A resource optimization method and apparatus for a cellular-free massive MIMO system integrating communication, sensing, and energy.

Abstract

This application provides a resource optimization method and apparatus for a cellular-free massive MIMO system integrating communication, sensing, and energy. The method includes: acquiring current system state information to determine the performance indicators of each communication user, each sensing target, and each energy user; constructing an optimization objective based on the performance indicators of the sensing targets and energy users; constructing an optimization problem using AP mode selection variables and global precoding variables as decision variables; configuring the operating mode of each AP as either a sensing mode or an energy mode; and constraining the optimization problem to ensure that the performance indicator of each communication user is not lower than a preset quality of service threshold, the transmit power of each AP does not exceed its maximum transmit power, and the total number of APs configured in sensing mode in the system is within a set range; solving the optimization problem to obtain the optimal operating mode and optimal precoding vector for each AP; and configuring APs based on the optimal precoding vector to improve the overall system performance.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and specifically to a method and apparatus for resource optimization in a non-cellular massive MIMO system that integrates communication, sensing, and performance. Background Technology

[0002] In recent years, non-cellular massive multiple input multiple output (MIMO) technology, as a user-centric network architecture, has become one of the key technologies for realizing future sixth-generation (6G) networks. In this network architecture, a large number of access points (APs) are distributed across the service area and connected to the central processing unit (CPU) via fronthaul links. The CPU coordinates and processes multiple APs to achieve continuous coverage and collaborative gains.

[0003] However, traditional non-cellular systems are typically centered on communication services, making it difficult to effectively address interference management and resource coupling issues when multiple services coexist. With the rapid development of applications such as low-altitude economy, intelligent transportation, and intelligent manufacturing, communication networks not only need to support data transmission but also require target detection, positioning and tracking, and environmental awareness capabilities, driving the development of Integrated Sensing and Communication (ISAC) technology. ISAC achieves communication and sensing synergy through shared spectrum, hardware, and signal processing links, improving spectrum utilization. However, in a non-cellular distributed architecture, the spatial distribution relationships between communication users, sensing targets, and multiple access points (APs) result in high-dimensional precoding variables and complex interference coupling. Traditional centralized or fixed-structure precoding schemes struggle to consistently meet sensing performance constraints while satisfying the Quality of Service (QoS) requirements.

[0004] On the other hand, with the development of the Internet of Things and edge intelligence, the demand for continuous power supply for low-power terminals is becoming increasingly prominent. Simultaneous Wireless Information and Power Transfer (SWIPT) technology provides wireless power to low-power terminals through radio frequency energy harvesting, but there is resource competition between energy transmission and information transmission, and the energy signal may cause additional interference to communication users. In multi-AP collaborative scenarios, if each AP uses the same operating mode for a long time, it will restrict the freedom of energy transmission and exacerbate the contradiction between power allocation and interference management.

[0005] Therefore, while integrating ISAC and SWIPT in a cellular-free massive MIMO system architecture to build a multi-service collaborative network that simultaneously serves communication users, sensing targets, and energy receiving users can improve spectrum utilization efficiency and power supply capacity, it also brings high-dimensional and strongly coupled optimization challenges. On the one hand, communication, sensing, and energy signals share time-frequency resources and transmit power, leading to mutual constraints among multiple performance indicators, making it difficult to balance sensing and power supply needs while ensuring communication QoS. On the other hand, AP mode selection is a discrete decision, while precoding design is a continuous optimization. The two are tightly coupled under per-AP power budget and switching power constraints, making the problem exhibit mixed integer non-convex characteristics and high solution complexity.

[0006] Most related studies employ fixed AP modes or optimize precoding only for a single service, making it difficult to achieve a controllable trade-off between sensing and energy harvesting performance, and also failing to effectively suppress interference from sensing and energy signals on the communication link. Therefore, under the premise of meeting the QoS of communication users and the transmit power limitations of each AP, how to jointly optimize AP mode selection and precoding matrix to improve multi-service collaborative gains while reducing system interference coupling remains a key technical challenge that urgently needs to be overcome in this field. Summary of the Invention

[0007] In view of this, this application provides a method and apparatus for resource optimization of a cellular-free massive MIMO system that integrates communication, sensing, and energy.

[0008] Specifically, this application is implemented through the following technical solution:

[0009] According to a first aspect of the embodiments of this specification, a resource optimization method for a non-cellular massive MIMO system integrating communication, sensing, and energy is provided. The system includes a central processing unit and multiple distributed access points (APs) for providing services to communication users, sensing targets, and energy users. The method is executed by the central processing unit and includes:

[0010] Step S1: Obtain the current system status information, and determine the performance indicators of each communication user, each sensing target, and each energy user based on the system status information;

[0011] Step S2: Based on the sensing target and the performance indicators of the energy user, an optimization target is constructed. An optimization problem is built using AP mode selection variables and global precoding variables as decision variables. Each AP's operating mode is configured as either a sensing mode or an energy mode. The constraints of the optimization problem include:

[0012] The performance indicators of each communication user shall not be lower than the preset service quality threshold;

[0013] The transmit power of each AP shall not exceed its maximum transmit power;

[0014] The total number of APs configured in synergy mode in the system is within the set range;

[0015] Step S3: Solve the optimization problem to obtain the optimal working mode and optimal precoding vector for each AP.

[0016] According to a second aspect of the embodiments of this specification, a resource optimization device for a non-cellular massive MIMO system integrating communication, sensing, and energy is provided. The system includes a central processing unit and multiple distributed access points (APs) for providing services to communication users, sensing targets, and energy users. The device is applied to the central processing unit and includes:

[0017] The indicator calculation unit is used to obtain the current system status information and determine the performance indicators of each communication user, each sensing target, and each energy user based on the system status information.

