Energy efficiency optimization method for active RIS-assisted UAV mobile edge computing system

Abstract

The invention discloses an energy efficiency optimization method for an active RIS-assisted UAV mobile edge computing system, and the method comprises the steps: constructing a mobile edge computing system model based on an active RIS-assisted UAV; constructing a communication model of a mobile edge computing system of the active RIS-assisted UAV; constructing an energy consumption model of the mobile edge computing system of the active RIS-assisted UAV; establishing a multivariable joint optimization model by taking maximization of the energy efficiency of the system as a target; solving the target optimization model by using a convex optimization method; according to the invention, through the active RIS, the resource allocation is optimized in a combined manner, the maximum system energy efficiency is finally obtained, and the performance of the unmanned aerial vehicle network system is improved.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to an energy efficiency optimization method for an active RIS-assisted UAV mobile edge computing system. Background Art

[0002] Mobile Edge Computing (MEC) has become an effective solution to help ground terminal (GT) devices handle computationally intensive and delay-sensitive tasks. Compared with the traditional cloud computing model, the advantage of MEC is that by deploying edge servers at the edge of the network topology, such as base stations or access points (APs) close to mobile IoT devices, GT devices can offload their computing tasks to these edge servers, which effectively reduces the computing burden of GT devices.

[0003] The servers of ground MEC networks are usually fixed and easily affected by obstacles in urban environments, making it difficult to adapt to dynamic and heterogeneous environments, and unable to quickly and reliably support hotspots and temporary computing. The combination of UAV (Unmanned Aerial Vehicle) and MEC technology can provide more flexible computing offload services.

[0004] Although UAV-assisted MEC networks have solved many problems of traditional MEC networks, they still face complex and time-varying fading channels. This can lead to unstable signal quality of UAVs, reduced signal-to-noise ratio, increased bit error rate, signal strength fluctuations, data transmission delays and packet loss. In addition, the endurance of UAVs is also limited by their flight energy consumption, and they cannot support long-term hovering or long-distance maneuvers. Reconfigurable Intelligent Surface (RIS) can dynamically control the amplitude, phase, frequency and other characteristics of electromagnetic waves, thereby controlling the propagation behavior of electromagnetic waves in free space, breaking through the limitation that traditional wireless channels cannot be actively regulated. Combining RIS with the UAV-MEC system is a win-win strategy that can greatly improve the wireless transmission environment, increase the received signal power, and thus improve the energy efficiency of the UAV system.

[0005] At present, most of the research on RIS UAV mobile edge computing systems uses a large number of low-cost passive reflective elements, and most of them fix the RIS on buildings to assist the uplink. However, traditional passive RIS (Passive RIS, PRIS) is difficult to overcome the double fading problem in the RIS-assisted cascade link. In contrast, active RIS (Active RIS, ARIS) solves the double fading problem by adjusting the signal phase and amplifying the signal amplitude, thereby further improving the channel quality. Summary of the invention

[0006] In order to solve the problems existing in the above-mentioned prior art, the present invention proposes an energy efficiency optimization method for an active RIS-assisted UAV mobile edge computing system. This method combines the high flexibility of UAV and the characteristics that active RIS can significantly improve the quality of communication channels compared to passive RIS, and achieves the goal of maximizing the energy efficiency of the system by reasonably optimizing the receiving beam vector, unloading amount, active RIS reflection coefficient, UAV flight trajectory and flight speed at MEC.

[0007] An energy efficiency optimization method for an active RIS-assisted UAV mobile edge computing system comprises the following steps:

[0008] S1: Establish system model, communication model and energy consumption model based on ground terminal equipment, active RIS and UAV operation parameters;

[0009] S2: Based on the total number of tasks completed by the ground terminal devices and MEC servers and the weighted total energy consumption of the system, an optimization problem is constructed with the goal of maximizing the system energy efficiency.

[0010] S3: The maximum energy efficiency function is optimized and solved using the linear minimum mean square error estimation, continuous convex approximation method and Dinkelbach algorithm. The local optimal solution of the original problem is obtained through continuous iteration to achieve the maximum system energy efficiency.

