CN106211198A - Inferior obliqued overaction and the method for community attachment is combined under a kind of cloud Radio Access Network - Google Patents

Abstract

The invention discloses and under a kind of cloud Radio Access Network, combine Inferior obliqued overaction and the method for community attachment, belong to moving communicating field；Specifically include: step one, for certain descending cloud wireless access network, set up user, the system model of RRH and BBU pool；Step 2, the network information of BBU pool sensory perceptual system model；Step 3, each user, according to nearby principle, send service request by RRH to BBU pool respectively；Step 4, BBU pool are according to service request type, and the most each user sets minimum SINR thresholding；And measure each RRH channel information to each user；Step 5, BBU pool, according to minimum SINR thresholding, channel information and the network information of perception, use heuritic approach to calculate Inferior obliqued overaction vector and community attachment.Advantage is: associating Inferior obliqued overaction and community attachment can effectively reduce network energy consumption, reach energy-conservation target, ensure the service quality of user simultaneously.

Description

Inferior obliqued overaction and the method for community attachment is combined under a kind of cloud Radio Access Network

Technical field

The invention belongs to moving communicating field, be specifically related under a kind of cloud Radio Access Network to combine Inferior obliqued overaction and The method of community attachment.

Background technology

Cloud wireless access network (Cloud Radio Access Network, C-RAN) is the strong candidate network architecture of 5G, As it is shown in figure 1, by traditional base station is separated into Remote Radio Unit (RRH, Remote Radio closer to the user Head) and the baseband processing unit pond (BBU pool, Baseband Processing Unit pool) that concentrates in together, its In, RRH is responsible for sending, to user, the data that BBU pool processes, or the data receiving user transfer to BBU pool process；BBU Pool is the center processing unit in C-RAN, is responsible for the work such as the collection of information, process；Achieve the centralized place of base-band information Reason, cooperation radio and real-time cloud calculate, thus are greatly improved system spectral efficiency, reduction cost, shared processing resources, Improve infrastructure utilization rate.

But, along with increase and the continuous lifting of user's request of number of users, need to dispose substantial amounts of RRH so that Network energy consumption (including transmitting energy consumption and the circuit energy consumption of RRH) sharply increases.The transmitting energy consumption of RRH is by Inferior obliqued overaction vector Determining, the circuit energy consumption (i.e. RRH is to activate or dormancy) of RRH is closely related with community attachment state.Additionally, use along with mobile The increase at family, due to fronthaul, (forward link, connects the link between BBU pool and RRH, is responsible for carrying BBU pool And the baseband signal between RRH) in transmission be digital baseband signal, quantity of information is relatively big, and therefore, fronthaul capacity constraint becomes For restricting a key factor of C-RAN performance boost.

It should be noted that the quantity of information in fronthaul is relevant to the user that this RRH services, Ji You community attachment shape State determines.It is therefore desirable to combine consideration Inferior obliqued overaction and community attachment in the C-RAN that fronthaul is limited, to reduce Network energy consumption.

Document 1: for the group rarefaction wave beam forming method of green cloud wireless access network, by minimizing weighting l₁/l₂Model Number, it is proposed that a kind of united beam sparse based on group shapes and community adherence method, to reduce the network energy consumption of C-RAN；But It is, the problem failing to consider fronthaul capacity constraint so that the program cannot be applied in the network of reality；

Document 2: the united beam for the limited cloud wireless access network of forward link shapes and Access Control scheme, proposes A kind of compromise proposal launching energy consumption and fronthaul transmission quantity of information, and propose a kind of based on weighting l₁Opening of norm Hairdo derivation algorithm；The program is only absorbed in reduction transmission energy consumption, does not considers circuit Energy Consumption Factors, i.e. community attachment state, because of This, the program cannot really effectively reduce network energy consumption.

Summary of the invention

The present invention is in order to preferably reduce transmitting energy consumption and the circuit energy consumption of RRH, and considers in realistic communication network There is the factor of fronthaul capacity limit, propose to combine Inferior obliqued overaction under a kind of cloud Radio Access Network and community is attached The method.

Specifically comprise the following steps that

Step one, for certain descending cloud wireless access network, set up user, the system model of RRH and BBU pool；

System model includes: N number of RRH, I user and a BBU pool；Each RRH is respectively by one Fronthaul link connects in BBU pool；Each RRH provides service at least one user or does not takes for any user Business, an at least corresponding RRH of user；

RRH collection be combined into 1,2 ..., n ..., N}；User collection be combined into 1,2 ..., i ..., I}；N number of fronthaul chain The capacity set on road is combined into { C₁,C₂,...,C_n,...,C_N}。

Step 2, the network information of BBU pool sensory perceptual system model；

The network information includes the quantity of RRH, and the maximum number of user that fronthaul link corresponding to each RRH can service Amount；

Each fronthaul link is different according to medium, and the maximum number of user amount that can service is different.

Step 3, each user, according to nearby principle, send service request by RRH to BBU pool respectively.

