Flexible scheduling method and system for communication sensing spectrum resources for limited block length transmission

Abstract

The invention discloses a communication sensing spectrum resource flexible scheduling method and system for limited block length transmission, and relates to the technical field of communication. The method comprises the following steps: 1, constructing a system model; 2, analyzing the performance under different resource allocation schemes based on a system model, measuring the communication performance by a reachable limited block length transmission rate, and measuring the sensing performance by a reachable sensing rate; 3, determining an optimization problem, and constructing a weighting and rate maximization problem based on the communication rate and the sensing rate; and 4, designing an optimization scheme, decomposing a highly-coupled non-convex optimization problem into three sub-problems, converting each sub-problem into a solvable convex optimization problem through an approximation method, respectively solving the solvable convex optimization problem, and obtaining a solution of an original problem through an alternate iteration method until the algorithm converges or reaches the maximum number of iterations. According to the invention, sharing of a communication function and a sensing function on spectrum resources is realized, adaptive improvement of systems aiming at different emphasized targets is realized, and ultra-reliable low-delay transmission is supported.

Description

Technical Field

[0001] The present invention belongs to the field of communication technologies, and more specifically, relates to a method and system for flexible scheduling of communication-aware spectrum resources for finite block length transmission. Background Art

[0002] The Internet of Everything (IoE) technology bridges the gap between virtual networks and the real physical world, promoting wide applications in fields such as smart home, smart city, industrial automation, and healthcare. However, on the one hand, the surge in IoE devices has led to a huge increase in wireless spectrum demand, causing channel congestion and affecting communication quality; on the other hand, various IoE applications have put forward stringent requirements for reliability, latency, and high-precision environmental perception capabilities, posing significant challenges to network design. In addition, for the uplink short data packet transmission requirements generated by a large number of IoE devices, how to ensure their efficient and reliable transmission remains an unsolved problem.

[0003] Integrated Sensing and Communication (ISAC) technology has become a revolutionary paradigm for future wireless networks, which can meet the urgent need for coexistence of information interaction and environmental perception capabilities. By sharing spectrum resources, the communication and sensing functions are coordinated. Therefore, ISAC can significantly improve spectrum efficiency, and further optimize the performance by combining Non-Orthogonal Multiple Access (NOMA) technology. Its environmental perception ability can also provide real-time information such as device location and channel state for the communication system, effectively improving transmission reliability. In addition, finite block length transmission naturally fits the requirements of low-latency IoE applications, but due to the fact that channel distortion cannot be fully averaged, the decoding error probability cannot be ignored. How to achieve fast decision-making, precise resource allocation, and environmental perception through ISAC technology, and strike a balance among transmission rate, latency, and reliability, has become the key to optimizing the performance of IoE networks. Summary of the Invention

[0004] To achieve flexible scheduling of communication and sensing spectrum resources and further improve the low-latency communication and high-reliability sensing capabilities of wireless networks, the present invention aims to propose a finite blocklength transmission framework for an ISAC system based on uplink NOMA. Three spectrum resource allocation methods for NOMA-based communication signals and sensing signals, as well as two successive interference cancellation (SIC) orders, are designed. By establishing the achievable finite blocklength transmission rate and sensing rate, the communication performance and sensing performance are quantified respectively. Further, to achieve the balance and adaptive coordination of communication capabilities and sensing capabilities, a weighted sum rate maximization problem is designed, and by adjusting the weight parameters of communication performance indicators and sensing performance indicators, flexible improvement of the system for different emphasis targets is realized.

[0005] To solve the above technical problems, the technical solution adopted by the present invention is: a method for flexible scheduling of communication and sensing spectrum resources for finite blocklength transmission, including the following steps:

[0006] Step 1, construct an ISAC system model based on uplink NOMA. The ISAC system model includes a dual-functional base station, device A and device B equipped with single antennas, and a sensing target;

[0007] Step 2, quantify communication performance indicators and sensing performance indicators. Based on the ISAC system model constructed in Step 1, analyze the performance under several different resource allocation schemes. Among them, the communication performance is measured by the achievable finite blocklength transmission rate, and the sensing performance is measured by the achievable sensing rate;

[0008] Step 3, determine the optimization problem. Based on the communication rate and sensing rate obtained in Step 2, construct a weighted sum rate maximization problem to achieve the trade-off between communication performance and sensing performance;

[0009] Step 4, design an optimization scheme. Decompose the highly coupled non-convex optimization problem constructed in Step 3 into three sub-problems: transmit power allocation, receive combining matrix optimization, and sensing signal design. Transform each sub-problem into a solvable convex optimization problem through an approximation method and solve them respectively, and use the alternating iteration method until the algorithm converges or reaches the maximum number of iterations to obtain the solution of the original problem.

