WO2012073118A1

WO2012073118A1 - Method and local node for coordinated power allocation

Info

Publication number: WO2012073118A1
Application number: PCT/IB2011/003190
Authority: WO
Inventors: Zhenning Shi; Yajuan Luo; Lin Huang; Daqing Gu
Original assignee: France Telecom
Priority date: 2010-12-01
Filing date: 2011-11-30
Publication date: 2012-06-07
Also published as: WO2012071727A1

Abstract

The present invention relates to a method for coordinated power allocation within a first cell (C0) associated with a first local node (e NB0), the method comprising, for each of at least one resource block (B_j) used in the first cell, receiving (100), at the first local node, at least one power coordinating value (ξ_j ^(i→0)) depending on an average data throughput value (R_j ^一(i)) determined for the resource block (j) in a neighbouring cell (Ci) of said first cell and assigning (200) a transmit power (P_j ^(o)) to said one resource block (B_j) in accordance with the received power coordinating value (ξ_j ^(i→0). The present invention relates also to a corresponding computer program, a local node configured to perform the steps of this method as well as to a network including such a local node.

Description

METHOD AND LOCAL NODE FOR COORDINATED POWER ALLOCATION

The invention relates to the field of power allocation in wireless communication systems, and more specifically in multicarrier based communication networks.

Wireless communication networks usually define a plurality of cells each covered by a local node, also called base station, which transmits and receives radio frequency signals to user equipments located within the cell.

When user equipments located in different neighbouring cells use the same frequencies for transmission, a detrimental effect called intercell interference can occur and affect the signal to noise ratio of their transmission.

Figure 1 shows a wireless communication network affected by such an intercell interference effect.

In this figure, three cells C1-C3 associated respectively with local nodes eNBl- eNB3 and user equipments UE1-UE3 are illustrated.

In each of these cells, a "serving link" connect the local nodes with the user equipments. However, when some of the user equipments are located in an area near the border of these cells (represented by a gray area on figure 1), these user equipments may also receive "interference link" signals emitted from local nodes of neighbouring cells, which causes intercell interference. In figure 1, this is the case for user equipments UE2 and UE3 which receive interference link signals from local node eNBl, such signals being detrimental to their own transmission.

In order to avoid such intercell interferences between users of neighbouring cells, several schemes have been suggested so far.

For instance, it is possible to use a frequency reuse scheme within a cell to avoid severe intercell interference. Such a scheme is usually called SFFR (for Soft Fractional Frequency Reuse).

This scheme consists in dividing, within one cell, the available spectrum into resource block subsets and to allow the users located substantially at the center of the cell to access to all resource blocks while the users located at the cell edges (thus close to the users located in neighbouring cells) are only provided access to a portion of the whole spectrum. The subchannels occupied by cell edge users are orthogonal between cells.

In such a scheme, it is typical to allocate less transmit power to cell center users and more transmit power to cell edge users, and the power ratio between the two user groups can be used to generate frequency reuse patterns ranging from frequency reuse factor one to loose reuse.

Other schemes, such as TMU-based channel allocation, aims at maximizing the system throughput improvement by channel assignment. This is enabled by evaluating the achievable throughput with the user of interest as well as the throughput without this user, and choosing the user that maximizes the system throughput.

The above-discussed frequency reuse scheme presents nevertheless several drawbacks.

First of all, since this scheme employs orthogonal frequency assignment, the spectrum accessible by the users is limited. Furthermore, the frequency reuse patterns employed in SFFR is static or semi-static (according to LTE-Rel 8) and thus not adaptive to a time- varying radio environment.

Moreover, the split of the spectrum into two subsets for the use around the cell center and at the cell edges, and the power ratio between the corresponding two user groups are performed empirically in the existing techniques. This is suboptimal and less efficient for a variety of scenarios.

On the other hand, US Patent Application No. US2010/009710, "Distributed Inter- cell interference Mitigation in OFDMA Multi-Carrier Wireless Data Networks," H. Zhang et al., discloses a coordinated downlink power allocation scheme for an OFDMA based network.

In this scheme, the power update at local cells is performed to maximize the modified weighted data rate summed over the local cells as well as their in-neighbor sets, which are defined as being the set of cells that are subject to the interference from the considered cells.

This coordinated power allocation scheme, which takes into account the weighted sum of modified rates of a number of co-channel cells, presents nevertheless some drawbacks as well.

First, since it maximizes the weighted sum of modified instantaneous data rates of co-channel cells, it does not lead automatically to the maximization of throughput achieved by users in a long term.

Furthermore, such a scheme is not foreseen to be used with a scheduling scheme where cell edge users get more scheduling opportunities. Finally, this scheme operates in an iterative way where the base stations need to take turns to update power, which causes long delay and substantial signaling overhead when many base stations are involved.

