Security enhancement and transmission optimization method and system for URLLC services

Abstract

A method and system for security enhancement and transmission optimization of URLLC services, including: constructing a URLLC service security transmission model; constructing a URLLC service security transmission performance characterization framework based on short packet information theory to obtain an equivalent achievable security transmission rate (AESR); constructing a security performance optimization problem based on the proposed security transmission performance characterization framework, and enhancing the security performance of the URLLC service while ensuring the delay and reliability performance of the URLLC service by combining the channel estimation pilot length with the artificial noise injection strategy; and quickly converging to an approximate optimal solution to the above optimization problem with low time complexity through a two-dimensional search algorithm based on the golden section idea. The present invention constructs a security performance analysis framework that conforms to the short packet transmission characteristics of the URLLC service, provides a theoretical basis for the security performance analysis of the URLLC system, and uses artificial noise technology to counter eavesdropping attacks from malicious nodes to achieve security enhancement of the URLLC service.

Description

Technical Field

[0001] The present invention belongs to the field of wireless communication technology, and in particular relates to a method and system for security enhancement and transmission optimization for URLLC services. Background Art

[0002] With the rapid development of wireless communication technology, the continued rapid growth of mobile communication demand has become an irresistible trend and law. In recent years, the ubiquitous use of smart terminals, diverse new services, and the widespread adoption of the Internet of Things (IoT) have placed new demands on mobile communication systems, including extreme speeds, ultra-low latency, ultra-high reliability, and support for massive connections. Given the continued growth of user demand, increasingly diverse application scenarios, and increasingly differentiated service types, traditional 4G systems are clearly unable to continue to support today's communication needs. Consequently, the fifth-generation (5G) mobile communication system has emerged. Unlike previous generations, which only provide interpersonal mobile broadband communications, 5G systems are designed to meet the information needs of human society after 2020. They will penetrate mobile communications into emerging areas such as the Internet of Things (IoT), deeply integrating them with applications such as industrial facilities, medical equipment, and transportation, effectively meeting the information service needs of vertical industries such as industry, healthcare, and transportation. In other words, 5G systems must not only meet people's demand for high-quality digital communications, but also support IoT-related applications within their system architecture, building a digital world where everything is connected. In view of the diverse application scenarios and differentiated service types of 5G networks, the International Telecommunication Union (ITU) divided 5G network applications into the following three scenarios in a white paper published in September 2015 [1]: 1) Enhanced Mobile Broadband (eMBB), as an evolution of traditional mobile broadband services, this scenario corresponds to human-to-human communication applications such as high-definition video and real-time social networking, which mainly pursues high speed and wide coverage; 2) Massive Machine-Type Communications (mMTC), this scenario corresponds to massively connected IoT applications such as smart cities and environmental monitoring, which mainly pursues large connections and low energy consumption; 3) Ultra-Reliable and Low-Latency Communications (URLLC), this scenario corresponds to touch-level applications such as industrial control, autonomous driving, and telemedicine, which mainly pursues ultra-low latency and ultra-reliable wireless transmission.

[0003] As one of the three major application scenarios of 5G networks, URLLC mainly corresponds to mission-critical IoT applications represented by industrial control and autonomous driving, and human-computer interaction applications represented by telemedicine and tactile Internet. The basic characteristics of these URLLC applications are extremely low end-to-end latency and extremely high transmission reliability. Different URLLC applications have different requirements for latency and reliability based on actual needs. According to the typical URLLC application defined in the 3GPP TR 38.913 document, it requires a 32-byte data packet to be sent to the destination with 99.999% transmission reliability within a 1ms user plane end-to-end latency. In particular, the user plane end-to-end latency refers to the total delay experienced by the data packet from the transmitting end data link layer to the receiving end data link layer, including transmission delay, queuing delay, processing delay, etc.; transmission reliability represents the probability of a data packet being successfully received and decoded within a given end-to-end latency. Clearly, traditional 4G networks struggle to support the demanding performance requirements of URLLC services. The main reasons are as follows: First, the basic time scheduling unit specified in the 4G network protocol, the Transmission Time Interval (TTI), is 1ms. Furthermore, 4G networks employ a grant-based transmission method, meaning that data packets must undergo several signaling interactions before transmission can be performed. This protocol configuration means that the end-to-end latency of 4G networks is typically much greater than 1ms. This means that the user plane end-to-end latency requirement of 1ms for URLLC services is unattainable under 4G network configurations. Second, traditional 4G networks generally use diversity technologies to improve transmission reliability, such as spatial diversity through multiple antennas, time diversity through fast automatic repeat request (HARQ), and frequency diversity through multiple connections. However, the 1ms end-to-end latency constraint for URLLC services makes time diversity difficult to implement and its benefits are limited. In summary, 5G networks require a redesign of physical layer parameters, transceiver architecture, and radio access protocols based on 4G networks to ensure the performance requirements of URLLC services.

[0004] To address the above challenges, existing technologies only focus on the latency and reliability indicators of URLLC services. However, performance indicators other than latency and reliability of URLLC service-related application scenarios have not been fully explored and lack effective definitions. Summary of the Invention

[0005] The purpose of the present invention is to provide a method and system for security enhancement and transmission optimization of URLLC services to solve the above problems.

[0006] To achieve the above object, the present invention adopts the following technical solutions:

[0007] The method for enhancing security and optimizing transmission for URLLC services includes the following steps:

[0008] Construct a URLLC service security transmission model based on the URLLC service security transmission system in the industrial Internet of Things scenario;

[0009] Based on the short packet information theory, a URLLC service security transmission performance characterization framework is constructed to obtain the equivalent achievable secure transmission rate (AESR).

[0010] Based on the proposed secure transmission performance characterization framework, a security performance optimization problem is constructed. By combining the channel estimation pilot length with the artificial noise injection strategy, the security performance of URLLC services is enhanced while ensuring the delay and reliability performance.

[0011] Through a two-dimensional search algorithm based on the golden section idea, it can quickly converge to the approximate optimal solution of the above optimization problem with low time complexity.

[0012] Furthermore, the URLLC service security transmission system in the industrial Internet of Things scenario includes a central controller, an actuator, and several passive eavesdropping nodes; the central controller acts as a transmitter, sending URLLC signaling containing critical mission information via downlink to remotely control the actuator nodes to complete corresponding actions; several passive eavesdropping nodes are distributed around the central controller.