[0018] The problem modeling unit is used to construct an optimization objective based on the sensing target and the performance indicators of the energy users. The optimization problem is constructed using AP mode selection variables and global precoding variables as decision variables. The working mode of each AP is configured as either sensing mode or energy mode. The constraints of the optimization problem include: the performance indicators of each communication user are not lower than the preset service quality threshold; the transmission power of each AP does not exceed its maximum transmission power; and the total number of APs configured in sensing mode in the system is within a set range.

[0019] The model solving unit is used to solve the optimization problem and obtain the optimal working mode and optimal precoding vector for each AP.

[0020] According to a third aspect of the embodiments of this specification, a non-cellular massive MIMO system for communication, sensing, and energy integration is provided, including a central processing unit and multiple distributed access points (APs), wherein the multiple APs are connected to the central processing unit via a fronthaul link, and wherein the central processing unit is configured to perform the method described in the first aspect to provide services to communication users, sensing targets, and energy users.

[0021] According to a fourth aspect of the embodiments of this specification, an electronic device is provided, including a processor; and a computer-readable storage medium storing computer program instructions that, when executed by the processor, cause the processor to perform the method described in the first aspect.

[0022] According to a fifth aspect of the embodiments of this specification, a computer-readable storage medium is provided having a computer program stored thereon, the computer program being executed by a processor of the method described in the first aspect.

[0023] The resource optimization scheme provided in this application has at least the following advantages:

[0024] (1) The embodiments of this application can improve the system's detection performance for sensing targets and the wireless power supply performance for energy users by jointly optimizing the AP working mode and the global precoding vector, while ensuring the service quality of all communication users;

[0025] (2) The embodiments of this application can resolve the power competition and mutual interference of different service signals at the AP level through power constraints and AP mode selection mechanisms. For example, when the AP is optimized to the sensing mode, its energy signal power is forced to zero to avoid the energy signal from interfering with the communication and sensing signals sent by the same AP. When the AP is optimized to the energy mode, its communication and sensing signal power is zero, so that the AP's transmit power budget and antenna degrees of freedom are concentrated on the precoding design of energy transmission, thereby avoiding the power sharing and spatial orientation degree of freedom competition between the sensing service and the power supply service within the same AP, and improving the directional gain and power supply controllability for energy users.

[0026] (3) The embodiments of this application allow the central processing unit to dynamically divide the AP cluster into two functional subsets that focus on sensing and power supply respectively according to real-time channel conditions, user distribution and service requirements. This enables the limited transmission power and antenna resources to be adaptively concentrated for the most needed services, thereby achieving efficient coordination and optimal balance of communication, sensing and power transmission resources in the spatial and power dimensions at the system level.

[0027] (4) The embodiments of this application provide a unified resource optimization framework for non-cellular massive MIMO systems, enabling them to simultaneously meet the diversified and differentiated needs for communication, sensing, and power supply in scenarios such as low-altitude economy and intelligent transportation in future 6G networks, thereby enhancing the system's support capabilities for integrated services and its level of intelligence. Attached Figure Description

[0028] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Some specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings in an exemplary and non-limiting manner. The same reference numerals in the drawings indicate the same or similar parts or components. Those skilled in the art should understand that these drawings are not necessarily drawn to scale. In the drawings:

[0029] Figure 1 This is a flowchart illustrating an exemplary embodiment of the present application of a resource optimization method for a cellular-free massive MIMO system with integrated communication, sensing, and energy capabilities.

[0030] Figure 2 This is a schematic diagram illustrating the complete process of a resource optimization method for a cellular-free massive MIMO system, as shown in an exemplary embodiment of this application.

[0031] Figure 3 This is a schematic diagram illustrating a performance comparison of precoded vectors based on precoded covariance matrix and rank recovery, as shown in an exemplary embodiment of this application.

[0032] Figure 4 This is a schematic diagram illustrating a system performance comparison according to an exemplary embodiment of this application;

[0033] Figure 5 This is a block diagram illustrating a cellular-free massive MIMO system in an exemplary embodiment of this application;

[0034] Figure 6 This is a block diagram illustrating an electronic device according to an exemplary embodiment of this application;

[0035] Figure 7 This is a block diagram illustrating an exemplary embodiment of a cellular-free massive MIMO system resource optimization device for integrated communication, sensing, and energy. Detailed Implementation

[0036] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0037] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0038] The embodiments described in this specification will now be described in detail.

[0039] This application provides a resource optimization method for a cellular-free massive MIMO system integrating communication, sensing, and energy. The system includes a central processing unit and multiple distributed access points (APs) for providing services to communication users, sensing targets, and energy users. The method is executed by the central processing unit.

[0040] The non-cellular massive MIMO system serves communication users, sensing targets, and energy users. The system provides high-speed data services to communication users, achieves high-precision detection of sensing targets, and provides efficient wireless power supply to energy users. The central processing unit collects global information such as channel status and target location, and runs the resource optimization method of this embodiment to dynamically determine the optimal operating mode for each access point (AP) and calculate the optimal precoding scheme, thereby maximizing the overall system performance while meeting the basic requirements of various services.

[0041] This embodiment of the non-cellular massive MIMO system includes M access points (APs) controlled collaboratively by a central processing unit. In a certain scenario, the system provides services to U communication users, T energy users, and Q sensing targets. The sets of communication users, energy users, and sensing targets are denoted as follows: , as well as .

[0042] Each AP is equipped with Root transmitting antenna and The root receiving antenna introduces a binary AP mode selection variable. Used to indicate the first The operating modes of each AP, among which Indicates the first Each AP operates in a sensing mode, used to send communication signals and sensing signals; Then it means the first Each AP operates in energy mode and is used to send energy signals.

[0043] Figure 1 This is a flowchart illustrating an exemplary embodiment of a resource optimization method for a cellular-free massive MIMO system with integrated communication, sensing, and performance. Figure 1 As shown, the system resource optimization method 100 includes at least the following steps:

[0044] Step S1: Obtain the current system status information, and determine the performance indicators of each communication user, each sensing target, and each energy user based on the system status information.

