阅读说明：本技术 多用户网络集中式自适应功率分配设计方法及装置 (Centralized self-adaptive power distribution design method and device for multi-user network ) 是由 王劲涛 陈宇超 汤皓玥 潘长勇 于 2021-07-27 设计创作，主要内容包括：本发明公开了一种多用户网络集中式自适应功率分配设计方法及装置,该方法基于历史的信道状态和干扰噪声信息,基于李雅普诺夫优化框架以及虚拟队列的设计,利用在线学习的算法实现自适应的功率分配算法。能够在当前信道状态信息和干扰噪声未知的情况下进行功率分配,同时满足平均的功率预算约束,且与信道状态和干扰噪声分布已知的固定最优功率分配策略相比,具有渐近最优的特性。(The invention discloses a method and a device for designing centralized self-adaptive power distribution of a multi-user network. The power distribution can be carried out under the condition that the current channel state information and the interference noise are unknown, the average power budget constraint is met, and the power distribution has the characteristic of asymptotic optimization compared with a fixed optimal power distribution strategy with known channel state and interference noise distribution.)

1. A centralized self-adaptive power distribution design method for a multi-user network is characterized by comprising the following steps:

updating the virtual queue according to a power distribution strategy at the last moment in the multi-user network;

and designing a power distribution strategy at the current moment according to the channel state and the interference noise information at the previous moment, the power budget constraint of the multi-user network and an optimization target.

2. The method of claim 1, wherein the virtual queue q (t) is:

Q(t)＝max{0，Q(t-1)+g(p(t))}，

g(p(t))＝r(p(t)；c)-B，

wherein r (·; c) is a power pricing function, and B is a power budget upper limit.

3. The method of claim 1, wherein the power budget constraint for the multi-user network is:

wherein T is total duration, r (·;) is power pricing function, c is pricing parameter, B is power budget upper limit, and p (T) is power allocation decision vector of central controller of multi-user network at time T.

4. The method of claim 1, wherein the optimization objective comprises optimizing a utility function of the multi-user network:

wherein p (t) is a power allocation decision vector at time t for a central controller of a multi-user network, p_m(t) power allocated to antenna m, N is the number of base stations, a_nIs a base stationn number of antennas, h_m(t) is the channel gain between antenna m and the user at time t, σ²Is the white noise of the user's receiver,interference noise power received by the user serving antenna m at time t.

5. The method of claim 2, wherein the power allocation policy for the current time is:

wherein Q (t) is a virtual queue, and V and alpha are hyper-parameters of the algorithm, setFor the feasible set of power allocations, p (t) is the power allocation decision vector at time t for the central controller of the multi-user network.

6. An apparatus for centralized adaptive power allocation design for a multi-user network, comprising:

the first design module is used for updating the virtual queue according to a power distribution strategy at the last moment in the multi-user network;

and the second design module is used for designing the power distribution strategy at the current moment according to the channel state and the interference noise information at the previous moment, the power budget constraint of the multi-user network and the optimization target.

7. The apparatus of claim 6, wherein the virtual queue Q (t) is:

Q(t)＝max{0，Q(t-1)+g(p(t))}，

g(p(t))＝r(p(t)；c)-B，

wherein r (·; c) is a power pricing function, and B is a power budget upper limit.

8. The apparatus of claim 6, wherein the power budget constraint for the multi-user network is:

wherein T is total duration, r (·;) is power pricing function, c is pricing parameter, B is power budget upper limit, and p (T) is power allocation decision vector of central controller of multi-user network at time T.

9. The apparatus of claim 6, wherein the optimization objective comprises optimizing a utility function of the multi-user network:

wherein p (t) is a power allocation decision vector at time t for a central controller of a multi-user network, p_m(t) power allocated to antenna m, N is the number of base stations, a_nIs the number of antennas of base station n, h_m(t) is the channel gain between antenna m and the user at time t, σ²Is the white noise of the user's receiver,interference noise power received by the user serving antenna m at time t.

10. The apparatus of claim 7, wherein the power allocation policy for the current time is:

wherein Q (t) is a virtual teamColumn, V and alpha are hyper-parameters, sets, of the algorithmFor the feasible set of power allocations, p (t) is the power allocation decision vector at time t for the central controller of the multi-user network.