[0013] Furthermore, the URLLC service security transmission model:

[0014] The total transmission delay of URLLC signaling is limited to m, and its corresponding transmission reliability is limited to ∈; there are several passive eavesdroppers {E k , k=1,2,…}; the passive eavesdropper {E k , k=1, 2, ...} Its spatial distribution in two-dimensional space is modeled as a density λ E Homogeneous Poisson point process Φ E ; The legal link between the central controller and the actuator is recorded as link 0, and the link between the central controller and each passive eavesdropper E is recorded as link 0. k The eavesdropping links between them are denoted as links k=1, 2, ...; the channel response matrices of the legitimate link 0 and the eavesdropping links k=1, 2, ... are expressed as and

[0015] Where i∈{u,d} represents uplink transmission and downlink transmission respectively, α represents the path loss parameter, d0 represents the straight-line distance between the central controller and the actuator, and d k Represents the central controller and any passive eavesdropper E k The straight-line distance between0，i and g k represent the small-scale fading components of the legitimate link and the eavesdropping link respectively; g 0，i and g k Each element in obeys a zero-mean unit-variance complex Gaussian distribution, that is, The response matrix between the uplink and downlink channels of the legal link is reciprocal, that is,

[0016] Further, legal links:

[0017] Before the central controller broadcasts URLLC signaling downlink, it obtains the instantaneous channel state information about link 0; the unit transmission cycle of URLLC signaling is divided into two stages: channel training and data transmission. In the channel training stage, the actuator uploads a symbol length of m to the central controller. t The pilot sequence Used to perform channel estimation; the received signal at the central controller is expressed as:

[0018]

[0019] where ρ t Represents the transmission power of the actuator end; Represents additive white Gaussian noise, whose elements all obey zero mean and variance Ω n Complex Gaussian distribution; when the central controller receives the signal After that, it will use the minimum mean square error algorithm to calculate the legal link uplink channel h 0，u The instantaneous value of is estimated; the channel estimation value is expressed as The corresponding estimation error is and are independent of each other, and the expressions are as follows:

[0020]

[0021] in Represents the transmission signal-to-noise ratio at the actuator; after completing the channel estimation, the central controller combines the h 0，u The beamforming matrix is ​​generated based on the prior knowledge of uplink and downlink channel reciprocity and the maximum ratio combining principle. The central controller injects artificial noise into the null space of the data signal during the data transmission phase. Specifically, the private data signal transmitted by the central controller is expressed as:

[0022]

[0023] in Represents a symbol length of m s =mm t The data signal, ρs Represents the transmission power of the central controller. represents the power distribution ratio between data transmission and artificial noise; the URLLC signaling received by the actuator through the legal link is expressed as:

[0024]

[0025] in Represents additive white Gaussian noise, whose elements all obey zero mean and variance Ω n The equivalent received signal-to-noise ratio of the actuator after decoding is:

[0026]

[0027] in Represents the transmission signal-to-noise ratio at the central controller.

[0028] Further, eavesdropping link:

[0029] Each eavesdropper E k The URLLC signaling intercepted by eavesdropping on the link is represented as follows:

[0030]

[0031] in Represents additive white Gaussian noise, whose elements all obey zero mean and variance Ω n complex Gaussian distribution; each eavesdropper E k The equivalent received signal-to-noise ratio is:

[0032]

[0033] The interception capability of an eavesdropper for URLLC signaling is quantified as

[0034] Furthermore, a URLLC service security transmission performance characterization framework is constructed:

[0035] The privacy capacity that can achieve absolute security is defined as the upper limit of the number of data bits that a transmitter can send within a unit symbol duration while ensuring that the probability of decoding errors on the legitimate end and the probability of information interception on the eavesdropping end are arbitrarily low. The privacy capacity based on the assumption of asymptotically infinite code length is expressed as:

[0036] C sec =C0-C E =log2(1+γ0)-log2(1+γ E ), (8)

[0037] where γ0 and Represents the instantaneous received signal-to-noise ratio of the legitimate link and the eavesdropping link respectively; the upper limit of the private rate of the short packet communication service is within the traditional private capacity index C sec On the basis of , two penalty terms are added, and the mathematical expression is:

[0038]

[0039] in and represents the channel dispersion coefficient of the legitimate link and the eavesdropping link, ∈∈(0,1) represents the target decoding error probability of the legitimate receiving end, δ∈(0,1) represents the information interception probability of the eavesdropping end, Q -1 (z) represents the inverse function of the Gaussian Q-function, where the mathematical expression of the Gaussian Q-function is

[0040] According to formula (9), as long as the current data transmission rate of URLLC signaling does not exceed its private rate upper limit R sec , then there must be a corresponding wiretap short packet channel coding structure that makes the transmission reliability limit ∈∈(0,1) and security limit δ∈(0,1) of URLLC signaling under a given transmission delay;

[0041] The transmission performance upper limit of URLLC services with security requirements is evaluated based on the equivalent achievable secure transmission rate (AESR). Under the premise of given transmission reliability and security constraints, the mathematical expectation of the number of URLLC service bits that can be transmitted per unit symbol time is expressed as follows:

[0042]

[0043] The mathematical expressions of the integral terms Φ0(γ0) and Ψ0(γ0,∈) are:

[0044]

[0045] and

[0046]

[0047] in is the probability density function of the instantaneous signal-to-noise ratio γ0 of the legal link, and the mathematical expression of the random variable γ0 is shown in (5); the random variable γ0 obeys the shape parameter A T , the scale parameter is Gamma distribution, that is The PDF expression of the random variable γ0 is as follows:

[0048]

[0049] Integral term Φ E (γ E ) and Ψ E (γ E ,δ) are:

[0050]

[0051] and

[0052]

[0053] in is the instantaneous signal-to-noise ratio of the eavesdropping link γ E PDF of Before the mathematical expression of E The cumulative distribution function of is:

[0054]

[0055] in Then, by performing the derivation operation, we get The mathematical expression is:

[0056]

[0057] in

[0058] According to the above expression, the above expression is simplified to calculate the AESR.

[0059] Further simplification is divided into:

[0060] No artificial noise injection In this case, and Based on the equation and The integral term Φ0(γ0) shown in formula (11) is simplified to:

[0061]

[0062] Based on the Laplace approximation method, the closed-form approximation of the integral term Ψ0(γ0,∈) shown in formula (12) is obtained as follows:

[0063]

[0064] in and The formula (17) shows The mathematical expression is simplified to Integral term Φ E (γ E) is simplified to:

[0065]

[0066] Define a sufficiently large positive real number It meets the following conditions: hour, The integral term Ψ shown in formula (15) E (γ E ,δ) is approximately:

[0067]

[0068] By introducing the Gaussian-Chebyshev integration method and changing the polynomial, we can get the integral term Ψ E (γ E , the closed-form approximation of δ) is:

[0069]

[0070] in and The mathematical expressions of are shown in formulas (16) and (17) respectively; finally, we get AESR in The closed-form approximation for this case is:

[0071]

[0072] Secure transmission based on artificial noise injection In this case, and Using the same theoretical derivation method as equations (18) and (19), the integral terms Φ0(γ0) and Ψ0(γ0,∈) in this case are simplified to:

[0073]

[0074] and

[0075]

[0076] Based on the method of integration by parts, the integral term Φ E (γ E ) is converted to:

[0077]

[0078] By introducing the Gaussian-Chebyshev integration method and changing the polynomial, we can finally get the integral term Ψ in this case: E (γ E , δ) is a closed-form approximation:

[0079]

[0080] Similarly, by using the same theoretical derivation method as formulas (21) and (22), we can obtain the integral term Ψ in this case: E (γ E , δ) is a closed-form approximation:

[0081]

[0082] Finally, we get AESR in The closed-form approximation for this case is:

[0083]

[0084] Furthermore, based on the secure transmission performance characterization framework, a security performance optimization problem is constructed. By jointly designing the channel estimation pilot length and the artificial noise injection strategy, the security performance of URLLC is enhanced while ensuring the service delay and reliability performance. Specifically, by adjusting the channel estimation pilot length m t and artificial noise injection power ratio Perform joint optimization to maximize AESR while ensuring the latency, reliability, and security performance constraints of URLLC services. The optimization problem is formulated as follows:

[0085]

[0086] stm t ={1,2,…,m-1}, (30a)

[0087]

[0088] The objective function The mathematical expressions in different cases are shown in formulas (23) and (29) respectively.

[0089] Furthermore, the security performance optimization algorithm for URLLC services is as follows:

[0090] The two-dimensional search algorithm based on the golden section idea quickly converges to the approximate optimal solution of the optimization problem proposed in step 3. It is divided into the following two stages:

[0091] First, this algorithm iteratively reduces the optimization variable m based on the golden section search method. t The search space of And optimize the variable m t The initial search space is denoted as [m t，L , m t，U ], where m t，Land m t，U The initial values ​​are m t，L =1 and m t，U =m-1; in each iteration, with a fixed golden ratio to [m t，L , m t，U ] is shrunk by setting the following two coefficients: and

[0092] By using the one-dimensional search method, we can calculate the t =m1 and m t =m2, which is the ratio of artificial noise injection power that maximizes AESR. The following judgment conditions are used to determine m t，L and m t，U Iteratively update the value of: If The optimal solution is inferred The value is in the interval [m1,m t，U ]In this case, we make Then the search space [m t，L , m t，U ]Updated to And update the values ​​of m1 and m2 to and if The optimal solution is inferred The value is in the interval [m t，L , m2]; in this case, let Then the search space [m t，L , m t，U ]Updated to And update the values ​​of m1 and m2 to and When m t，U -m t，L The iteration is completed when ≤0.5; and is the approximate optimal solution output by the algorithm, and the corresponding AESR optimal value is

[0093] Furthermore, the security enhancement and transmission optimization system for URLLC services includes:

[0094] A secure transmission model building module is used to build a URLLC service secure transmission model based on the URLLC service secure transmission system in the industrial Internet of Things scenario;

[0095] The equivalent achievable secure transmission rate acquisition module is used to build a URLLC service secure transmission performance characterization framework based on short packet information theory to obtain the equivalent achievable secure transmission rate (AESR).

[0096] The security performance optimization problem construction module is used to construct a security performance optimization problem based on the proposed security transmission performance characterization framework. By combining the channel estimation pilot length with the artificial noise injection strategy, the security performance of URLLC services can be enhanced while ensuring the delay and reliability performance of the services.

[0097] The optimal solution acquisition module is used to quickly converge to the approximate optimal solution of the above optimization problem with low time complexity through a two-dimensional search algorithm based on the golden section idea.

[0098] Compared with the prior art, the present invention has the following technical effects:

[0099] In view of the current situation that the existing URLLC-related research only focuses on delay and reliability indicators, while ignoring the security requirements of URLLC service transmission. The present invention considers the downlink URLLC service transmission system for mission-critical industrial Internet of Things scenarios, and proposes a new security performance analysis framework based on full consideration of the short packet transmission characteristics of the URLLC service and the random distribution characteristics of the eavesdropper's geographical location. Specifically, due to the short packet transmission characteristics of the URLLC service, the channel estimation is inaccurate and data decoding errors are inevitable, which leads to the traditional performance indicators based on Shannon theory no longer being applicable to describe the transmission performance of the URLLC service. After comprehensively considering the impact of the short packet transmission characteristics and the random distribution characteristics of the eavesdropper's geographical location on the transmission performance of the URLLC service, the present invention proposes a new indicator called the equivalent achievable secure transmission rate (AESR) to describe the transmission performance upper limit of the URLLC service with security requirements, that is, the average number of URLLC service bits that can be transmitted per unit symbol time under the premise of ensuring given transmission reliability and security constraints. According to the above definition, the present invention obtains a closed-form expression for AESR based on the Laplace approximation method and the Gaussian-Chebyshev integration method, which provides a theoretical basis for the safety performance analysis of the URLLC system.

[0100] At the same time, the openness of the wireless channel and the vulnerability of the URLLC service to eavesdropping attacks are taken into account. The present invention designs a new security enhancement and performance optimization scheme based on the proposed performance analysis framework, which is as follows: On the one hand, the present invention injects artificial noise into the null space of the transmitted signal matrix to effectively suppress the eavesdropper's signal interception capability without interfering with the normal decoding of the legitimate receiver, so as to combat eavesdropping attacks from malicious nodes. On the other hand, the present invention proposes a two-dimensional search algorithm based on the golden section search method, which jointly optimizes the channel estimation pilot length and the artificial noise injection strategy to achieve an approximately optimal transmission design under the premise of ensuring the delay, reliability and security constraints of the URLLC service, that is, to maximize the AESR. In summary, the scheme proposed in the present invention can effectively suppress the potential information leakage risk caused by randomly distributed eavesdroppers while ensuring the transmission delay and reliability performance of the URLLC service, thereby achieving security enhancement and transmission optimization of the URLLC service. BRIEF DESCRIPTION OF THE DRAWINGS

[0101] Figure 1 The figure shows a model diagram of the URLLC service downlink secure transmission system considered in the present invention.

[0102] Figure 2 The figure shows the comparison between different evaluation models for URLLC service AESR performance under different transmission delays (channel uses).

[0103] Figure 3 The figure shows the comparison of AESR performance between the algorithm proposed in the present invention and the existing scheme under different transmission delays (channel uses).

[0104] Figure 4 Shown is the AESR performance for a given channel estimation pilot length (channel uses) and different artificial noise injection power ratios.