[0010] Preferably, in Step 1, the ISAC system model includes a dual-functional base station (DFBS) equipped with N t uniform linear array (ULA) transmit antennas and N r ULA receive antennas, device A and device B equipped with single antennas, and a sensing target; the device set is defined as

[0011] From a communication perspective, the device sends a finite block length signal, and the DFBS is responsible for receiving it. The communication symbol sent by device k is x k , where k ∈ {A, B}, and its transmit power is p k , and it satisfies (zero mean) and (unit power); The communication channel is characterized by a Rayleigh fading model, and its mathematical expression is:

[0012]

[0013] where, represents the small-scale fading component, characterizes the large-scale path loss; It is assumed that the DFBS can obtain the complete channel state information;

[0014] From a sensing perspective, the DFBS transmits a sensing waveform and receives the radar echo reflected by the target. The power of the sensing signal satisfies the constraint s H s ≤ p s , p s is the maximum power of the sensing signal; The target response matrix is modeled as where represents the target response from the transmitting array to the nth r receiving antenna; According to the basic theoretical research of the widely distributed antenna array, the inter-column correlation of the target response matrix is ignored; This matrix can be further decomposed into where is the target reflection coefficient that follows a complex Gaussian distribution, characterizing the combined effect of the two-way path loss and the target radar cross-section area, represents the average intensity of the target reflection, and are the transmit steering vector and the receive steering vector respectively, and θ is the target azimuth angle; The target sensing process can be equivalently regarded as a target response matrix estimation problem; The received signal of the DFBS is:

[0015]

[0016] where represents additive white Gaussian noise with variance σ 2 .

[0017] Preferably, in step two,

[0018] I. Quantification of communication performance: Given the code block length L k , the tolerable decoding error probability ε k, when receiving the signal-to-interference-plus-noise ratio (SINR) γ k , the achievable finite blocklength transmission rate of the device is

[0019]

[0020] where is the channel dispersion, which is used to measure the random variability of the channel relative to a deterministic channel with the same capacity, and Q -1 [·] represents the inverse function of the Gaussian Q-function ; assuming that the receive combining vector when the base station decodes the signal x k is w k , then the detailed analysis of the received SINR γ k is as follows:

[0021] · OMA-ISAC mode: The communication signal and the sensing signal occupy different frequency band resources, and there is no interference between the two signals with different functions. Since the NOMA method is used between communication signals, the undecoded signal will be used as the interference of the currently decoded signal. Therefore,

[0022] where represents the interference caused by the undecoded communication signal;

[0023] · NOMA-ISAC, communication-centered SIC order: All signals are transmitted in the NOMA mode, the communication and sensing functions share all spectrum resources, and SIC is performed at the receiving end. First, the sensing signal is decoded and removed, and the sensing signal will not interfere with the communication signal; therefore,

[0024] where represents the interference caused by the undecoded communication signal;

[0025] · NOMA-ISAC, sensing-centered SIC order: First, the communication signal is decoded, and there is interference caused by the sensing signal when decoding the communication signal; therefore, where represents the interference caused by the undecoded communication signal, represents the interference plus noise caused by the sensing signal;

[0026] · Semi-NOMA-ISAC, communication-centered SIC order: NOMA is used for signals with different functions, while OMA is used between different communication signals. There is no interference between communication signals, and the sensing signal is decoded and removed first, and the sensing signal will not interfere with the communication signal; therefore,

[0027] · Semi - NOMA - ISAC, sensing - centered SIC order: First decode the communication signal, and there is interference caused by the sensing signal when decoding the communication signal; thus, where represents the interference plus noise caused by the sensing signal;

[0028] II. Sensing performance quantization; set (positive definite matrix); Use the achievable sensing rate as the metric to evaluate the sensing performance; the achievable sensing rate is related to both the spectrum resource allocation method and the SIC order. The detailed analysis is as follows:

[0029] · OMA - ISAC method: where B represents all available bandwidth, β represents the bandwidth allocation coefficient, and (1 - β)B represents the bandwidth occupied by the sensing signal;

[0030] · NOMA - ISAC, communication - centered SIC order: where represents the interference plus noise experienced by the base station when decoding the sensing signal, characterizes the large - scale path loss;

[0031] · NOMA - ISAC, sensing - centered SIC order:

[0032] · Semi - NOMA - ISAC, communication - centered SIC order:

[0033] where represents the interference plus noise experienced by the base station when decoding the sensing signal, β k B represents the partial bandwidth occupied by device k, characterizes the large - scale path loss;

[0034] · Semi - NOMA - ISAC, sensing - centered SIC order:

[0035] Preferably, in step three, by jointly optimizing the device transmit power, the combining vector at the DFBS side, and the sensing signal design, maximize the weighted sum of the communication rate and the sensing rate; this optimization problem is formulated in the following mathematical form:

[0036]

[0037] where, (C1) and (C2) are the maximum power constraint conditions, P k,max and p sare the maximum transmission powers of the device and the base station respectively, (C3) and (C4) represent the communication QoS constraint and the sensing QoS constraint respectively, and Γ k and Γ s represent the minimum communication rate threshold and the minimum sensing rate threshold respectively, and (C5) and (C6) are the preset IDI SIC order and IFI SIC order constraints respectively.