It is thus an object of the present invention to overcome the drawbacks of the existing prior art and to propose a solution which can achieve long-term maximization of throughput achieved by users, offers a large dynamic range and does not need to be performed in an iterative way.

According to the present invention, there is provided a method for coordinated power allocation within a first cell associated with a first local node, the method comprising, for at least one resource block used in the first cell, the steps of receiving, at the first local node, at least one power coordinating value depending on an average user data throughput value determined for the resource block in a neighbouring cell of said first cell and assigning a transmit power to said one resource block in accordance with the received power coordinating value.

Advantageously, the power coordinating value further depends on a power coordination parameter such that the power coordinating value is proportional to the average data throughput value raised to the power of the opposite of the power coordination parameter, in order to choose an appropriate trade-off between the gain increase for cell edge users and the performance of average user throughput.

In an advantageous embodiment, the power coordinating value is defined as follows:

— H ⁽ⁱ⁾

—(0

wherein Rj is the average data throughput value determined for the resource block

B_j in the neighbouring cell Ci, γ is the power coordination parameter, Η^{. ^} is the channel gain H^ between the first cell CO and the scheduled user on the j-th resource block B_j in the neighbouring cell Ci and Nj° is the noise-plus-interference factor Ν^{. ^} on the j-th resource block B_j in the neighbouring cell Ci.

In a particularly advantageous embodiment, the value of the power coordination parameter is comprised between 1 and 3, which offers a good trade-off between the gain increase for cell edge users and the performance of average user throughput.

In a further embodiment, an a-PF scheduling scheme with a scheduling parameter is used in the first cell. The use of such an a-PF scheduling scheme allows enhancing further the coordinated power allocation by choosing an appropriate scheduling parameter a.

In such an embodiment, it is particularly advantageous that the value of the scheduling parameter is comprised between 0,75 and 2, in order to obtain a good trade-off between fairness among scheduled cell edge users and efficiency in terms of average data throughput.

In another advantageous embodiment, the step of assigning the transmit power to the resource block comprises:

- calculating a transmit power for said resource block in accordance with said power coordinating value and a system-defined parameter;

- computing the total transmit power assigned to all resource blocks in the first cell; and

- if a condition on the computed total assigned transmit power is verified, assigning said calculated transmit power to said resource block, otherwise adjusting the system-defined parameter in accordance with the computed total assigned transmit power.

In a further embodiment, the system-defined parameter is adjusted by replacing system-defined parameter β with the value Αβ^■ sgn(P_T0T - P_MAX ) + β , wherein Αβ is a predefined variation value, PMAX is the maximum transmit power assignable by the first local node and Ρχοτ is the total assigned transmit power for all resource blocks in the first cell CO.

Advantageously, the calculation of the transmit power for the resource block, the computation of the total transmit power and the adjustment of the system-defined parameter are performed iteratively until the condition on the computed total transmit power is verified.

In an advantageous embodiment, the condition on the computed total transmit power is verified if the computed total transmit power is less than the maximum transmit power of the first local node.

In a further advantageous embodiment, the condition on the computed total transmit power is verified if the difference between the maximum transmit power of the first local node and the computed total transmit power is less than a predetermined threshold value.

In accordance with a further aspect of the present invention, there is further provided a computer program product comprising instruction codes for implementing the steps of a method for coordinated power allocation as discussed before, when loaded and run on processing means of a local node in a wireless communication network.

In accordance with a further aspect of the present invention, there is provided a local node comprising transmission means for transmitting a signal to at least one user equipment scheduled with a resource block in a first cell and processing means for assigning transmit power to said resource block, characterised in that the transmissions means and the processing means are adapted to perform the steps of the method for coordinated power allocation as discussed before.

In accordance with a further aspect of the present invention, there is provided a wireless communication network comprising at least first and second local nodes as discussed before, these first and second local nodes being configured to perform the steps of the method for coordinated power allocation as discussed before.

Other features and advantages of the invention will become apparent from the following description of non-limiting exemplary embodiments, with reference to the appended drawings, in which:

- Figure 1, already discussed, shows a wireless communication network comprising a plurality of cells covered by local nodes;

- Figure 2 illustrates a flow chart of the steps of the method for coordinated power allocation according to the present invention; - Figure 3 illustrates a flow chart of the substeps of the power allocation step employed in the coordinated power allocation according to the present invention;

- Figure 4 illustrates a wireless communication network comprising at least one local node adapted to perform the steps of the coordinated power allocation according to the present invention; and

- Figures 5 A, 5B, 5C and 5D show simulated distribution curves for different types of parameters, in order to show the advantage of using the method according to the present invention.

