Joint trajectory and resource allocation method for otfs-based secure uav isac system

Abstract

The present application relates to the OTFS-based safe unmanned aerial vehicle ISAC system joint trajectory and resource allocation method, belongs to the field of integrated sensing. The system model is constructed by using time delay-Doppler domain orthogonal time-frequency space modulation technology to suppress the serious Doppler spread caused by high mobility, and the safety and sensing cooperation problem is expressed as a joint optimization problem with the maximum average secrecy rate as the target, and meeting the sensing performance constraint, power constraint and unmanned aerial vehicle mobility constraint, an efficient alternating iteration solution algorithm is designed by fusing multiple strategies, which efficiently processes the non-convex optimization problem, and considers the security communication and sensing performance. The advantage is to solve the Doppler effect interference problem existing in the traditional orthogonal frequency division multiplexing reference scheme in the high mobility environment, and the collaborative optimization of security communication and sensing performance can be realized, while effectively supporting the reliable operation of the safe ISAC system, the security-sensing performance trade-off performance is significantly improved.

Description

Technical Field

[0001] This invention belongs to the field of Sensor-Communication Integration (ISAC) technology, and relates to the physical layer security problem of UAVs based on Orthogonal Time-Frequency Space OTFS in ISAC systems. In particular, it relates to a joint trajectory and resource allocation of a secure UAV ISAC system based on OTFS. Background Technology

[0002] Integrated Sensing and Communication (ISAC) simultaneously carries wireless communication and environmental sensing functions on the same spectrum resources, hardware platform, and signal waveform. This improves spectrum utilization, reduces system deployment costs, and provides crucial support for applications in 6G mobile communication, such as intelligent transportation, low-altitude networks, emergency communications, target monitoring, and integrated air-space-ground networks. With the continuous growth in the number of wireless terminals and the demand for sensing services, the problems of spectrum occupation, hardware redundancy, and insufficient coordination efficiency caused by the separate deployment of traditional communication and radar sensing systems are becoming increasingly prominent. Therefore, the deep integration of communication and sensing functions has become an important research direction for the development of wireless networks.

[0003] Among various ISAC platforms, drones possess advantages such as high mobility, flexible deployment, adjustable coverage, and easy access to line-of-sight links, making them suitable as airborne base stations, airborne relays, or airborne sensing nodes to serve ground users and sensing targets. Compared to fixed ground base stations, drones can improve link geometry by adjusting their flight trajectories, enabling rapid establishment of communication and sensing services in complex obstructed environments or temporary mission scenarios. Therefore, introducing drones into ISAC systems is beneficial for improving service coverage, sensing flexibility, and network robustness in dynamic environments. However, drone-based ISAC systems also face more prominent security challenges. Due to the broadcast characteristics of wireless channels, confidential information sent by drones to legitimate users may be received by unauthorized ground nodes or potential eavesdropping nodes. Simultaneously, drones need to transmit detection signals into space when performing sensing tasks, a process that may expose the system's operational status and target-related information. In integrated sensing and communication security scenarios, the system not only needs to improve the data transmission rate of legitimate links but also needs to suppress the information acquisition capabilities of eavesdropping links while simultaneously ensuring the detection quality of sensing tasks. Therefore, there is a natural resource competition between communication confidentiality and sensing performance.

[0004] Existing physical layer security techniques usually enhance the advantage of the legitimate link over the eavesdropping link through beamforming, artificial noise, power allocation, cooperative jamming or trajectory optimization, etc. These methods can improve the secrecy rate to some extent, but there are still deficiencies in the UAV ISAC system. The reason is that the UAV trajectory will simultaneously affect the path loss, channel gain and observation geometry of the legitimate user, eavesdropping node and sensing target; the configuration of communication resources, sensing resources and transmit power will also jointly determine the secrecy rate and sensing accuracy. Therefore, if only a single dimension is optimized, it is difficult to fully coordinate the coupling relationship between trajectory, power, communication resources and sensing resources. In addition, the UAV platform is in a state of continuous motion, and the communication and sensing channels usually have obvious time-varying characteristics. Under high mobility conditions, the channel will produce significant Doppler shift and double selective fading. The traditional Orthogonal Frequency Division Multiplexing (OFDM) system mainly modulates and allocates resources in the time-frequency domain, and its subcarrier orthogonality is easily destroyed by Doppler spread, which in turn produces inter-carrier interference, resulting in a decline in communication reliability and degradation of sensing parameter estimation performance. Especially in the scene where the UAV speed is high, the carrier frequency is high, or the sensing target changes significantly, the adaptability of the OFDM system to the Doppler effect is limited.

[0005] Orthogonal Time Frequency Space (OTFS) modulation is a new modulation method for 6G systems, which maps information symbols to the Delay-Doppler (DD) domain, so that the fast-changing time-frequency channel can present a relatively stable and sparse structure in the equivalent representation in the Delay-Doppler domain. Compared with traditional time-frequency domain transmission methods, OTFS can more effectively characterize the characteristics of multipath delay and Doppler shift, and has strong robustness to channel changes in high mobility scenarios. Therefore, applying OTFS to the UAV-assisted ISAC system can improve the stability of communication and the reliability of sensing in a strong Doppler environment, and provide a new design space for resource allocation based on the structure of the Delay-Doppler domain.