[0045] The system status information refers to the real-time data set reflecting the current physical state of the network and service requirements, which is needed for the central processing unit to perform joint optimization. The central processing unit acquires the system status information periodically or on demand through the fronthaul link with its coordinated distributed access points (APs).

[0046] In some embodiments, the system state information includes channel state information, network topology, and service information. The channel state information characterizes the propagation environment of radio waves, including, for example, the communication channel vector from each AP to each communication user, the energy transmission channel vector to each energy user, and the equivalent array direction vector of the transmitter pointing to each sensing target. Each AP obtains the communication channel vector and the energy transmission channel vector through its receiving antenna by performing channel estimation, and obtains the equivalent array direction vector of the transmitter pointing to each sensing target through echo processing of the sensing signal or prior information, and then sends it to the central processing unit. The network topology and service information include the central processing unit maintaining the currently active set of communication users, the set of sensing targets, and the set of energy users, including the number, identification, and service requirements of each service object set, such as the interference signal-to-noise ratio of communication users.

[0047] In response to preset optimization events, such as reaching a preset periodic scheduling time, or detecting events like new user access or target movement, the central processing unit (CPU) collects the current system state information and determines whether the channel state or the set of service objects has changed. If a change has occurred, the CPU triggers a new round of complete optimization, executing steps S2 and S3 based on the new system state information. If no change has occurred, the current network state is determined to be relatively stable. In this case, the CPU will not trigger re-optimization but will directly use the previous optimization calculation results to avoid unnecessary computational overhead and improve system efficiency.

[0048] Step S2: Based on the perception target and the performance indicators of the energy user, construct an optimization target, and construct an optimization problem using the AP mode selection vector and the global precoding vector as decision variables.

[0049] Each AP is configured to operate in either a sensing mode or an energy mode. The constraints of the optimization problem include:

[0050] The performance indicators of each communication user shall not be lower than the preset service quality threshold;

[0051] Each AP's transmit power does not exceed its maximum transmit power; that is, if an AP is optimized for the sensing mode, the component of its transmit power allocated to the energy signal is zero; if an AP is optimized for the energy mode, the component of its transmit power allocated to the communication and sensing signals is zero.

[0052] The total number of APs configured in the sensing mode in the system is within the set range.

[0053] Step S3: Solve the optimization problem to obtain the optimal working mode and optimal precoding vector for each AP.

[0054] like Figure 1 As shown in the system resource optimization method 100, this embodiment, by jointly optimizing the AP operating mode and the global precoding vector, can improve the system's detection performance for sensing targets and its wireless power supply performance for energy users while ensuring the quality of service for all communication users. Through power constraints and AP mode selection mechanisms, power competition and mutual interference between different service signals at the AP level can be resolved. For example, when an AP is optimized for sensing mode, its energy signal power is forced to zero, avoiding self-interference of the energy signal with communication and sensing signals transmitted by the same AP. Conversely, when an AP is optimized for energy mode, its communication and sensing signal power is zero, allowing the AP's transmit power budget and antenna degrees of freedom to be concentrated on the precoding design for energy transmission. This avoids power sharing and spatial directional degree of freedom competition between sensing and power supply services within the same AP, improving the directional gain and power supply controllability for energy users.

[0055] This embodiment allows the central processing unit to dynamically divide the AP cluster into two functional subsets, one focusing on sensing and the other on power supply, based on real-time channel conditions, user distribution, and service requirements. This enables limited transmit power and antenna resources to be adaptively concentrated on the most needed services, thereby achieving efficient coordination and optimal trade-offs of communication, sensing, and power transmission resources in spatial and power dimensions at the system level. Furthermore, it provides a unified resource optimization framework for non-cellular massive MIMO systems, enabling them to simultaneously meet the diversified and differentiated needs of communication, sensing, and power supply in future 6G network scenarios such as low-altitude economy and intelligent transportation, enhancing the system's support capabilities for integrated services and its level of intelligence.

[0056] In some embodiments, the performance metrics for each communication user include signal-to-interference noise ratio (SINR), the performance metrics for each sensing target include signal-to-noise ratio (SNR), and the performance metrics for each energy user include energy harvesting power.

[0057] The system status information includes the aggregated communication channel vectors from all APs to each communication user, the aggregated energy transmission channel vectors from all APs to each energy user, and the equivalent covariance weight matrix of the transmitter of each sensing target.

[0058] Next, combine Figure 2 Steps S1 to S3 are explained in detail.

[0059] In some embodiments, step S1 specifically includes:

[0060] In this embodiment, the communication link, energy link, and sensing link all adopt the Ricean fading model, and the three types of links use the same Ricean factor K. Therefore, the line-of-sight (LoS) propagation weights are... Weights for Non-Line of Sight (NLoS) propagation for

[0061] (1)

[0062] For any communication user The m-th AP to the m-th AP can be obtained through formula (2). Communication channel vectors of a communication user :

[0063] (2)

[0064] in, Represents the field of complex numbers. The dimension is The complex vector space, This represents the large-scale fading coefficient. For the Loss component, For NLoS Rayleigh components. Transfer all APs to the... The aggregated communication channel vector is obtained by stacking the channels of each communication user. :

[0065] (3)

[0066] in , It is the transpose operator. This is an operator that is completely equivalent to .

[0067] Similarly, for any energy user The energy transmission channel vector from the m-th AP to the t-th energy user is: The aggregated energy transmission channel vector is obtained by stacking all APs to the t-th energy user's channels. .

[0068] For any perceived target The equivalent array direction of the transmitter of the m-th AP pointing to the q-th sensing target is... Embedding it into the global dimension yields:

[0069] (4)

[0070] in, This represents the equivalent gain of a two-way link transmitted by the m-th AP, scattered by the q-th sensing target, and received by the receiving antenna. The echo signal, after being received by the AP's receiving antenna, is sent to the central processing unit for centralized processing via the fronthaul link.