[0011] Furthermore, the system model, communication model and energy consumption model in S1 specifically include:

[0012] The system model includes: the system consists of a ground MEC server, a drone and K ground terminal devices in the area. The set {1,2,…,K} represents all ground terminal devices. The drone is equipped with an active RIS, which acts as a relay to help the ground terminal devices forward tasks to the ground MEC server for calculation.

[0013] Assume that K ground terminal devices and MEC servers are all located on the ground and have a fixed position, that is, the height is 0. The horizontal coordinate of the kth ground terminal device is represented by r k =[x k ,yk ] T , the horizontal coordinate of the ground MEC server is represented by r e =[x e ,y e ] T , assuming that UAV-ARIS maintains a fixed altitude H and its horizontal position remains unchanged. The flight time is evenly divided into T time slots, and the duration of each time slot is δ t Therefore, at time slot t, the position and flight speed of UAV-ARIS are expressed as q[t] = [x[t], y[t]] T ,v[t]=[v x [t],v y [t]] T , This results in the following flight constraints:

[0014] q[1]=q0,q[T+1]=q F

[0015] ||v[t]||≤v max

[0016]

[0017] where v max Indicates the maximum flight speed of the UAV, q0,q F They represent the starting and ending positions of the flight respectively. The distance from the kth ground terminal device to ARIS in the tth time slot can be expressed as The distance between ARIS and MEC server at the tth time slot can be expressed as Since the UAV has a good line-of-sight path, considering the pure LoS path, the channel from the kth ground terminal device to ARIS in the tth time slot can be expressed as Where β represents the path loss per unit distance, a k [t] represents the steering vector. Similarly, the channel from ARIS to MEC in the tth time slot can be expressed as where a M [t] and a R [t] represents the receive and transmit steering vectors of MEC and ARIS respectively.

[0018] The communication model includes: assuming that ARIS is configured with N units, the ground MEC server is configured with M receiving antennas, and K ground terminal devices are configured with single antennas, the channel from ARIS to the ground edge server in the tth time slot can be expressed as The channel from the kth ground terminal device to ARIS in the tth time slot is expressed as

[0019] Consider that the ground terminal equipment performs task offloading in a TDMA manner, and each ground terminal equipment is allocated equal-time sub-slots, that is, Assume that the transmission power of the kth ground terminal device is P k , the emission symbol is s k , then the received signal at MEC is expressed as

[0020] in β k,n [t] and φ k,n [t]∈[0,2π) represent the reflection amplitude and phase from the kth ground terminal device to the nth reflection unit of ARIS in the tth time slot; s k [t] represents the transmission symbol of the kth ground terminal device in the tth time slot, and denote the thermal noise at the ARIS and MEC receiving ends, respectively, and I M and I N denote the M×M and N×N identity matrices respectively, and denote the Gaussian white noise power at the receiving end of ARIS and MEC respectively.

[0021] Assume that the receiving beam vector of the kth ground terminal device at the MEC in the tth time slot is The signal recovered by the kth ground terminal device at the tth time slot on the MEC server is expressed as Then the rate of the tth time slot of the kth ground terminal device can be expressed as:

[0022]

[0023] Where B represents the transmission bandwidth.

[0024] This paper considers partial offloading, that is, the computational task of the kth ground terminal device in the tth time slot is divided into two parts, which are expressed as as well as And meet in is the amount of local computation, is the offloaded computational effort, l k [t] is the total computational effort of the kth ground terminal device. The computational effort of local computing should satisfy the following conditions:

[0025]

[0026] Where b represents the number of CPU cycles required to process a 1-bit task, and f L,krepresents the maximum computing capacity of the kth ground terminal device. The offloading constraint can be expressed as

[0027]

[0028] The above constraint states that the offload amount cannot exceed the product of the maximum rate and the allocated time.