[0006] Although OTFS provides an effective waveform basis for high-speed moving UAV ISAC systems, there is still a lack of joint design methods in existing research that meet the collaborative needs of secure communication and target perception. Specifically, in a secure UAV ISAC system, time-delay-Doppler domain resource units can be used for both secure communication and target perception; the allocation of transmission power will simultaneously affect the signal-to-noise ratio of the legitimate link, the signal-to-noise ratio of the eavesdropping link, and the signal-to-noise ratio of the perception; and the trajectory of the UAV will change the link distance, the channel strength, and the dominant Doppler position. The above factors are coupled with each other, so that the system optimization problem usually appears as a mixed integer non-convex problem containing continuous variables, binary variables, and nonlinear constraints. The direct use of exhaustive search or traditional global optimization methods has high computational complexity and is difficult to meet the real-time requirements of actual systems. On the other hand, the perception constraints in the secure UAV ISAC system usually need to be dynamically met under different working conditions. When the perception threshold is strict, the system needs to allocate more resources for the perception task; when the perception requirement is relatively loose, too much perception resource will compress the secure communication resource and reduce the average secure rate. The existing fixed communication / perception resource ratio or fixed power ratio method is difficult to adapt to different perception constraints and channel state changes, and is prone to problems such as not meeting the requirements of perception performance or excessive loss of secure performance. Therefore, a mechanism is needed that can adaptively adjust the communication and perception resource ratio according to the system operating state.

[0007] In summary, the existing technology still has the following problems in the high-speed moving secure UAV ISAC scenario: first, the traditional OFDM system is sensitive to the Doppler spread caused by the movement of the UAV, and it is difficult to fully guarantee the communication and perception performance; second, existing security designs mainly focus on a single link such as trajectory, power, or beam, and lack joint optimization schemes combined with the time-delay-Doppler domain resource structure of OTFS; third, there is strong coupling between communication security, perception constraints, UAV mobility, and power limitations, resulting in high direct solution complexity; fourth, a fixed perception ratio is difficult to adapt to the secure-perception trade-off under different perception requirements. Therefore, it is necessary to propose a joint trajectory and resource allocation method for a secure UAV ISAC system based on OTFS, which jointly considers the UAV trajectory, resource allocation, transmission power, and perception ratio adjustment in the time-delay-Doppler domain, to improve the average secure rate and high mobility robustness while guaranteeing the perception performance. SUMMARY

[0008] The present application provides a joint trajectory and resource allocation method for a secure UAV ISAC system based on OTFS, to solve the problems of severe Doppler effect, mutual restriction of communication security and perception performance, strong coupling of trajectory and resource allocation, and high direct optimization solution complexity in high-speed moving UAV ISAC systems.

[0009] The technical scheme adopted by the present application is as follows:

[0010] Step 1, build a secure unmanned aerial vehicle ISAC system scene, make the unmanned aerial vehicle as an aerial ISAC platform, send secret information to the legal ground user and perform sensing on the target at the same time under the condition that there is a potential eavesdropping node;

[0011] Step 2, establish a time delay-Doppler domain signal model based on an orthogonal time frequency space (OTFS), organize the transmission resources into a time delay-Doppler domain resource grid, and set a communication allocation variable, a sensing allocation variable and a transmission power variable for each time delay-Doppler domain resource unit;

[0012] Step 3, establish a geometric channel model and a radial Doppler model related to the trajectory of the unmanned aerial vehicle, and map the Doppler shift generated by the movement of the unmanned aerial vehicle into a dominant Doppler grid index;

[0013] Step 4, based on the time delay-Doppler domain effective channel gain of the legal link and the eavesdropping link, build a secrecy rate model on each time delay-Doppler domain resource unit, including the signal-to-noise ratio, the secrecy gain, the time slot secrecy rate and the long-term average secrecy rate, and use the sensing performance index based on the Cramer-Rao bound (CRB) to represent the sensing quality;

[0014] Step 5, establish a joint optimization problem with the goal of maximizing the long-term average secrecy rate, and the constraint conditions of the joint optimization problem include the sensing performance constraint, the total power constraint of the time slot, the power constraint of a single resource unit, the communication and sensing resource mutual exclusion constraint and the unmanned aerial vehicle mobility constraint;

[0015] Step 6, decompose the joint optimization problem into four sub-problems: trajectory optimization, time delay-Doppler domain resource allocation, transmission power control and sensing proportion update;

[0016] Step 7, solve by using an alternating iteration method, output the optimized unmanned aerial vehicle trajectory, time delay-Doppler domain communication resource allocation matrix, time delay-Doppler domain sensing resource allocation matrix and transmission power matrix, and use them to realize the secure communication and sensing integrated transmission in a high mobility scenario.

[0017] In step 1 of the application, the specific process of building a secure unmanned aerial vehicle ISAC system scene includes:

[0018] Suppose that the unmanned aerial vehicle flies at a fixed height H, and the horizontal trajectory of the unmanned aerial vehicle is discretized into N time slots, and the horizontal position of the unmanned aerial vehicle in the nth time slot is represented as ;

[0019] Suppose that the horizontal positions of the legal ground user Bob and the potential eavesdropping node Eve are represented as , The drone sends confidential information to Bob, a legitimate ground user, and performs target perception tasks in the presence of a potential eavesdropping node, Eve.