[0071] In this embodiment, the equivalent covariance weight matrix of the q-th sensing target at the transmitting end for:

[0072] (5)

[0073] in, This is the conjugate transpose operator. Let represent the spatial correlation matrix of the m-th AP transmitter antenna, and .

[0074] In this way, the aggregated communication channel vectors from all APs to each communication user can be obtained. Aggregated energy transfer channel vectors from all APs to each energy user and the equivalent covariance weight matrix of the transmitter of each sensing target. .

[0075] The global downlink transmit signal vector of the system at a certain moment It carries communication signals, sensing signals, and energy signals, among which the communication signals are specific to each communication user. The precoded variables are , symbol is ;Sensing signals are shared A perceptual stream, whose precoded variables are , conform to The energy signal has a total of Energy flow, precoded variables are , conform to Then the global downlink transmission signal This can be expressed by formula (6):

[0076] (6)

[0077] Then variable , , and These are the decision variables for the optimization problem.

[0078] Assumption , , All are zero-mean and independent random variables, for any communication user Its received signal is Among them, noise , This represents a complex Gaussian distribution. Since communication users consider both the sensed signal and the energy signal as interference, their expected signal power is:

[0079] (7)

[0080] The interference power is caused by interference from other communication users. Perceived interference With energy interference It consists of three parts, as shown in formula (8), and the interference power is as follows:

[0081] (8)

[0082] Therefore, the first The SINR of a communication user can be written as

[0083] (9)

[0084] This embodiment uses the sensing SNR of the sensing target as the sensing performance index. The echo signal power of the sensing target q is... for:

[0085] (10)

[0086] Based on this, the SINR of the q-th perceived target is:

[0087] (12)

[0088] in, This indicates the number of global sensing receiving channels. Indicates the first Noise power of each sensing and receiving channel, This indicates the coherent processing gain.

[0089] For any energy user t, the received signal is Ignoring noise contributions, the received RF power of energy user t is:

[0090] (13)

[0091] For energy harvesting power, considering the different conversion efficiencies of communication, sensing, and energy signals in the rectifier, the energy harvesting power of the t-th energy user is:

[0092] (14)

[0093] Among them, nonnegative parameters The conversion efficiencies of communication, sensing, and energy signals in the rectifier are, in order.

[0094] In some embodiments, step S2 includes: calculating a normalized average signal-to-noise ratio based on the signal-to-noise ratio of all sensing targets, and calculating a normalized average energy harvesting power based on the energy harvesting power of all energy users; calculating a weighted sum of the normalized average signal-to-noise ratio and the normalized average energy harvesting power; and constructing the optimization target by maximizing the weighted sum.

[0095] In some embodiments, step S2 specifically includes:

[0096] To simultaneously consider both sensing performance and energy harvesting performance, this patent constructs the following optimization objective function:

[0097] (15)

[0098] , These are non-negative weighted systems for perception and energy, respectively. The normalized average SNR, Normalized average energy harvesting power:

[0099] (16)

[0100] in, , These are normalization coefficients used to balance the numerical scales of indices with different dimensions. To characterize the power budget and mode coupling of each AP, the global precoding variables are divided into blocks according to AP.

[0101] (17)

[0102] in, Let represent the precoded components of communication, sensing, and energy on the corresponding stream for the m-th AP, respectively. Then, the transmit power allocated to the communication and sensing components and the transmit power allocated to the energy component for the m-th AP are respectively:

[0103] (18)

[0104] In formula (18) For norm budget symbol.

[0105] To maximize As the optimization objective, the AP mode selection variable is used. and the precoded variables of communication users Pre-encoded variables of the perceived target Precoded variables of energy users As parameters to be optimized, the solution is performed under the following constraints:

[0106] The first constraint is that for any communication user Its SINR is required to be no less than the threshold. ,Right now .

[0107] The second constraint is, .

[0108] in, Let m be the maximum transmit power budget for the m-th AP. Forced ,when Forced This enables switch-coupled operation between the AP mode and the transmittable signal components.

[0109] The third constraint is, .

[0110] That is, under the condition of satisfying the binary constraints, the AP mode selection variable also needs to be reduced to the number of APs in synesthesia mode, where and These are the lower and upper bounds of the total number of people in synesthesia mode, respectively.

[0111] Based on this embodiment, it can be seen that the optimization problem of step S2 is... for:

[0112]

[0113] Here, "st" means "subject to", that is, "satisfying the following constraints".

[0114] It can be seen that, due to this optimization problem It also includes binary variables. And the non-convex quadratic coupling terms introduced by the SINR constraint and the objective function, therefore the optimization problem This is a non-convex optimization problem with strong coupling between discrete and continuous variables. Therefore, in step S3 of this embodiment, the quadratic terms of the pre-encoded variables are equivalently transformed into covariance matrices and solved using a continuous relaxation and phased iterative strategy.

[0115] In some embodiments, step S3 includes: converting the quadratic terms corresponding to the global precoding variables into equivalent precoding covariance matrix form, and relaxing the AP mode selection variables into continuous relaxation variables; introducing a penalty term in the optimization objective to make the continuous relaxation variables approach 0 or 1, and constructing a convex upper bound approximation for the penalty term to form an iteratively solvable positive semidefinite programming subproblem; iteratively solving the positive semidefinite programming subproblem until the convergence condition is met to obtain an intermediate solution, and performing a binary decision on the continuous relaxation variables in the intermediate solution to determine the optimal working mode of AP; solving the positive semidefinite programming subproblem again under the optimal working mode to obtain the final precoding covariance matrix; performing eigenvalue decomposition and rank recovery operation on the final precoding covariance matrix to generate the optimal precoding vector.