Technical Field

The present invention relates to the field of communications technologies, and in particular, to a method and an apparatus for designing centralized adaptive power allocation for a multi-user network.

Background

Power allocation strategies in multi-user networks are one of the hot spots of research in the field of communications. Through theoretical derivation, it is easy to find that the optimal power allocation strategy is determined by the current channel state information of each channel. In a multi-user scenario where the time-invariant channel or channel state information is completely known, a conventional water-filling algorithm applied to a multi-antenna (MIMO) system can also be applied to a multi-user network and is optimal. In addition, the traditional cave filling algorithm has also proven to be the optimal algorithm when each device has a separate upper power bound, and may be applied to many communication scenarios, including multi-user scenarios.

The above algorithm relies on current timely and accurate channel state information. When a user moves at a high speed or a communication environment changes, channel state information is difficult to obtain in time, and therefore, an adaptive power allocation strategy needs to be designed in an online learning manner. Research has been conducted to design adaptive power allocation algorithms using algorithms based on-line gradient ascent to achieve power and throughput tradeoffs.

However, there have been studies based on online learning that rarely take into account long-term power constraints. Generally, the short-term power constraint is determined according to the hardware requirement of the communication device, and the long-term power constraint further considers constraints such as power efficiency and the like on the basis of the hardware constraint so as to achieve the purpose of increasing the resource utilization rate, and meanwhile, the long-term average power constraint also provides greater flexibility for adaptive power allocation.

Disclosure of Invention

The present invention is directed to solving, at least to some extent, one of the technical problems in the related art.

Therefore, an object of the present invention is to provide a centralized adaptive power allocation design method for a multi-user network, which can implement adaptive design of a power allocation strategy under the condition that current channel state information and interference noise are unknown, and achieve asymptotically optimal network total utility under the condition that a power budget constraint is satisfied.

Another objective of the present invention is to provide a centralized adaptive power allocation design apparatus for a multi-user network.

In order to achieve the above object, an embodiment of an aspect of the present invention provides a centralized adaptive power allocation design method for a multi-user network, including the following steps: updating the virtual queue according to a power distribution strategy at the last moment in the multi-user network; and designing a power distribution strategy at the current moment according to the channel state and the interference noise information at the previous moment, the power budget constraint of the multi-user network and an optimization target.

The centralized self-adaptive power distribution design method of the multi-user network of the embodiment of the invention updates the virtual queue according to the power distribution strategy at the last moment; and designing a power distribution strategy at the current moment according to the channel state and the interference noise information at the previous moment. Therefore, the power allocation strategy can be adaptively designed under the condition that the current channel state information and the interference noise are unknown, and the asymptotically optimal network total utility is achieved under the condition that the power budget constraint is met.

In addition, the method for designing centralized adaptive power allocation of a multi-user network according to the above embodiment of the present invention may further have the following additional technical features:

further, in an embodiment of the present invention, the virtual queue q (t) is:

Q(t)＝max{0,Q(t-1)+g(p(t))}，

g(p(t))＝r(p(t)；c)-B，

wherein r (·; c) is a power pricing function, and B is a power budget upper limit.

Further, in an embodiment of the present invention, the power budget constraint of the multi-user network is:

wherein T is total duration, r (·;) is power pricing function, c is pricing parameter, B is power budget upper limit, and p (T) is power allocation decision vector of central controller of multi-user network at time T.

Further, in one embodiment of the invention, the optimization objective includes optimizing a utility function of the multi-user network:

wherein p (t) is a power allocation decision vector at time t for a central controller of a multi-user network, p_m(t) power allocated to antenna m, N is the number of base stations, a_nIs the number of antennas of base station n, h_m(t) is the channel gain between antenna m and the user at time t, σ²Is the white noise of the user's receiver,interference noise power received by the user serving antenna m at time t.

Further, in an embodiment of the present invention, the power allocation policy at the current time is:

wherein Q (t) is a virtual queue, and V and alpha are hyper-parameters of the algorithm, setFor the feasible set of power allocations, p (t) is the power allocation decision vector at time t for the central controller of the multi-user network.