[0020] In the nth time slot, the transmit power on the delay-Doppler domain resource cell (k,l) is expressed as: Where k represents the Doppler index, l represents the time delay index, and the communication allocation variable is represented as... Perceptual assigned variables are represented as ;

[0021] To prevent the same latency-Doppler domain resource unit from simultaneously undertaking communication and sensing tasks, resource mutual exclusion constraints are applied:

[0022] .

[0023] In step 2 of this invention, the specific process of establishing a time-delay-Doppler domain signal model based on OTFS includes:

[0024] Let x[k,l] be the transmission symbol on the time-delay-Doppler domain resource grid;

[0025] After OTFS modulation, wireless channel transmission, and OTFS demodulation, the time-delay-Doppler domain signal at the receiving end is represented as follows:

[0026] ;

[0027] Where k' and l' represent the Doppler index and delay index of the transmitted symbols during the summation process, respectively; The equivalent delay from transmitting resource unit (k',l') to receiving resource unit (k,l) is represented by the Doppler domain channel response. This represents additive white noise;

[0028] The time-delay-Doppler domain signal model is used to describe the time-delay-Doppler coupling relationship caused by UAV motion, and serves as the basis for subsequent time-delay-Doppler domain resource allocation.

[0029] In step 3 of this invention, the specific process of establishing the geometric channel model and radial Doppler model related to the UAV trajectory includes:

[0030] For node i, b and e represent the legitimate ground user node Bob and the potential eavesdropper node Eve, respectively, w i For the corresponding node location, the large-scale channel gain from the UAV to node i in the nth time slot is expressed as:

[0031] ;

[0032] wherein, β0 represents a reference channel gain;

[0033] The radial Doppler shift of the UAV relative to the legitimate ground user Bob in the nth time slot is represented as:

[0034] ;

[0035] wherein, λ represents a carrier wavelength, the radial Doppler shift is mapped to a dominant Doppler grid index:

[0036] ;

[0037] wherein, represents a delay-Doppler domain Doppler resolution.

[0038] In step 4 of the application, the specific process of constructing the secrecy rate model includes:

[0039] In the nth time slot, the signal-to-noise ratio of the resource unit (k, l) at node i is represented as:

[0040] ;

[0041] wherein, σ² represents noise power, represents a delay-Doppler domain effective channel gain;

[0042] The delay-Doppler domain effective channel gain is represented as:

[0043] ;

[0044] wherein, L i [n] the large-scale channel gain of the UAV to node i in the nth time slot, represents a nominal delay-Doppler domain channel coefficient, represents a channel degradation factor caused by mobility;

[0045] The secrecy gain of the resource unit (k, l) in the nth time slot is represented as:

[0046] ;

[0047] The system secrecy rate in the nth time slot is represented as:

[0048] ;

[0049] wherein, ; the long-term average secrecy rate is represented as:

[0050] .

[0051] In step 4 of the application, the sensing performance index based on the Cramer-Rao bound (CRB) is used to represent the sensing quality, and specifically includes:

[0052] The sensing performance index in the nth time slot is represented as:

[0053] ;

[0054] Wherein C represents a proportional constant, represents a coherence time, represents a sensing signal-to-noise ratio (SNR);

[0055] The sensing SNR is determined by the time-delay-Doppler domain resource unit allocated for the sensing task and its transmission power, and is represented as:

[0056] ;

[0057] In step 5 of the application, the joint optimization problem is represented as:

[0058] ;

[0059] Wherein represents a sensing CRB threshold, represents a total power upper limit of a single time slot, represents a power upper limit of a single time-delay-Doppler domain resource unit, represents a maximum speed of the UAV, represents a time slot length.

[0060] In step 6 of the application, the trajectory optimization sub-problem optimizes the UAV trajectory q under the condition that the time-delay-Doppler domain resource allocation and the transmission power are fixed;

[0061] Since the path loss term and the Doppler term are non-convex functions with respect to q[n], the non-convex expression is locally linearized around the current iteration point, a convex approximation problem is constructed, and the successive convex approximation (SCA) method is used to update the UAV trajectory;

[0062] Through the trajectory optimization, the UAV gradually moves to a position that is beneficial to the legal link and far away from the eavesdropping node, so as to enhance the channel asymmetry between the legal link and the eavesdropping link;

[0063] The time-delay-Doppler domain resource allocation sub-problem is executed under the condition that the UAV trajectory and the transmission power are fixed, and specifically includes:

[0064] According to the dominant Doppler grid index A Doppler sensing weight function is introduced:

[0065] ;

[0066] wherein, denotes the weight strength, denotes the weight distribution width;

[0067] define the effective communication utility of the delay-Doppler domain resource unit based on the Doppler-aware weight function:

[0068] ;

[0069] In the resource allocation, the delay-Doppler domain resource unit with higher effective communication utility and conducive to meeting the awareness constraint is preferentially selected, and the necessary resources are reserved for the awareness task according to the CRB constraint, so that the communication and awareness resource allocation simultaneously meet the mutual exclusion constraint and the awareness performance constraint.