[0105] Figure 5 Shown is the AESR performance under different channel estimation pilot lengths (channel uses). DETAILED DESCRIPTION

[0106] The present invention is further described below with reference to the accompanying drawings:

[0107] See also Figures 1 to 5 :

[0108] The method for security enhancement and transmission optimization for URLLC services proposed in the present invention includes the following steps:

[0109] Build a URLLC service security transmission model;

[0110] Construct a URLLC service security transmission performance characterization framework based on short packet information theory;

[0111] Based on the proposed secure transmission performance characterization framework, a security performance optimization problem is constructed. By jointly designing the channel estimation pilot length and artificial noise injection strategy, the security performance of URLLC services is enhanced while ensuring the latency and reliability performance.

[0112] Through a two-dimensional search algorithm based on the golden section idea, it quickly converges to the approximate optimal solution of the optimization problem proposed in step 3 with low time complexity.

[0113] Next, we will elaborate on the specific content of each step:

[0114] Construction of URLLC service secure transmission model

[0115] like Figure 1 As shown in Figure 1, we consider a URLLC service security transmission system for industrial IoT scenarios, which includes a central controller (controller), an actuator (actuator) that performs related tasks, and several passive eavesdroppers. In particular, the central controller acts as a transmitter, which remotely controls the actuator nodes to complete related actions by sending URLLC signaling containing key task information downlink. In general, we assume that the total transmission delay of URLLC signaling is limited to m (channel uses), and its corresponding transmission reliability (transmission error probability under a given delay limit) is limited to ∈. At the same time, we assume that there are several passive eavesdroppers {E k , k=1,2,…} try to intercept the URLLC control signaling sent by the central control. For realistic considerations, we assume that the central controller is equipped with A T ≥2 transmitting antennas, actuator and each passive eavesdropper E k Since the eavesdropper will not actively send any radiation signal that may reveal its position when trying to intercept URLLC control signaling, we usually assume that the passive eavesdropper {E k , k = 1, 2, ...} is unknown in terms of its specific number and geographical location, and its spatial distribution in two-dimensional space is modeled as a density λ E Homogeneous Poisson point process (PPP)Φ E For the convenience of distinction, we will record the legitimate link between the central controller and the actuator as link 0 in the following analysis, and the link between the central controller and each passive eavesdropper E as kThe eavesdropping links between them are denoted as links k = 1, 2, .... Based on the above description, we may express the channel response matrices of the legitimate link 0 and the eavesdropping links k = 1, 2, ... as and where i∈{u,d} represents uplink transmission (from the actuator to the central controller) and downlink transmission (from the central controller to the actuator), α represents the path loss parameter, d0 represents the straight-line distance between the central controller and the actuator, and d k Represents the central controller and any passive eavesdropper E k The straight-line distance between 0，i and g k represent the small-scale fading components of the legitimate link and the eavesdropped link respectively. In particular, we assume that g 0，i and g k Each element in obeys a zero-mean unit-variance complex Gaussian distribution, that is, In addition, we also assume that the response matrix between the uplink and downlink channels of the legal link is reciprocal, that is, Next, we introduce the receiving signal models of the legitimate link and the eavesdropping link respectively:

[0116] 1) Legal link: Before the central controller broadcasts URLLC signaling downlink, it first needs to obtain the instantaneous channel state information (CSI) about link 0. Therefore, the unit transmission cycle of URLLC signaling is divided into two stages: channel training and data transmission. In the channel training stage, the actuator uploads a symbol length of m to the central controller. t (channel uses) pilot sequence Used to perform channel estimation. Accordingly, the received signal at the central controller can be expressed as:

[0117]

[0118] where ρ t Represents the transmission power of the actuator end; Represents additive white Gaussian noise, whose elements all obey zero mean and variance Ω n When the central controller receives the signal After that, it will use the minimum mean square error (MMSE) algorithm to calculate the legal link uplink channel h 0，u The instantaneous value of is estimated. Accordingly, the channel estimation value can be expressed as The corresponding estimation error is According to the general properties of the MMSE estimation algorithm, it is not difficult to know that and are independent of each other, and we can get the following expressions:

[0119] in Represents the transmission signal-to-noise ratio at the actuator. After completing the channel estimation, the central controller will combine the h 0，u The beamforming matrix is ​​generated based on the prior knowledge of uplink and downlink channel reciprocity and the maximum ratio combining principle. At the same time, the central controller injects artificial noise (AN) into the null space of the data signal during the data transmission phase to suppress the eavesdropper's receiving signal-to-noise ratio. Specifically, the private data signal transmitted by the central controller can be expressed as:

[0120]

[0121] in Represents a symbol length of m s =mm t (channel uses) data signal, ρ s Represents the transmission power of the central controller. represents the power distribution ratio between data transmission and artificial noise. Therefore, the URLLC signaling received by the actuator through the legal link can be expressed as:

[0122]

[0123] in Represents additive white Gaussian noise, whose elements all obey zero mean and variance Ω n Finally, it is not difficult to deduce that the equivalent received signal-to-noise ratio of the actuator after decoding is:

[0124]

[0125] in Represents the transmission signal-to-noise ratio at the central controller.

[0126] Eavesdropping Link: As mentioned above, when an eavesdropper attempts to intercept URLLC control signaling, it will not actively send any radiation signals that may reveal its location. k , k=1,2,…} are unknown to the legitimate end. In order to consider the extreme case, we may as well assume that all passive eavesdroppers have no channel estimation errors and their channel state information is only known to the legitimate end through statistical distribution. Based on the above assumptions, each eavesdropper E k The URLLC signaling intercepted by eavesdropping on the link can be expressed as:

[0127]

[0128] in Represents additive white Gaussian noise, whose elements all obey zero mean and variance Ω n Finally, it is not difficult to deduce that each eavesdropper E k The equivalent received signal-to-noise ratio is:

[0129]

[0130] In particular, we consider the case of non-colluding eavesdroppers in this invention, that is, all passive eavesdroppers {E k , k=1,2,…} all independently try to intercept URLLC signaling. Based on this assumption, we can further quantify the interception capability of the eavesdropper for URLLC signaling as

[0131] URLLC service secure transmission performance characterization framework based on short packet information theory

[0132] In traditional real-time communication services such as video and voice, the duration of their arrival streams is relatively long (usually exceeding 10 seconds), so the channel coding blocks of their data packets are long enough. Given this characteristic, the Shannon capacity C = log2(1 + γ) with the assumption of asymptotically infinite code length is widely used to evaluate the transmission performance limit of wireless systems, where γ represents the instantaneous signal-to-noise ratio at the receiving end. From the perspective of information theory, we can also draw the following conclusion: as long as the current transmission rate of the service does not exceed the Shannon capacity, there must be a corresponding Guassian channel coding scheme that allows the data packet to be correctly decoded by the receiving end. Based on this principle, we can further define the secrecy capacity that can achieve absolute security, that is, the upper limit of the number of data bits that the transmitter can send within the unit symbol duration while ensuring that the probability of decoding errors at the legitimate end is arbitrarily low and the probability of information interception at the eavesdropping end is arbitrarily low. Therefore, the secrecy capacity based on the assumption of asymptotically infinite code length can be expressed as:

[0133] C sec =C0-C E =log2(1+γ0)-log2(1+γ E ), (8)

[0134] where γ0 and Represent the instantaneous received signal-to-noise ratio of the legitimate link and the eavesdropping link respectively. From the perspective of information theory, we can also draw the following conclusion from formula (8): As long as the current transmission rate of the service does not exceed the private capacity C sec, then there must be a corresponding wiretap channel coding structure so that the data packet can be correctly decoded by the receiving end without any information leakage. However, this conclusion is no longer applicable to short packet services represented by URLLC. This is because the discrete small data packet transmission characteristics of URLLC services mean that its channel coding length is extremely limited, which in turn makes the traditional Shannon capacity upper limit unattainable in short packet communication scenarios. Specifically, the privacy rate upper limit of short packet communication services is within the traditional privacy capacity index C sec On the basis of , two penalty terms are added, and the mathematical expression is:

[0135]

[0136] in and represents the channel dispersion coefficient of the legitimate link and the eavesdropping link, ∈∈(0,1) represents the target decoding error probability of the legitimate receiving end, δ∈(0,1) represents the information interception probability of the eavesdropping end, Q -1 (z) represents the inverse function of the Gaussian Q-function, where the mathematical expression of the Gaussian Q-function is From the perspective of information theory, we can draw the following conclusion from formula (9): As long as the current data transmission rate of URLLC signaling does not exceed its private rate upper limit R sec , then there must be a corresponding wiretap short packet channel coding structure that ensures that the transmission reliability limit ∈∈(0, 1) and the security limit δ∈(0, 1) of URLLC signaling under a given transmission delay are met. Otherwise, the transmission reliability limit and security limit of URLLC signaling cannot be guaranteed at the same time. Note that the specific design of the short packet secure channel coding scheme is beyond the scope of this invention and will not be discussed here.

[0137] Based on the above analysis, we propose the concept of the achievable effective secrecy rate (EER) for the first time in this paper to evaluate the transmission performance upper limit of URLLC services with security requirements. EER is defined as the mathematical expectation of the number of URLLC service bits that can be transmitted per symbol time under given transmission reliability and security constraints. Its expression is:

[0138]

[0139] On the one hand, the mathematical expressions of the integral terms Φ0(γ0) and Ψ0(γ0,∈) are:

[0140]

[0141] and

[0142]

[0143] in is the probability density function (PDF) of the instantaneous signal-to-noise ratio γ0 of the legal link, and the mathematical expression of the random variable γ0 is shown in (5). Through a series of mathematical derivations, it is not difficult to draw the following conclusion: the random variable γ0 obeys the shape parameter A T , scale parameter (scaleparameter) is (in ) is the Gamma distribution, that is Correspondingly, the PDF expression of the random variable γ0 is as follows:

[0144]

[0145] On the one hand, the integral term Φ E (γ E ) and Ψ E (γ E ,δ) are:

[0146]

[0147] and

[0148]

[0149] in is the instantaneous signal-to-noise ratio of the eavesdropping link γ E In the solution Before we get the mathematical expression of random variable γ, we first use the basic theory of random geometry to get the random variable γ E The cumulative distribution function (CDF) is:

[0150]

[0151] in Then by taking the derivative, we can get The mathematical expression is:

[0152]

[0153] in In fact, calculating AESR using integral expressions such as those shown in formulas (10)-(17) is obviously more complicated. Next, we use a series of approximation tools to simplify the above mathematical expressions, which are divided into the following two cases:

[0154] No artificial noise injection In this case, we have and Based on the equation and The integral term Φ0(γ0) shown in formula (11) can be simplified to:

[0155]

[0156] Based on the Laplace approximation method, we can obtain the closed-form approximation of the integral term Ψ0(γ0,∈) as shown in formula (12), as follows:

[0157]

[0158] in and As shown in formula (17) The mathematical expression of can be simplified to Integral term Φ E (γ E ) can be simplified accordingly:

[0159]

[0160] To simplify the integral term Ψ E (γ E ,δ), we first define a sufficiently large positive real number It meets the following conditions: We can approximately think that Based on this definition, the integral term Ψ shown in formula (15) E (γ E ,δ) can be approximated as:

[0161]

[0162] By introducing the Gaussian-Chebyshev integration method and changing the polynomial, we can finally get the integral term Ψ E (γ E , the closed-form approximation of δ) is:

[0163]

[0164] in and and The mathematical expressions of are shown in formulas (16) and (17). Finally, it is not difficult to obtain AESR in The closed-form approximation for this case is:

[0165]

[0166] Secure transmission based on artificial noise injection In this case, we have and Using the same theoretical derivation method as equations (18) and (19), we can simplify the integral terms Φ0(γ0) and Ψ0(γ0,∈) in this case to:

[0167]

[0168] and

[0169]

[0170] Based on the method of integration by parts, we can in this case convert the integral term Φ E (γ E ) is converted to:

[0171]

[0172] By introducing the Gaussian-Chebyshev integration method and changing the polynomial, we can finally get the integral term Ψ in this case E (γ E , δ) is a closed-form approximation:

[0173]

[0174] Similarly, by using the same theoretical derivation method as formulas (21) and (22), we can finally obtain the integral term Ψ in this case E (γ E , δ) is a closed-form approximation:

[0175]

[0176] Finally, it is not difficult to get AESR in The closed-form approximation for this case is:

[0177]

[0178] Construction of security enhancement strategy and transmission optimization problem for URLLC services

[0179] According to the above analysis of the mathematical expression of AESR, it is not difficult to see that the AESR performance of URLLC service depends not only on the inherent system parameters such as the communication distance of the legal transceiver, the density of the eavesdropping space, the transmission reliability limit, the security limit, but also on the channel estimation pilot length m. t and artificial noise injection power ratio Therefore, based on the proposed security transmission performance characterization framework, we construct a security performance optimization problem and jointly design the channel estimation pilot length and artificial noise injection strategy to enhance the security performance of URLLC while ensuring the delay and reliability performance of the service. Specifically, we try to improve the channel estimation pilot length m t and artificial noise injection power ratio Perform joint optimization to maximize AESR while ensuring the latency, reliability, and security performance constraints of URLLC services. In general, the optimization problem in this invention is constructed as follows:

[0180]

[0181] stm t ={1,2,…,m-1}, (30a)

[0182]

[0183] The objective function The mathematical expressions in different cases are shown in formulas (23) and (29) respectively.