[0038] Preferably, in step four, rewrite the finite blocklength transmission rate in formula (3) into the following form:

[0039]

[0040] where the parameter can be regarded as a constant, and there is ι k > 0;

[0041] Using the first-order Taylor expansion, the achievable finite blocklength transmission rate can be approximated as

[0042]

[0043] where is an auxiliary function constructed for the convenience of writing, represents the first derivative of g(γ k ), represents the first-order Taylor expansion of g(γ k ), represents the feasible point. It can be proved that is an increasing concave function of γ k ;

[0044] Preferably, the solutions of the three sub-problems are as follows:

[0045] The first sub-problem: Transmission power allocation

[0046] Establish a concave lower bound of log2(1 + γ k ) as α k log2γ k + η k ≤ log2(1 + γ k ), where the auxiliary variable Then the lower bound of log2(1 + γ k,OMA ) is as follows

[0047]

[0048] C k,OMA is a concave function of q k ; The optimization problem is solved using the CVX toolbox;

[0049] The second sub-problem: Receive combining matrix optimization

[0050] Given the device power p k , let satisfy W k ≥ 0, and Rank(W k ) = 1; Further, define Then there is, can be rewritten as

[0051]

[0052] First, deal with the non - convexity of the F k,1 term; Expand at the feasible point , and we can get

[0053]

[0054] To deal with the non - convexity of the F k,2 term, introduce the slack variables v = {v1,..., v K}, where each element satisfies Then there is

[0055]

[0056] We can get the lower bound of

[0057]

[0058] is a concave function of W k ; By ignoring the Rank(W k ) = 1 rank - one constraint, transform the optimization problem into a solvable convex optimization problem, and use the CVX toolbox to solve it; Use the Gaussian randomization method to restore the rank - one constraint;

[0059] The third sub - problem: Sensing signal design

[0060] Define to satisfy S ≥ 0, S = S H , Tr(S) ≤ p s and Rank(S) = 1; Introduce the following equivalent formula ||S|| * - ||S||2 = 0, and introduce it as a penalty term into the objective function;

[0061]

[0062] Use CVX to solve the optimization problem; By gradually reducing the penalty term until ||S|| * - ||S||2 ≤ ζ s, a rank-one solution to the optimization problem can be obtained.

[0063] Preferably, the joint optimization algorithm based on alternating iteration is as follows:

[0064] In each iteration, the transmit power allocation, sensing signal design, and receive beamforming design of the IoE device are alternately optimized, where the initial point of each iteration is the solution of the previous iteration; finally, when the algorithm converges, a high-quality sub-optimal solution to the initial optimization problem can be obtained; the summary is as follows:

[0065]

[0066] A communication-aware spectrum resource flexible scheduling device for finite blocklength transmission includes a dual-functional base station, IoE devices, and sensing targets; the dual-functional base station receives uplink transmission signals, transmits sensing signals, and receives sensing echo signals to detect sensing targets; the uplink transmission signals received by the dual-functional base station specifically include signals from IoE devices and sensing echo signals; the IoE devices transmit finite blocklength coded signals to the base station, and the sensing targets do not actively transmit any signals.

[0067] The beneficial effects of adopting the above technical solutions are as follows: This method solves the problem of flexible allocation of spectrum resources in a communication-sensing integrated system. Based on the ISAC framework of uplink NOMA technology, it realizes the sharing of communication functions and sensing functions on spectrum resources. Through the design of different spectrum resource allocation methods, the design of SIC order, and the design of the weights of communication functions and sensing functions, adaptive improvements to systems with different emphasis targets are realized. At the same time, this method is based on the finite blocklength coding theory and supports ultra-reliable low-latency transmission. On this basis, through the collaborative scheduling of multi-dimensional resources, the efficient integration of communication functions and sensing functions is realized, and the utilization rate of spectrum resources is improved. Description of the Drawings

[0068] Figure 1 is the spectrum resource allocation method for communication signals and sensing signals;

[0069] Figure 2 is the serial interference cancellation order;

[0070] Figure 3 is the technical structure diagram;

[0071] Figure 4 is the ISAC model based on uplink NOMA;

[0072] Figure 5 is the simulation Figure 1 ;

[0073] Figure 6 is the simulation Figure 2 ;

[0074] Figure 7 is a simulation Figure 3 。 Specific embodiments

[0075] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0076] The system of the present invention includes a dual-functional base station, an IoE device, and a sensing target. The dual functions that the base station can perform simultaneously include: receiving an uplink transmission signal, transmitting a sensing signal, and receiving a sensing echo signal to detect the sensing target. The function of the base station to receive the uplink transmission signal includes receiving signals from the IoE device and receiving the sensing echo signal. Through the joint design of the dual functions of the base station and considering the joint optimization of the transmission power of the IoE device at the same time, the weighted sum of the achievable finite blocklength transmission rate and the sensing rate can be maximized, where the communication performance weight coefficient is ω c , the sensing performance weight coefficient is ω s , satisfying ω c +ω s = 1. By flexibly adjusting ω c and ω s , the adaptive improvement of the system for different focused targets can be achieved.