[0070] In step 6 of the application, the transmit power control sub-problem and the awareness ratio update sub-problem include:

[0071] Under the condition that the UAV trajectory and the delay-Doppler domain resource allocation are fixed, the projection gradient ascent method is used to update the transmit power of each delay-Doppler domain resource unit; the projection gradient ascent method first adjusts the transmit power in the direction of improving the secrecy rate, and then projects the update result back to the feasible region that meets the single-time-slot total power constraint and the single-resource-unit power constraint;

[0072] Introducing an awareness ratio variable When it is detected that the awareness constraint in the nth time slot is not met, the awareness ratio is increased according to the following rules:

[0073] ;

[0074] wherein, denotes the awareness ratio update step, denotes the upper limit of the awareness ratio;

[0075] Through the awareness ratio update, the system can adaptively adjust the ratio between the communication resource and the awareness resource according to the awareness constraint satisfaction situation.

[0076] The alternating iteration solving process in step 7 of the application includes:

[0077] Initialize the system parameters, the delay-Doppler domain channel, the UAV trajectory, the communication resource allocation matrix, the awareness resource allocation matrix, the transmit power matrix and the iteration index;

[0078] Update the geometric channel gain, the radial Doppler frequency shift, the dominant Doppler grid index and the delay-Doppler domain equivalent channel gain according to the current UAV trajectory;

[0079] Update the communication resource allocation matrix and the sensing resource allocation matrix based on CRB sensing constraints and Doppler sensing utility;

[0080] Under the condition of fixed resource allocation and trajectory, the transmit power matrix is ​​updated by projective gradient ascent;

[0081] Under the condition of fixed resource allocation and transmission power, the drone trajectory is updated via SCA;

[0082] The perception ratio or perception weight is adaptively updated based on the satisfaction of perception constraints.

[0083] Calculate the average security rate and the sensing CRB. If the convergence condition is met or the maximum number of iterations is reached, output the updated UAV trajectory, communication resource allocation matrix, sensing resource allocation matrix, and transmit power matrix; otherwise, proceed to the next iteration.

[0084] The present invention has the following beneficial effects: First, the present invention utilizes the structured channel representation of OTFS in the delay-Doppler domain to improve the robustness of UAVs to Doppler spread in high-mobility scenarios; Second, the present invention jointly designs UAV trajectory, delay-Doppler domain resource allocation, and power control, which can simultaneously enhance the advantages of legitimate links over eavesdropping links in spatial geometry, resource grid, and power dimensions; Third, the present invention introduces an adaptive sensing ratio update mechanism, which can dynamically increase sensing resources when sensing constraints become stricter, thereby avoiding the failure of sensing performance to meet requirements due to the pursuit of confidentiality rates; Fourth, the present invention adopts a combination of alternating iteration and low-complexity heuristic allocation, which reduces the solution complexity of mixed integer non-convex problems and facilitates engineering implementation. Attached Figure Description

[0085] Figure 1 This is a schematic diagram of the ISAC system for secure unmanned aerial vehicles (UAVs) according to the present invention.

[0086] Figure 2 This is a schematic diagram of the delay-Doppler domain resource organization and signal processing based on OTFS in this invention;

[0087] Figure 3 This is a flowchart of the alternating iterative solution algorithm of the present invention;

[0088] Figure 4 This is a convergence behavior diagram of the hybrid joint design method proposed in this invention;

[0089] Figure 5 This is a comparison chart of ablation experiment results using the hybrid joint design method of this invention;

[0090] Figure 6 This is a comparison chart of the optimized trajectories of drones under different schemes of the present invention;

[0091] Figure 7 This is a comparison diagram of the confidentiality-perception trade-off between the adaptive perception ratio strategy of this invention and the fixed ratio baseline;

[0092] Figure 8 This is a comparison chart of the average security rate of the OTFS scheme and OFDM scheme of this invention at different UAV speeds;

[0093] Figure 9 This is a comparison chart of the average security rate of the OTFS scheme and OFDM scheme under different transmit power budgets. Detailed Implementation

[0094] The technical solution of the present invention will be described in detail below with reference to specific embodiments. This embodiment takes a rotary-wing UAV as an aerial ISAC platform as an example. The UAV flies at a fixed altitude and simultaneously performs secure communication and target perception tasks within a discrete time period. The legitimate ground communication user is denoted as Bob, and the potential eavesdropping node is denoted as Eve. Eve's location can be considered known according to the worst-case security baseline, thus facilitating a conservative design of the system's security performance.

[0095] Step 1: Construct the secure communication scenario and time-delay Doppler domain resource variables for the UAV ISAC system. Let the UAV's flight altitude be H, and its horizontal trajectory be discretized into N time slots. The horizontal position of the UAV in the nth time slot is represented as... (R is the set of horizontal real numbers). The horizontal positions of legitimate users and eavesdropping nodes are represented as follows: and OTFS transmission resources are divided into several delay-Doppler domain resource units in the delay-Doppler domain. The transmit power of resource unit (k,l) in the nth time slot is expressed as... The communication allocation variable is represented as Perceptual assigned variables are represented as .

