A ris-assisted rsma ultra-dense network and coverage performance analysis method

Abstract

The application discloses a RIS-assisted RSMA super-dense network and a coverage performance analysis method. The RIS-assisted RSMA super-dense network comprises: a plurality of micro base stations arranged in a hotspot area with dense users; the spatial positions of the micro base stations are modeled by using a Poisson cluster process, and the micro base stations are independently and identically distributed around a cluster center; a plurality of macro base stations, the spatial positions of which are modeled by using a Poisson point process, and the spatial positions of the macro base stations are independent and uniformly distributed; wherein each micro base station and macro base station is respectively equipped with RIS. The RIS-assisted RSMA super-dense network provided by the application can improve the coverage rate of the system and the regional spectrum efficiency of the system.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of wireless communication, and particularly relates to an RIS-aided RSMA ultra-dense network and a coverage performance analysis method. BACKGROUND

[0002] With the rapid development of mobile communication technology, users' demand for high-speed and low-latency data transmission is increasing. Especially in densely populated urban areas, traditional cellular networks have been difficult to meet the growing traffic demand. Therefore, ultra-dense networks (UDN) have emerged as an effective solution. By increasing the number and density of base stations, UDN can significantly improve network capacity and coverage quality, thereby better meeting user demand. Although UDN can significantly improve network transmission performance, its dense deployment also brings new challenges.

[0003] The main problems existing in the current ultra-dense network include:

[0004] Increased interference leads to decreased spectrum reuse efficiency:

[0005] The dense deployment of base stations causes adjacent base stations to frequently share spectrum resources, resulting in serious co-channel interference and reducing the effectiveness of spectrum reuse. Spectrum reuse and interference management become increasingly complex.

[0006] Traditional spectrum allocation and scheduling mechanisms fail:

[0007] Traditional spectrum utilization methods often fail to achieve efficient data transmission in high-density environments, resulting in uneven resource allocation or spectrum waste. Not only does this limit the overall performance of the network, but it can also lead to a decline in user experience.

[0008] There are signal blind areas:

[0009] Although the base stations are denser, the actual signal propagation is blocked by obstacles, which can form coverage blind areas in some areas.

[0010] Hotspot area load imbalance:

[0011] Although there are multiple base stations in some user-intensive areas, due to unreasonable deployment or improper scheduling, there may still be coverage deficiencies, and problems such as poor communication quality for users at the edge of the cell may occur.

[0012] Orthogonal Multiple Access (OMA) and Non-Orthogonal Multiple Access (NOMA) are commonly used multi-user access technologies.

[0013] Problems with Orthogonal Multiple Access (OMA) include:

[0014] Low spectrum efficiency: only orthogonal division (TDMA / FDMA) can be used to use resources; cannot take advantage of the sharing gain brought by non-orthogonal access. Limited system capacity: the number of users is limited by the granularity of physical resources; when the number of users is large, the scheduling resources are insufficient, and the system throughput is reduced. Lack of interference coordination capability: resources must be "hardly separated"; cannot flexibly schedule in high interference scenarios. Weak support for heterogeneous services: it is difficult to meet the multi-class QoS requirements of high / low speed, low latency / high capacity, etc.; service strategy is single and lacks flexibility. Lack of universality and evolution: cannot be used as a unified access architecture compatible with other solutions; not suitable for natural evolution to future 6G requirements.

[0015] Problems of non-orthogonal multiple access (NOMA) include:

[0016] Weak interference management capability: dependent on fixed power ordering; cannot effectively handle non-ideal channels or asymmetric scenarios. Poor robustness: strong dependence on CSI estimation; sensitive to user location ordering. Strong architecture limitations: does not support information splitting; cannot unify multiple access modes. Poor scalability: as the number of users increases, power allocation and SIC costs increase dramatically; not suitable for large-scale MIMO systems. High pressure on the receiving end: heavy computational burden for multi-stage SIC; weak fault tolerance, error rate easily affected. Limited QoS support: does not support dynamic service types; difficult to achieve fine-grained resource control. SUMMARY

[0017] The purpose of the present application is to overcome the defects of the prior art and provide an RIS-assisted RSMA ultra-dense network which can improve the coverage of the system and the area spectrum efficiency of the system.

[0018] The present application also provides a coverage performance analysis method of an RIS-assisted RSMA ultra-dense network, which can realize simplified solving of the coverage and area spectrum efficiency of the RIS-assisted RSMA ultra-dense network, and improve the accuracy of the solving result.

[0019] The technical scheme provided by the present application is:

[0020] An RIS-assisted RSMA ultra-dense network comprises:

[0021] A plurality of micro base stations are deployed in a user-intensive hotspot area; the spatial positions of the micro base stations are modeled by a Poisson cluster process, and the micro base stations are independently and identically distributed around the cluster center;

[0022] A plurality of macro base stations are modeled by a Poisson point process, and the spatial positions of the macro base stations are independently and uniformly distributed;

[0023] Each of the micro base stations and the macro base stations is respectively equipped with an RIS.

[0024] Preferably, the phase offset of the reflecting unit of the RIS is set as:

[0025]

[0026] wherein φ n denotes the phase offset of the nth reflecting unit of the RIS, denotes the channel phase from the serving base station to the nth reflecting unit of the RIS, denotes the channel phase from the nth reflecting unit of the RIS to the user, denotes the phase offset caused by the serving base station to user link.