[0077] According to the actual requirements of the system, the spectrum resource allocation method for communication signals and sensing signals can be flexibly selected. Figure 1 Is the spectrum resource allocation method for communication signals and sensing signals.

[0078] Spectrum resource allocation method 1: NOMA-ISAC. In this scheme, all signals are transmitted in the NOMA mode. The communication and sensing functions share all the spectrum resources, and SIC is performed at the receiving end. The communication signal and the sensing signal simultaneously occupy the entire available bandwidth B.

[0079] Spectrum resource allocation method 2: OMA-ISAC. In this scheme, OMA is adopted between different functional signals, which respectively occupy independent spectrum resources; while NOMA is used between multiple communication signals to achieve time-frequency resource reuse. This means that the communication and sensing functions operate in orthogonal frequency bands without interfering with each other. Among them, the communication signal occupies a bandwidth of βB, and the radar echo signal occupies a bandwidth of (1-β)B, where β∈(0,1) is the bandwidth allocation coefficient. At this time, there is no Inter-Function-Interference (IFI) in the system, and the Inter-Device-Interference (IDI) is eliminated by the SIC method.

[0080] Spectrum resource allocation method 3: Semi-NOMA-ISAC. In this scheme, NOMA is adopted for different functional signals, while OMA is used between different communication signals. The communication signals x A and x B each occupy bandwidths of βB and (1-β)B respectively, and the sensing signal occupies the entire available bandwidth B. At this time, there is no IDI in the system, and the IFI is eliminated by the SIC method.

[0081] Among them, the SIC method for eliminating IFI can be divided into two orders: communication-centered and sensing-centered. In the former, the communication signal is first regarded as an interference signal, the sensing echo signal is decoded and eliminated, and then the communication signal is decoded; in the latter, the sensing echo signal is first regarded as an interference signal, the communication signal is decoded and eliminated, and then the sensing echo signal is decoded. According to the actual requirements and emphases of the system, the SIC order can be flexibly selected. Figure 2 is the serial interference cancellation order.

[0082] The IoE device transmits a finite blocklength coded signal to the base station to reduce the transmission delay.

[0083] The communication performance is quantified by the achievable finite blocklength transmission rate, and the sensing performance is quantified by the achievable sensing rate. The performance loss term introduced by the finite blocklength transmission makes the objective function highly coupled and strictly non-convex. Therefore, the present invention aims to propose an approximation method to transform the finite blocklength transmission rate into a form that is easy to handle. In addition, for each spectrum resource allocation scheme and SIC order, a joint solution method based on alternating optimization is developed until the algorithm converges or reaches the maximum number of iterations. Figure 3 is the technical structure diagram.

[0084] The method of the present invention is specifically divided into the following four steps:

[0085] Step 1, construct a system model.

[0086] Consider an uplink NOMA-based ISAC system model, which includes a base station equipped with N tA dual-functional base station (DFBS) with M uniform linear array (ULA) transmit antennas and N r ULA receive antennas, devices A and B equipped with single antennas, and a sensing target, as Figure 4 shown. The set of devices is defined as

[0087] From a communication perspective, the devices send finite blocklength signals and the DFBS is responsible for receiving. The communication symbol sent by device k is x k , k ∈ {A, B}, with transmit power p k , and satisfying (zero mean) and (unit power). The communication channel is characterized by a Rayleigh fading model, and its mathematical expression is:

[0088]

[0089] where represents the small-scale fading component, and characterizes the large-scale path loss. It is assumed that the DFBS can obtain the complete channel state information.

[0090] From a sensing perspective, the DFBS transmits a sensing waveform and receives the radar echo reflected by the target. The power of the sensing signal satisfies the constraint s H s ≤ p s , p s being the maximum power of the sensing signal. The target response matrix is modeled as where represents the target response from the transmit array to the nth r receive antenna. Based on the basic theoretical research of widely-spaced antenna arrays, we neglect the inter-column correlation of the target response matrix. This matrix can be further decomposed as where is the target reflection coefficient following a complex Gaussian distribution, characterizing the combined effect of the two-way path loss and the target radar cross-section, represents the average intensity of the target reflection, and are the transmit steering vector and the receive steering vector respectively, and θ is the target azimuth angle. It should be noted in particular that the target sensing process can be equivalently regarded as a problem of estimating the target response matrix. The received signal at the DFBS is:

[0091]

[0092] where represents the variance as σ 2Additive white Gaussian noise.

[0093] Step 2, quantization of communication performance and sensing performance.

[0094] 1) Quantization of communication performance. To effectively reduce the transmission delay, finite blocklength transmission is considered. However, due to the average deficiency of thermal noise and channel distortion, there will be a performance loss compared with long packet transmission. Given the code block length L k , the tolerable decoding error probability ε k , and the received signal-to-interference-plus-noise ratio (SINR) γ k , the achievable finite blocklength transmission rate of the device is

[0095]

[0096] where is the channel dispersion, which is used to measure the random variability of the channel relative to a deterministic channel with the same capacity, and Q -1 [·] represents the Gaussian Q-function the inverse function of dt. Among them, the received SINR γ k is related to both the spectrum resource allocation method and the SIC order. Let the received combining vector when the base station decodes the signal x k be w k , then the detailed analysis of the received SINR γ k is as follows.

