An embedded CR-assisted NOMA resource scheduling method, device, system and medium

Abstract

The application discloses an embedded CR-assisted NOMA resource scheduling method, device, system and medium, and belongs to the technical field of wireless communication system spectrum resource scheduling, and comprises the following steps: a secondary user (SU) in a cognitive radio network senses a spectrum state of a primary user (PU), and initiates a spectrum access request to a fusion center (FC); the FC obtains all spectrum hole sets at the current moment, and determines an optimal spectrum hole access strategy and feeds back to each requesting SU according to the maximum system and rate criterion after each requesting SU accesses; each requesting SU accesses the spectrum hole in an embedded mode according to the feedback. Compared with the current two CR modes, the application proposes an embedded CR mode, which allows SUs to access the spectrum holes released by PUs to access the spectrum. This method can avoid the shortcomings brought by underlay and overlay modes, can significantly improve the overall spectrum utilization of the system, and has the advantages of small calculation complexity.

Description

Technical Field

[0001] This invention belongs to the field of spectrum resource scheduling technology for wireless communication systems, and more specifically, relates to an embedded CR-assisted NOMA resource scheduling method, device, system, and medium. Background Technology

[0002] The Internet of Things (IoT) is a key technology enabler for smart cities, where intelligent objects communicate with each other. With the increase in connected devices, spectrum shortages become increasingly prominent. Cognitive radio (CR) networks have been proposed as a promising solution to improve spectrum resources for IoT communications. In CR networks, high-priority primary users (PUs) and low-priority secondary users (SUs) coexist in two main modes: underlay and overlay. In underlay mode, SUs are allowed to transmit without interfering with PUs. That is, SUs can share the same spectrum as PUs, but they must ensure that the interference they cause to PUs is limited to acceptable levels. In overlay mode, SUs can only transmit when the licensed spectrum is not occupied by any PU. That is, SUs must wait until PUs are not using the spectrum before using it to ensure that they do not interfere with PU communications.

[0003] These two paradigms of CR are widely used in existing spectrum resource scheduling schemes. The underlay paradigm introduces additional interference to the primary user, making it difficult to guarantee the primary user's own QoS requirements; while the overlay paradigm cannot share spectrum with the PU and can only use completely idle frequency bands, resulting in low spectrum utilization.

[0004] Furthermore, studying resource scheduling under shared spectrum conditions involving multiple PUs and multiple SUs within the same spectrum is quite complex. Therefore, existing NOMA resource scheduling studies assisted by the CR paradigm all assume that there is at most one PU in the same frequency band, considering the access performance of SU users. However, while this model is indeed applicable to some user scenarios, it cannot cover the vast majority of cases where the number of active PUs in the same frequency band is greater than two. Summary of the Invention

[0005] To address the shortcomings and improvement needs of existing technologies, this invention provides an embedded CR-assisted NOMA resource scheduling method, device, system, and medium. It aims to solve the technical problems existing in the prior art, such as additional interference to PUs, inability to guarantee their own QoS, low overall system spectrum utilization, high probability of SUs user interruption, single research model, low application value due to ideal CSI assumptions, and high computational complexity.

[0006] To achieve the above objectives, in a first aspect, the present invention provides an embedded CR-assisted NOMA resource scheduling method, comprising:

[0007] In a cognitive radio network, a secondary user (SU) senses the spectrum status of a primary user (PU) and initiates a spectrum access request to the convergence center (FC). The licensed spectrum is divided into multiple orthogonal sub-bands, and PUs share the spectrum in a NOMA (Normally Orthogonal Array) configuration within a single sub-band.

[0008] The FC obtains the set of all spectrum holes at the current time, and determines the optimal spectrum hole access strategy based on the criterion of maximizing the system and rate after each requesting SUs accesses the system, and feeds it back to each requesting SUs.

[0009] Each requesting SUs accesses the spectrum holes in an embedded manner based on the feedback.

[0010] Furthermore, the system and rate R are expressed as follows:

[0011]

[0012] Where C represents the number of sub-bands, N c This represents the total number of users in the c-th sub-band. This indicates the nth frequency band under the c-th sub-band. c individual users The signal-to-interference-to-noise ratio.