Step 4, BBU pool are according to service request type, and the most each user sets minimum SINR thresholding；And survey Measure each RRH channel information to each user；

h_n,iRepresent n-th RRH channel information to i-th user.

Step 5, BBU pool, according to minimum SINR thresholding, channel information and the network information of perception, use heuristic Algorithm calculates Inferior obliqued overaction vector and community attachment.

Specifically comprise the following steps that

N number of RRH is initialized by step 501, BBU pool；

Initialization includes: each RRH is active, and provides service for all of user；It is in connection time initial The RRH-user of state is to V；And network lowest energy consumption F time initial^(min)And the network beamforming vectors of correspondence And meet F^(min)=+∞,

Particularly as follows:

N-th RRH is active, and is designated as a_n=1；Otherwise, RRH in a dormant state, is designated as a_n=0；

Each RRH provides service for all of user: b_n,i=1 represents that the n-th RRH is that user i transmits data；b_n,i=0 Represent that the n-th RRH does not transmits data for user i；

RRH-user toRepresent；

Network energy consumption F (Φ) is expressed as:

It it is circuit energy consumption during the n-th RRH resting state；It it is circuit energy consumption during the n-th RRH state of activation；η_nIt it is the efficiency of RF power amplifier of the n-th RRH；Represent the transmitting power maximum of the n-th RRH； w_n,iIt it is the beamforming vectors of user i in the n-th RRH；WhereinRepresent vector w_n,iEuclid norm square；

All of RRH-user is according to priority carried out ascending order arrangement by step 502, BBU pool to V；

Priority β of user i in n-th RRH_n,iFormula is calculated as follows:

${\mathrm{\β}}_{n,i}=\frac{|h_{n,i}^H{w}_{n,i}^i{|}^2}{\underset{j\≠i}{\Σ}|h_{n,j}^H{w}_{n,i}^i{|}^2}\·\frac{C_n{\mathrm{\η}}_n}{P_n^a}$

Represent channel information h_n,iConjugate transpose；Represent lax beamforming vectors；Represent RRH-user to (n, available signal power i),Represent that RRH-user is to (other all users i) are done by n Disturb sum；Illustrate the situation of the n-th RRH.

Lax beamforming vectorsMeet and make network energy consumption F (Φ) minimum, and meet following constraints:

$\underset{\mathrm{\Φ}=\{a_n,b_{n,i},w_{n,i},\∀n\∈N,\∀i\∈I\}}{min}F\left(\mathrm{\Φ}\right)$

$s.t.\underset{i=1}{\overset{I}{\Σ}}\left|\right|w_{n,i}|{|}^2\≤P_n^{\max },n\∈N---\left(S1\right)$

SINR_i≥γ_i,i∈I (S2)

$\underset{i=1}{\overset{I}{\Σ}}{b}_{n,i}^{\left(k\right)}\≤a_n^{\left(k\right)}{C}_n,\∀n\∈N---\left(S3\right)$

$\left|\right|w_{n,i}|{|}_2\≤b_{n,i}\sqrt{P_n^{max}},\∀n\∈N,\∀i\∈I---\left(S4\right)$

$a_n\∈\[0,1\],b_{n,i}\∈\[0,1\],\∀n\∈N,\∀i\∈I---\left(S5\right)$

Constraints (S1) represents: in the n-th RRH, all users' launches the power transmitting merit less than or equal to the n-th RRH Rate maximum；

Constraints (S2) represents that the Signal to Interference plus Noise Ratio of i-th user is the minimum of this user setting more than or equal to BBU Pool Threshold value γ_i；

Constraints (S3) represents that the number of users that the n-th RRH is serviced is less than or equal to the n-th fronthaul link Capacity；

Constraints (S4) represents that the n-th RRH is less than the transmitting power of the n-th RRH to the power of launching of i-th user Maximum；

Constraints (S5) represents variable a_n,b_n,iValue is the real number between 0 to 1；

V is iterated solving by the RRH-user after ascending order is arranged by step 503, BBU pool, obtains final network energy Consumption F (Φ) and beamforming vectors w_n,i；

Specifically comprise the following steps that

Step 5031, judge whether RRH-user is empty set to V, if it is, iteration terminates, otherwise enter step 5032；

Step 5032, first the RRH-user chosen in set V are to (p q), makes b_p,q=0, remaining RRH-user simultaneously Corresponding b value is kept constant；

Step 5033, according to the b of RRH-user couple in set V_p,qThe a of correspondence is set_pValue, and RRH-user in V Deletion element (p, q)；

Step 5034, judge whether each RRH meets constraints:If be unsatisfactory for, currently change Generation number k=k+1, returns step 5031 and carries out next iteration；Otherwise, step 5035 is entered；

Step 5035, according to all b values of current all RRH-users couple and a value of correspondence, calculate kth time iteration Network energy consumption network energy consumption F^(k)And the beamforming vectors of correspondence

$\underset{\mathrm{\Φ}=\{a_n,b_{n,i},w_{n,i},\∀n\∈N,\∀i\∈I\}}{min}F\left(\mathrm{\Φ}\right)$