[0116] Since the AP mode selection variable is discrete and coupled with the global precoding variable under power constraints, the optimization problem is a non-convex optimization problem with strong coupling between discrete and continuous variables. To address this problem, this embodiment employs a phased iterative solution algorithm to obtain the AP mode selection and precoding scheme; the phased solution includes the following three stages:

[0117] In Phase 1, the quadratic terms of the quadratic terms corresponding to the precoded variables are equivalently transformed in the form of covariance matrices, and the rank structure constraints introduced by the precoded variables are relaxed in a positive semidefinite manner; and the binary AP mode selection variables are... Relaxation is a continuous relaxation variable And introduce a mechanism in the objective function that promotes... For penalty terms that approach 0 or 1, construct a first-order upper bound approximation for the non-convex part of the penalty term to obtain the corresponding positive semidefinite programming subproblem.

[0118] Phase Two: Before the iteration begins, an initial reference point is selected. , exemplary And set the penalty coefficient, maximum number of iterations, and convergence threshold. For the ... The next iteration uses the previous round's reference point. For the linearization point, calculate the equivalent linear bias parameter of the penalty term. Based on this, the corresponding convex semi-positive definite programming subproblem for this round is constructed and the covariance matrix is ​​obtained by solving it. With continuous slack variables Then Use this as a reference point for the next iteration and repeat the above process until the convergence condition is met, resulting in a continuous relaxation solution. And its corresponding precoded covariance matrix solution. Through continuous relaxation solutions... Binary transformation yields the optimal operating mode of AP. and in Under the given conditions, solve the corresponding positive semidefinite programming problem again to obtain the final covariance matrix solution.

[0119] In Phase 3, rank recovery is performed based on the final covariance matrix solution to generate an implementable precoding vector. Compliant scaling and power redistribution are then applied to the APs to ensure that the resulting precoding scheme satisfies the switch-type power upper bound constraint and the maximum transmit power budget constraint for each AP, while also satisfying the SINR threshold constraint for communication users, thereby outputting the optimal precoding vector.

[0120] In some embodiments, step S3 specifically includes:

[0121] Phase 1: Covariance equivalent transformation and semidefinite convex optimization modeling.

[0122] Optimization problem The rank of quadratic terms of precoded variables is converted into covariance matrix form:

[0123] (19)

[0124] Here , Represents the trace of a matrix.

[0125] because as well as Then the quadratic terms related to the precoded variables can be uniformly rewritten in matrix trace form, that is:

[0126] (20)

[0127] Therefore, the SINR constraint for the u-th communication user can be equivalently written as:

[0128] (twenty one)

[0129] Similarly, the SNR of the q-th sensing target and the energy harvesting power of the t-th energy user can be equivalently written as:

[0130] (twenty two)

[0131] In this embodiment, a selection matrix is ​​introduced to express the transmit power constraints of each AP in a unified matrix form. The selection matrix is ​​used to extract the block diagonal submatrix corresponding to the m-th AP.

[0132] (twenty three)

[0133] in, This is an operator that arranges a given matrix in order on the diagonal to form a block diagonal matrix.

[0134] Due to arbitrary covariance matrix The transmit power of the m-th AP is Then the switching power constraint for each AP can be expressed as:

[0135] (twenty four)

[0136] Since the covariance matrix is ​​a positive semi-definite matrix, it must satisfy... , here express It is a positive semi-definite matrix. To ensure that the output has a finite number of feasible precoding vectors, this embodiment imposes a rank constraint:

[0137] (25)

[0138] in, This represents the rank of the matrix.

[0139] And, the binary constraints Relaxation is a continuous relaxation constraint over a continuous interval. ,exist Under constraints, a relaxation problem that can be continuously optimized is obtained.

[0140] This embodiment aims to promote continuous slack variables. The iteration converges to 0 and 1, and a penalty term is introduced into the objective function:

[0141] (26)

[0142] in, .

[0143] Thus, this embodiment can address the optimization problem of step S2. Transform into an optimization problem :

[0144]

[0145] Due to this optimization problem It also includes non-convex terms. And the rank constraint, therefore this optimization problem This is a non-convex optimization problem.

[0146] To obtain a computable relaxation model, a positive semidefinite relaxation is applied to the aforementioned rank structure constraints, i.e., the rank constraints are removed, retaining only the positive semidefiniteness. Simultaneously, to address the non-convex penalty term... This is transformed into a form solvable by convex optimization tools. This embodiment is based on a successive principalization solution framework of differential convex decomposition, given a reference point. At that time, The non-convex part is approximated by a first-order linear upper bound based on the following convex inequality:

[0147] (27)

[0148] The The convex upper bound is:

[0149] (28)

[0150] in, To and Irrelevant constant terms.

[0151] Therefore, the objective function is ,in This is the linear bias coefficient. Then the reference point... The corresponding positive semidefinite programming subproblem is:

[0152]

[0153] The above semidefinite programming subproblem This is a standard convex optimization problem, which can be solved using convex optimization solvers such as CVX.

[0154] Phase Two: Iteratively solve the semi-positive definite programming subproblem using successive convex approximations.

[0155] Due to the semidefinite programming subproblem linear bias coefficients in Dependence on reference point If reference point If it remains fixed, then only that reference point can be obtained. The convex approximation solution is located nearby. Based on this, this embodiment adopts a successive convex approximation iterative solution strategy. That is, in each iteration, the continuous solution obtained in the previous round is used as the new reference point to repeatedly construct and solve the convex subproblem, so that the upper bound approximation is continuously tightened as the reference point is updated, thereby forming an iterative update sequence and driving the process. It tends to move towards a binary structure.

[0156] For example, let the iteration count be... Before the iteration begins, initialization is performed first. Then it enters the iterative process: in the first... In each iteration, the current reference point is taken as... Based on this, the linear bias coefficient is calculated. Substitute it into the template of a semidefinite programming subproblem. The optimal relaxation solution can then be obtained by using the CVX isoconvex optimization solver. and the corresponding covariance matrix solution , , Finally, As a reference point for the next round and let Repeat the above process to make It gradually approaches 0 and 1.