[0096] To prevent the same latency-Doppler domain resource unit from simultaneously undertaking communication and sensing tasks, the following mutual exclusion constraint is applied to each resource unit:

[0097]

[0098] Step 2: Establish the OTFS delay-Doppler domain signal model. Let x[k,l] be the transmit symbol on the delay-Doppler domain resource grid, where k and l represent the Doppler index and delay index, respectively, and k' and l' are the Doppler and delay index traversals. After OTFS modulation, wireless channel transmission, and OTFS demodulation, the delay-Doppler domain signal at the receiving end can be expressed as:

[0099]

[0100] Where k' and l' represent the Doppler index and delay index of the transmitted symbols during the summation process, respectively; The equivalent delay from transmitting resource element (k',l') to receiving resource element (k,l) – the Doppler domain channel response – is represented. This represents additive white noise. The above model can describe the time-delay-Doppler coupling relationship caused by UAV motion and provide a basis for subsequent resource allocation in the time-delay-Doppler domain.

[0101] Step 3: Establish channel gain and Doppler models related to the UAV trajectory. For nodes... Let b and e represent the legitimate ground user node Bob and the potential eavesdropper node Eve, respectively. The large-scale channel gain from the UAV to node i in the nth time slot is:

[0102]

[0103] Where β0 represents the reference channel gain. The above equation shows that the drone trajectory alters the geometric distance between the legitimate link and the eavesdropping link, thereby changing the channel strength difference between the two links, which is an important basis for physical layer security design.

[0104] The drone's movement also generates a radial Doppler component relative to the legitimate user, and the Doppler frequency shift relative to Bob in the nth time slot is approximately...

[0105]

[0106] Where λ is the carrier wavelength. To correlate UAV motion with time-delay-Doppler domain resource allocation, the Doppler frequency shift is mapped to the following dominant Doppler raster index:

[0107]

[0108] in, Let be the time-delay-Doppler domain Doppler resolution. According to the above equation, the UAV trajectory not only affects path loss but also influences the Doppler position in the time-delay-Doppler domain that is more suitable for allocating communication or sensing tasks.

[0109] Step 4: Based on the effective channel gain in the delay-Doppler domain for both legitimate and eavesdropping links, establish a confidentiality rate and sensing performance model for each resource unit in the delay-Doppler domain. In the nth time slot, the signal-to-noise ratio of resource unit (k,l) at node i is expressed as...

[0110]

[0111] in, For noise power, The time-delay-Doppler domain equivalent channel gain is expressed as follows:

[0112]

[0113] in, Represents the nominal delay-Doppler domain channel coefficients. This represents the channel degradation factor caused by mobility. The security gain of resource element (k,l) in the nth time slot is defined as...

[0114]

[0115] Only time-delay-Doppler domain resource units allocated for communication purposes participate in the transmission of secure information; therefore, the system security rate in the nth time slot is...

[0116]

[0117] in, The long-term average rate of secrecy is

[0118]

[0119] In terms of perception, this invention uses a surrogate index based on the Cramer-Rao boundary to describe perception performance. The perception index in the nth time slot is...

[0120]

[0121] Where C is a proportionality constant. For coherent time, The sensing signal-to-noise ratio (SNR) is determined by the delay-Doppler domain resource elements allocated to the sensing task and their transmit power, and can be expressed as:

[0122]

[0123] Step 5: Construct the joint optimization problem. Based on the above model, this invention aims to maximize the long-term average security rate while simultaneously satisfying sensing performance, transmission power, resource exclusivity, and UAV motion constraints. The joint optimization problem can be expressed as:

[0124]

[0125] in, To sense the CRB threshold, This represents the upper limit of total power in a single time slot. This represents the upper limit of power for a single-delay-Doppler domain resource unit. The maximum speed of the drone, Let be the time slot length. The trajectory variable is nonlinearly coupled to the objective function and constraints through path loss and Doppler terms. The resource allocation variable is a binary variable, and the power variable is coupled with the security rate and sensing performance. Therefore, this problem belongs to the mixed integer nonconvex optimization problem.

[0126] Step 6: To reduce the complexity of the solution, this invention decomposes the problem into four modules based on the physical effects of the variables: trajectory optimization, time-delay-Doppler domain resource allocation, power control, and sensing ratio update.

[0127] Step 6.1, Trajectory Optimization. Under the conditions of fixed time-delay-Doppler domain resource allocation and transmit power, the UAV trajectory q is optimized. The path loss and Doppler term in this subproblem remain non-convex functions. This invention locally linearizes the non-convex expression near the current iteration point, constructing a convex approximation problem, and updates the trajectory through a successive convex approximation method. Through this step, the UAV can gradually move to a position more favorable to legitimate links while moving away from eavesdropping nodes, thereby enhancing channel asymmetry.

[0128] Step 6.2, Delay-Doppler Domain Resource Allocation. Under the condition of fixed UAV trajectory and power, obtain the dominant Doppler grid index. To utilize the Doppler structure, this invention introduces a Doppler sensing weighting function:

[0129]

[0130] in, Indicates weight strength. This represents the width of the weight distribution. Based on this weight, the effective communication utility of a time-delay-Doppler domain resource unit is defined as:

[0131]

[0132] During resource allocation, priority is given to time-delay-Doppler domain resource units with high effective communication utility and those conducive to satisfying perception constraints, while reserving necessary resources for perception tasks based on CRB constraints. This allocation rule simultaneously satisfies the mutual exclusion constraints of communication and perception, and avoids complex global binary search.