$s.t.\underset{i=1}{\overset{I}{\Σ}}\left|\right|w_{n,i}|{|}^2\≤P_n^{\max },n\∈N---\left(S1\right)$

The constraints met is: SINR_i≥γ_i,i∈I (S2)

$\underset{i=1}{\overset{I}{\Σ}}{b}_{n,i}\≤a_n{C}_n,\∀n\∈N---\left(S3\right)$

$\left|\right|w_{n,i}|{|}_2\≤b_{n,i}\sqrt{P_n^{max}},\∀n\∈N,\∀i\∈I---\left(S4\right)$

Step 5036, judge network energy consumption F^(k)Whether less than F^(min), if it is, F^(k)For current lowest energy consumption, record F^min=F^(k),Enter step 5037；Otherwise, b is set_p,q=1, a_p=1, and return step 5031.

Step 5037 is until by complete to whole iteration for the RRH-user in set V, obtaining final F (Φ) and right The beamforming vectors w answered_n,i；

The value of final F (Φ) is F^(min), and the beamforming vectors w of correspondence_n,iValue be

Step 504, for the n-th RRH, C-RAN according to iteration beamforming vectors w_n,i, calculate the transmission data of this RRH x_n；

$x_n=\underset{i=1}{\overset{I}{\Σ}}{w}_{n,i}{s}_i,\∀n\∈N$

s_iRepresent the data of user i request；

Step 505, fronthaul link are by data x_nBeing transferred to user corresponding to RRH, user obtains request data s_i；

Step 506, user community are adhered to according to RRH, ensure that system model energy consumption is minimum in conjunction with beamforming vectors.

It is an advantage of the current invention that:

1), Inferior obliqued overaction and the method for community attachment, associating downlink wave beam are combined under a kind of cloud Radio Access Network Shape and community attachment can effectively reduce network energy consumption, reach energy-conservation target, ensure the service quality of user simultaneously.

2), Inferior obliqued overaction and the method for community attachment are combined under a kind of cloud Radio Access Network, abundant when design Consider limited, the fronthaul capacity constraint of the problem being likely encountered in reality, such as peak power etc., it is easy in reality Network is applied.

3), Inferior obliqued overaction and the method for community attachment are combined under a kind of cloud Radio Access Network, can be at multinomial Completing in time to solve, processing speed is fast.

Accompanying drawing explanation

Fig. 1 is to combine Inferior obliqued overaction and the system model of community attachment under the present invention a kind of cloud Radio Access Network Figure；

Fig. 2 is to combine Inferior obliqued overaction and the method flow of community attachment under the present invention a kind of cloud Radio Access Network Figure；

Fig. 3 is that BBU pool of the present invention calculates Inferior obliqued overaction vector and the method flow diagram of community attachment；

Fig. 4 is the method flow diagram that the iterative that BBU pool of the present invention carries out tightening obtains beamforming vectors；

Fig. 5 is the present invention with averaging network energy consumption under LTE-A algorithm with γ_iVariation diagram；

Fig. 6 is that the present invention averagely activates RRH quantity with γ under LTE-A algorithm_iVariation diagram.

Detailed description of the invention

Below in conjunction with accompanying drawing, the present invention is described in further detail.

In existing network energy consumption, circuit energy consumption account for significant proportion, therefore necessary when considering network energy-saving problem Introduce community attachment techniques, to control the state (activating or dormancy) of base station (or RRH).Compared to existing scheme, the present invention Propose a kind of associating Inferior obliqued overaction and the method for community attachment, first introduce network system model, then describe Associating Inferior obliqued overaction and the concrete grammar of community attachment, finally provide the performance simulation of the method；Send out to effectively reduce Penetrate energy consumption and circuit energy consumption, take into account the QoS requirement of user, maximum transmission power restriction and fronthaul simultaneously Capacity limit, effectively reduces network energy consumption；And propose a kind of based on the low complex degree method for solving tightened, can be multinomial Complete to solve in the formula time, accelerate processing speed.

As in figure 2 it is shown, specifically comprise the following steps that

Step one, for certain descending cloud wireless access network, set up user, the system model of RRH and BBU pool；

As it is shown in figure 1, descending C-RAN system model includes: N number of multiple antennas RRH, I single-antenna subscriber and a BBU pool；Each RRH is respectively connected in BBU pool by a fronthaul link；During each RRH state of activation permissible Servicing one or more user, be not any user service during resting state, a user at least receives the service of a RRH；

RRH collection be combined into 1,2 ..., n ..., N}；User collection be combined into 1,2 ..., i ..., I}；N number of fronthaul chain The capacity set on road is combined into { C₁,C₂,...,C_n,...,C_N}；Each capacity is the number of users of each fronthaul link carrying, User's maximum number of the most each RRH service, is constant.