Some embodiments of the present invention are now described in more detail with reference to the attached figures.

Figure 2 illustrates a flow chart of the steps of the method for coordinated power allocation according to the present invention.

For the sake of illustrating the invention, such a method is performed here within a first cell CO covered by a first local node eNBO, but may be performed in parallel in any number of cells Ci covered by their respective local nodes eNBi.

In this cell CO, a set of J of resource blocks { B_j } i_<j<j are used by the local node eNBO to communicate with user equipments within the first cell CO. Such resource blocks B_j represent more particularly blocks of frequencies in an OFDMA Network.

For the j-th resource block B_j of this set of resource blocks, the method for coordinated power allocation comprises the reception (step 100), by the first local node eNBO, of at least one power coordinating value ξ^^(ι→0) depending at least on an average data throughput value R determined for the j-th resource block B_j in one of the neighbouring cells Ci of said first cell CO.

Once the first local node eNBO has received the power coordinating value ξ^^(ι→0) sent from the second local node eNBi, the first local node eNBO assigns (assignment step 200) a transmit power P.^0> to the j-th resource block B_j in accordance with the received power coordinating value ξ₎ ^(ι→0).

In order to illustrate the invention, a single power coordinating value ^_j ^(l→0) from a single neighbouring cell Ci is received here. However, several power coordinating value ξ₎ ^(ι→0) can also be received from different neighbouring cells Cl,...Ci,...,CN during the receiving step 200.

This first power coordinating value ξ^^(ι→0) can be obtained via a user-channel pairing using a standard scheduling method, for instance a round robin method or an a-PF scheduler method. Alternatively, this first power coordinating value ξ^^(ι→0) can be estimated statistically and sent to neighbouring cells on a super-frame basis to reduce signalling overhead. It is also noted that reception of coordinating value ξ^^(ι→0) from different neighbouring cells Ci can be asynchronous.

Here again, when several power coordinating values ξ ^1→0),...,ξ ^1→0) are received respectively from different neighbouring cells Cl,...Ci of the first cell CO, the assignment of the transmit power P ^ is carried out in accordance with these different power coordinating values ξ ^1→0),...,ξ ^1→0).

With such an assignment procedure, the transmit power of the signals transmitted from the first local node eNBO to the user equipments within the first cell CO in the j-th resource block B_j is thus calculated in accordance with the average data throughput determined within the neighbouring cells Cl,..,Ci and can thus be adjusted to minimize the impact of intercell interferences in the long term with the users of this neighbouring cell while maximizing the long term throughput of users in cell CO.

The above-mentioned receiving and assigning steps can be performed in parallel for one, several or all of the J resource blocks B_j of the set of resource blocks { B_j } i_<j<j that are involved in the multi-cell coordination.

Once a transmit power P.^0> is assigned to one, several or all of the J resource blocks

B_j, the first local node eNBO can then apply the specifically assigned transmit power value P^ for transmitting (transmission step 300) signals to the user equipments with the corresponding resource blocks B_j.

Subsequently, the first local node eNBO can update its own channel statistic and user SINR, compute new power coordinating values ξ^^0→ι ...,ξ^^0→ι) based on these updated channel statistics and transmit these new power coordinating values ^_j ^(0→1),...,^_j ^(0→l) respectively to local nodes eNB l,...eNBi, in order for them to update in turn their own transmit power values (power coordinating value updating step 400). The frequency to update and distribute coordination values ξ^^0→ι ...,ξ^^0→ι) is not necessarily the same as the power scheduling cycle in those cells.

In advantageous embodiment, the above-mentioned Dower coordinating value ξ ^1→0) depends also on a power coordination parameter γ such that the power coordinating value ξ^^(ι→0) is proportional to the average data throughput value raised to the power of the opposite of the power coordination parameter γ, that is to say : ξ^{'^→0) <x ) ⁷

Such a power coordination parameter γ is a strictly positive value and modulates therefore the influence of average data throughput value on the power coordinated allocation scheme, in order to give more flexibility to the network operator to choose an appropriate trade-off according to the network context.

Such a power coordination parameter γ is commonly allocated to all the cells involved with this coordinated power allocation scheme in the network, that is to say to at least all the neighbouring cells Ci of the first cell CO or even all the cells of the network, and is advantageously adjustable and distributed by a central controller in a semi-static manner to the different cells involved.

In order to also take into account the transmission factors, such as the noise, affecting the channels in the j-th resource block B_j the above-mentioned power coordinating value ξ ^1→0) η^ also depend on an average interference value N determined for the j-th resource block B_j in one of the neighbouring cells Ci of said first cell CO and a channel gain value uf determined between the said first cell CO and the user in one of neighbouring cells Ci. An example of such a power coordinating value ξ^^(ι→0) is given below.