[0029] The energy consumption model includes: in the UAV-ARIS assisted MEC communication system, the total energy consumption is mainly composed of the energy consumption of local computing and offloading tasks of ground terminal devices, the computing energy consumption of MEC servers, the flight energy consumption of UAVs, and the energy consumption of ARIS. The energy consumption of all ground terminal devices in the tth time slot can be modeled as

[0030]

[0031] in represents the switched capacitance coefficient of the ground terminal device. The energy consumption of the MEC server in the tth time slot can be expressed as

[0032]

[0033] in Represents the switch capacitance coefficient of the MEC server. As for the flight energy consumption of the UAV, a new energy consumption model for rotary-wing UAVs is adopted, which takes into account the actual thrust-to-weight ratio, as shown below:

[0034]

[0035] Where P0 is the blade power in the hovering state, U tip is the blade angular velocity, d o is the fuselage drag ratio, ρ is the air density, s is the rotor volume, A is the rotor area, P i is the induced power in the hovering state, and v0 is the induced speed when the rotor flies forward.

[0036] Since ARIS is an active unit and consumes energy, the power consumption of the nth unit is modeled as follows:

[0037] P ARIS,n [t] = μp out,n [t]+P c +P DC

[0038] Where μ represents the amplifier efficiency, P c and P DC Respectively represent the static circuit power consumption and DC bias power consumption, p out,n[t] represents the output power of the nth unit, which is related to the incident signal power at the tth time slot and can be expressed as

[0039]

[0040] in g k,n [t] represents the channel from the kth user ground terminal device to the nth ARIS reflection unit in the tth time slot. Therefore, the energy consumption of ARIS in the tth time slot can be expressed as

[0041]

[0042] Since the energy consumption of the drone is much greater than the unloading energy consumption, the following weighted energy consumption is considered as follows:

[0043]

[0044] Where α is the weighting coefficient.

[0045] Furthermore, the optimization problem constructed in S2 with the goal of maximizing the system energy efficiency specifically includes: maximizing the system energy efficiency by jointly optimizing the UAV trajectory Q, the UAV flight speed v, the unloading amount l, the ARIS reflection coefficient Θ, and the MEC receiving beam vector w, as shown below:

[0046]

[0047] C4:q1=q0,q[T+1]=q F

[0048] C5:||v[t]||≤v max

[0049]

[0050] Among them, C1 represents the maximum task volume constraint of local calculation, C2 represents the maximum unloadable task volume constraint, C3 represents the sum of the task volume of all time slots of each ground terminal device should be greater than a certain threshold, C4 represents the initial position and end position of the UAV flight, C5 represents the flight speed of the UAV cannot exceed the specified maximum flight speed, and C6 represents the position change of the UAV during the flight.

[0051] Furthermore, the process of optimizing the maximum energy efficiency function using linear minimum mean square error estimation, continuous convex approximation and Dinkelbach algorithm includes: optimizing the receiving beam vector w at the MEC, optimizing the unloading amount l, optimizing the ARIS reflection coefficient Θ, optimizing the UAV flight trajectory Q and the flight speed v, solving these sub-problems, and obtaining the local optimal solution of the original problem by continuously iterating the solutions to the sub-problems, ultimately achieving the goal of maximizing the system energy efficiency.

[0052] Beneficial effects of the present invention:

[0053] 1. The present invention takes into account the narrow coverage area of ​​ground terminals by fixed RIS in complex urban communication environments and the double fading problem that passive RIS cannot overcome, and dynamically combines UAV, MEC server and active RIS to greatly improve the quality of the communication channel.

[0054] 2. The present invention takes into account the impact of active RIS energy consumption on the objective function, and through reasonable optimization of the MEC receiving beam vector, unloading amount, reflection coefficient of active RIS, flight trajectory and flight speed of UAV, and continuously iterates the solutions of sub-problems to obtain the local optimal solution of the original problem, ultimately achieving the goal of maximizing the system energy efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 It is a basic flow chart of the energy efficiency optimization method of the active RIS-assisted UAV mobile edge computing system in the present invention;

[0056] Figure 2 It is a system model diagram of the mobile edge computing system of the active RIS assisted UAV in the present invention;

[0057] Figure 3 It is a graph showing the trajectory changes of the drone under different schemes of the present invention;

[0058] Figure 4 It is an iterative convergence diagram of the energy efficiency of the active RIS-assisted UAV-MEC system under different schemes of the present invention. DETAILED DESCRIPTION

[0059] The present invention is further explained below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. It should be noted that the words "front", "rear", "left", "right", "upper" and "lower" used in the following description refer to directions in the accompanying drawings, and the words "inner" and "outer" refer to directions toward or away from the geometric center of a specific component, respectively.