[0013] Furthermore, the system and rate R are expressed as follows:

[0014]

[0015]

[0016] Where C represents the number of sub-bands, N c R represents the total number of users in the c-th sub-band. c,k ρ represents the sum rate of the k-th user in the c-th sub-band; k [to] represents the time correlation coefficient of the k-th user between time slot t and time slot o, where o represents the initial time. ε c,k This represents the k-th user T in the c-th sub-band. c,k The transmission power, H c,k [o] represents the channel gain coefficient of the k-th user in the c-th sub-band at time o. β j N represents the large-scale fading coefficient, and N0 represents the noise variance.

[0017] Furthermore, the k-th SU user in the c-th sub-band The SINR must meet the following requirements:

[0018]

[0019] in, This represents the k-th SU user in the c-th sub-band. Transmission power, Φ represents the channel gain coefficient of the k-th SU user in the c-th sub-band. c Let c represent the set of users in the c-th sub-band. This represents the j-th user T in the c-th sub-band. c,j The channel gain coefficient, N0 represents the noise variance, τ su This represents the SINR threshold value of SU.

[0020] Secondly, the present invention provides an embedded CR-assisted NOMA resource scheduling device, comprising:

[0021] The request module is used by secondary users (SUs) in the cognitive radio network to perceive the spectrum status of primary users (PUs) and initiate spectrum access requests to the convergence center (FC). The licensed spectrum is divided into multiple orthogonal sub-bands, and PUs share the spectrum in the form of NOMA within a single sub-band.

[0022] The processing module is used by FC to obtain the set of all spectrum holes at the current time, and determine the optimal spectrum hole access strategy based on the criterion of maximizing the system and rate after each requesting SUs accesses the system, and feed it back to each requesting SUs.

[0023] The scheduling module is used for each requesting SUs to access the spectrum holes in an embedded manner based on the feedback.

[0024] Thirdly, the present invention provides an embedded CR-assisted NOMA resource scheduling system, comprising: a computer-readable storage medium and a processor;

[0025] The computer-readable storage medium is used to store executable instructions;

[0026] The processor is used to read executable instructions stored in the computer-readable storage medium and execute the embedded CR-assisted NOMA resource scheduling method as described in the first aspect.

[0027] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the embedded CR-assisted NOMA resource scheduling method as described in the first aspect.

[0028] In summary, the above-described technical solutions conceived in this invention can achieve the following beneficial effects:

[0029] 1. Compared with other CR paradigms, the design method of this invention has the advantages of not generating additional interference to existing PUs, ensuring QoS for both parties, making full use of idle spectrum resources, and significantly improving system performance and speed.

[0030] 2. Compared with other CR paradigm-assisted NOMA resource scheduling methods, which assume that there is at most one PU in the frequency band, the design method of this invention is more universal, as it does not limit the number of PUs in the same frequency band and has greater practical application value.

[0031] 3. In existing research on the Internet of Things (IoT) or the Internet of Vehicles (IoV), for the sake of research convenience, most studies assume that users are in a stationary state, that is, that CSI is time-invariant. However, in actual application scenarios, due to the dynamic characteristics of the channel and the speed of moving objects, CSI is time-varying. Therefore, the design method of this invention considers the channel aging characteristics, studies a real-time channel prediction method, and gives a closed expression for the channel CSI, making the research on resource scheduling algorithms more practically valuable.

[0032] 4. Unlike existing resource scheduling schemes that mostly adopt offline methods, the present invention provides an online learning-based spectrum resource scheduler. Compared with other offline, AI-based scheduling algorithms, the present invention has the advantages of lower computational complexity, real-time adaptability to changing environments, and online learning to adjust the allocation scheme. Attached Figure Description

[0033] Figure 1 A flowchart illustrating an embedded CR-assisted NOMA resource scheduling method provided in an embodiment of the present invention;

[0034] Figure 2(a) is a schematic diagram of resource scheduling for the CR underlay paradigm and the CR overlay paradigm provided in the embodiments of the present invention;

[0035] Figure 2(b) is a schematic diagram of resource scheduling for embedded CR-assisted NOMA provided in an embodiment of the present invention;

[0036] Figure 3 This is a schematic diagram of the uplink communication model of an Internet of Things (IoT) system (taking vehicle-to-everything (V2X) as an example) provided in an embodiment of the present invention.