[0097] · OMA-ISAC mode: The communication signal and the sensing signal occupy different frequency band resources, and there is no interference between the two signals with different functions. However, since the NOMA method is used between communication signals, the undecoded signal will be used as the interference of the currently decoded signal. Therefore,

[0098] where represents the interference caused by the undecoded communication signal.

[0099] · NOMA-ISAC, communication-centered SIC order: All signals are transmitted in the NOMA mode, the communication and sensing functions share all spectrum resources, and SIC is performed at the receiving end. First, the sensing signal is decoded and removed, and the sensing signal will not interfere with the communication signal. Therefore,

[0100] where represents the interference caused by the undecoded communication signal.

[0101] · NOMA-ISAC, perception-centric SIC order: Decode the communication signal first. When decoding the communication signal, there is interference caused by the sensing signal. Therefore, where represents the interference caused by the undecoded communication signal, represents the interference plus noise caused by the sensing signal.

[0102] · Semi-NOMA-ISAC, communication-centric SIC order: Apply NOMA to different functional signals and OMA to different communication signals. There is no interference between communication signals, and the sensing signal is decoded and removed first, so the sensing signal does not cause interference to the communication signal. Therefore,

[0103] · Semi-NOMA-ISAC, perception-centric SIC order: Decode the communication signal first. When decoding the communication signal, there is interference caused by the sensing signal. Therefore, where represents the interference plus noise caused by the sensing signal.

[0104] 2) Perception performance quantization. Different from the instantaneous target response matrix G, the correlation matrix remains relatively stable over a long time and can be approximated by a large amount of historical estimation data. To ensure the feasibility of theoretical analysis, further set (positive definite matrix). The achievable sensing rate is used as an index to evaluate the perception performance. This choice is based on the unified requirement for the performance evaluation of the dual-functional system, aiming to achieve a systematic trade-off analysis of the communication-sensing cooperative efficiency. The achievable sensing rate is related to both the spectrum resource allocation method and the SIC order. The detailed analysis is as follows, where the analysis method of SINR is similar to the above communication performance quantization analysis method.

[0105] · OMA-ISAC mode: where B represents all available bandwidth, β represents the bandwidth allocation coefficient, and (1-β)B represents the bandwidth occupied by the sensing signal.

[0106] · NOMA-ISAC, communication-centric SIC order: where represents the interference plus noise experienced by the base station when decoding the sensing signal, characterizes the large-scale path loss.

[0107] · NOMA-ISAC, perception-centric SIC order:

[0108] · Semi-NOMA-ISAC, communication-centric SIC order: where represents the interference plus noise experienced by the base station when decoding the sensing signal, β k B represents the partial bandwidth occupied by device k, characterizes the large-scale path loss.

[0109] ·Semi-NOMA-ISAC, sensing-centric SIC order:

[0110] Step 3, determine the optimization problem.

[0111] To achieve the co-optimization and trade-off of communication and sensing capabilities, this paper jointly optimizes the device transmit power, the combining vector at the DFBS side, and the sensing signal design to maximize the weighted sum of the communication rate and the sensing rate. This optimization problem can be formulated in the following mathematical form:

[0112]

[0113] where (C1) and (C2) are the maximum power constraint conditions, P k,max and p s are the maximum transmit powers of the device and the base station respectively, (C3) and (C4) represent the communication QoS constraint and the sensing QoS constraint respectively, Γ k and Γ s represent the minimum communication rate threshold and the minimum sensing rate threshold respectively, and (C5) and (C6) are the preset IDI SIC order and IFI SIC order constraints respectively.

[0114] Step 4, design the optimization scheme.

[0115] It can be seen that due to the non-convexity of the objective function and the constraint conditions, (P) is a strictly non-convex optimization problem. In addition, the highly coupled optimization variables {p, w k , s} make it difficult to obtain the global optimal solution. Therefore, this method develops an algorithm based on alternating optimization (AO) to find a high-quality suboptimal solution. First, an approximation method is proposed to rewrite the finite blocklength transmission rate in Equation (3) into the following more tractable form:

[0116]

[0117] where the parameter can be regarded as a constant, and ι k > 0.

[0118] Using the first-order Taylor expansion, the achievable finite blocklength transmission rate can be approximated as

[0119]

[0120] where An auxiliary function constructed for the convenience of writing, denotes the first derivative of g(γ k ). denotes the first-order Taylor expansion of g(γ k ). represents a feasible point. It can be proven that is an increasing concave function with respect to γ k .

[0121] Taking the optimization scheme design in the OMA-ISAC mode as an example, it is decomposed into the following three sub-problems.