[0060] Figure 1It is a basic flow chart of the energy efficiency optimization method of the active RIS-assisted UAV mobile edge computing system in the present invention, including the following steps:

[0061] S1: Establish system model, communication model and energy consumption model based on ground terminal equipment, active RIS and UAV operation parameters;

[0062] S2: Based on the total number of tasks completed by the ground terminal devices and MEC servers and the weighted total energy consumption of the system, an optimization problem is constructed with the goal of maximizing the system energy efficiency.

[0063] S3: The maximum energy efficiency function is optimized and solved using the linear minimum mean square error estimation, continuous convex approximation method and Dinkelbach algorithm. The local optimal solution of the original problem is obtained through continuous iteration to achieve the maximum system energy efficiency.

[0064] like Figure 2 As shown, further, the system model, communication model, and energy consumption model in S1 specifically include:

[0065] The system model includes: the system consists of a ground MEC server, a drone and K ground terminal devices in the area. The set {1,2,…,K} represents all ground terminal devices. The drone is equipped with an active RIS, which acts as a relay to help the ground terminal devices forward tasks to the ground MEC server for calculation.

[0066] Assume that K ground terminal devices and MEC servers are all located on the ground and have a fixed position, that is, the height is 0. The horizontal coordinate of the kth ground terminal device is represented by r k =[x k ,y k ] T , the horizontal coordinate of the ground MEC server is represented by r e =[x e ,y e ] T , assuming that UAV-ARIS maintains a fixed altitude H and its horizontal position remains unchanged. The flight time is evenly divided into T time slots, and the duration of each time slot is δ t Therefore, at time slot t, the position and flight speed of UAV-ARIS are expressed as q[t] = [x[t], y[t]] T ,v[t]=[v x [t],v y [t]] T , This results in the following flight constraints:

[0067] q[1]=q0,q[T+1]=q F

[0068] ||v[t]||≤v max

[0069]

[0070] where v max Indicates the maximum flight speed of the UAV, q0,q F They represent the starting and ending positions of the flight respectively. The distance from the kth ground terminal to ARIS in the tth time slot can be expressed as The distance between ARIS and MEC server at the tth time slot can be expressed as Since the UAV has a good line-of-sight path, considering the pure LoS path, the channel from the kth ground terminal device to ARIS in the tth time slot can be expressed as

[0071] , where β represents the path loss per unit distance, a k [t] represents the steering vector. Similarly, the channel from ARIS to MEC in the tth time slot can be expressed as where a M [t] and a R [t] represents the receive and transmit steering vectors of MEC and ARIS respectively.

[0072] The communication model includes: assuming that ARIS is configured with N units, the ground MEC server is configured with M receiving antennas, and K ground terminal devices are configured with single antennas, the channel from ARIS to the ground MEC server in the tth time slot can be expressed as The t-th time slot channel from the k-th ground terminal device to ARIS is expressed as

[0073] Consider that the ground terminal equipment performs task offloading in a TDMA manner, and each ground terminal equipment is allocated equal-time sub-slots, that is, Assume that the transmission power of the kth ground terminal device is P k , the emission symbol is s k , then the received signal at MEC can be expressed as

[0074] in β k,n [t] and φ k,n [t]∈[0,2π) represent the reflection amplitude and phase from the kth ground terminal device to the nth reflection unit of ARIS in the tth time slot. k [t] represents the transmission symbol of the kth ground terminal device in the tth time slot, and denote the thermal noise at the ARIS and MEC receiving ends, respectively, and I M and I N denote the M×M and N×N identity matrices respectively, and denote the Gaussian white noise power at the receiving end of ARIS and MEC respectively.

[0075] Assume that the receiving beam vector of the kth ground terminal device at the MEC in the tth time slot is The signal recovered by the kth ground terminal device at the tth time slot on the MEC server is expressed as Then the rate of the tth time slot of the kth ground terminal device can be expressed as:

[0076]

[0077] Where B represents the transmission bandwidth.