[0037] Figure 4 The graphs showing the system and rate variations with user speed under different methods provided in the embodiments of the present invention;

[0038] Figure 5 This is a process diagram showing the increase of system and rate with the number of iterations under different methods provided in the embodiments of the present invention;

[0039] Figure 6A graph showing the variation of the average instantaneous sum and rate values ​​with time index t under different methods provided in the embodiments of the present invention;

[0040] Figure 7 The graph shows the system and rate variations with user transmission power under different methods and speeds provided in the embodiments of the present invention. Detailed Implementation

[0041] To make the objectives, system composition, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0042] In this invention, the terms "first," "second," etc. (if present) in the invention and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0043] See Figure 1 , combined Figures 2(a) to 7 This invention provides an embedded CR-assisted NOMA resource scheduling method, including operations S1 to S3.

[0044] In operation S1, the secondary user (SU) in the cognitive radio network senses the spectrum status of the primary user (PU) and initiates a spectrum access request to the fusion center (FC); the licensed spectrum is divided into multiple orthogonal sub-bands, and PUs share the spectrum in the form of NOMA in a single sub-band.

[0045] In this embodiment, the SU initiates a perception request and simultaneously sends its own requirement data packet (including its own CSI and its own QoS) to the fusion center FC.

[0046] Since the PU (Power Utility) belongs to the licensed spectrum, it exists in two states: busy and idle. When the PU is not using its own spectrum resources, the PU is in an idle state, and its spectrum resources can be reused by the SU (Subscriber Unit) requesting access in the cognitive radio network.

[0047] The licensed spectrum is divided into multiple orthogonal sub-bands, which exist in an orthogonal relationship and do not interfere with each other; and multiple PUs within a sub-band reuse the spectrum through power domain NOMA. When there are PU spectrum holes, the SU access spectrum holes share the spectrum with other PUs in the same frequency band through NOMA.

[0048] Operation S2: FC obtains the set of all spectrum holes at the current time, and determines the optimal spectrum hole access strategy based on the criterion of maximizing the system and rate after each requesting SUs accesses the system, and feeds it back to each requesting SUs.

[0049] In this embodiment, the FC actively acquires the set of all spectrum holes at the current time (including: idle PU spectrum, CSI of other busy PUs in the same sub-band).

[0050] First, considering the dynamic channel, this invention, for the first time, applies a channel aging model to a CR paradigm-assisted NOMA system for resource scheduling algorithms. Based on the channel aging model, real-time channel prediction is studied, and a closed-form expression for the real-time predicted CSI is given, effectively compensating for the performance loss caused by channel aging.

[0051] The expression for real-time CSI channel prediction is as follows:

[0052]

[0053] Among them, H k,i [t+τ] represents the channel gain coefficient (CSI) from the k-th user to the i-th antenna at time t+τ; H k,i [t]~CN(0,β k,i ), β k,i Represents the large-scale fading coefficient; Ψ k,i [t+τ]~CN(0,β k,i ) represents the gain function representing the evolution of the channel coefficients at time t+τ after a time interval τ; ρ k [t+τ] represents the time correlation coefficient of the k-th user between time slot t and time slot t+τ. Assuming the channel evolves according to Jakes' model, then ρ k [t+τ]=J0(2πf D,k T s (t+τ)), where J0(·) is a Bessel function of the first kind of order 0, and T s Indicates the sampling time. This means that when the velocity is v k The Doppler frequency shift of the k-th user, f c and These represent the carrier frequency and the speed of light, respectively.

[0054] The resource scheduling algorithm criterion is to maximize both system performance and speed. The system performance and speed expressions are as follows:

[0055]

[0056] Where C represents the number of sub-bands, N c This represents the total number of users in the c-th sub-band. This indicates the nth frequency band under the c-th sub-band. c individual users The signal-to-interference-to-noise ratio.

[0057] Secondly, this invention proposes a low-complexity dynamic spectrum scheduler based on online learning. According to the dynamically changing CSI, our proposed scheduler adjusts the spectrum allocation scheme in real time during iterative learning to achieve optimal performance with lower complexity.

[0058] In the process of dynamic learning for optimal spectrum resource scheduling, the real-time changing characteristics of the channel are taken into account.