In a further advantageous embodiment, the coordinated power mitigation scheme of the present invention can be used jointly with a scheduling scheme wherein cell edge users get more scheduling opportunities.

Generic scheduling schemes are usually employed in communication systems, such as round robin (RR) scheduling schemes, max C/I scheduling schemes or proportional fair (PF) scheduling schemes.

In the case of a round robin scheduling scheme, in each scheduling cycle, the RR scheduler randomly picks users from a user set with non-empty queues for transmission. It guarantees fairness among all active users. Nevertheless, it is inefficient in realizing user throughputs.

The Max C/I scheduling always schedules the users in best channel conditions. It therefore maximizes the instantaneous sum throughput of the cell. However, users close to cell site that usually maintains a bore-sight link with serving base station are most likely to be scheduled while cell edge users starve.

The a-PF scheduling (wherein a>l), on the other hand, maximizes the logarithm of the average throughput of users by scheduling users in their own best channel conditions. That is to say, a user whose channel condition is close to its own peak channel quality is scheduled. It balances between throughput maximization and user fairness and proves its success in high data-rate cellular networks such as UMTS.

Generic alpha-PF schedulers are thus more advantageous in the present case than other types of schedulers in that it allows increasing alpha tradeoffs cell throughput for more fairness among users. Since the PF metric of neighboring cell users is considered, a cluster utility-maximizing power allocation scheme can automatically avoid generating excessive intercell interference to provide fairness to users in neighboring cells. Hence, fairness can be used in the present invention as an intercell interference coordination (ICIC) mechanism which goes beyond the conventional scope in scheduling.

An exemplary a-PF scheduling scheme with a scheduling parameter a is discussed for instance in "Packet Scheduling Algorithms with Fairness Control for CDMA Reverse Link", from J. Shin et al., wherein the scheduling parameter a > 1.

For these reasons, an a-PF scheduling scheme with a scheduling parameter a is advantageously used in the first cell CO, in addition to the power coordinated allocation method of the present invention.

The use of such an a-PF scheduling scheme allows enhancing further the coordinated power allocation by choosing an appropriate scheduling parameter a.

In order to further illustrate the present invention, the first power coordinating value ξ^^(ι→0) may be obtained according to the following calculation:

The power optimization value P⁽⁰⁾ for the first cell CO can be formulated as follows:

wherein:

- i (for i≠0) is the index of neighbouring cells Ci of the first cell CO taken into account by the algorithm;

- j is the index for the j-th resource block B_j considered; the total number of resource blocks;

P_j is the individual power value assigned to the j-th resource block B_j in cell

CO;

- R ;⁽j>^') i ·s the average throughput for a scheduled user on the j-th resource block B_j in the neighbouring cell Ci;

- r^{ is the corresponding instantaneous throughput for a scheduled user on the j-th resource block B_j in the neighbouring cell Ci;

—(0

- T is the time average window for the evaluation of the parameter Rj ;

- aPF is a generalized utility function for instance given by the following equation:

wherein R is the user average throughput and a is the scheduling parameter of the aPF generalized utility function.

A further condition applying here is that the sum of the individual transmit power values P_j ⁽⁰⁾, for each of the J resource blocks, remains lower or equal than the maximum transmit power P_MAX of the local node eNBO, that is to say that the following equation is verified:

The problem can be thus reformulated in a convex optimization by applying Lagrangian relaxation, as given by the following equation:

+ ΜΡ_ΜΑΧ -∑Ρ )

(4)

The individual transmit power value P_j ⁽⁰⁾ assigned to the j-th resource block B_j can then be determined as follows: , wherein:

- N⁽⁰⁾ and N⁽ⁱ⁾ are the noise-plus-interference factors on the j-th resource block B_j respectively in cell CO and in its neighbouring cell Ci;

- G_j ⁰⁾ is the channel gain between cell CO and the scheduled user;

- H® is the channel gain between cell CO and the user equipment in the neighbouring cell Ci;

- SINR® is the signal-to-noise ratio for the scheduled user on the j-th resource block B_j in the neighbouring cell Ci;

- β (>=0) is a cell-specific parameter.