[0078] This paper considers partial offloading, that is, the computational task of the kth ground terminal device in the tth time slot is divided into two parts, which are expressed as as well as And meet in is the amount of local computation, is the offloaded computational effort, l k [t] is the total computational effort of the kth ground terminal device. The computational effort of local computing should satisfy the following conditions:

[0079]

[0080] Where b represents the number of CPU cycles required to process a 1-bit task, and f L,k represents the maximum computing capacity of the kth ground terminal device. The offloading constraint can be expressed as

[0081]

[0082] The above constraint states that the offload amount cannot exceed the product of the maximum rate and the allocated time.

[0083] The energy consumption model includes: in the UAV-ARIS assisted MEC communication system, the total energy consumption is mainly composed of the energy consumption of local computing and offloading tasks of ground terminal devices, the computing energy consumption of MEC servers, the flight energy consumption of UAVs, and the energy consumption of ARIS. The energy consumption of all ground terminal devices in the tth time slot can be modeled as

[0084]

[0085] in represents the switch capacitance coefficient of the ground terminal device. The energy consumption of the MEC server in the tth time slot can be expressed as

[0086]

[0087] in Represents the switch capacitance coefficient of the MEC server. As for the flight energy consumption of the UAV, a new energy consumption model for rotary-wing UAVs is adopted, which takes into account the actual thrust-to-weight ratio, as shown below:

[0088]

[0089] Where P0 is the blade power in the hovering state, U tip is the blade angular velocity, d o is the fuselage drag ratio, ρ is the air density, s is the rotor volume, A is the rotor area, P i is the induced power in the hovering state, and v0 is the induced speed when the rotor flies forward.

[0090] Since ARIS is an active unit that consumes energy, the power consumption of the nth unit can be modeled as follows:

[0091] P ARIS,n [t] = μp out,n [t]+P c +P DC

[0092] Where μ represents the amplifier efficiency, P c and P DC Respectively represent the static circuit power consumption and DC bias power consumption, p out,n [t] represents the output power of the nth unit, which is related to the incident signal power at the tth time slot and can be expressed as

[0093]

[0094] in represents the channel from the kth ground terminal device to the nth ARIS reflection unit in the tth time slot. Therefore, the energy consumption of ARIS in the tth time slot can be expressed as

[0095]

[0096] Since the flight energy consumption of the drone is much greater than the unloading energy consumption, the following weighted energy consumption is considered as follows:

[0097]

[0098] Where α is the weighting coefficient.

[0099] Furthermore, the optimization problem constructed in S2 with the goal of maximizing the system energy efficiency specifically includes: maximizing the system energy efficiency by jointly optimizing the UAV trajectory Q, the UAV flight speed v, the unloading amount l, the ARIS reflection coefficient Θ, and the MEC receiving beam vector w, as shown below:

[0100]

[0101] C4:q1=q0,q[T+1]=q F

[0102] C5:||v[t]||≤v max

[0103]

[0104] Among them, C1 represents the maximum task volume constraint of local calculation, C2 represents the maximum unloadable task volume constraint, C3 represents the sum of the task volume of all time slots of each terminal should be greater than a certain threshold, C4 represents the initial position and end position of the UAV flight, C5 represents the flight speed of the UAV cannot exceed the specified maximum flight speed, and C6 represents the position change of the UAV during the flight.

[0105] Due to the tight coupling and non-convexity between variables in the optimization problem, it is difficult to solve it directly. In order to solve this problem, the present invention uses linear minimum mean square error estimation, continuous convex approximation and Dinkelbach algorithm to decouple the complex non-convex optimization problem into four easy-to-solve sub-problems, and then uses CVX to solve them and perform alternating optimization, and finally obtains the local optimal solution of the original problem.