Wherein, the n-th RRH has m_nRoot antenna, definitionIt it is the beamforming vectors of user i in the n-th RRH；

Step 2, the network information of BBU pool sensory perceptual system model；

The network information includes the quantity of RRH, and the maximum number of user that fronthaul link corresponding to each RRH can service Amount；

Each fronthaul link is different according to medium, and the maximum number of user amount that can service is different.

Owing to user's baseband signal data amount is relatively big, the use that therefore can be carried by the fronthaul of BBU pool to RRH Family information is limited.The number of users serviced by RRH is as the load of fronthaul.Assume the fronthaul of the n-th RRH At most service C_nIndividual user, then the fronthaul capacity-constrained of the n-th RRH is expressed as:

Step 3, each user, according to nearby principle, send service request by RRH to BBU pool respectively.

Step 4, BBU pool are according to service request type, and the most each user sets minimum SINR thresholding；And survey Measure each RRH channel information to each user；

h_n,iRepresent n-th RRH channel information to i-th user.

Assume that each user data noise is separate, then the received signal to noise ratio of user i is expressed as:

${\mathrm{SINR}}_i=\frac{|\underset{n=1}{\overset{N}{\Σ}}{h}_{n,i}^H{w}_{n,i}{|}^2}{\underset{j\∈I,j\≠i}{\Σ}|\underset{n=1}{\overset{N}{\Σ}}{h}_{n,i}^H{w}_{n,j}{|}^2+{\mathrm{\σ}}_i^2},\∀i\∈I$

Represent channel information h_n,iConjugate transpose；w_n,iIt it is the beamforming vectors of user i in the n-th RRH；w_n,jFor The beamforming vectors of user j in n-th RRH；For noise power, it is a constant in the case of bandwidth determines.

BBU Pool is the minimum threshold that sets of user i as γ_i；γ_iMinimum receiving terminal for i-th user believes dry making an uproar Ratio, is determined by the apllied type of service of this user, is a constant.

The guarantee of the service needed service quality of user's request, by the qos constraint SINR of user i_iMore than threshold value γ_i, then qos constraint is expressed as: SINR_i≥γ_i,i∈I。

Step 5, BBU pool, according to minimum SINR thresholding, channel information and the network information of perception, use heuristic Algorithm calculates Inferior obliqued overaction vector and community attachment.

As it is shown on figure 3, specifically comprise the following steps that

N number of RRH is initialized by step 501, BBU pool；

Initialization includes: each RRH is active, and provides service for all of user；It is in connection time initial The RRH-user of state is to V；And network lowest energy consumption F time initial^(min)And the network beamforming vectors of correspondence And meet F^(min)=+∞,F when understanding initial^(min)Solution more than or equal to any an iteration.

Particularly as follows:

a_nRepresent the state of the n-th RRH；When the n-th RRH is active, it is designated as a_n=1；Otherwise, RRH is in dormancy State, is designated as a_n=0；

Each RRH provides service for all of user；b_n,iRepresent the connection status of the n-th RRH and user i；b_n,i=1 table Show that the n-th RRH is that user i transmits data；b_n,i=0 represents that the n-th RRH does not transmits data for user i；

a_n, b_n,i, w_n,iThere is following logical relation: a_n=0 b_n,i=0 necessarily sets up；b_n,i=0 w_n,i=0 necessarily becomes Vertical.

RRH-user toRepresent；

Network energy consumption F (Φ) is expressed as:Φ refers to the set of variable:

It it is circuit energy consumption during the n-th RRH resting state；It it is circuit energy consumption during the n-th RRH state of activation；

Then the energy consumption of the n-th RRH is expressed as:

$P_n=a_n(P_n^a+\frac{1}{{\mathrm{\η}}_n}\underset{i=1}{\overset{I}{\Σ}}|\left|w_{n,i}\right|{|}_2^2)+(1-a_n)P_n^s=P_n^s+a_n{P}_n^{ams}+\frac{1}{{\mathrm{\η}}_n}\underset{i=1}{\overset{I}{\Σ}}\left|\right|w_{n,i}|{|}_2^2$

η_nIt it is the efficiency of RF power amplifier of the n-th RRH；It is n-th Total energy consumption when RRH is active,Be the n-th RRH in a dormant state time total energy consumption；

Represent the transmitting power maximum of the n-th RRH；w_n,iIt it is the beamforming vectors of user i in the n-th RRH；Its InRepresent vector w_n,iEuclid norm square；

For the n-th RRH, launch power maximumConstraint representation is following formula:

$\underset{i=1}{\overset{I}{\Σ}}\left|\right|w_{n,i}|{|}^2\≤P_n^{max},n\∈N$

All of RRH-user is according to priority carried out ascending order arrangement by step 502, BBU pool to V；

Priority β of user i in n-th RRH_n,iFormula is calculated as follows:

${\mathrm{\β}}_{n,i}=\frac{|h_{n,i}^H{w}_{n,i}^r{|}^2}{\underset{j\≠i}{\Σ}|h_{n,j}^H{w}_{n,i}^r{|}^2}\·\frac{C_n{\mathrm{\η}}_n}{P_n^a}$

Represent channel information h_n,iConjugate transpose；Represent lax beamforming vectors；Represent RRH-user to (n, available signal power i),Represent that RRH-user is to (other all users i) are done by n Disturb sum,Illustrate the situation of the n-th RRH.Obviously, useful signal is the strongest, disturbs the most weak, the fronthaul capacity of RRH The biggest, efficiency of RF power amplifier is the highest, and the lowest RRH-user of self circuit energy consumption is to having higher priority； Then, to RRH-users all in V to according to priority carrying out ascending order arrangement.