- γ is a power coordination parameter, commonly allocated to the first cell CO and all the neighbouring cells Ci, as discussed previously,

In order for the neighbouring cells in a coordinated cluster to generate and provide the cell CO with some information on the assignment and channel state condition, a power coordination value forwarded by neighbouring cell Ci to cell CO of interest is then given as follows:

(i→0) SINR (0 Λ H (0

(6)

SINR + 1 N (0

With a signal-to-noise ratio typically above 5 dB in a user

SINR

equipment, 1 , and the previous equation (6) can be approximated by the

SINR® + 1

following equation:

(i)

: (i→0) _ _{ί Έ>} ^{ί) -γ ¹¹

(7)

N

In this last equation, it appears that the power coordination value

forwarded by neighbouring cell Ci to the first cell CO of interest, depends only on: — (0

- the average throughput Rj for a scheduled user on the j-th resource block B_j in neighbouring cell Ci;

- the power coordination parameter γ;

- the channel gain between the first cell CO and the scheduled user on the j- th resource block B_j in the neighbouring cell Ci; and

- the noise-plus-interference factor Nj° on the j-th resource block B_j in the neighbouring cell Ci.

Hence, in an advantageous embodiment wherein not only the average data throughput is taken into account to perform coordinated power allocation but also the transmission conditions of channel used in the j-th resource block B_j, the power coordinating value ξ^^(ι→0) can be computed as follows :

Such a power coordination value ^^(,→0) , once computed, can be transmitted from the local node eNBi of neighbouring cell Ci to the local node eNBO of the first cell CO, where it can be used to calculate the transmit power P_j ⁽⁰⁾ to be assigned to the j-th resource block B_j in the first cell CO.

Figure 3 illustrates a flow chart for further describing an embodiment of the power assignment step 200 performed in the coordinated power allocation according to the present invention.

In this embodiment, the power assignment step 200 comprises the substep of computing (calculation step 201) at least one individual transmit power value P_j ⁽⁰⁾ for a j-th resource block B_j in accordance with the received first power coordinating value ξ^^(ι→0) and a system-defined parameter β.

Taking into account the above-discussed received first power coordinating value (^ι→0), the individual power value P ⁽⁰⁾ can be computed, for instance, as follows :

wherein:

- β is the system-defined parameter,

- R _j ⁽⁰⁾ is the average throughput for a scheduled user on the j-th resource block B_j in the first cell,

- γ is a power coordination parameter commonly allocated to the first cell CO and all the neighbouring cells Ci,

- ξ⁽'^→0> _S the power coordinating value determined for the j-th resource block B_j in neighbouring cell Ci,

- G^⁰⁾ is the channel gain between the first cell CO and the scheduled user and N^⁰⁾ is the noise-plus-interference on the j-th resource block B_j in the first cell CO.

Once the individual power value computing step 201 has been performed, the power assignment step 200 comprises the computation (step 203) of the total transmit power Ρχοτ assigned to all the resource blocks in the first cell CO.

Subsequently to this total transmit power computation step, a condition on the computed total transmit power Ρχοτ is verified during a verification step 205, in order to determine if the computed individual transmit power value P_j ⁽⁰⁾ can be assigned or not to the resource block B_j.

In order to obtain the total power value Ρχοτ assigned to all resource blocks by the first local node eNBO, the total power value Ρχοχ is computed for instance by summing the individual power values P_j ⁽⁰⁾ assigned to all of the J resource blocks j within the first cell CO, according to the following equation :

(9) p_T0T =t^pr^}

Once the total power value Ρχοχ is computed, a condition depending on this computed total transmit power during the verification step 205, and if this condition is verified, the computed individual transmit power P_j ⁽⁰⁾ is assigned to the j-th resource block

Bj during an assigning step 207 signal. Otherwise, if such a condition on the computed total transmit power Ρχοχ is not verified during this verification step 205, the system- defined parameter β is adjusted in accordance with the computed total transmit power Ρχοτ (adjustment step 209).

The above-discussed system-defined parameter β is initially defined as being β₀ (for instance βο=0) and represents a non-negative cell-specific parameter, which is numerically determined to keep the power constraint in (3) tight.

It can be shown that the total transmit power P_TOT monotonically decreases with respect to the system-defined parameter β .

There are a number of options to obtain β such that P_TOT = P_max .

One illustrative way is to initialize β as β ₀ = 0 , and update it to let the total transmit power Ρ_ΤΟτ approach the maximum transmit power P_max gradually.

An other illustrative way can be to use bisection search as follows:

1. Set maximum and minimum values as _y5_min = 0 , and max ^~~

max

2. Set β =

3. Perform proposed power allocation, compute "TOT ^~

4. If (10) and (11) (in the revised manuscript) is simultaneously satisfied, exit with {P_j ⁰⁾ } ; otherwise goto step 5.

5. If P-roT < ^max , ^max = β \ otherwise = β .

6. Go to step 2.

In a specific embodiment, the system-defined parameter β is adjusted by replacing β with the value Δβ^■ sgn(P_ror - P_MAX ) + β , wherein:

- Αβ (>=0) is a predefined small variation value, for instance Αβ(η)

n + 1 wherein n is an increment starting from n=0 and increased for each adjustment step 209 performed.