[0133] Step 6.3, Power Control. Under the condition that the UAV trajectory and delay-Doppler domain resource allocation are fixed, the transmit power of each delay-Doppler domain resource unit is optimized. Since the security rate is still non-convex with respect to the power variable, this invention adopts a projected gradient ascent method for updating: first, the power is adjusted along the gradient direction that increases the security rate, and then the updated result is projected back into the feasible region that satisfies the total power constraint and the power constraint of a single resource unit, thereby gradually improving the security performance while maintaining the feasibility of the constraints.

[0134] Step 6.4, Adaptive Perception Scale Update. To avoid the disruption of perception constraints due to excessive resource allocation to secure communication during the alternating iteration process, this invention introduces a perception scale variable. When a sensing constraint is not met in the nth time slot, the sensing ratio is increased according to the following rules:

[0135]

[0136] in, To sense the scale update step size, This represents the upper limit of the sensing ratio. Using the above formula, the system can adaptively adjust the resource allocation between communication and sensing when the sensing target or environmental conditions change.

[0137] Step 7: Solve the problem using an alternating iterative approach, outputting the optimized UAV trajectory, time-delay-Doppler domain communication resource allocation matrix, time-delay-Doppler domain sensing resource allocation matrix, and transmit power matrix. These are used to achieve secure integrated sensing and communication transmission in highly mobile scenarios. The overall iterative process is as follows:

[0138] (1) Initialize system parameters, including time-delay-Doppler domain grid size, UAV flight altitude H, and maximum speed. Average or peak transmit power, sensing threshold And the maximum number of iterations;

[0139] (2) Generate or estimate the delay-Doppler domain channels of the legitimate link, the eavesdropping link, and the target sensing link;

[0140] (3) Initialize the UAV trajectory Delay-Doppler domain communication allocation matrix Perception allocation matrix Transmit power matrix and the iteration index i=0;

[0141] (4) Based on the current trajectory Update the geometric channel gain, Doppler frequency shift, dominant Doppler grid index, and time-delay-Doppler domain equivalent channel gain;

[0142] (5) Update the delay-Doppler domain communication / sensing resource allocation matrix based on CRB sensing constraints and Doppler sensing utility. and ;

[0143] (6) Under the condition of fixed resource allocation and trajectory, update the transmit power matrix by projecting gradient ascent. ;

[0144] (7) Under the condition of fixed resource allocation and transmission power, update the UAV trajectory through SCA. ;

[0145] (8) Adaptively update the perception ratio or perception weight based on the satisfaction of perception constraints. ;

[0146] (9) Calculate the average secrecy rate If the convergence condition or the maximum number of iterations is reached, the UAV trajectory is output using the perception CRB. Delay-Doppler domain communication resource allocation matrix Delay-Doppler domain sensing resource allocation matrix and transmit power matrix Otherwise, let i = i+1 and return to step (4).

[0147] Simulation experiment example. To verify the effectiveness of the method of the present invention, the drone was set to fly at an altitude of 100 m (see scenario settings). Figure 1 The flight time is 50 seconds and discrete into 50 trajectory points; the OTFS system uses a 5 GHz carrier frequency, 20 MHz bandwidth, 15 kHz subcarrier spacing, and a 16×16 time-delay-Doppler domain resource grid (the time-delay-Doppler signal processing process of the OTFS system is described in [link to documentation]). Figure 2 The maximum speed of the UAV is 20 m / s, the peak transmit power is 30 dBm, and the noise power is −100 dBm; the location of the eavesdropping node is fixed at (200 m, 0 m); the maximum number of iterations for alternating optimization is 50, and the sensing power ratio is adaptively adjusted between 0.2 and 0.8; the legitimate link and the eavesdropping link adopt a hybrid channel model of line-of-sight and Rayleigh fading.

[0148] In the above embodiments, the proposed joint design method (see the flowchart of the alternating iterative solution) is as follows: Figure 3 It converges stably within a finite number of iterations (see convergence graph). Figure 4 The average security rate can reach approximately 83 bps / Hz, indicating that the method proposed in this invention is reasonable, and the ablation experiment results (see...) Figure 5 Compared to ablation schemes that only optimize trajectory, only optimize resources and power, or lack latency-Doppler domain resource selection, this approach achieves higher confidentiality rates and lower perceived agent CRB. Trajectory optimization results (see...) Figure 6 This indicates that the method of the present invention can guide drones to areas that are more conducive to legitimate communication and far away from eavesdropping nodes, thereby improving the channel difference between legitimate links and eavesdropping links.

[0149] Results of the confidentiality-awareness trade-off experiment (see...) Figure 7In the present invention, when the CRB threshold is set to 0.08, 0.06, 0.04 and 0.03 respectively, the adaptive sensing ratio strategy can maintain a high security rate while meeting the sensing requirements; the fixed sensing ratio strategy may fail to meet the sensing constraints due to insufficient sensing resources, or sacrifice the security rate due to excessive sensing resources.