Lax beamforming vectorsMeet and make network energy consumption F (Φ) minimum, and meet following constraints:

$\underset{\{a_n,b_{n,i},w_{n,i},\∀n\∈N,\∀i\∈I\}}{min}F\left(\mathrm{\Φ}\right)$

$s.t.\underset{i=1}{\overset{I}{\Σ}}\left|\right|w_{n,i}|{|}^2\≤P_n^{\max },n\∈N---\left(S1\right)$

SINR_i≥γ_i,i∈I (S2)

$\underset{i=1}{\overset{I}{\Σ}}{b}_{n,i}^{\left(k\right)}\≤a_n^{\left(k\right)}{C}_n,\∀n\∈N---\left(S3\right)$

$\left|\right|w_{n,i}|{|}_2\≤b_{n,i}\sqrt{P_n^{max}},\∀n\∈N,\∀i\∈I---\left(S4\right)$

$a_n\∈\[0,1\],b_{n,i}\∈\[0,1\],\∀n\∈N,\∀i\∈I---\left(S5\right)$

Constraints (S1) represents: in the n-th RRH, all users' launches the power transmitting merit less than or equal to the n-th RRH Rate maximum, whereinRepresent in the n-th RRH the Euclid norm square of the beamforming vectors of all users it With；

Constraints (S2) represents that the Signal to Interference plus Noise Ratio of i-th user is the minimum of this user setting more than or equal to BBU Pool Threshold value γ_i；By in the qos constraint expression formula of the energy consumption formula band access customer i of the n-th RRH, obtain the service of user i Quality constraint result is as follows:

$\sqrt{\underset{j\≠i}{\Σ}|\underset{n=1}{\overset{N}{\Σ}}{h}_{n,i}^H{w}_{n,j}{|}^2+{\mathrm{\σ}}_i^2}\≤\frac{1}{\sqrt{{\mathrm{\γ}}_i}}\mathrm{Re}\left\{\underset{n=1}{\overset{N}{\Σ}}{h}_{n,i}^H{w}_{n,i}\right\},\∀i\∈I$

$\mathrm{Im}\left\{\underset{n=1}{\overset{N}{\Σ}}{h}_{n,i}^H{w}_{n,i}\right\}=0,\∀i\∈I$

Constraints (S3) represents that the number of users that the n-th RRH is serviced is less than or equal to the n-th fronthaul link Capacity；

Constraints (S4) represents that the n-th RRH is less than the transmitting power of the n-th RRH to the power of launching of i-th user Maximum；

Constraints (S3) and (S4) ensure that binary variable a_n, b_n,i, w_n,iThere is following logical relation: if n-th In a dormant state, then the n-th RRH can not transmit data for any user to RRH；If the n-th RRH does not sends number to user i According to, then corresponding beamforming vectors is 0.

Constraints (S5) represents variable a_n,b_n,iValue is the real number between 0 to 1；

The CRAN network energy consumption minimization problem modeling of physical constraint, generally one MIXED INTEGER second order cone rule will be considered Drawing (MI-SOCP:Mixed-Integer Second-Order Cone Programming), this problem is a NP-Hard (non-deterministic polynomial hard) problem, i.e. cannot try to achieve the optimal solution of problem in polynomial time, Solving complexity can sharply increase along with the lifting of problem variables number.

By binary variable a_n,b_n,iSpan to relax be the real number between 0 to 1, the theory of convex optimization understand this and ask An entitled convex problem, uses interior point method to solve SOCP Second-order cone programming, belongs to convex optimization problem, try to achieve in polynomial time The suboptimal solution of problem；

The iterative that V is tightened by step 503, BBU pool by the RRH-user after ascending order arrangement, obtains wave beam Shape vector；

First, it is assumed that each RRH is active, and provide service for all of user.Then, according to certain Criterion carries out ascending order arrangement to RRH-user to (i.e. this RRH is this user service).Then, iteration remove RRH-from front to back User couple, even this RRH no longer services for this user, and records the result of each iteration.Until finally removing all of RRH- User couple, the most all RRH are the most no longer any user service.