[0106] The specific calculation process includes:

[0107] 1. Optimize the receiving beam vector w at MEC

[0108] For the objective function of the optimization problem P1, only the transmission rate of the ground terminal device is related to the receiving beam. Therefore, maximizing the energy efficiency, that is, maximizing the receiving rate of MEC, for the TDMA access mode, the maximum achievable rate of the kth ground terminal device in the tth time slot can be expressed as

[0109]

[0110] It has been proven that the linear minimum-mean-square error (LMMSE) detector is the optimal receiving beam vector. The normalized receiving beam based on LMMSE can be expressed as

[0111]

[0112] in

[0113] 2. Optimize uninstall volume

[0114] According to the optimization problem P1, we get the sub-problem of optimizing the unloading amount:

[0115]

[0116] Using the Dinkelbach algorithm, the auxiliary variable ξ is introduced to represent the optimal objective function value of the previous iteration, and the terms related to the unloading amount are retained. The objective function of the above optimization problem can be transformed into:

[0117]

[0118] It can be observed that the objective function is a linear function of the unloading amount, and the constraint condition is a linear constraint on the unloading amount, so it can be solved directly, where the update of the auxiliary variable ξ is as follows:

[0119]

[0120] 3. Optimize ARIS reflection coefficient θ

[0121] According to the optimization problem P1, the sub-problem of optimizing the reflection coefficient is as follows:

[0122]

[0123] According to the energy consumption expression of ARIS in the tth time slot, the objective function

[0124]

[0125] It can be observed that the objective function has nothing to do with the phase of ARIS, so it is only necessary to optimize the phase of ARIS to maximize the emission rate, that is,

[0126]

[0127] because Consider adjusting the ARIS phase to align the cascaded channels, that is,

[0128]

[0129] Based on the above results, the rate of the kth user in the tth time slot can be expressed as

[0130]

[0131] A simple transformation yields:

[0132]

[0133] Use SCA to transform the above constraints into convex constraints as follows:

[0134]

[0135] The transformed optimization problem is as follows:

[0136]

[0137] This optimization problem is a convex optimization problem and can be solved using CVX.

[0138] 4. Optimize UAV flight trajectory Q and flight speed v

[0139] According to the optimization problem P1, the sub-problems of optimizing the flight trajectory and flight speed are obtained:

[0140]

[0141] C2:q1=q0,q[T+1]=q F

[0142] C3:||v[t]||≤v max

[0143]

[0144] First, for the UAV flight energy consumption, use the inequality Therefore, the upper bound of the UAV flight energy consumption can be obtained as follows:

[0145]

[0146] It can be seen that this part is a convex function of speed v. For the objective function part, due to definition So

[0147]

[0148] Introduce auxiliary variables || q[t]-r k || 2 +H 2 ≥ω k [t], then the above formula can be transformed into

[0149]

[0150] Therefore, the objective function is transformed into

[0151]

[0152] This formula is a joint convex function of ω and velocity v. Further constraints ||q[t]-r k || 2 +H 2 ≥ω k [t] is transformed into a convex constraint as follows:

[0153]

[0154] For ground terminal equipment rate

[0155]

[0156] in For the above function, introduce auxiliary variables SCA can be used to transform it into the following form:

[0157]

[0158] At this point, the above formula is a linear function. In summary, the subproblem is transformed into the following form

[0159]

[0160] C5:q1=q0,q[T+1]=q F

[0161] C6:||v[t]||≤v max

[0162]

[0163] The above optimization problem is a convex optimization problem about the trajectory and velocity of the UAV, which can be solved using CVX.

[0164] The present invention is further described below in combination with simulation conditions and results, and the simulation parameters are set as follows:

[0165] The total bandwidth of the channel is 1MHz, the number of ground terminal devices is 4, the flight altitude of the drone is 50m, the time period is 50s, divided into 50 time slots, the maximum flight speed of the drone is 50m / s, the initial and final positions of the drone are [0,200]m and [0,-200]m, and the transmission power of the ground terminal device is 500mW. The computing power of the local computing of the ground terminal device and the MEC end are 5×10 8 Hz and 10 10 Hz and its effective capacitance coefficient 1×10 -27 and 1×10 -28, the number of CPU cycles required to process a unit bit task is 1×10 3 , the noise power at the MEC receiving end and the noise power at ARIS are -100dBm and -60dBm respectively, and the weight of the UAV flight energy consumption is 0.0001.