[0184] Security performance optimization algorithm for URLLC services

[0185] In fact, the AESR maximization problem constructed in the previous step is a typical mixed integer linear programming problem, so it is impossible to solve the optimal solution based on the KKT condition. In view of this, we propose a two-dimensional search algorithm based on the golden section idea to quickly converge to the approximate optimal solution of the optimization problem proposed in step 3. Generally, the algorithm proposed in this invention is divided into the following two stages:

[0186] First, the algorithm iteratively reduces the optimization variable m based on the Golden-section search method (GSSM). t To facilitate readers’ understanding, we record And optimize the variable m t The initial search space is denoted as [m t，L , m t，U ], where m t，L and m t，UThe initial values ​​are m t，L =1 and m t，U =m-1. In each iteration, we use a fixed golden ratio to [m t，L , m t，U ] is shrunk by setting the following two coefficients: and

[0187] Next, we use the one-dimensional search method to calculate the t =m1 and m t =m2, which is the ratio of artificial noise injection power that maximizes AESR. At the same time, we use the following judgment conditions to determine m t，L and m t，U Iteratively update the value of: If Then we can infer the optimal solution The value is in the interval [m1,m t，U ]. In this case, we make Then the search space [m t，L , m t，U ]Updated to And update the values ​​of m1 and m2 to and if Then we can infer the optimal solution The value is in the interval [m t，L , m2]. In this case, we let Then the search space [m t，L , m t，U ]Updated to And update the values ​​of m1 and m2 to and Since the optimization variable m t In formula (30a) it is restricted to integers, so when m t，U -m t，L When ≤0.5, it means the iteration is completed. and is the approximate optimal solution output by the algorithm, and the corresponding AESR optimal value is

[0188] In the present invention, unless otherwise specified, the simulation parameters are set as follows: Figure 1As shown in Figure 1, we consider a URLLC service downlink secure transmission system, which includes a central controller (controller), an actuator (actuator) that performs related tasks, and several passive eavesdroppers. In particular, the central controller acts as a transmitter and remotely controls the actuator nodes to complete related actions by sending URLLC signaling containing key task information downlink. The straight-line distance between the central controller and the actuator is d0 = 100m, and the passive eavesdropper {E k , k = 1, 2, ...} is unknown in terms of its specific number and geographical location, and its spatial distribution in two-dimensional space is modeled as a density λ E =1(units / m 2 ) of the homogeneous Poisson point process (PPP)Φ E In terms of wireless link parameters, we assume that the path loss parameter α = 4, and the transmit power for transmitting channel estimation pilot and URLLC signaling is ρ t =5dBm and ρ s =10dBm. In addition, the system bandwidth is W = 360kHz, which means that the total number of symbols in 1ms duration is 360 (channel uses) and the noise power spectral density is N0 = -174dBm / Hz. In addition, the reliability and security constraints of URLLC services are ∈ = 10 -7 and δ = 10 -3 .

[0189] Figure 1 The figure shows a model diagram of the URLLC service downlink secure transmission system considered in the present invention.

[0190] Figure 2 The figure shows the comparison between different evaluation models for URLLC service AESR performance under different transmission delays m (channel uses), where the relevant system parameters are set as follows: T =8, N=20, from Figure 2 It is not difficult to draw the following conclusions: 1) The AESR defined in the present invention based on the short packet transmission theory is always faster than the ergodic secrecy rate based on the traditional asymptotic infinite code length assumption. This indicates that the traditional security performance indicators represented by the ergodic security rate overestimate the security transmission performance of the URLLC service. 2) The integral expression of AESR shown in formula (10), the closed-form approximation of AESR shown in formulas (23) and (29), and the Monte-Carlo simulation curves in actual scenarios are highly consistent, which verifies that the closed-form approximation of AESR proposed in this invention is very accurate. 3) The AESR is obviously much larger than This shows that the security performance of URLLC services can be significantly improved by injecting artificial noise.

[0191] Figure 3 The figure shows the comparison of AESR performance between the algorithm proposed in this invention and the existing scheme under different transmission delays m (channel uses), where the relevant system parameters are set as follows: T =8, N=20, from Figure 3 It is not difficult to draw the following conclusions: 1) The local optimal solution output by the two-dimensional search algorithm proposed in this invention is highly consistent with the global optimal solution output by the exhaustive search algorithm, which shows that the two-dimensional search algorithm proposed in this invention can quickly converge to the global optimal solution with lower time complexity; 2) The pilot length m of the channel estimation is t and artificial noise injection power ratio The AESR of URLLC service can be improved by optimizing them separately, which shows that by designing the corresponding algorithm, t and Joint optimization is necessary to maximize the AESR of URLLC services.

[0192] Figure 4 The pilot length m is shown for a given channel estimation t =0.25m (channel uses) and different artificial noise injection power ratios AESR performance under the following conditions, where the relevant system parameters are set as follows: T =8, m=360, λ E ={0.01, 0.1, 1, 10}. Figure 4 It is not difficult to draw the following conclusions: 1) The AESR performance of URLLC service decreases with the ratio of artificial noise injection power to The increase of α first increases and then decreases, that is, designing a suitable artificial noise injection strategy is essential to achieve AESR maximization; 2) The AESR performance of URLLC service decreases with the increase of the eavesdropper spatial distribution density λ E This is due to the fact thatE The increase in the number of messages means that eavesdroppers have a stronger ability to intercept information, thereby increasing the risk of information leakage.

[0193] Figure 5 Shown are different channel estimation pilot lengths m t AESR performance under (channel uses), where the relevant system parameters are set as follows: A T ={4,8},m=360. Figure 5 It is not difficult to draw the following conclusions: 1) The AESR performance of URLLC service decreases with the channel estimation pilot length m t The increase first increases and then decreases, that is, under a given delay limit m, the channel estimation pilot length m t and data transmission symbol length m s =mm t Designing a suitable compromise between the two helps to maximize AESR; 2) When the transmission reliability constraint ∈∈(0, 1) or the security constraint δ∈(0, 1) of the URLLC service becomes stricter, its AESR performance will inevitably decline. Intuitively, when the receiver's requirements for decoding capability become higher, the transmitter can only meet this requirement by reducing the number of bits transmitted per unit time, which leads to a decline in AESR; 3) The AESR performance of the URLLC service decreases with the number of transmitting antennas A. T This shows that additional spatial gain can be obtained by increasing the number of transmitting antennas, and combined with the artificial noise auxiliary strategy, the security performance of the URLLC system can be significantly improved.

[0194] In summary, the present invention constructs a security performance analysis framework that conforms to the short packet transmission characteristics of URLLC services, providing a theoretical basis for the security performance analysis of URLLC systems. Furthermore, the present invention proposes a security enhancement framework based on artificial noise, and achieves security enhancement and transmission optimization for URLLC services by jointly optimizing the channel estimation pilot length and artificial noise injection strategy.