- sgn(x) is the function outputting the sign of x. - P_MAX is the maximum transmit power of the first local node eNBO, that is to say the maximum transmit power than can assigned to resource blocks by this first local node eNBO,

- Ρχοτ is the computed total transmit power assigned by furst local node eNBO to all resource blocks Bi to Bj in the first cell CO.

In other words, β ± Αβ→ β depending on whether the total assigned power value

P_TOT is higher or lower than the maximum total power value P_MAX-

The assignment of the individual transmit power P ⁰⁾ to the j-th resource block Bj is thus advantageously conditional to the total transmit power Ρ_ΤΟτ of the local node in order to take into account the limits imposed on said local node.

In an embodiment of the present invention, the condition on the computed total transmit power Ρχοτ is verified if the value of the computed total transmit power Ρχοτ is less than the value of the maximum transmit power P_MAX of the first local node, as follows:

(10) Pxox < PMAX

Using such a condition ensures that the assigned individual transmit powers P_j do not exceed the total transmit power capacity of the local node eNBO.

In an advantageous embodiment of the present invention, a further condition can be taken into account in addition to the above-mentioned condition defined by equation (10). Such a further condition is that the difference between the maximum total power value P_MAX and the total power value Ρχοτ is below a predetermined threshold value ΔΡ, as follows:

(I D PMAX - Ρτοτ < ΔΡ

With this second condition, one tries to optimise the individual transmit power values P ⁰⁾ by maximising their values while keeping below a level where they would exceed the total transmit power capacity of the local node eNBO.

In this embodiment, if the two conditions Ρχοτ < P_MAX and P_MAX - Ρτοτ < ΔΡ are fulfilled during the verification step 205, the assignment step 207 is performed. If one of these two condition is not fulfilled, the system-defined parameter β is adjusted in accordance with the computed total transmit power Ρχοχ.

In other words, β ± Αβ→ β depending on whether the total assigned power value Ρτοτ is higher or lower than the maximum total power value P_MAX-

Figure 4 illustrates a wireless communication network comprising at least one local node adapted to perform the steps of the coordinated power allocation according to the present invention.

In figure 4, a wireless communication network comprising a first cell CO, a second cell CI and a third cell Ci, covered respectively by local nodes eNB0,eNBl and eNBi, is described.

The different local nodes transmit power coordinating values to each other, that is to say:

- first local node eNBO transmit power coordinating values ξ^^0→1) and ξ^^(0→ι) respectively to local node eNBl and eNBi;

- second local node eNBl transmit power coordinating values ξ^^(1→0) and ξ ^1→ι) respectively to local node eNBO and eNBi;

- third local node eNBi transmit power coordinating values ξ^^(ι→0) and ξ ^1→1) respectively to local node eNBO and eNBl;

Here, the first local node eNBO is arranged to perform steps 100' to 400' which are similar to steps 100 to 400 of the previously discussed coordinated power allocation method.

More specifically, the first local node eNBO receives power coordinating values ξ ^1→0) and ξ^^(ι→0) from local nodes eNBl and eNBi during a receiving step 100' similar to previously discussed receiving step 100.

With these received power coordinating values ξ^^(1→0) and ξ ^1→0), the first local node eNBO can assign a transmit power value P ⁰⁾ to the j-th resource block B_j during an assignment step 200' similar to the assignment step 200 described previously, for instance by using equation (8) with these new power coordinating values ξ^^1→0) and ξ ^1→0).

This assigned transmit power value P ⁰⁾ can then be used for transmitting signals to the scheduled user in the j-th resource block B_j during a transmission step 300' similar to the transmission step 300.

The first local node eNBO can subsequently update its own channel and user SINR statistics, compute updated power coordinating values ξ^^(0→1) and ξ^^(0→ι) in accordance with equation (7) and transmit these updated power coordinating values ξ^^(0→1) and ξ^^(0→ι) respectively to local nodes eNBl and eNBi, so that they can in turn update their own transmit power values in accordance with the present invention. Figures 5A-5D shows simulated distribution results for different types of parameters, in order to show the advantage of the present invention.

Static network simulations have been carried out for a microcell scenario, with the simulation parameters specified in the following Table 1 :

Table 1: Simulation assumptions Figure 5A shows the user SINR distribution curve achieved with the method of the present invention, when the power coordination parameter γ is chosen to be 1 and compared to a SINR distribution curve obtained with a usual equal power assignment scheme.

It is apparent on this figure 5A that the SINR has been greatly improved due to the enhance interference coordination achieved with the method of the present invention.