As shown in Figure 4, specifically comprise the following steps that

Step 5031, judge whether RRH-user is empty set to V, if it is, iteration terminates, otherwise enter step 5032；

Step 5032, first the RRH-user chosen in set V are to (p q), makes b_p,q=0, gather its of V with season The b of remaining element keeps constant；

State of activation according to RRH and the user provided the user, the user couple that in set of computations V, priority level is minimum, Namely first RRH-user to (p, q)；

OrderWherein k is current iteration number of times, chooses first RRH-user couple in set V (p, q), order

Step 5033, according to the b of RRH-user couple in set V_p,qThe a of correspondence is set_pValue, and RRH-user in V Deletion element (p, q)；

Step 5034, judge whether each RRH meets constraints:If be unsatisfactory for, make k=k+ 1, return step 5031 and carry out next iteration；Otherwise, step 5035 is entered；

Step 5035, according to all b values of current all RRH-users couple and a value of correspondence, calculate kth time iteration Network energy consumption network energy consumption F^(k)And the beamforming vectors of correspondence

$\underset{\mathrm{\Φ}=\{a_n,b_{n,i},w_{n,i},\∀n\∈N,\∀i\∈I\}}{min}F\left(\mathrm{\Φ}\right)$

$s.t.\underset{i=1}{\overset{I}{\Σ}}\left|\right|w_{n,i}|{|}^2\≤P_n^{\max },n\∈N---\left(S1\right)$

The constraints met is: SINR_i≥γ_i,i∈I (S2)

$\underset{i=1}{\overset{I}{\Σ}}{b}_{n,i}^{\left(k\right)}\≤a_n^{\left(k\right)}{C}_n,\∀n\∈N---\left(S3\right)$

$\left|\right|w_{n,i}|{|}_2\≤b_{n,i}\sqrt{P_n^{max}},\∀n\∈N,\∀i\∈I---\left(S4\right)$

Step 5036, judge network energy consumption F^(k)Whether less than F^(min), if it is, F^(k)For current lowest energy consumption, calculate Corresponding beamforming vectors isF^(min)=F^(k),Enter step 5037；Otherwise, arrange And return step 5031；

F^(min)For from network energy consumption minimum known to the 1st time to kth iteration；For from the 1st time to kth iteration The network beamforming vectors that known minimal network energy consumption is corresponding.

Step 5037 is until by complete to whole iteration for the RRH-user in set V, obtaining final F (Φ) and right The beamforming vectors w answered_n,i；

The value of final F (Φ) is F^(min), and the beamforming vectors w of correspondence_n,iValue be

Step 504, for the n-th RRH, C-RAN is according to iteration result w_n,iCalculate transmission data x of this RRH_n；

$x_n=\underset{i=1}{\overset{I}{\Σ}}{w}_{n,i}{s}_i,\∀n\∈N$

s_iRepresenting the data of user i request, average is 0, and variance is 1.；

Step 505, fronthaul link are by data x_nBeing transferred to user corresponding to RRH, user obtains request data s_i；

Due to the known whole channel informations of BBU pool The reception signal table of user i It is shown as

$y_i=\underset{n=1}{\overset{N}{\Σ}}{h}_{n,i}^H{w}_{n.i}{s}_i+\underset{j\∈I,j\≠i}{\Σ}\underset{n=1}{\overset{N}{\Σ}}{h}_{n,i}^H{w}_{n,j}{s}_j+z_i,\∀i\∈I$

WhereinFor useful signal,For interference, z_iFor additive white Gaussian noise.

Step 506, user community are adhered to according to RRH, ensure that system model energy consumption is minimum in conjunction with beamforming vectors.

Iterative algorithm example:

Assuming to exist in network 3 RRH, 3 users, each RRH can service 2 users, i.e. C_n=2, n=1,2,3. The ascending result of calculation of priority of RRH-user couple is as follows: (1,2), (3,2), (2,3), (1,3), (1,1), (3,1), (2,1), (2,2), (3,3), therefore V={ (1,2), (3,2), (2,3), (1,3), (1,1), (3,1), (2,1), (2,2), (3 3)}。

For the 1st iteration:

1), RRH-user be not empty set to V, carry out next step；

2) the minimum user of priority level, is selected to (1,2), renewalB with remaining element of set in season V It is set to 1；I.e.Arrange simultaneouslyDelete element (1,2), obtain V={ (3,2), (2,3), (1,3), (1,1), (3,1), (2,1), (2,2),(3,3)}；

3), judge the 2nd, 3 RRH, be all unsatisfactory forBecause limiting each RRH can service 2 users, return Return step 1) carry out the 2nd iteration；

4) the minimum user of priority level, is selected to (3,2), renewalSet with the b of remaining element of set in season V It is set to 1；I.e.Arrange simultaneously Delete element (3,2), obtain V={ (2,3), (1,3), (1,1), (3,1), (2,1), (2,2), (3,3) }；

5), judge the 2nd RRH, be unsatisfactory forBecause limiting each RRH can service 2 users, return step Rapid 1) the 3rd iteration is carried out；

6) the minimum user of priority level, is selected to (2,3), renewalB with remaining element of set in season V It is set to 1；I.e.Arrange simultaneouslyDelete unit Element (2,3), obtains V={ (1,3), (1,1), (3,1), (2,1), (2,2), (3,3) }；

7), judge each RRH, be satisfied byBecause limiting each RRH can service 2 users, knownUnder conditions of, Solve problems, it is assumed that Xie WeiIf F⁽³⁾≤F^(min), then F^(min)=F⁽³⁾For current lowest energy consumption, Corresponding beamforming vectors isIf problem is without solving, or has solution, but F⁽³⁾＞ F^(min), then make Return step 1), carry out next iteration.