[0059] During uplink data transmission, the signal y received by the i-th antenna at the BS end in sub-band c at time t is... c,i [t] is as follows:

[0060]

[0061]

[0062] Among them, S c,k,i [t]~CN(0,1) represents the signal transmitted by the k-th user to the i-th antenna of the base station in sub-band c at time t; θ k,i [t]~CN(0,β k,i ) represents the channel gain coefficient from the k-th user to the i-th antenna at time t.

[0063] Substituting (4) into (3), the y received by the base station under sub-frequency band c c,i [t] is:

[0064]

[0065] The first term is the desired received signal, I. cki,n,1 For interference caused by channel estimation errors, I cki,n,2 Interference caused by channel timeout, I cki,n,3 Interference caused by other users in sub-band c of the communication. This is the estimated channel gain at time 0. The estimation error at time 0 is denoted as .

[0066]

[0067] Therefore, the sum rate of the k-th user in sub-band c at time t can be written as:

[0068]

[0069] In this invention, the performance impact caused by channel timeout is mainly considered. Therefore, there is no interference caused by channel estimation errors. Therefore, there is I cki,n,1 =0.

[0070] Continuing to derive equation (7), the sum rate expression of the system at time t is:

[0071]

[0072] Therefore, the sum rate of user k under sub-band c is expressed as follows:

[0073]

[0074] The system and rate R are expressed as follows:

[0075]

[0076] In operation S3, each requesting SUs communicates using the indicated hole spectrum based on the feedback.

[0077] Specific examples:

[0078] Figure 3 The diagram illustrates the uplink communication model of a CR-assisted NOMA vehicle-to-everything (V2X) network according to an embodiment of the present invention. There are M primary users (PUs), and spectrum resources are allocated to each PU according to the power domain NOMA criterion. For ease of study, the channel gains of the PUs are arranged in ascending order. Each PU has two states: active and idle. A fusion center (FC) is present in the diagram, and its main function is resource scheduling. Assume that at a certain moment, there are N sub-users (SUs), and the FC can obtain the CSI of the PUs and SUs.

[0079] Unlike traditional SU access methods based on the underlay paradigm, the proposed method in this invention directly embeds the SU into the spectral holes of idle PUs. Therefore, it can be assumed that before and after SU access, the interference generated by the accessing SU on other active PUs in the same frequency band is equal to or less than the interference generated by the idle PU in its active state. This assumption is considered acceptable. Therefore, in this design, only the QoS of the accessing SU itself needs to be considered, denoted by SINR, i.e., the SINR of the SU needs to satisfy a threshold value τ. su .

[0080]

[0081] in, This represents the k-th SU user in the c-th sub-band. Transmission power, Φ represents the channel gain coefficient of the k-th SU user in the c-th sub-band. c Let c represent the set of users in the c-th sub-band. This represents the j-th user T in the c-th sub-band. c,jThe channel gain coefficient, N0 represents the noise variance, τ su This represents the SINR threshold value of SU.

[0082] Based on the above analysis, the expression for the system optimization problem can be derived as follows: Where B = {B0, B1, ..., B} Q-1} represents a set of scheduling policies for SUs access. This represents the set of sub-frequency bands involving SU access under the q-th access strategy. This represents the m-th sub-band involving SU access under the q-th access strategy; This indicates that under the q-th access strategy, a total of C are involved. q Each sub-frequency band has SU access, f q,m This represents the m-th sub-band accessed under the q-th access strategy. and They represent the cth time. q Access under each sub-band Each SUs and existence One active PU. Additionally, Indicates the cth time q The set of active PUs and access SUs in each sub-band.

[0083] Figure 4 The relationship between system speed and vehicle mobility is shown. The performance of the resource allocation scheme proposed in this invention is compared with that of traditional NOMA and random allocation schemes. When the speed is set to 20 km / h, the resource allocation scheme proposed in this invention increases the speed by 21% compared with the random allocation scheme, and achieves a speed gain of 331% compared with the traditional NOMA method. It can also be seen that the speed decreases rapidly with increasing vehicle speed, which is due to the channel aging problem, which is exacerbated by increasing vehicle speed.

[0084] Figure 5 The system and rate are shown under time-varying channel conditions with a Doppler frequency shift of 0.0005, equivalent to a speed of approximately 54 km / h. Iterative indices represent the online learning process. It can be seen that continuous learning significantly improves the system and rate, and optimal performance is rapidly achieved within 40 iterations. Furthermore, it can be observed that the system and rate obtained by the proposed scheme are significantly higher than those of the traditional NOMA method.