Claims

1. A method for security enhancement and transmission optimization for URLLC services, characterized in that: The following steps are involved: Construct a URLLC service security transmission model based on the URLLC service security transmission system in the industrial Internet of Things scenario; Based on the short packet information theory, a URLLC service security transmission performance characterization framework is constructed to obtain the equivalent achievable secure transmission rate (AESR). Based on the proposed secure transmission performance characterization framework, a security performance optimization problem is constructed. By combining the channel estimation pilot length with the artificial noise injection strategy, the security performance of URLLC services is enhanced while ensuring the delay and reliability performance. Through a two-dimensional search algorithm based on the golden section idea, it quickly converges to the approximate optimal solution of the above optimization problem with low time complexity; Constructing a URLLC service security transmission performance characterization framework: The privacy capacity that can achieve absolute security is defined as the upper limit of the number of data bits that a transmitter can send within a unit symbol duration while ensuring that the probability of decoding errors on the legitimate end and the probability of information interception on the eavesdropping end are arbitrarily low. The privacy capacity based on the assumption of asymptotically infinite code length is expressed as: C sec =C0-C E =log2(1+γ0)-log2(1+γ E ), (8) where γ0 and Represents the instantaneous received signal-to-noise ratio of the legitimate link and the eavesdropping link respectively; the upper limit of the private rate of the short packet communication service is within the traditional private capacity index C sec On the basis of , two penalty terms are added, and the mathematical expression is: in represents the channel dispersion coefficient of the legitimate link and the eavesdropping link, ∈∈(0,1) represents the target decoding error probability of the legitimate receiving end, δ∈(0,1) represents the information interception probability of the eavesdropping end, Q -1 (z) represents the inverse function of the Gaussian Q-function, where the mathematical expression of the Gaussian Q-function is According to formula (9), as long as the current data transmission rate of URLLC signaling does not exceed its private rate upper limit R sec , then there must be a corresponding wiretap short packet channel coding structure that makes the transmission reliability limit ∈∈(0,1) and security limit δ∈(0,1) of URLLC signaling under a given transmission delay; The transmission performance upper limit of URLLC services with security requirements is evaluated based on the equivalent achievable secure transmission rate (AESR). Under the premise of given transmission reliability and security constraints, the mathematical expectation of the number of URLLC service bits that can be transmitted per unit symbol time is expressed as follows: The mathematical expressions of the integral terms Φ0(γ0) and Ψ0(γ0,∈) are: and in is the probability density function of the instantaneous signal-to-noise ratio γ0 of the legal link, and the mathematical expression of the random variable γ0 is shown in (5); the random variable γ0 obeys the shape parameter A T , the scale parameter is Gamma distribution, that is The PDF expression of the random variable γ0 is as follows: Integral term Φ E (γ E ) and Ψ E (γ E ,δ) are: and in ) is the instantaneous signal-to-noise ratio γ of the eavesdropping link E PDF of Before the mathematical expression of E The cumulative distribution function of is: in Then, by performing the derivation operation, we get The mathematical expression is: in According to the above expression, the above expression is simplified to calculate the AESR.

2. The method for security enhancement and transmission optimization for URLLC services according to claim 1, wherein: The URLLC service security transmission system in the industrial Internet of Things scenario includes a central controller, actuators, and several passive eavesdropping nodes. The central controller acts as a transmitter, sending URLLC signaling containing critical mission information via downlink to remotely control the actuator nodes to complete corresponding actions. Several passive eavesdropping nodes are distributed around the central controller.

3. The method for security enhancement and transmission optimization for URLLC services according to claim 2, wherein: URLLC service security transmission model: The total transmission delay of URLLC signaling is limited to m, and its corresponding transmission reliability is limited to ∈; there are several passive eavesdroppers {E k , k=1,2,…}; the passive eavesdropper {E k , k=1, 2, ...} Its spatial distribution in two-dimensional space is modeled as a density λ E Homogeneous Poisson point process Φ E ; The legal link between the central controller and the actuator is recorded as link 0, and the link between the central controller and each passive eavesdropper E is recorded as link 0. k The eavesdropping links between them are denoted as links k = 1, 2, ...; the channel response matrices of the legitimate link 0 and the eavesdropping links k = 1, 2, ... are expressed as and Where i∈{u,d} represents uplink transmission and downlink transmission respectively, α represents the path loss parameter, d0 represents the straight-line distance between the central controller and the actuator, and d k Represents the central controller and any passive eavesdropper E k The straight-line distance between 0，i and g k represent the small-scale fading components of the legitimate link and the eavesdropping link respectively; g 0，i and g k Each element in obeys a zero-mean unit-variance complex Gaussian distribution, that is, The response matrix between the uplink and downlink channels of the legal link is reciprocal, that is, 4. The method for security enhancement and transmission optimization for URLLC services according to claim 3, wherein: Legal links: Before the central controller broadcasts URLLC signaling downlink, it obtains the instantaneous channel state information about link 0; the unit transmission cycle of URLLC signaling is divided into two stages: channel training and data transmission. In the channel training stage, the actuator uploads a symbol length of m to the central controller. t The pilot sequence Used to perform channel estimation; the received signal at the central controller is expressed as: where ρ t Represents the transmission power of the actuator end; Represents additive white Gaussian noise, whose elements all obey zero mean and variance Ω n Complex Gaussian distribution; when the central controller receives the signal After that, it will use the minimum mean square error algorithm to calculate the legal link uplink channel h 0，u The instantaneous value of is estimated; the channel estimation value is expressed as The corresponding estimation error is and are independent of each other, and the expressions are as follows: in represents the transmission signal-to-noise ratio at the actuator; After completing the channel estimation, the central controller combines the h 0，u The beamforming matrix is ​​generated based on the prior knowledge of uplink and downlink channel reciprocity and the maximum ratio combining principle. The central controller injects artificial noise into the null space of the data signal during the data transmission phase. Specifically, the private data signal transmitted by the central controller is expressed as: in Represents a symbol length of m s =mm t The data signal, ρ s Represents the transmission power of the central controller. represents the power distribution ratio between data transmission and artificial noise; the URLLC signaling received by the actuator through the legal link is expressed as: in Represents additive white Gaussian noise, whose elements all obey zero mean and variance Ω n The equivalent received signal-to-noise ratio of the actuator after decoding is: in Represents the transmission signal-to-noise ratio at the central controller.