In order to achieve the above object, an embodiment of another aspect of the present invention provides an apparatus for centralized adaptive power allocation design for a multi-user network, including: the first design module is used for updating the virtual queue according to a power distribution strategy at the last moment in the multi-user network; and the second design module is used for designing the power distribution strategy at the current moment according to the channel state and the interference noise information at the previous moment, the power budget constraint of the multi-user network and the optimization target.

The centralized self-adaptive power distribution design device of the multi-user network of the embodiment of the invention updates the virtual queue according to the power distribution strategy at the last moment; and designing a power distribution strategy at the current moment according to the channel state and the interference noise information at the previous moment. Therefore, the power allocation strategy can be adaptively designed under the condition that the current channel state information and the interference noise are unknown, and the asymptotically optimal network total utility is achieved under the condition that the power budget constraint is met.

In addition, the centralized adaptive power allocation design apparatus for a multi-user network according to the above embodiment of the present invention may further have the following additional technical features:

further, in an embodiment of the present invention, the virtual queue q (t) is:

Q(t)＝max{0,Q(t-1)+g(p(t))}，

g(p(t))＝r(p(t)；c)-B，

wherein r (·; c) is a power pricing function, and B is a power budget upper limit.

Further, in an embodiment of the present invention, the power budget constraint of the multi-user network is:

wherein T is total duration, r (·;) is power pricing function, c is pricing parameter, B is power budget upper limit, and p (T) is power allocation decision vector of central controller of multi-user network at time T.

Further, in one embodiment of the invention, the optimization objective includes optimizing a utility function of the multi-user network:

wherein p (t) is a power allocation decision vector at time t for a central controller of a multi-user network, p_m(t) power allocated to antenna m, N is the number of base stations, a_nIs the number of antennas of base station n, h_m(t) is the channel gain between antenna m and the user at time t, σ²Is the white noise of the user's receiver,interference noise power received by the user serving antenna m at time t.

Further, in an embodiment of the present invention, the power allocation policy at the current time is:

wherein Q (t) is a virtual queue, and V and alpha are hyper-parameters of the algorithm, setFor the feasible set of power allocations, p (t) is the power allocation decision vector at time t for the central controller of the multi-user network.

Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.

Drawings

The foregoing and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:

FIG. 1 is a schematic diagram of a multi-user network model according to one embodiment of the invention;

FIG. 2 is a flow chart of a design method for centralized adaptive power allocation for a multi-user network according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of an average network utility performance simulation according to one embodiment of the invention;

FIG. 4 is a diagram of an average budget violation performance simulation, according to one embodiment of the invention;

FIG. 5 is a diagram illustrating simulation of average network utility performance under different parameters, according to an embodiment of the present invention;

fig. 6 is a schematic structural diagram of a centralized adaptive power allocation design apparatus for a multi-user network according to an embodiment of the present invention.

Detailed Description

Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are illustrative and intended to be illustrative of the invention and are not to be construed as limiting the invention.

In a multi-user network, the channel state information of each user is time-varying at a high speed (possibly caused by user movement or environmental change), and the power of interference noise is random or even hostile, so that the current channel state and interference noise information cannot be obtained quickly and accurately by methods such as channel estimation. The power allocation strategy can only be based on historical channel conditions and interference noise information.

Provision is made for centralized adaptive power allocation in a multi-user network, as shown in fig. 1, which is arranged throughout the network with a central controller controlling N base stations, each of which N is equipped with a, serving a plurality of users_nThe root antenna thus serves a corresponding number of users, each receiving signal received at the receiving end by a user may be mixed with interference noise in addition to the transmission signal transmitted by the antenna and the white noise of the receiver. Since the user may move at a high speed, and the uncertainty of the interference noise, the channel state and the interference noise information at the current time cannot be accurately obtained, the central controller needs to design a current power allocation strategy according to the historical channel state and the interference noise information to optimize the utility function of the entire network, and at the same time, the average power pricing budget constraint at the central controller end must be satisfied.

The following describes a method and an apparatus for designing centralized adaptive power allocation for a multi-user network according to an embodiment of the present invention with reference to the accompanying drawings.

First, a method for designing a centralized adaptive power allocation for a multi-user network according to an embodiment of the present invention will be described with reference to the accompanying drawings.

Fig. 2 is a flowchart of a design method of centralized adaptive power allocation for a multi-user network according to an embodiment of the present invention.