[0157] To obtain feasible AP partitioning and eliminate structural uncertainties caused by continuous relaxation, a convergence determination is made for the iterative process, and a binary decision and re-solution of the fixed binary structure are performed when the conditions are met.

[0158] For example, the criterion shown in formula (29) is used as the convergence criterion:

[0159] (29)

[0160] in Indicates the first The objective function value obtained in the nth iteration Let be the convergence threshold. Iteration stops when any convergence criterion is met; otherwise, it continues to the next iteration. Let the continuous solution at convergence be... To obtain a feasible AP partition, construct a binary vector:

[0161]

[0162] Thus, based on Obtain the optimal operating mode of AP.

[0163] Due to the semidefinite programming subproblem China adopts Even with continuous relaxation and binary penalty, the convergent solution may still have relaxation errors in its structure. To obtain a semidefinite relaxed solution that is strictly consistent with the final AP partition, after determining... After that, set And order Solve the following optimization problem. :

[0164]

[0165] The above optimization problem This is a positive semidefinite programming problem, which can be solved using CVX tools to obtain the solution of the precoding covariance matrix under positive semidefinite relaxation. .

[0166] Phase 3: After the rank recovery operation, output the AP selection and precoding matrix.

[0167] Since the optimal solution obtained in stage two is a precoding covariance matrix under positive semidefinite relaxation, it generally does not satisfy the rank structure required for a finite number of precoding vectors. Therefore, this embodiment performs rank recovery on the covariance matrix solution obtained in stage two to generate an implementable global precoding vector, and combines the optimal operating mode of the AP to perform component uniformity and AP power compliance scaling on the global precoding vector, so that the final optimal precoding vector satisfies the on / off power constraints for each AP and the SINR threshold constraints for communication users.

[0168] For any communication user For its covariance matrix Perform eigenvalue decomposition and extract the eigenvector corresponding to the largest eigenvalue. With eigenvalues Constructing communication precoding vectors :

[0169] (30)

[0170] Similarly, for the perceptual covariance matrix Perform eigenvalue decomposition and take the first eigenvalue. Key features Constructing perceptual precoding vectors ; and the energy covariance matrix Perform eigenvalue decomposition and take the first eigenvalue. Key features Constructing energy precoding vectors .

[0171] (31)

[0172] To ensure that APs optimized for sensing mode carry only communication and sensing components, and APs optimized for energy mode carry only energy components, the binary optimal operating mode obtained in Phase 2 is used. Construct the selection matrix:

[0173] (32)

[0174] Based on the selection matrix, the precoded vector is subjected to component unification processing:

[0175] (33)

[0176] The uniformized covariance components are obtained using formula (33). , , .

[0177] Calculate the power of each AP in the communication and sensing components, as well as its power in the energy component:

[0178] (34)

[0179] Even after rank restoration and component uniformity, the transmit power of each AP may still exceed the budget. Therefore, the AP power scaling factor is obtained as shown in equation (35):

[0180] (35)

[0181] And calculate the block diagonal scaling matrix of the AP power scaling factor:

[0182] (36)

[0183] The precoding vector is power-compliantly scaled based on the AP power scaling factor as described in formula (36) to obtain the final output precoding vector:

[0184] (37)

[0185] Each AP satisfies the constraints. and

[0186] .

[0187] Optionally, if the SINR of a communication user falls below the threshold due to power compliance scaling, a power redistribution subproblem can be further solved to restore SINR feasibility, and the final feasible AP selection matrix and precoding matrix can be output.

[0188] To illustrate the resource optimization process of the non-cellular massive MIMO system in this embodiment in detail, the following simulation examples are provided.

[0189] The simulation parameters for this embodiment are set as follows:

[0190] The system comprises nine access points (APs), each equipped with four transmit antennas and two receive antennas. The system serves two sensing targets and two energy users, with the number of communication users varying between two and twelve. Each AP has a maximum transmit power of 1W, and the Rice factor is set to... 6dB, noise power is −108dBm, communication SINR threshold is set to 6dB, energy conversion efficiency parameters The values ​​were set to 0.2, 0.2, and 1 respectively. All schemes were optimized under the same channel conditions and constraints.

[0191] Figure 3This paper presents a comparison of the precoding covariance matrix solution obtained through semidefinite relaxation and the precoding vector solution after rank recovery operation, in terms of average SNR and total energy harvesting power, when the number of communication users increases from 2 to 12. (Reference) Figure 3 It can be seen that when the number of communication users increases from 2 to 12, the curves of the precoding covariance matrix solution and the rank-recovered precoding vector solution basically coincide on the two indicators of average SNR and total energy harvesting power. This indicates that the recovery process from the covariance domain solution to the implementable vector solution is approximately lossless. This verifies that the output process of covariance optimization to rank recovery adopted in the embodiments of this application can maintain system performance close to the covariance domain solution while satisfying the given constraints, and has good engineering feasibility.

[0192] Figure 4 The paper presents a comparison of the average SNR and total energy harvesting power of the resource optimization scheme of this application embodiment, the traditional random energy AP mode selection scheme, and the traditional all-inductive AP model selection scheme when the number of communication users increases from 2 to 12. (Reference) Figure 4 As can be seen, when the number of communication users increases from 2 to 12, the resource optimization scheme of this application embodiment consistently outperforms the other two traditional schemes in both sensing and energy metrics. For example, in terms of sensing, the average SNR is improved by 0.56~0.70dB compared to the random energy AP mode selection scheme, with an average improvement of 0.65dB; compared to the full-sensory AP mode selection scheme, it is improved by 0.58~1.12dB, with an average improvement of 0.80dB. In terms of energy, the total energy harvesting power is improved by 8.5%~60.5% compared to the random energy AP mode selection scheme, with an average improvement of 36.1%; compared to the full-sensory AP mode selection scheme, it is improved by 22.4%~61.4%, with an average improvement of 41.2%.