[0085] Figure 6 This shows the average uplink instantaneous and rate variations over the first 400 slots within an infinitely long resource block (averaging PUs and SUs across all sub-bands). From Figure 6As can be seen, the superiority of the proposed scheme is evident in the first 300 time slots at a speed of 54 km / h. However, when the speed increases to 108 km / h, the superiority of the spectrum allocation is no longer sufficient to offset the performance degradation caused by increased speed and channel aging, thus the two curves eventually overlap. Therefore, it is necessary to design an appropriate resource block length to reduce the impact of channel aging.

[0086] Figure 7 The graph shows the system's sum and rate as a function of user transmission power. It can be seen that, at any transmission power, the sum and rate of the spectrum hole allocation scheme proposed in this invention are significantly better than the random allocation scheme. It can also be seen that increasing transmission power can increase the total capacity. However, there is a capacity ceiling in the high transmission power region (>45dBm). This is because multi-user interference (MUI) increases with increasing transmission power, thus limiting system performance. Therefore, effective MUI elimination is needed to further improve capacity.

[0087] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An embedded CR-assisted NOMA resource scheduling method, characterized in that, include: In a cognitive radio network, a secondary user (SU) senses the spectrum status of a primary user (PU) and initiates a spectrum access request to the convergence center (FC). The licensed spectrum is divided into multiple orthogonal sub-bands, and PUs share the spectrum in a NOMA (Normally Orthogonal Array) configuration within a single sub-band. The FC obtains the set of all spectrum holes at the current moment, and determines the optimal spectrum hole access strategy based on maximizing the system and rate after each requesting SUs accesses the system, and feeds it back to each requesting SUs; c Sub-band k SU users The SINR must meet the following requirements: in, Indicates the first c Sub-band k SU users Transmission power, Indicates the first c Sub-band k Channel gain coefficients for each SU user Indicates the first c User set under each sub-band Indicates the first c Sub-band j individual users The channel gain coefficient, Indicates the noise variance. This represents the SINR threshold value of SU; Each requesting SUs accesses the spectrum holes in an embedded manner based on the feedback.

2. The embedded CR-assisted NOMA resource scheduling method according to claim 1, characterized in that, System and rate R The expression is as follows: in, C Indicates the number of sub-bands. N c Indicates the first c Total number of users per sub-band Indicates the first c Sub-band n c individual users The signal-to-interference-to-noise ratio.

3. The embedded CR-assisted NOMA resource scheduling method according to claim 1, characterized in that, System and rate R The expression is as follows: in, C Indicates the number of sub-bands. N c Indicates the first c Total number of users per sub-band Indicates the first c Sub-band k The sum and rate of each user; Indicates the first k Individual users in time slots t and time slot o The time correlation coefficient between them o Indicates the initial time. Indicates the first c Sub-band k individual users Transmission power, Indicates the first c Sub-band k Individual users o Channel gain coefficient at time 10:00 , Represents the large-scale fading coefficient. This represents the noise variance.

4. An embedded CR-assisted NOMA resource scheduling device, characterized in that, The method for executing the embedded CR-assisted NOMA resource scheduling method according to any one of claims 1-3 includes: The request module is used to allow secondary users (SUs) in a cognitive radio network to perceive the spectrum status of primary users (PUs) and initiate spectrum access requests to the convergence center (FC). The licensed spectrum is divided into multiple orthogonal sub-bands, and PUs share the spectrum in a NOMA (Normally Orthogonal Array) configuration within a single sub-band. The processing module is used by FC to obtain the set of all spectrum holes at the current time, and determine the optimal spectrum hole access strategy based on the criterion of maximizing the system and rate after each requesting SUs accesses the system, and feed it back to each requesting SUs. The scheduling module is used for each requesting SUs to access the spectrum holes in an embedded manner based on the feedback.

5. An embedded CR-assisted NOMA resource scheduling system, characterized in that, include: Computer-readable storage media and processors; The computer-readable storage medium is used to store executable instructions; The processor is used to read executable instructions stored in the computer-readable storage medium and execute the embedded CR-assisted NOMA resource scheduling method as described in any one of claims 1-3.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the embedded CR-assisted NOMA resource scheduling method as described in any one of claims 1-3.

Images