As shown in fig. 2, the method for designing centralized adaptive power allocation for a multi-user network includes the following steps:

in step S1, the virtual queue is updated according to the power allocation policy at the previous time in the multi-user network.

Let the virtual queue be Q (t), the virtual queue is updated as follows:

Q(t)＝max{0,Q(t-1)+g(p(t))}，

wherein the content of the first and second substances,

g(p(t))＝r(p(t)；c)-B，

is the violation of the power budget at the current time.

The maintenance of the virtual queue is completed through the updating of the virtual queue, and in the algorithm, the virtual queue has the function of recording the size of the accumulated budget violation, and the value of the budget violation is regarded as the waiting number in the virtual queue. Therefore, as long as the queue is kept stable, the long-term budget constraint can be ensured to meet the requirement.

It should be noted that, in the embodiment of the present invention, there is no number of power budget constraints, and if there are multiple power budget constraints for different base stations, when constructing a virtual queue, only multiple queues need to be constructed, and only multiple queues need to be satisfied to be stable at the same time, so that all long-term budget constraints can meet the requirements. In the formula, the scalar q (t) will be changed into the vector q (t), and the processing method is similar, which is not described herein.

In step S2, the power allocation strategy at the current time is designed according to the channel state and the interference noise information at the previous time, the power budget constraint of the multi-user network and the optimization target.

Under the condition that the current channel state and the interference noise information are unknown, the central controller takes the total utility function of the whole network as a target to optimize, and designs a self-adaptive power allocation strategy, and meanwhile, the average power budget constraint of the central controller end needs to be met.

Specifically, in a multi-user network, there is a central controller that centrally schedules the resource (i.e., power) allocation of each base station to serve multiple users. The central controller needs to pay for the required transmit power of each base station. Thus, in addition to the transmit power upper bound on each antenna side of each base station, the central controller side has an average power budget constraint:

wherein T is total duration, r (·;) is power pricing function, c is pricing parameter, B is power budget upper limit, and p (T) is power allocation decision vector of central controller of multi-user network at time T.

The central controller adaptively adjusts the power distribution among the antennas of each base station to optimize the utility function of the whole network, the total operating time of the system is set as T, at each time T, the power distribution strategy of the central controller is represented by a vector p (T), and the network utility at the current time is as follows:

wherein p (t) is the power allocation decision vector of the central controller at time t, p_m(t) power allocated to antenna m, N is the number of base stations, a_nIs the number of antennas of base station n, h_m(t) is the channel gain between antenna m and the user at time t, σ²Is the white noise of the user's receiver,interference noise power received by the user serving antenna m at time t.

Each antenna of each base station has a power constraint upper limit P:

the power allocation strategy at the current moment is as follows:

wherein V and alpha are hyper-parameters of the algorithm, and the size and the set are determined subsequentlyFor a feasible set of power allocations, i.e. the allocated power per antenna cannot exceed the upper limit P:

in summary, through steps S1 and S2, a complete adaptive power allocation scheme can be obtained.

For algorithm initialization, Q (0) is required to be 0,and (4) finishing. The power distribution strategy actually solves a convex optimization problem, and further a quadratic programming problem, so that the existing convex optimization theory has a good algorithm to quickly solve the current power distribution, and the algorithm has a good practical value. Through theoretical derivation, the current hyper-parameter is selected asIn time, the algorithm can not only ensure convergence, but also ensure asymptotically optimal property compared with a fixed optimal power allocation strategy with known channel state information and interference noise distribution, and simultaneously can meet long-term power budget constraint.

The simulation verification results given in fig. 3 and fig. 4 can prove that the algorithm has the property of progressive optimization compared with the upper bound under random interference noise, wherein the simulation parameters in fig. 3 and fig. 4 are set to 3 base stations, each base station is provided with 1 antenna, and the power pricing function is a linear model, namely r (p; c) ═ c^Tp, wherein the parameter c ═ 0.65,0.4,0.3]^TThe upper bound of the power budget B is 0.75, the upper bound of the transmission power of each antenna is P1, and the channel gain h is_m(t) Gaussian distributions obeying independent equal distributions, i.e. | h_m(t)|²White noise power σ subject to a negative exponential distribution with the parameter μ ═ 1²The interference noise is subject to uniform distributionThrough simulation, the traditional cave type water injection algorithm can achieve a fixed optimal power allocation decision; meanwhile, the curves in fig. 3 and fig. 4 also show that the algorithm gradually approaches the optimal average network utility as the operation time increases, and meanwhile, the average budget violation tends to zero, thus verifying the characteristics of gradual optimization and satisfaction of the average power budget constraint of the algorithm.