[0166] Figure 3 The comparison of UAV trajectories under random RIS phase shift, fixed UAV trajectory, and the proposed algorithm is shown when the number of ground terminal devices is 4 and the period is T = 50s.

[0167] Figure 4 It shows the iterative convergence diagram of the energy efficiency of the active RIS-assisted UAV-MEC system under different schemes. It can be seen from the figure that the energy efficiency of the UAV system will increase rapidly at the beginning until it begins to converge after about 6-7 iterations. The algorithm proposed in the present invention can achieve higher energy efficiency than the other two benchmark schemes.

[0168] The technical means disclosed in the scheme of the present invention are not limited to the technical means disclosed in the above-mentioned implementation mode, but also include technical schemes composed of any combination of the above technical features.

Claims

1. An energy efficiency optimization method for an active RIS-assisted UAV mobile edge computing system, characterized in that: The specific steps include: S1: Establish system model, communication model and energy consumption model based on ground terminal equipment, active RIS and UAV operation parameters; S2: Based on the total number of tasks completed by the ground terminal equipment and MEC server and the weighted total energy consumption of the system, an optimization problem is constructed with the goal of maximizing the system energy efficiency; S3: The maximum energy efficiency function is optimized and solved using the linear minimum mean square error estimation, continuous convex approximation method and Dinkelbach algorithm. The local optimal solution of the original problem is obtained through continuous iteration to achieve the maximum system energy efficiency.

2. The energy efficiency optimization method of an active RIS-assisted UAV mobile edge computing system according to claim 1 is characterized in that: The system model established in S1 specifically includes: The system model includes a ground MEC server, a drone and K ground terminal devices in the area; the set {1,2,…,K} represents all ground terminal devices; the drone is equipped with an active RIS, which acts as a relay to help ground terminal devices forward tasks to the ground MEC server for calculation; Assume that K ground terminal devices and MEC servers are all located on the ground and have a fixed position, that is, the height is 0. The horizontal coordinate of the kth ground terminal device is represented by r k =[x k ,y k ] T , the horizontal coordinate of the MEC server is represented by r e =[x e ,y e ] T , assuming that UAV-ARIS maintains a fixed altitude H and its horizontal position remains unchanged; the flight time is evenly divided into T time slots, and the duration of each time slot is δ t , so the position and flight speed of UAV-ARIS at time slot t are expressed as This results in the following flight constraints: q[1]=q0,q[T+1]=q F ||v[t]||≤v max where v max Indicates the maximum flight speed of the UAV, q0,q F They represent the starting and ending positions of the flight respectively; the distance from the kth ground terminal to ARIS in the tth time slot is expressed as The distance between ARIS and MEC server at the tth time slot can be expressed as Since the UAV has a good line-of-sight path, considering the pure LoS path, the channel from the kth ground terminal device to ARIS in the tth time slot can be expressed as Where β represents the path loss per unit distance, a k [t] represents the steering vector; similarly, the channel from ARIS to MEC in the tth time slot is expressed as where a M [t] and a R [t] represents the receive and transmit steering vectors of MEC and ARIS respectively.

3. The energy efficiency optimization method of an active RIS-assisted UAV mobile edge computing system according to claim 1 is characterized in that: The communication model established in S1 includes: assuming that ARIS is configured with N units, the ground MEC server is configured with M receiving antennas, and K ground terminal devices are configured with single antennas, the channel from ARIS to the ground MEC server in the tth time slot is expressed as The t-th time slot channel from the k-th ground terminal device to ARIS is expressed as Consider that the ground terminal equipment performs task offloading in a TDMA manner, and each ground terminal equipment is allocated equal-time sub-slots, that is, Assume that the transmission power of the kth ground terminal device is P k , the emission symbol is s k , then the received signal at MEC is expressed as in β k,n [t] and φ k,n [t]∈[0,2π) represent the reflection amplitude and phase from the kth ground terminal device to the nth reflection unit of ARIS in the tth time slot; s k [t] represents the transmission symbol of the kth ground terminal device in the tth time slot, and denote the thermal noise at the ARIS and MEC receiving ends, respectively, and I M and I N denote the M×M and N×N identity matrices respectively, and denote the Gaussian white noise power at the receiving end of ARIS and MEC respectively. Assume that the receiving beam vector of the kth ground terminal device at the MEC in the tth time slot is The signal recovered by the kth ground terminal device at the tth time slot on the MEC server is expressed as Then the rate of the tth time slot of the kth ground terminal device is expressed as: Where B represents the transmission bandwidth; This paper considers partial offloading, that is, the computational task of the kth ground terminal device in the tth time slot is divided into two parts, which are expressed as as well as And meet in is the amount of local computation, is the offloaded computational effort, l k [t] is the total computational effort of the kth ground terminal device. The computational effort of local computing should satisfy the following conditions: Where b represents the number of CPU cycles required to process a 1-bit task, and f L,k represents the maximum computing capacity of the kth ground terminal device; the constraint on the offloading amount is expressed as The above constraint states that the offload amount cannot exceed the product of the maximum rate and the allocated time.