5. The method for security enhancement and transmission optimization for URLLC services according to claim 3, wherein: Eavesdropping link: Each eavesdropper E k The URLLC signaling intercepted by eavesdropping on the link is represented as follows: in Represents additive white Gaussian noise, whose elements all obey zero mean and variance Ω n complex Gaussian distribution; each eavesdropper E k The equivalent received signal-to-noise ratio is: The interception capability of an eavesdropper for URLLC signaling is quantified as 6. The method for security enhancement and transmission optimization for URLLC services according to claim 1, wherein: Simplified into: No artificial noise injection In this case, and Based on the equation and The integral term Φ0(γ0) shown in formula (11) is simplified to: Based on the Laplace approximation method, the closed-form approximation of the integral term Ψ0(γ0,∈) shown in formula (12) is obtained as follows: in and The formula (17) shows The mathematical expression is simplified to Integral term Φ E (γ E ) is simplified to: Define a sufficiently large positive real number It meets the following conditions: hour, The integral term Ψ shown in formula (15) E (γ E ,δ) is approximately: By introducing the Gaussian-Chebyshev integration method and changing the polynomial, we can get the integral term Ψ E (γ E , the closed-form approximation of δ) is: in and The mathematical expressions of are shown in formulas (16) and (17) respectively; finally, we get AESR in The closed-form approximation for this case is: Secure transmission based on artificial noise injection In this case, and Using the same theoretical derivation method as equations (18) and (19), the integral terms Φ0(γ0) and Ψ0(γ0,∈) in this case are simplified to: and Based on the method of integration by parts, the integral term Φ E (γ E ) is converted to: By introducing the Gaussian-Chebyshev integration method and changing the polynomial, we can finally get the integral term Ψ in this case: E (γ E , δ) is a closed-form approximation: Similarly, by using the same theoretical derivation method as formulas (21) and (22), we can obtain the integral term Ψ in this case: E (γ E , δ) is a closed-form approximation: Finally, we get AESR in The closed-form approximation for this case is:

7. The method for security enhancement and transmission optimization for URLLC services according to claim 1, wherein: Based on the secure transmission performance characterization framework, a security performance optimization problem is constructed. By jointly designing the channel estimation pilot length and artificial noise injection strategy, the security performance of URLLC is enhanced while ensuring the delay and reliability performance of the service. Specifically, by adjusting the channel estimation pilot length m t and artificial noise injection power ratio Perform joint optimization to maximize AESR while ensuring the latency, reliability, and security performance constraints of URLLC services. The optimization problem is formulated as follows: s.t.m t ={1,2,…,m-1}, (30a) The objective function The mathematical expressions in different cases are shown in formulas (23) and (29) respectively.

8. The method for security enhancement and transmission optimization for URLLC services according to claim 1, wherein: Security performance optimization algorithm for URLLC services: The two-dimensional search algorithm based on the golden section idea quickly converges to the approximate optimal solution of the optimization problem proposed in step 3. It is divided into the following two stages: First, this algorithm iteratively reduces the optimization variable m based on the golden section search method. t The search space of And optimize the variable m t The initial search space is denoted as [m t，L , m t，U ], where m t，L and m t，U The initial values ​​are m t，L =1 and m t，U =m-1; In each iteration, with a fixed golden ratio to [m t，L , m t，U ] is shrunk by setting the following two coefficients: and By using the one-dimensional search method, we can calculate the t =m1 and m t =m2, which is the ratio of artificial noise injection power that maximizes AESR. The following judgment conditions are used to determine m t,L and m t，U Iteratively update the value of: If The optimal solution is inferred The value is in the interval [m1,m t,U ]In this case, we make Then the search space [m t,L ,m t,U ]Updated to And update the values ​​of m1 and m2 to and if The optimal solution is inferred The value is in the interval [m t，L , m2]; in this case, let Then the search space [m t，L , m t，U ]Updated to And update the values ​​of m1 and m2 to and When m t,U -m t,L The iteration is completed when ≤0.5; and is the approximate optimal solution output by the algorithm, and the corresponding AESR optimal value is 9. The security enhancement and transmission optimization system for URLLC services is characterized by: include: A secure transmission model building module is used to build a URLLC service secure transmission model based on the URLLC service secure transmission system in the industrial Internet of Things scenario; The equivalent achievable secure transmission rate acquisition module is used to build a URLLC service secure transmission performance characterization framework based on short packet information theory to obtain the equivalent achievable secure transmission rate (AESR). The security performance optimization problem construction module is used to construct a security performance optimization problem based on the proposed security transmission performance characterization framework. By combining the channel estimation pilot length with the artificial noise injection strategy, the security performance of URLLC services can be enhanced while ensuring the delay and reliability performance of the services. The optimal solution acquisition module is used to quickly converge to the approximate optimal solution of the above optimization problem with low time complexity through a two-dimensional search algorithm based on the golden section idea; Constructing a URLLC service security transmission performance characterization framework: The privacy capacity that can achieve absolute security is defined as the upper limit of the number of data bits that a transmitter can send within a unit symbol duration while ensuring that the probability of decoding errors on the legitimate end and the probability of information interception on the eavesdropping end are arbitrarily low. The privacy capacity based on the assumption of asymptotically infinite code length is expressed as: C sec =C0-C E =log2(1+γ0)-log2(1+γ E ), (8) where γ0 and Represents the instantaneous received signal-to-noise ratio of the legitimate link and the eavesdropping link respectively; the upper limit of the private rate of the short packet communication service is within the traditional private capacity index C sec On the basis of , two penalty terms are added, and the mathematical expression is: in and represents the channel dispersion coefficient of the legitimate link and the eavesdropping link, ∈∈(0,1) represents the target decoding error probability of the legitimate receiving end, δ∈(0,1) represents the information interception probability of the eavesdropping end, Q -1 (z) represents the inverse function of the Gaussian Q-function, where the mathematical expression of the Gaussian Q-function is According to formula (9), as long as the current data transmission rate of URLLC signaling does not exceed its private rate upper limit R sec , then there must be a corresponding wiretap short packet channel coding structure that makes the transmission reliability limit ∈∈(0,1) and security limit δ∈(0,1) of URLLC signaling under a given transmission delay; The transmission performance upper limit of URLLC services with security requirements is evaluated based on the equivalent achievable secure transmission rate (AESR). Under the premise of given transmission reliability and security constraints, the mathematical expectation of the number of URLLC service bits that can be transmitted per unit symbol time is expressed as follows: The mathematical expressions of the integral terms Φ0(γ0) and Ψ0(γ0,∈) are: and in is the probability density function of the instantaneous signal-to-noise ratio γ0 of the legal link, and the mathematical expression of the random variable γ0 is shown in (5); the random variable γ0 obeys the shape parameter A T , the scale parameter is Gamma distribution, that is The PDF expression of the random variable γ0 is as follows: Integral term Φ E (γ E ) and Ψ E (γ E ,δ) are: and in is the instantaneous signal-to-noise ratio of the eavesdropping link γ E PDF of Before the mathematical expression of E The cumulative distribution function of is: in Then, by performing the derivation operation, we get The mathematical expression is: in According to the above expression, the above expression is simplified to calculate the AESR.

Images