[0122] 1) Transmit power allocation

[0123] Since the achievable sensing rate R s,OMA is independent of the device transmit power p k , the sensing performance metric does not need to be considered when performing power allocation. Establish a concave lower bound of log2(1 + γ k ) as α k log2γ k + η k ≤ log2(1 + γ k ), where the auxiliary variable Then the lower bound of log2(1 + γ k,OMA ) is as follows

[0124]

[0125] C k,OMA is a concave function with respect to q k . The optimization problem can be solved using the CVX toolkit.

[0126] 2) Receive combining matrix optimization

[0127] When performing receive combining matrix optimization, only the communication performance needs to be considered. Given the device power p k , the optimization problem is reformulated into a more tractable form. Let satisfy W k ≥ 0, and Rank(W k ) = 1. Further, define Then there is

[0128] Accordingly, can be rewritten as

[0129]

[0130] First, deal with the non-convexity of the F k,1 term. At the feasible point Expanding at this point gives

[0131]

[0132] To handle the non - convexity of term F k,2 a slack variable \(v=\{v_1,\ldots,v K \}\) is introduced, where each element satisfies Then we have

[0133]

[0134] We can obtain the lower bound of

[0135]

[0136] which is a concave function of \(W k By ignoring the rank - one constraint \(\text{Rank}(W k ) = 1\), the optimization problem can be transformed into a solvable convex optimization problem and can be solved using the CVX toolbox. The rank - one constraint can be recovered using the Gaussian randomization method.

[0137] 3) Sensing signal design

[0138] It should be noted that the formulas related to communication performance are independent of the sensing signal \(s\). Therefore, when designing the sensing signal, only how to maximize the achievable sensing rate needs to be considered. Define satisfying \(S\geq0\), \(S = S H \), \(\text{Tr}(S)\leq p s and \(\text{Rank}(S)=1\). To handle the non - convex rank - one constraint, the following equivalent formula \(\|S\| * -\|S\|^2 = 0\) is introduced and used as a penalty term in the objective function.

[0139]

[0140] The optimization problem can be solved using CVX. By gradually reducing the penalty term until \(\|S\| * -\|S\|^2\leq\zeta s is satisfied, a rank - one solution to the optimization problem can be obtained.

[0141] For joint optimization, we propose a joint optimization algorithm based on alternating iteration to maximize the communication - sensing weighted sum rate. More specifically, in each iteration, the transmit power allocation, sensing signal design, and receive beamforming design of the IoE device are alternately optimized, where the initial point of each iteration is the solution of the previous iteration. Finally, when the algorithm converges, a high - quality sub - optimal solution to the initial optimization problem can be obtained. Summarized as follows:

[0142]

[0143] Note that for the solutions under other power allocation schemes and SIC orders, this method is still applicable, and only adaptive improvements need to be made to it. This will not be elaborated here.

[0144] The effects of the embodiments of the present invention can be further illustrated by simulation.

[0145] In the simulation, the parameters are set as follows: the carrier frequency is 4 GHz, the available bandwidth is 10 MHz, the noise power is -104 dBm, the path loss coefficient is -20 dB, and the average target response intensity is 1 / 20. The number of IoE devices is 2, and the number of transmitting and receiving antennas of the base station is 16 each. The transmission block length is 1000, and the tolerable decoding error probability is 10 -5 。

[0146] Figure 5 Shows the influence of the communication weight coefficient on the achievable weighted sum rate of the system, and the following phenomena can be observed. 1) For the OMA-ISAC scheme and the sensing-centered SIC order, the achievable weighted sum rate decreases monotonically with the increase of the communication weight. This is because in this scheme, the sensing signal is not interfered by the communication signal. When the communication weight increases (i.e., the sensing weight decreases), the decreasing effect of the weighted sensing rate is more significant; 2) When the communication-centered SIC order is adopted, the weighted sum rate first decreases and then increases with the communication weight, and reaches the minimum value at ω c =ω s =0.5. This balance point corresponds to a working state where the communication and sensing performances are relatively balanced; 3) There is a significant bilateral effect in the setting of the weight coefficient. When the communication weight is too small, although the total weighted rate is high, it will lead to the deterioration of the communication performance. Similarly, too small a sensing weight will also cause damage to the sensing performance. This conclusion shows that by flexibly designing the weight coefficient and the SIC order, the expected performance trade-off between communication and sensing functions can be achieved.

[0147] Figure 6 Depicts the relationship between the achievable weighted sum rate of the system and the number of transmitting and receiving antennas. The number of transmitting antennas and receiving antennas is always equal. As the number of antennas increases, the achievable transmission rate performance of the system is significantly improved. First, the increase in the number of antennas can significantly improve the system capacity and spectral efficiency, which is mainly due to the enhancement of the spatial multiplexing gain; second, more antennas help to improve the sensing accuracy and reliability, and the improvement of the sensing performance is beneficial to further improve the overall system performance; finally, through the comparative analysis of the schemes, it is found that when the OMA-ISAC scheme is adopted, since the communication and sensing functions are independent of each other, the increase in the number of antennas brings more degrees of freedom, so its achievable rate growth rate is the most significant.