[0193] The resource optimization scheme of this application, under the premise of satisfying communication QoS and the transmit power budget and switching power constraints of each AP, optimizes AP mode selection and precoding matrix, effectively alleviating the interference coupling and resource contention problems caused by the parallel operation of multiple communication, sensing, and energy services. Based on the simulation results described in this embodiment, this application embodiment, by jointly optimizing AP mode selection and precoding matrix, can significantly improve the sensing target SNR and energy harvesting power of energy users compared to traditional schemes, while ensuring the minimum SINR for communication users. After determining the optimal operating mode of the AP, the precoding covariance matrix solution and the precoding vector solution obtained by rank recovery maintain a high degree of consistency in performance, verifying the effectiveness of this application embodiment.

[0194] This application also provides a cellular-free massive MIMO system for the integration of sensing and communication capabilities. Figure 5A block diagram of a cellular-free massive MIMO system is shown as an exemplary embodiment of this application, such as Figure 5 As shown, the system includes a central processing unit and multiple distributed access points (APs). The APs are connected to the central processing unit via fronthaul links. The central processing unit is configured to execute... Figure 1 The resource optimization method shown is intended to provide services to communication users, sensing targets, and energy users.

[0195] Figure 6 This is a schematic diagram of an electronic device illustrated in this specification according to an exemplary embodiment. Please refer to... Figure 6 At the hardware level, the device includes a processor 602, an internal bus 604, a network interface 606, memory 608, a hardware acceleration device 610, and non-volatile memory 612, and may also include other hardware required for its functions. One or more embodiments of this application can be implemented in software, for example, the processor 602 reads the corresponding computer program from the non-volatile memory 612 into memory 608 and then runs it. Of course, in addition to software implementation, one or more embodiments of this application do not exclude other implementation methods, such as logic devices or a combination of hardware and software, etc. That is to say, the execution subject of the above processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0196] Figure 7 This is a block diagram illustrating an exemplary embodiment of a cellular-free massive MIMO system resource optimization device for integrated communication, sensing, and performance. The system resource optimization device can be applied to, for example... Figure 6 The electronic device shown implements the technical solution of this application. The system resource optimization device includes: an index calculation unit 710, a problem modeling unit 720, and a model solving unit 730, wherein:

[0197] The indicator calculation unit 710 is used to obtain the current system status information and determine the performance indicators of each communication user, each sensing target and each energy user based on the system status information.

[0198] The problem modeling unit 720 is used to construct an optimization objective based on the sensing target and the performance indicators of the energy users. The optimization problem is constructed using AP mode selection variables and global precoding variables as decision variables. The working mode of each AP is configured as either a sensing mode or an energy mode. The constraints of the optimization problem include: the performance indicators of each communication user are not lower than a preset service quality threshold; the transmission power of each AP does not exceed its maximum transmission power; and the total number of APs configured in sensing mode in the system is within a set range.

[0199] The model solving unit 730 is used to solve the optimization problem and obtain the optimal working mode and optimal precoding vector for each AP.

[0200] In some embodiments, the performance metrics for each communication user include interference signal-to-noise ratio (SINR), the performance metrics for each sensing target include signal-to-noise ratio (SNR), and the performance metrics for each energy user include energy harvesting power.

[0201] In some embodiments, the system state information includes aggregated communication channel vectors from all APs to each communication user, aggregated energy transmission channel vectors from all APs to each energy user, and the transmitter equivalent covariance weight matrix of each sensing target.

[0202] In some embodiments, the problem modeling unit 720 is configured to calculate a normalized average signal-to-noise ratio based on the signal-to-noise ratio of all sensing targets, and a normalized average energy harvesting power based on the energy harvesting power of all energy users; calculate a weighted sum of the normalized average signal-to-noise ratio and the normalized average energy harvesting power; and construct the optimization objective by maximizing the weighted sum.

[0203] In some embodiments, the model solving unit 730 is used to convert the quadratic terms corresponding to the global precoded variables into equivalent precoded covariance matrix form, and relax the AP mode selection variables into continuous relaxation variables; introduce a penalty term in the optimization objective to make the continuous relaxation variables approach 0 or 1, and construct a convex upper bound approximation for the penalty term to form an iteratively solvable positive semidefinite programming subproblem; iteratively solve the positive semidefinite programming subproblem until the convergence condition is met to obtain an intermediate solution, and perform a binary decision on the continuous relaxation variables in the intermediate solution to determine the optimal working mode of AP; solve the positive semidefinite programming subproblem again in the optimal working mode to obtain the final precoded covariance matrix; perform eigenvalue decomposition and rank recovery operation on the final precoded covariance matrix to generate the optimal precoded vector.

[0204] For the device embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0205] Accordingly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the methods described in any of the above embodiments.

[0206] Accordingly, embodiments of this application also provide a computer program product configured to perform the methods described in any of the above embodiments.

[0207] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer, which can take the form of a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email sending and receiving device, game console, tablet computer, wearable device, or any combination of these devices.

[0208] In a typical configuration, a computer includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0209] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0210] Computer-readable media, including both permanent and non-permanent, removable and non-removable media, can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random-access memory (SRAM), dynamic random-access memory (DRAM), other types of random-access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage, quantum memory, graphene-based storage media or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0211] While this specification contains numerous specific implementation details, these should not be construed as limiting the scope of any invention or the scope of the claims, but rather are primarily intended to describe features of specific embodiments of a particular invention. Certain features described in the various embodiments herein may also be implemented in combination in a single embodiment. Conversely, various features described in a single embodiment may also be implemented separately in various embodiments or in any suitable sub-combination. Furthermore, while features may function in certain combinations as described above and even initially claimed in this way, one or more features from a claimed combination may be removed from that combination in some cases, and a claimed combination may refer to a sub-combination or a variation thereof.