[0150] In the high mobility robustness experiment, the results of 60 Monte Carlo experiments (see...) Figure 8 As can be seen, at different drone speeds, the OTFS-based secure ISAC scheme outperforms the OFDM-based baseline scheme, and the performance gap widens further with increasing drone speed, indicating that the OTFS delay-Doppler domain representation is more suitable for capturing and utilizing Doppler structures in high-mobility scenarios. In experiments with different transmit power budgets, the results (see...) Figure 9 This indicates that increasing the transmit power can improve the security rate of OTFS and OFDM schemes, but the design based on OTFS in this invention maintains a significant advantage at all power levels.

[0151] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Equivalent substitutions or improvements made by those skilled in the art to system parameters, channel models, resource allocation rules, or trajectory constraint forms without departing from the principles of the present invention should all fall within the protection scope of the present invention.

Claims

1. A joint trajectory and resource allocation method for a secure unmanned aerial vehicle (UAV) ISAC system based on OTFS, characterized in that, Includes the following steps: Step 1: Construct a secure drone integrated sensory IC system scenario, enabling the drone to act as an aerial IC platform, simultaneously sending confidential information to legitimate ground users and sensing targets in the presence of potential eavesdropping nodes. Step 2: Establish a time-delay-Doppler domain signal model based on orthogonal time-frequency space OTFS, organize the transmission resources into a time-delay-Doppler domain resource grid, and set communication allocation variables, sensing allocation variables, and transmit power variables for each time-delay-Doppler domain resource unit; Step 3: Establish a geometric channel model and a radial Doppler model related to the UAV trajectory, and map the Doppler frequency shift generated by the UAV motion into a dominant Doppler grid index; Step 4: Based on the effective channel gain in the delay-Doppler domain of the legitimate link and the eavesdropping link, construct a security rate model on each delay-Doppler domain resource unit, including signal-to-noise ratio, security gain, time slot security rate and long-term average security rate, and use the sensing performance index based on Cramer-Rao Boundary (CRB) to characterize the sensing quality. Step 5: Establish a joint optimization problem with the goal of maximizing the long-term average security rate. The constraints of the joint optimization problem include perception performance constraints, total power constraints of time slots, power constraints of single resource units, mutual exclusion constraints between communication and perception resources, and UAV mobility constraints. Step 6: Decompose the joint optimization problem into four sub-problems: trajectory optimization, time-delay-Doppler domain resource allocation, transmit power control, and sensing ratio update. Step 7: Solve the problem using an alternating iterative method, and output the optimized UAV trajectory, time-delay-Doppler domain communication resource allocation matrix, time-delay-Doppler domain sensing resource allocation matrix, and transmit power matrix to achieve secure integrated sensing transmission in high-mobility scenarios.

2. The method for joint trajectory and resource allocation of a secure UAV ISAC system based on OTFS according to claim 1, characterized in that, In step 1, the specific process of constructing the secure drone ISAC system scenario includes: Suppose the drone flies at a fixed altitude H. Discretize the drone's horizontal trajectory into N time slots. The horizontal position of the drone in the nth time slot is represented as... ; Let the horizontal positions of the legitimate ground user Bob and the potential eavesdropping node Eve be denoted as follows: , The drone sends confidential information to Bob, a legitimate ground user, and performs target perception tasks in the presence of a potential eavesdropping node, Eve. In the nth time slot, the transmit power on the delay-Doppler domain resource cell (k,l) is expressed as: Where k represents the Doppler index, l represents the time delay index, and the communication allocation variable is represented as... Perceptual assigned variables are represented as ; To prevent the same latency-Doppler domain resource unit from simultaneously undertaking communication and sensing tasks, resource mutual exclusion constraints are applied: 。 3. The method for joint trajectory and resource allocation of a secure UAV ISAC system based on OTFS according to claim 1, characterized in that, In step 2, the specific process of establishing the time-delay-Doppler domain signal model based on OTFS includes: Let x[k,l] be the transmission symbol on the time-delay-Doppler domain resource grid; After OTFS modulation, wireless channel transmission, and OTFS demodulation, the time-delay-Doppler domain signal at the receiving end is represented as follows: ； Where k' and l' represent the Doppler index and delay index of the transmitted symbols during the summation process, respectively; The equivalent delay from transmitting resource unit (k',l') to receiving resource unit (k,l) is represented by the Doppler domain channel response. This represents additive white noise; The time-delay-Doppler domain signal model is used to describe the time-delay-Doppler coupling relationship caused by UAV motion, and serves as the basis for subsequent time-delay-Doppler domain resource allocation.

4. The method for joint trajectory and resource allocation of a secure UAV ISAC system based on OTFS according to claim 1, characterized in that, In step 3, the specific process of establishing the geometric channel model and radial Doppler model related to the UAV trajectory includes: For node i, b and e represent the legitimate ground user node Bob and the potential eavesdropper node Eve, respectively, w i For the corresponding node location, the large-scale channel gain from the UAV to node i in the nth time slot is expressed as: ； Where β0 represents the reference channel gain; The radial Doppler shift of the UAV relative to the legitimate ground user Bob in the nth time slot is expressed as: ； Where λ represents the carrier wavelength, the radial Doppler frequency shift is mapped to the dominant Doppler grid index: ； in, This indicates the time-delay-Doppler domain Doppler resolution.