Further, the simulation results given in fig. 5 can verify the influence of the hyper-parameters of the algorithm on the convergence of the algorithm. Three different sets of hyper-parameters are set in fig. 5, which are:defining constants at the same timeSo thatThe other parameter settings are consistent with fig. 3 and 4. The graph of fig. 5 shows that the magnitude of the hyperparameter V measures the trade-off of the algorithm for network utility performance convergence and budget violation convergence: the larger the hyperparameter V is, the higher the average network utility convergence of the algorithm is, but the higher the cost is the higher budget violation probability; conversely, the smaller V, the slower the convergence of the average network utility of the algorithm, but the smaller the probability of exceeding the budget. Therefore, the reasonable selection of the hyper-parameters is the key for the algorithm to reach the optimal working point.

Further, the embodiment of the present invention has very good performance when the channel is a gaussian channel and the interference noise has unknown distribution, but the method proposed by the present invention is not dependent on this, and can still work normally under other situations, including hostile channels and hostile interference noise.

According to the centralized self-adaptive power distribution design method of the multi-user network provided by the embodiment of the invention, the virtual queue is updated according to the power distribution strategy at the last moment; and designing a power distribution strategy at the current moment according to the channel state and the interference noise information at the previous moment. Therefore, the power allocation strategy can be adaptively designed under the condition that the current channel state information and the interference noise are unknown, and the asymptotically optimal network total utility is achieved under the condition that the power budget constraint is met.

Next, a centralized adaptive power allocation design apparatus for a multi-user network according to an embodiment of the present invention will be described with reference to the drawings.

Fig. 6 is a schematic structural diagram of a centralized adaptive power allocation design apparatus for a multi-user network according to an embodiment of the present invention.

As shown in fig. 6, the multi-user network centralized adaptive power allocation designing apparatus 10 includes: a first design module 100 and a second design module 200.

The first design module 100 is configured to update the virtual queue according to a power allocation policy at a previous time in the multi-user network. And a second design module 200, configured to design a power allocation policy at the current time according to the channel state and the interference noise information at the previous time, as well as the power budget constraint and the optimization target of the multi-user network.

Further, in an embodiment of the present invention, the virtual queue q (t) is:

Q(t)＝max{0,Q(t-1)+g(p(t))}，

g(p(t))＝r(p(t)；c)-B，

wherein r (·; c) is a power pricing function, and B is a power budget upper limit.

Further, in one embodiment of the present invention, the power budget constraint for the multi-user network is:

wherein T is total duration, r (·;) is power pricing function, c is pricing parameter, B is power budget upper limit, and p (T) is power allocation decision vector of central controller of multi-user network at time T.

Further, in one embodiment of the invention, the optimization objective comprises a utility function of the optimal multi-user network:

wherein p (t) is a power allocation decision vector at time t for a central controller of a multi-user network, p_m(t) power allocated to antenna m, N is the number of base stations, a_nIs the number of antennas of base station n, h_m(t) is the channel gain between antenna m and the user at time t, σ²Is the white noise of the user's receiver,interference noise power received by the user serving antenna m at time t.

Further, in an embodiment of the present invention, the power allocation policy at the current time is:

wherein Q (t) is a virtual queue, and V and alpha are hyper-parameters of the algorithm, setFor the feasible set of power allocations, p (t) is the power allocation decision vector at time t for the central controller of the multi-user network.

It should be noted that the foregoing explanation of the method embodiment is also applicable to the apparatus of this embodiment, and is not repeated herein.

According to the centralized self-adaptive power distribution design device of the multi-user network provided by the embodiment of the invention, the virtual queue is updated according to the power distribution strategy at the last moment; and designing a power distribution strategy at the current moment according to the channel state and the interference noise information at the previous moment. Therefore, the power allocation strategy can be adaptively designed under the condition that the current channel state information and the interference noise are unknown, and the asymptotically optimal network total utility is achieved under the condition that the power budget constraint is met.

Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "a plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.

In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.

Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.