Figure 5B shows the user throughput distribution curve achieved with the method of the present invention, when the power coordination parameter γ is chosen to be 1 and compared to a user throughput distribution curve obtained with a usual equal power assignment scheme.

It is apparent on this figure 5B that the invented scheme outperforms the baseline system substantially. In particular, a 50 % increase is achieved for five-percentile users, justifying the use of the present invention in the intercell-interference limited scenario.

Figure 5C shows the user throughput distribution curve achieved with the method of the present invention for different power coordination parameters γ.

On this figure 5C, it can be seen that different level of tradeoffs can be obtained between cell center users and cell edge users, by varying the power coordination parameter

Y-

In particular, when the power coordination parameter γ is increased, further gains can be achieved for cell edge users. This gain increase for cell edge users is however obtained at the cost of degrading the average user throughput.

It is thus advantageous to select a power coordination parameter γ which allows achieving a favorable trade-off between the gain increase for cell edge users and the performance of average user throughput.

Table 2 below reports the average user throughput and the five-percentile user throughput for various power coordination parameter γ values, compared to those for a non-cooperative equal-power allocation scheme. Power Average user Increase on Cell edge Increase on cell coordination throughput Average user throughput edge throughput parameter γ throughput

γ = 0.25 3.33 Mbps 9.0 % 994 kbps 21.1 % γ = 0.50 3.37 Mbps 10.2 % 1101 kbps 35.3 % γ = 0.75 3.38 Mbps 10.7 % 1208 kbps 47.2 % γ = 1.00 3.38 Mbps 10.6 % 1288 kbps 57.0 % γ = 2.00 3.27 Mbps 7.0 % 1506 kbps 83.5 % γ = 3.00 3.10 Mbps 1.5 % 1606 kbps 95.7 % γ = 4.00 2.93 Mbps -4.1 % 1626 kbps 98.1 %

Table 2: Average user throughput and five-percentile throughput for various γ values

It can be seen on Table 2 that the five-percentile user throughput gain monotonically increases as γ becomes large. Nevertheless, the average user throughput decreases as γ > 3.

From this Table 2, it can be derived that a favorable tradeoff between, on the one hand, the gain increase for cell edge users and, on the other hand, the performance of average user throughput can be achieved for power coordination parameter values comprised between 1 and 3, that is to say for ye [l,3] .

Figure 5D shows the average user throughput versus increase on the cell edge throughput (five percentile user throughput) in an embodiment of the invention wherein both the coordinated power allocation method of the present invention and an a-PF scheduling scheme are used in the first cell CO, for different values of the scheduling parameter a and the power coordination parameter γ is used in the coordinated power allocation scheme.

On this figure, when one considers the scheduling parameter a, it clearly appears that when the scheduling parameter a is increased, the cell edge users gain in terms of fairness, at the cost of an efficiency decrease of the average data throughput once a maximum values has been reached.

On the contrary, when one considers the power coordination parameter γ, it clearly appears that when the power coordination parameter γ is increased (i.e. when changing from one curve to another in the upward direction of figure 5D), the cell edge users gain in terms of fairness at the cost of an efficiency decrease of the total average throughput. Consequently, changing the power coordination parameter γ here allows providing the network operator with a broader range of possible trade-off between fairness and efficiency, as seen by the different curves shown in figure 5D.

Therefore, by choosing proper parameters a and γ, it is possible to offer a larger dynamic range to the network operator. With such a combined scheme, more cell edge users can be scheduled and more benefits can be achieved in mitigating intercell interference and enhancing the user performance on the cell border.

Though the embodiment is put forward for a single-input single-output (SISO) channel, the proposed invention can be applied to multiple-input multiple-output (MIMO) systems as well. In the MIMO scenario, the transmitted power for the eigenmodes of the spatial channel matrix on a number of resource blocks is optimized to maximize the long term user throughputs in a number of co-channel cells, following the above-mentioned strategy.

The example is presented with a generic network structure where a number of overlapping cells reuses the resources in frequency, time and spatial domains. The network can be either a classic cellular network, a heterogeneous network comprising underlaid macrocells and overlaid low-power small access points, e.g., home eNodeB (HeNB) or indoor relays, or cognitive radio network.

The invention also relates to a computer program product that is able to implement any of the method steps as described above when loaded and run on the processing means of a local node of a wireless communication network as described previously.

The computer program may be stored or distributed on a suitable medium supplied together with or as a part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems.

While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive, the invention being not restricted to the disclosed embodiments. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. A single processor or other unit may fulfil the functions of several items recited in the claims. The mere fact that different features are recited in mutually different dependent claims does not indicate that a combination of these features cannot be advantageously used. Any reference signs in the claims should not be construed as limiting the scope of the invention.