8), be not empty set for the 4th iteration: V, carry out next step；UpdateArrange simultaneously Delete element (1,3), obtains V={ (1,1), (3,1), (2,1), (2,2), (3,3) }；All RRH are satisfied byKnownUnder conditions of, Solve problems, it is assumed that Xie WeiIf F⁽⁴⁾≤F^(min), then F^(min)=F⁽⁴⁾For current mental retardation Consumption, corresponding beamforming vectors isIf problem is without solving, or has solution, but F⁽⁴⁾>F^(min), then make Return step 1), carry out next iteration.

9), be not empty set for the 5th iteration: V, carry out next step；UpdateArrange simultaneously Delete element (1,1), obtains V={ (3,1), (2,1), (2,2), (3,3) }；All RRH are satisfied byKnownUnder conditions of, Solve problems assumes that solution is set toIf F⁽⁵⁾≤F^(min), then F^(min)=F⁽⁵⁾For current mental retardation Consumption, corresponding beamforming vectors isIf problem is without solving, or has solution, but F⁽⁵⁾＞ F^(min), then make Return step 1), carry out next iteration.

Until proceeding to the 9th iteration, the final result obtained is F^(min), corresponding beamforming vectors is

The present invention is suggested plans and is carried out performance simulation compare with LTE-A suggests plans, the geographical position of each RRH of the latter Being divided into disjoint bunch, user in the range of it of bunch interior RRH cooperation sends data.Emulation uses Rayleigh fading letter Road model, system bandwidth is 10MHz, disposes 10 2 antennas in the range of [-1500m, 1500m] × [-1500m, 1500m] The maximum transmission power of RRH, RRH is 10W, η_n=0.25.

As it is shown in figure 5, the present invention carried algorithm network energy consumption is less than the used algorithm of LTE-A, this is because can be overlapping RRH cooperation mode can preferably reduce interference, thus ensures the service quality of user in the case of using relatively low energy consumption.This Outward, number of users the most energy consumption also be can be observed the biggest, reason is to need higher transmitting power and more activate RRH.

As shown in Figure 6, in the used algorithm of LTE-A, all RRH are all activated, and the carried algorithm of the present invention can be On the premise of ensureing same QoS of customer, dormancy part RRH, thus networking energy consumption is greatly reduced.

The novelty of the present invention is that not only allowing for beam shaping also contemplates community attachment, can be greatly reduced network Energy expenditure.

Claims (3)

1. combine Inferior obliqued overaction and the method for community attachment under a cloud Radio Access Network, it is characterised in that specifically walk Rapid as follows:

Step one, for certain descending cloud wireless access network, set up user, the system model of RRH and BBU pool；

Step 2, the network information of BBU pool sensory perceptual system model；

The network information includes the quantity of RRH, and the maximum number of user amount that fronthaul link corresponding to each RRH can service；

Step 3, each user, according to nearby principle, send service request by RRH to BBU pool respectively；

Step 4, BBU pool are according to service request type, and the most each user sets minimum SINR thresholding；And measure every Individual RRH is to the channel information of each user；

h_n,iRepresent n-th RRH channel information to i-th user；

Step 5, BBU pool, according to minimum SINR thresholding, channel information and the network information of perception, use heuritic approach Calculate Inferior obliqued overaction vector and community attachment；

Specifically comprise the following steps that

N number of RRH is initialized by step 501, BBU pool；

Initialization includes: each RRH is active, and provides service for all of user；It is in connection status time initial RRH-user to V；And network lowest energy consumption F time initial^(min)And the network beamforming vectors of correspondenceAnd it is full Foot F^(min)=+∞,

Particularly as follows:

N-th RRH is active, and is designated as a_n=1；Otherwise, RRH in a dormant state, is designated as a_n=0；

Each RRH provides service for all of user: b_n,i=1 represents that the n-th RRH is that user i transmits data；b_n,i=0 represents N-th RRH does not transmits data for user i；

RRH-user toRepresent；

Network energy consumption F (Φ) is expressed as:

It it is circuit energy consumption during the n-th RRH resting state；It it is circuit energy consumption during the n-th RRH state of activation；η_nIt it is the efficiency of RF power amplifier of the n-th RRH；Represent the transmitting power maximum of the n-th RRH； w_n,iIt it is the beamforming vectors of user i in the n-th RRH；WhereinRepresent vector w_n,iEuclid norm square；

All of RRH-user is according to priority carried out ascending order arrangement by step 502, BBU pool to V；

Priority β of user i in n-th RRH_n,iFormula is calculated as follows:

${\mathrm{\β}}_{n,i}=\frac{|h_{n,i}^H{w}_{n,i}^r{|}^2}{\underset{j\≠i}{\Σ}|h_{n,j}^H{w}_{n,i}^r{|}^2}\·\frac{C_n{\mathrm{\η}}_n}{P_n^a}$

Represent channel information h_n,iConjugate transpose；Represent lax beamforming vectors；Represent that RRH-uses Family to (n, available signal power i),Represent RRH-user to (n, i) to the interference of other all users it With；Illustrate the situation of the n-th RRH；

Lax beamforming vectorsMeet and make network energy consumption F (Φ) minimum, and meet following constraints:

$\underset{\mathrm{\Φ}=\{a_n,b_{n,i},w_{n,i},\∀n\∈N,\∀i\∈I\}}{min}F\left(\mathrm{\Φ}\right)$

$s.t.\underset{i=1}{\overset{I}{\Σ}}\left|\right|w_{n,i}|{|}^2\≤P_n^{\max },n\∈N---\left(S1\right)$

SINR_i≥γ_i,i∈I (S2)

$\underset{i=1}{\overset{I}{\Σ}}{b}_{n,i}^{\left(k\right)}\≤a_n^{\left(k\right)}{C}_n,\∀n\∈N---\left(S3\right)$

$\left|\right|w_{n,i}|{|}_2\≤b_{n,i}\sqrt{P_n^{max}},\∀n\∈N,\∀i\∈I---\left(S4\right)$

$a_n\∈\[0,1\],b_{n,i}\∈\[0,1\],\∀n\∈N,\∀i\∈I---\left(S5\right)$

Constraints (S1) represents: in the n-th RRH, all users' launches the power transmitting power less than or equal to the n-th RRH Big value；

Constraints (S2) represents that the Signal to Interference plus Noise Ratio of i-th user is the minimum threshold that this user sets more than or equal to BBU Pool Value γ_i；

Constraints (S3) represents that the number of users that the n-th RRH is serviced is less than or equal to the capacity of the n-th fronthaul link；

Constraints (S4) represents that the n-th RRH is less than the transmitting power maximum of the n-th RRH to the power of launching of i-th user Value；

Constraints (S5) represents variable a_n,b_n,iValue is the real number between 0 to 1；

V is iterated solving by the RRH-user after ascending order is arranged by step 503, BBU pool, obtains final network energy consumption F (Φ) and beamforming vectors w_n,i；

Step 504, for the n-th RRH, C-RAN according to iteration beamforming vectors w_n,i, calculate transmission data x of this RRH_n；

$x_n=\underset{i=1}{\overset{I}{\Σ}}{w}_{n,i}{s}_i,\∀n\∈N$

s_iRepresent the data of user i request；

Step 505, fronthaul link are by data x_nBeing transferred to user corresponding to RRH, user obtains request data s_i；

Step 506, user community are adhered to according to RRH, ensure that system model energy consumption is minimum in conjunction with beamforming vectors.

Inferior obliqued overaction and the method for community attachment is combined under a kind of cloud Radio Access Network the most as claimed in claim 1, It is characterized in that, the system model described in step one includes: N number of RRH, I user and a BBU pool；Each RRH divides Respectively connected in BBU pool by a fronthaul link；Each RRH at least one user provide service or not Service for any user, an at least corresponding RRH of user；

RRH collection be combined into 1,2 ..., n ..., N}；User collection be combined into 1,2 ..., i ..., I}；N number of fronthaul link Capacity set is combined into { C₁,C₂,...,C_n,...,C_N}。

Inferior obliqued overaction and the method for community attachment is combined under a kind of cloud Radio Access Network the most as claimed in claim 1, It is characterized in that, described step 503, particularly as follows:

Step 5031, judge whether RRH-user is empty set to V, if it is, iteration terminates, otherwise enter step 5032；

Step 5032, first the RRH-user chosen in set V are to (p q), makes b_p,q=0, remaining RRH-user is to right simultaneously The b value answered keeps constant；

Step 5033, according to the b of RRH-user couple in set V_p,qThe a of correspondence is set_pValue, and delete in RRH-user is to V Element (p, q)；

Step 5034, judge whether each RRH meets constraints:If be unsatisfactory for, current iteration number of times K=k+1, returns step 5031 and carries out next iteration；Otherwise, step 5035 is entered；

Step 5035, according to all b values of current all RRH-users couple and a value of correspondence, calculate the network of kth time iteration Energy consumption network energy consumption F^(k)And the beamforming vectors of correspondence

The constraints met is:

Step 5036, judge network energy consumption F^(k)Whether less than F^(min), if it is, F^(k)For current lowest energy consumption, record F^min= F^(k),Enter step 5037；Otherwise, b is set_p,q=1, a_p=1, and return step 5031；

Step 5037 is until by complete to whole iteration for the RRH-user in set V, obtaining final F (Φ) and correspondence Beamforming vectors w_n,i；

The value of final F (Φ) is F^min, and the beamforming vectors w of correspondence_n,iValue be