method of ACO that checking proposes among the present invention, the below has compared itself and best practice-exhaustive search (Brute Force Search) and the another kind of extensively power system capacity that second best measure-greedy method obtains of employing.For avoiding the interference to PU, above three kinds of methods have all been used the dual orthogonal transmission space multiplexing technique that proposes among the present invention.All n among N the CMS are compared in exhaustive search _TThe power system capacity that the combination of-M CMS obtains is selected corresponding to peaked combination.Greediness method compares the channel quality between CBS and all CMS, selects optimum front n _T-M access, it is widely used in the business communications system, such as 3G1X and Qualcomm HDR.

Shown in Fig. 3 a ~ Fig. 3 c is the obtained real-time system capacity of the lower above three kinds of methods of different PU numbers, wherein among Fig. 3 a, Fig. 3 b, Fig. 3 c respectively corresponding PU number be 1,2 and 3 o'clock situation.Shown in Figure 4 be respectively 1,2 and 3 o'clock above three kinds of average system capacity that method obtains for the PU number.The antenna number that adds up to 50, CBS of setting CMS is 4, and all CMS and PU equate and are normalized to 1 apart from the distance of CBS with season,

With

Be the average received signal to noise ratio of CMS end,

Its definition is respectively

For the inventive method, set the ant number and be 5, PU number and be 1,2 and the maximum iteration time of 3 o'clock the inventive method be set as respectively 25,10 and 5.

Because complexity mainly is positioned at the formula mistake! Do not find Reference source.In target function, thereby the assessment number of times of our based target function comes the complexity of three kinds of methods of comparison.The computation complexity of greediness method, exhaustive search method and the inventive method is compared as follows shown in the table:

The PU number	Greediness method	The exhaustive search method	The inventive method
				3	1	50	5*5=25
2	1	1225	5*10=50
				1	1	19600	5*25=125

By Fig. 3 a ~ Fig. 3 c, Fig. 4 and upper table as can be known, with respect to traditional greedy computational methods, the inventive method can be improved the shortcoming of its poor-performing; And with respect to the exhaustive search method, the inventive method can greatly reduce amount of calculation when obtaining approaching systematic function.The inventive method can effectively reduce conflicting between dispatching method performance and computation complexity, adapts in the cognitive Cellular Networks requirement to the dispatching method performance.

The invention provides a kind of down link of cognitive Cellular Networks dispatching method when empty; method and the approach of this technical scheme of specific implementation are a lot; the above only is preferred implementation of the present invention; should be understood that; for those skilled in the art; under the prerequisite that does not break away from the principle of the invention, can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.In the present embodiment not clear and definite each part all available prior art realized.

Claims (5)

1. dispatching method when the down link of a cognitive Cellular Networks is empty is characterized in that, may further comprise the steps:

Step (1) gathers the vectorial h of channel condition information between the cognitive base station of many antennas and N the single antenna cognition portable terminal _i, i=1,2 ..., the channel condition information vector g between N and cognitive base station and M the single antenna primary user _j, j=1,2 ..., M;

Step (2), every limit e in the initialization design of graphics _{S, s+1}(j) the upper pheromones τ that distributes _{S, s+1}(j)=(τ _Max+ τ _Min)/2, and heuristic information η _{S, s+1}(j)=| h _j| ², s=0,1 ..., n _T-M, j=1,2 ..., N, wherein τ _MaxWith τ _MinBe respectively the upper bound and the lower bound of pheromones content, τ _MaxRange of set value be 10 ~ 20, τ _MinRange of set value be 0 ~ 10, | h _j| channel condition information vector h is asked in expression _jAmplitude, n _TNumber of antennas for cognitive base station;

Step (3) generates the only artificial ant of m, and wherein the range of set value of m is 5 ~ 30, places ant vertex v in design of graphics ₀The place, every ant is by probability P r (e _{S, s+1}(j)) select limit e _{S, s+1}(j) from vertex v _sMove to vertex v _S+1, s=0,1..., n _T-M; Record ant n _TThe path that moves through for-M time, the corresponding cognitive portable terminal in the limit in the selecting paths forms access set

Step (4), cognitive base station are calculated the dual orthogonal beams figuration vector corresponding to k cognitive portable terminal

Step (5), allocation of transmit power are sent to the data flow of cognitive portable terminal for cognitive base station; Calculate the access set that every ant is selected

Power system capacity As follows:

Wherein,

For giving access set In the transmitted power of distribution of flows of k CMS,

The expression access set

Channel condition information vector between k cognitive portable terminal and the cognitive base station,

Expression is to vector

Carry out matrix transpose operation, Noise power for the receiver of cognitive portable terminal; Selecting access set corresponding to maximum system capacity is optimum access set Step (6) judges whether the maximum iteration time that reaches default, then stops and exporting optimum access set if reach

Maximum iteration time is got natural number;

Step (7) is upgraded the pheromones that distributes on all limits in the design of graphics and is returned step (3); Scheduling when finishing thus the down link sky.

2. dispatching method when the down link of cognitive Cellular Networks according to claim 1 is empty is characterized in that, ant is selected limit e in the step (3) _{S, s+1}(j) probability P r (e _{S, s+1}(j)) calculate by the following method:

Wherein

Expression connect Vertex v _sWith vertex v _S+1The set on all limits, α and β are the weight coefficient corresponding to pheromones and heuristic information, the range of set value of α and β is 0 ~ 10.

3. dispatching method when the down link of cognitive Cellular Networks according to claim 1 is empty is characterized in that, calculates the dual orthogonal beams figuration vector of k cognitive base station in the step (4) according to following methods

To k cognitive base station, make up following matrix

Compute matrix

The corresponding characteristic vector of 0 characteristic value Normalization

Obtain

4. dispatching method when the down link of cognitive Cellular Networks according to claim 1 is empty is characterized in that, the following water filling computing of foundation is to access set in the step (5)

In user k allocation of transmit power

Wherein λ is the water level of water filling computing, λ〉0, λ satisfies

P _TotBe the total transmitted power in base station, P _MaxBe the maximum transmit power that distribute for single cognitive mobile terminal data stream, symbol Be defined as:

Wherein Corresponding x, the corresponding P of b _Max, a=0.

5. dispatching method when the down link of cognitive Cellular Networks according to claim 1 is empty is characterized in that, limit e in the step (7) _{S, s+1}(j) the pheromones τ on _{S, s+1}(j) upgrade by the following method:

${\mathrm{\τ}}_{s,s+1}\left(j\right)\←[(1-\mathrm{\ρ})\·{\mathrm{\τ}}_{s,s+1}\left(j\right)+\mathrm{\Δ}{\mathrm{\τ}}_{s,s+1}^{\mathrm{best}}{]}_{{\mathrm{\τ}}_{\min }}^{{\mathrm{\τ}}_{\max }},$

Wherein

The maximum system capacity of finding for ant up to now,

For

Corresponding path, ρ is the pheromones volatility coefficient, and its range of set value is that 0 ~ 1, Q is the constant of adjusting pheromones increment size, and its range of set value is 0.01 ~ 0.1.

Images