[0212] Similarly, although the operations are depicted in a specific order in the accompanying drawings, this should not be construed as requiring these operations to be performed in the specific order shown or sequentially, or requiring all illustrated operations to be performed to achieve the desired result. In some cases, multitasking and parallel processing may be advantageous. Furthermore, the separation of various system modules and components in the above embodiments should not be construed as requiring such separation in all embodiments, and it should be understood that the described program components and systems can generally be integrated together in a single software product or packaged into multiple software products.

[0213] Thus, specific embodiments of the subject matter have been described. Other embodiments are within the scope of the appended claims. In some cases, the actions recited in the claims may be performed in a different order and still achieve the desired result. Furthermore, the processes depicted in the drawings are not necessarily shown in a specific order or sequence to achieve the desired result. In some implementations, multitasking and parallel processing may be advantageous.

[0214] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0215] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A resource optimization method for a cellular-free massive MIMO system integrating communication, sensing, and energy, characterized in that, The system includes a central processing unit and multiple distributed access points (APs) for providing services to communication users, sensing targets, and energy users. The method is executed by the central processing unit and includes: Step S1: Obtain the current system status information, and determine the performance indicators of each communication user, each sensing target, and each energy user based on the system status information; the system status information includes the aggregated communication channel vector from all APs to each communication user, the aggregated energy transmission channel vector from all APs to each energy user, and the transmitter equivalent covariance weight matrix of each sensing target. Step S2: Based on the sensing target and the performance indicators of the energy user, an optimization target is constructed. An optimization problem is built using AP mode selection variables and global precoding variables as decision variables. Each AP's operating mode is configured as either a sensing mode or an energy mode. The constraints of the optimization problem include: The performance indicators of each communication user shall not be lower than the preset service quality threshold; The transmit power of each AP shall not exceed its maximum transmit power; The total number of APs configured in synergy mode in the system is within the set range; Step S3 involves solving the optimization problem to obtain the optimal operating mode and optimal precoding vector for each AP; specifically including: The quadratic terms corresponding to the global precoded variables are equivalently converted into the form of precoded covariance matrices, and the AP mode selection variables are relaxed into continuous relaxation variables. A penalty term is introduced into the optimization objective to make the continuous slack variables approach 0 or 1, and a convex upper bound approximation is constructed for the penalty term to form an iteratively solvable positive definite programming subproblem. The positive semidefinite programming subproblem is solved iteratively until the convergence condition is met, and an intermediate solution is obtained. The continuous slack variables in the intermediate solution are then subjected to binary decision to determine the optimal working mode of AP. The positive semidefinite programming subproblem is solved again under the optimal working mode to obtain the final precoding covariance matrix. The final precoding covariance matrix is ​​subjected to eigenvalue decomposition and rank recovery operation to generate the optimal precoding vector.

2. The method according to claim 1, characterized in that, The performance metrics for each communication user include the interference signal-to-noise ratio (SINR), the performance metrics for each sensing target include the signal-to-noise ratio (SNR), and the performance metrics for each energy user include the energy harvesting power.

3. The method according to claim 2, characterized in that, Step S2 includes: The normalized average signal-to-noise ratio is calculated based on the signal-to-noise ratio of all sensing targets, and the normalized average energy harvesting power is calculated based on the energy harvesting power of all energy users. Calculate the weighted sum of the normalized average signal-to-noise ratio and the normalized average energy harvesting power; The optimization objective is constructed to maximize the weighted sum.

4. A resource optimization device for a cellular-free massive MIMO system integrating communication, sensing, and energy, characterized in that, The system includes a central processing unit and multiple distributed access points (APs) for providing services to communication users, sensing targets, and energy users. The device is applied to the central processing unit and includes: The indicator calculation unit is used to obtain the current system status information and determine the performance indicators of each communication user, each sensing target, and each energy user based on the system status information. The system status information includes the aggregated communication channel vector from all APs to each communication user, the aggregated energy transmission channel vector from all APs to each energy user, and the transmitter equivalent covariance weight matrix of each sensing target. The problem modeling unit is used to construct an optimization objective based on the sensing target and the performance indicators of the energy users. The optimization problem is constructed using AP mode selection variables and global precoding variables as decision variables. The working mode of each AP is configured as either sensing mode or energy mode. The constraints of the optimization problem include: the performance indicators of each communication user are not lower than the preset service quality threshold; the transmission power of each AP does not exceed its maximum transmission power; and the total number of APs configured in sensing mode in the system is within a set range. The model solving unit is used to solve the optimization problem and obtain the optimal working mode and optimal precoding vector for each AP. Specifically, the model solving unit is used to convert the quadratic terms corresponding to the global precoded variables into precoded covariance matrix form, and relax the AP mode selection variables into continuous relaxation variables; introduce a penalty term into the optimization objective to make the continuous relaxation variables approach 0 or 1, and construct a convex upper bound approximation for the penalty term to form an iteratively solvable positive semidefinite programming subproblem; iteratively solve the positive semidefinite programming subproblem until the convergence condition is met to obtain an intermediate solution, and perform a binary decision on the continuous relaxation variables in the intermediate solution to determine the optimal working mode of AP; solve the positive semidefinite programming subproblem again under the optimal working mode to obtain the final precoded covariance matrix; perform eigenvalue decomposition and rank recovery operation on the final precoded covariance matrix to generate the optimal precoded vector.

5. A cellular-free massive MIMO system for integrated communication, sensing, and energy transfer, characterized in that: The non-cellular massive MIMO system includes a central processing unit and multiple distributed access points (APs), with the APs connected to the central processing unit via fronthaul links; wherein: The central processing unit is configured to perform the method as described in any one of claims 1 to 3 to provide services to communication users, sensing targets, and energy users.

6. An electronic device, characterized in that, include: processor; as well as A computer-readable storage medium storing computer program instructions that, when executed by the processor, cause the processor to perform the method as described in any one of claims 1 to 3.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which is executed by a processor according to any one of claims 1 to 3.

Images