Claims

1. A method for coordinated power allocation within a first cell (CO) associated with a first local node (eNBO), the method comprising, for at least one resource block (B_j) used in the first cell:

- receiving (100), at the first local node, at least one power coordinating value

(ξ ^1→0)) depending on an average data throughput value ( R^ ) determined for the resource block (B_j) in a neighbouring cell (Ci) of said first cell;

- assigning (200) a transmit power ( V.^0> ) to said one resource block (B_j) in accordance with the received power coordinating value ¾^(l→0)).

2. The method according to claim 1, wherein the power coordinating value

¾^(l→0)) further depends on a power coordination parameter (γ) such that the power coordinating value ¾^(l→0)) is proportional to the average data throughput value ( - ? ) raised to the power of the opposite of the power coordination parameter (γ).

3. The method according to claim 2, wherein the power coordinating value

¾^(l→0)) is computed as follows :

H

(0

— (0

wherein Rj is the average data throughput value determined for the resource block B_j in the neighbouring cell Ci, γ is the power coordination parameter, is the channel gain between the first cell CO and the scheduled user on the j-th resource block B_j in the neighbouring cell Ci and Nj° is the noise-plus-interference factor on the j-th resource block B_j in the neighbouring cell Ci.

4. The method according to claim 2 or 3, wherein the value of the power coordination parameter (γ) is comprised between 1 and 3.

5. The method according to any one of claims 1 to 4, wherein an a-PF scheduling scheme with a scheduling parameter (a) is used in the first cell (CO).

6. The method according to any one of claim 5, wherein the value of the scheduling parameter (a) is comprised between 0,75 and 2.

7. The method according to any one of claims 1 to 6, characterised in that the step of assigning the transmit power to the resource block (B_j) comprises:

- calculating (201) a transmit power ( V.^0> ) for said resource block (B_j) in accordance with said power coordinating value ¾^(l→0)) and a system-defined parameter

(β);

- computing (203) the total transmit power (Ρτοτ) assigned to all resource blocks (Bi,B_j,Bj) in the first cell;

- if a condition on the computed total assigned transmit power is verified (205), assigning (207) said calculated transmit power to said resource block, otherwise adjusting (211) the system-defined parameter (β) in accordance with the computed total assigned transmit power.

8. The method according to claim 7, characterised in that the transmit power

( V.^0> ) for said resource block (B_j) is computed according to the following formula:

wherein β is the system-defined parameter, R _j ⁽⁰⁾ is the average throughput for a scheduled user on the j-th resource block B_j in the first cell, γ is the power coordination parameter, ξ^{'^→0) is the power coordinating value determined for the j-th resource block B_j in neighbouring cell Ci, ^⁰⁾ is the channel gain between the first cell CO and the scheduled user and N^⁰⁾ is the noise-plus-interference on the j-th resource block B_j in the first cell CO.

9. The method according to claim 7 or 8, characterised in that the system-defined parameter β is adjusted by replacing β with the value Αβ^■ sgn(P_T0T - P_MAX ) + β , wherein Αβ is a predefined variation value, PMAX is the maximum transmit power assignable by the first local node (eNBl) and Ρχοτ is the total assigned transmit power for all resource blocks (Bi_,... Bj) in the first cell CO.

10. The method according to any one of claims 6 to 9, characterized in that the calculation (201) of the transmit power ( P.^0> ) for the resource block (B_j), the computation

(203) of the total transmit power (Ρχοτ) and the adjustment (211) of the system-defined parameter (β) are performed iteratively until the condition on the computed total transmit power is verified.

11. The method according to any one of claims 6 to 10, characterized in that the condition on the computed total transmit power is verified (205) if the computed total transmit power (Ρτοτ) is less than the maximum transmit power (PMAX) of the first local node.

12. The method according to claim 11, characterized in that the condition on the computed total transmit power is verified (205) if the difference between the maximum transmit power (PMAX) of the first local node and the computed total transmit power (Ρτοτ) is less than a predetermined threshold value (ΔΡ).

13. A computer program product comprising instruction codes for implementing the steps of a method for coordinated power allocation according to any one of claims 1 to 12 when loaded and run on processing means of a local node in a wireless communication network.

14. A local node (eNBO) comprising transmission means for transmitting a signal to at least one user equipment scheduled with a resource block (B_j) in a first cell (CO) and processing means for assigning transmit power to said resource block, characterised in that the transmissions means and the processing means are adapted to perform the steps of the method for coordinated power allocation according to any one of claims 1 to 12.

15. A wireless communication network comprising at least first and second local nodes (eNB0,eNBl) according to claim 14, characterized in that said first and second local nodes are configured to perform the steps of the method for coordinated power allocation according to claims 1 to 12.