[0148] Figure 7It shows the impact of the transport block length on the weighted sum rate performance, and the following phenomena can be observed. 1) As the block length increases, the achievable weighted sum rate is significantly improved. This is because a longer transport block can effectively reduce the impact of channel dispersion, thereby improving the achievable transmission rate; 2) Progressive convergence characteristic. When the block length is large enough, the joint optimization design based on finite block length transmission will gradually approach the performance of infinite block length transmission; 3) Delay-performance trade-off. The block length is closely related to the transmission delay. A larger block length will necessarily lead to a higher delay. This means that achieving real-time communication through finite block length transmission requires sacrificing rate and reliability. This finding provides an important basis for future research on the multi-dimensional trade-off between block length, achievable rate, delay, and reliability.

[0149] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes, shall be covered by the protection scope of the present invention.

Claims

1. A communication-aware spectrum resource flexible scheduling method for finite blocklength transmission, characterized in that It includes the following steps: Step 1: Construct an ISAC system model based on uplink NOMA. The ISAC system model includes a dual-functional base station, Device A and Device B equipped with single antennas, and a sensing target; Step 2: Quantify the communication performance metrics and sensing performance metrics. Based on the ISAC system model constructed in Step 1, analyze the performance under several different resource allocation schemes. Among them, the communication performance is measured by the achievable finite blocklength transmission rate, and the sensing performance is measured by the achievable sensing rate; Step 3: Determine the optimization problem. Based on the communication rate and sensing rate analyzed in Step 2, construct a weighted sum rate maximization problem to achieve the trade-off between communication performance and sensing performance; Step 4: Design an optimization scheme. Decompose the highly coupled non-convex optimization problem constructed in Step 3 into three sub-problems: transmit power allocation, receive combining matrix optimization, and sensing signal design; transform each sub-problem into a solvable convex optimization problem through approximation methods and solve them respectively, and use the alternating iteration method until the algorithm converges or reaches the maximum number of iterations to obtain the solution of the original problem.

2. A flexible scheduling method for communication-aware spectrum resources for finite blocklength transmission according to claim 1, characterized in that In step one, the ISAC system model includes a dual-functional base station DFBS equipped with N t uniform linear array ULA transmit antennas and N r ULA receive antennas, device A and device B equipped with single antennas, and a sensing target; the device set is defined as From a communication perspective, the device sends a finite block length signal, and the DFBS is responsible for receiving it; the communication symbol sent by device k is x k , where k ∈ {A, B}, and its transmit power is p k , and it satisfies zero mean, and unit power; the communication channel is characterized by a Rayleigh fading model, and its mathematical expression is: Among them, represents the small-scale fading component, characterizes the large-scale path loss; assume that the DFBS obtains the complete channel state information; From a sensing perspective, the DFBS transmits a sensing waveform and receives the radar echo reflected by the target. The sensing signal power satisfies the constraint s H s ≤ p s , where p s is the maximum power of the sensing signal; the target response matrix is modeled as where represents the target response from the transmitting array to the nth r receiving antenna; according to the basic theoretical research of the distributed antenna array, the column - to - column correlation of the target response matrix is ignored; this matrix is further decomposed into where is the target reflection coefficient that follows a complex Gaussian distribution, representing the combined effect of the two - way path loss and the target radar cross - section area, represents the average intensity of the target reflection, and are the transmitting steering vector and the receiving steering vector respectively, and θ is the target azimuth angle; the target sensing process can be equivalently regarded as a target response matrix estimation problem; the received signal of the DFBS is: wherein represents additive white Gaussian noise with variance σ 2 .

3. A communication-aware spectrum resource flexible scheduling method for finite blocklength transmission according to claim 2, characterized in that In Step 2, Communication performance quantification: Given the code encoding block length L k , the tolerable decoding error probability ε k , the received signal-to-interference-plus-noise ratio SINRγ k , in the case of , the achievable finite blocklength transmission rate of the device is Among them, is the channel dispersion, which is used to measure the random variability of the channel relative to a deterministic channel with the same capacity, and Q -1 [·] represents the inverse function of the Gaussian Q-function ; assume that the receive combining vector when the base station decodes the signal x k is w k , then the analysis of the received SINR γ k is as follows: OMA-ISAC method: Communication signals and sensing signals occupy different frequency band resources, and there is no mutual interference between the two signals with different functions. Since the NOMA method is adopted between communication signals, the undecoded signals will act as interference to the currently decoded signals. where represents the interference caused by undecoded communication signals. NOMA-ISAC, communication-centric SIC order: all signals are transmitted in NOMA mode, the communication and sensing functions share all spectrum resources, and SIC is performed at the receiver. The sensing signal is decoded and removed first, and the sensing signal does not interfere with the communication signal; therefore, where represents the interference caused by the undecoded communication signal; NOMA-ISAC, perception-centered SIC order: first decode the communication signal, and there is interference caused by the sensing signal during the decoding of the communication signal; therefore, where represents the interference caused by the undecoded communication signal, represents the interference caused by the sensing signal plus noise; Semi-NOMA-ISAC, communication-centered SIC order: NOMA is used for different functional signals, while OMA is used for different communication signals. The communication signals do not interfere with each other, and the sensing signal is decoded and removed first, so the sensing signal does not interfere with the communication signals. Therefore, Semi-NOMA-ISAC, sensing-centered SIC order: First, decode the communication signal, and there is interference caused by the sensing signal during the decoding of the communication signal; therefore, where represents the interference plus noise caused by the sensing signal; II. Sensing performance quantification; Set R > 0 (positive definite matrix); Use the achievable sensing rate as the index to evaluate the sensing performance; The achievable sensing rate is related to both the spectrum resource allocation method and the SIC order; The detailed analysis is as follows: OMA-ISAC mode: where B represents the total available bandwidth, β represents the bandwidth allocation coefficient, and (1-β)B represents the bandwidth occupied by the sensing signal; NOMA-ISAC, communication-centered SIC order: where represents the interference plus noise experienced by the base station when decoding the sensing signal, characterizes the large-scale path loss; NOMA-ISAC, sensing-centered SIC order: Semi-NOMA-ISAC, communication-centric SIC order: where represents the interference plus noise experienced by the base station when decoding the sensing signal, and β k B represents the partial bandwidth occupied by device k, characterizes the large-scale path loss; Semi-NOMA-ISAC, sensing-centric SIC order:

4. A communication-aware spectrum resource flexible scheduling method for finite blocklength transmission according to claim 3, characterized in that In Step 3, maximize the weighted sum of the communication rate and the sensing rate by jointly optimizing the device transmit power, the combining vector at the DFBS, and the sensing signal design; This optimization problem is formulated in the following mathematical form: Among them, (C1) and (C2) are the maximum power constraint conditions, P k,max and p s are the maximum transmission powers of the device and the base station respectively. (C3) and (C4) represent the communication QoS constraint and the sensing QoS constraint respectively. Γ k and Γ s represent the minimum communication rate threshold and the minimum sensing rate threshold respectively. (C5) and (C6) are the preset IDI SIC order and IFI SIC order constraints respectively.

5. A communication-aware spectrum resource flexible scheduling method for finite blocklength transmission according to claim 4, characterized in that In Step 4, rewrite the finite blocklength transmission rate in Equation (3) as the following form: where the parameter is regarded as a constant, and there is ι k > 0; Using the first-order Taylor expansion, the achievable finite blocklength transmission rate is approximated as Among them is an auxiliary function constructed for ease of writing, represents the first derivative of g(γ k ), represents the first-order Taylor expansion of g(γ k ), represents a feasible point, is an increasing concave function with respect to γ k .

6. A communication-aware spectrum resource flexible scheduling method for finite blocklength transmission according to claim 5, characterized in that The solutions of the three sub-problems are as follows: The first sub-problem: Transmit power allocation Establish a concave lower bound of log2(1 + γ k ) as α k log2γ k + η k ≤ log2(1 + γ k ), where the auxiliary variable Then the lower bound of log2(1 + γ k,OMA ) is as follows C k,OMA is a concave function with respect to q k The optimization problem is solved using the CVX toolbox; The second sub-problem: Receive combining matrix optimization Given the device power p k , let satisfy and Rank(W k ) = 1; Further, define Then there is can be rewritten as First, deal with the non-convexity of term F k,1 ; expand at the feasible point to obtain To handle the non-convexity of term F k,2 a slack variable \(v = \{v_1,\ldots,v K \}\) is introduced, where each element satisfies Then we have It can be obtained The lower bound of is a concave function with respect to W k ; by ignoring the rank-one constraint of Rank(W k ) = 1, the optimization problem is transformed into a solvable convex optimization problem and solved using the CVX toolkit; the rank-one constraint is restored using the Gaussian randomization method; The third sub-problem: Sensing signal design Definition S ≥ 0, S = S H , Tr(S) ≤ p s and Rank(S) = 1; introduce the following equivalent formula ||S|| * -||S||² = 0, and introduce it as a penalty term into the objective function; Solve the optimization problem using CVX; by gradually reducing the penalty term until ||S|| * -||S||2 ≤ ζ s is satisfied, a rank-one solution to the optimization problem can be obtained.

7. A communication-aware spectrum resource flexible scheduling method for finite blocklength transmission according to claim 6, characterized in that The joint optimization algorithm based on alternating iteration is as follows: In each iteration, alternately optimize the transmit power allocation, sensing signal design, and receive beamforming design of the IoE device, where the initial point of each iteration is the solution of the previous iteration; Finally, when the algorithm converges, obtain a high-quality sub-optimal solution to the initial optimization problem.

8. A communication-aware spectrum resource flexible scheduling system for finite blocklength transmission, characterized in that It includes a dual-functional base station, IoE devices, and a sensing target; The dual-functional base station receives the uplink transmission signal, transmits the sensing signal, and receives the sensing echo signal to detect the sensing target; The uplink transmission signal received by the dual-functional base station includes the signal from the IoE device and the sensing echo signal; The IoE device transmits a finite blocklength coded signal to the base station, and the sensing target does not actively transmit any signal.

Images