5. The method for joint trajectory and resource allocation of a secure unmanned aerial vehicle (UAV) ISAC system based on OTFS according to claim 1, characterized in that, In step 4, the specific process of constructing the confidentiality rate model includes: In the nth time slot, the signal-to-noise ratio of resource unit (k,l) at node i is expressed as: ； Where σ² represents the noise power, Indicates the effective channel gain in the delay-Doppler domain; The time-delay-Doppler domain effective channel gain is expressed as: ； Among them, L i [n] Large-scale channel gain from UAV to node i in the nth time slot. Represents the nominal delay-Doppler domain channel coefficients. This represents the channel degradation factor caused by mobility. The security gain of resource unit (k,l) in the nth time slot is expressed as: ； The system security rate within the nth time slot is expressed as: ； in, The long-term average rate of secrecy is expressed as: 。 6. The method for joint trajectory and resource allocation of a secure UAV ISAC system based on OTFS according to claim 1, characterized in that, In step 4, the perception quality is characterized by a perception performance index based on Cramer-Rhodes boundary (CRB), specifically including: The sensing performance index in the nth time slot is expressed as: ； Where C represents the proportionality constant, Indicates coherence time. Indicates the perceived signal-to-noise ratio; The sensing signal-to-noise ratio is jointly determined by the delay-Doppler domain resource units assigned to the sensing task and their transmit power, expressed as: 。 7. The method for joint trajectory and resource allocation of a secure UAV ISAC system based on OTFS according to claim 1, characterized in that, In step 5, the joint optimization problem is expressed as: ； in, This represents the perception CRB threshold. This indicates the upper limit of total power in a single time slot. This represents the upper limit of power for a single delay-Doppler domain resource unit. Indicates the maximum speed of the drone. Indicates the time slot length.

8. The method for joint trajectory and resource allocation of a secure UAV ISAC system based on OTFS according to claim 1, characterized in that, In step 6, the trajectory optimization subproblem optimizes the UAV trajectory q under the conditions of time-delay-Doppler domain resource allocation and fixed transmission power; Since the path loss term and the Doppler term are non-convex functions with respect to q[n], the non-convex expression is locally linearized near the current iteration point to construct a convex approximation problem, and the successive convex approximation SCA method is used to update the UAV trajectory. Through the aforementioned trajectory optimization, the drone is gradually moved to a position that is favorable to the legitimate link and far away from the eavesdropping node, thereby enhancing the channel asymmetry between the legitimate link and the eavesdropping link; The time-delay-Doppler domain resource allocation sub-problem is performed under the condition that the UAV trajectory and transmit power are fixed, and specifically includes: According to the dominant Doppler raster index Introducing the Doppler sensing weighting function: ； in, Indicates weight strength. Indicates the width of the weight distribution; Based on the Doppler sensing weight function, the effective communication utility of the time-delay-Doppler domain resource unit is defined as follows: ； When allocating resources, priority is given to time-delay-Doppler domain resource units with high effective communication utility and that are conducive to meeting perception constraints. Necessary resources are reserved for perception tasks based on CRB constraints, so that the allocation of communication and perception resources simultaneously meets mutual exclusion constraints and perception performance constraints.

9. The method for joint trajectory and resource allocation of a secure unmanned aerial vehicle (UAV) ISAC system based on OTFS according to claim 1, characterized in that, In step 6, the transmit power control sub-problem and the sensing ratio update sub-problem include: Under the condition that the UAV trajectory and the time-delay-Doppler domain resource allocation are fixed, the transmission power of each time-delay-Doppler domain resource unit is updated by the projection gradient ascent method. The projection gradient ascent method first adjusts the transmission power along the gradient direction that increases the security rate, and then projects the update result back into the feasible domain that satisfies the total power constraint of a single time slot and the power constraint of a single resource unit. Introducing the perceived scale variable When it is detected that the sensing constraint is not met in the nth time slot, the sensing ratio is increased according to the following rules: ； in, This indicates the perceived scale update step size. This indicates the upper limit of the perceived proportion; The perception ratio update enables the system to adaptively adjust the ratio between communication resources and perception resources based on the satisfaction of perception constraints.

10. The method for joint trajectory and resource allocation of a secure unmanned aerial vehicle (UAV) ISAC system based on OTFS according to claim 1, characterized in that, The alternating iterative solution process in step 7 includes: Initialize system parameters, delay-Doppler domain channel, UAV trajectory, communication resource allocation matrix, sensing resource allocation matrix, transmit power matrix, and iteration index; Update the geometric channel gain, radial Doppler frequency shift, dominant Doppler grid index, and time-delay-Doppler domain equivalent channel gain based on the current UAV trajectory. Update the communication resource allocation matrix and the sensing resource allocation matrix based on CRB sensing constraints and Doppler sensing utility; Under the condition of fixed resource allocation and trajectory, the transmit power matrix is ​​updated by projective gradient ascent; Under the condition of fixed resource allocation and transmission power, the drone trajectory is updated via SCA; The perception ratio or perception weight is adaptively updated based on the satisfaction of perception constraints. Calculate the average security rate and the sensing CRB. If the convergence condition is met or the maximum number of iterations is reached, output the updated UAV trajectory, communication resource allocation matrix, sensing resource allocation matrix, and transmit power matrix; otherwise, proceed to the next iteration.

Images