4. The energy efficiency optimization method of an active RIS-assisted UAV mobile edge computing system according to claim 1 is characterized in that: The energy consumption model established in S1 includes that in the UAV-ARIS-assisted MEC communication system, the total energy consumption is mainly composed of the energy consumption of local computing and offloading tasks of ground terminal devices, the computing energy consumption of MEC servers, the flight energy consumption of UAVs, and the energy consumption of ARIS; the energy consumption of all ground terminal devices in the tth time slot is modeled as in represents the switch capacitance coefficient of the ground terminal device; the energy consumption of the MEC server in the tth time slot is expressed as in represents the switch capacitance coefficient of the MEC server; as for the flight energy consumption of the UAV, a new energy consumption model for rotary-wing UAVs is adopted, which takes into account the actual thrust-to-weight ratio, as shown below: Where P0 is the blade power in the hovering state, U tip is the blade angular velocity, d o is the fuselage drag ratio, ρ is the air density, s is the rotor volume, A is the rotor area, P i is the induced power in the hovering state, v0 is the induced speed when the rotor flies forward; Since ARIS is an active unit and consumes energy, the power consumption of the nth unit is modeled as follows: P ARIS,n [t]=μp out,n [t]+P c +P DC Where μ represents the amplifier efficiency, P c and P DC Respectively represent the static circuit power consumption and DC bias power consumption, p out,n [t] represents the output power of the nth unit, which is related to the incident signal power at the tth time slot and is expressed as in g k,n [t] represents the channel from the kth ground terminal device to the nth ARIS reflection unit in the tth time slot; therefore, the energy consumption of ARIS in the tth time slot is expressed as Since the flight energy consumption of the drone is much greater than the unloading energy consumption, the following weighted energy consumption is considered as follows: Where α is the weighting coefficient.

5. The energy efficiency optimization method of an active RIS-assisted UAV mobile edge computing system according to claim 1 is characterized in that: The optimization problem constructed in S2 with the goal of maximizing the system energy efficiency specifically includes: maximizing the system energy efficiency by jointly optimizing the UAV trajectory Q, the UAV flight speed v, the unloading amount l, the ARIS reflection coefficient Θ, and the MEC receiving beam vector w, as shown below: C4:q1=q0,q[T+1]=q F C5:||v[t]||≤v max Among them, C1 represents the maximum task volume constraint of local calculation, C2 represents the maximum unloadable task volume constraint, C3 represents the sum of the task volume of all time slots of each ground terminal device should be greater than a certain threshold, C4 represents the initial position and end position of the UAV flight, C5 represents the flight speed of the UAV cannot exceed the specified maximum flight speed, and C6 represents the position change of the UAV during the flight. The energy efficiency optimization method of an active RIS-assisted UAV mobile edge computing system according to claim 1 is characterized in that: The process of S3 optimizing the maximum energy efficiency function by using linear minimum mean square error estimation, continuous convex approximation and Dinkelbach algorithm includes: optimizing the receiving beam vector w at MEC, optimizing the unloading amount l, optimizing the ARIS reflection coefficient Θ, optimizing the UAV flight trajectory Q and the flight speed v, solving these sub-problems, and obtaining the local optimal solution of the original problem by continuously iterating the solutions to the sub-problems to achieve the maximum system energy efficiency.

Images