[0027] A method for analyzing the coverage performance of an RIS-assisted RSMA hyperdense network, for analyzing the performance of the RIS-assisted RSMA hyperdense network, comprising:

[0028] constructing a calculation model of the useful signal power distribution and a calculation model of the interference signal power;

[0029] wherein the calculation model of the useful signal power distribution is in the form of a gamma random variable, and the calculation model of the interference signal power is in the form of a Laplace transform expression;

[0030] obtaining the conditional coverage rate of the RIS-assisted RSMA hyperdense network according to the calculation model of the useful signal power distribution and the calculation model of the interference signal power;

[0031] deconditioning the conditional coverage rate to obtain the coverage rate of the RIS-assisted RSMA hyperdense network;

[0032] determining the ergodic rate of the user at the macro base station layer and the micro base station layer according to the conditional coverage rate of the RIS-assisted RSMA hyperdense network, and obtaining the area spectral efficiency of the RIS-assisted RSMA hyperdense network according to the ergodic rate;

[0033] wherein the ergodic rate includes a public ergodic rate and a private ergodic rate.

[0034] Preferably, the calculation model of the useful signal power distribution is:

[0035]

[0036] wherein κ s and θ s respectively denote the shape parameter and the scale parameter of the gamma distribution of the equivalent approximation of the useful signal power, E[S] and E 2 [S] respectively denote the first moment and the second moment of the gamma distribution of the equivalent approximation of the useful signal power, and S denotes the gamma distribution of the equivalent approximation of the useful signal power.

[0037] Preferably, the interference signal power calculation model is:

[0038]

[0039] wherein, s represents the ratio of the threshold to the scale parameter and the power; represents the ratio of the interference power to the service power; τ i represents the SIR threshold value of the typical user when associated with the i-th layer base station for successfully demodulating and decoding the received signal; θ s represents the scale parameter of the gamma distribution of the equivalent approximation of the useful signal power; P i represents the service base station transmit power; r represents the service base station to typical user distance; Φ j represents the base station location set of the layer where the interference base station is located; N represents the number of reflection units of each RIS, represents the intercept factor, f c is the carrier frequency, v c is the speed of light; α represents the path loss coefficient; represents the set of all base stations; η x,r represents the large-scale fading of the base station to the RIS in the cascaded link of the RIS-assisted base station and the typical user; η x represents the large-scale fading of the direct link between the base station and the typical user; represents the average value of the sub-point number of the cluster of the micro base station layer interference base station; P j represents the interference BS transmit power; represents the transmit power of the micro base station layer interference base station; represents the transmit power of the macro base station layer interference base station; represents the base station density of the macro base station layer interference base station; represents the conditional probability density function of the distance y of the interference base station to the typical user.

[0040] Preferably, the RIS-assisted RSMA ultra-dense network coverage is:

[0041]

[0042] wherein, represents the coverage when the typical user subject to the PPP distribution is associated with the i-th layer base station in the RIS-RSMA system, represents the coverage when the typical user subject to the PCP distribution is associated with the i-th layer base station in the RIS-RSMA system, SIR i represents the signal-to-interference ratio of the typical user when associated with the i-th layer base station, τ represents the SIR threshold value of the receiving end for successfully demodulating and decoding the received signal, respectively represent the probability that a typical user belongs to represent macro base station users, represent micro base station users; represent the set of all base stations.

[0043] Preferably, the area spectral efficiency of the RIS-assisted RSMA ultra-dense network is:

[0044]

[0045] wherein, respectively represent the ergodic rate of users following a Poisson point process distribution and users following a Poisson cluster process at the i-th tier base station; representing the micro base station tier, representing the macro base station tier, respectively represent the probability that a typical user belongs to represent macro base station users, represent micro base station users; λ i represent the density of the i-th tier base station, represent the set of all base stations.

[0046] Preferably, based on the two-user model of RSMA, the ergodic rate of users at each base station tier is calculated:

[0047]

[0048] wherein, represent the ergodic rate of users at the i-th tier base station; represent the rate component of the near-zone common flow, respectively represent the rate component of the near-zone typical user as a near user, the fixed user as a far user private flow, represent the rate component of the far-zone common flow, respectively represent the rate component of the far-zone typical user as a far user, the fixed user as a near user private flow.

[0049] The beneficial effects of the present application are:

[0050] The RIS-assisted RSMA ultra-dense network provided by the present application can improve the coverage of the system and the area spectral efficiency of the system.

[0051] The coverage performance analysis method of the RIS-assisted RSMA ultra-dense network provided by the present application can realize simplified solving of the coverage rate and the area spectral efficiency of the RIS-assisted RSMA ultra-dense network, and improve the accuracy of the solving result, so as to quickly obtain the coverage performance of the RIS-assisted RSMA ultra-dense network. ​​BRIEF DESCRIPTION OF DRAWINGS

[0052] Figure 1 The downlink system model diagram of the RIS-assisted RSMA hyper dense network described in the present application.

[0053] Figure 2 The relationship between the coverage of different network systems and the transmission power P2 of MBS described in the present application.

[0054] Figure 3 The relationship between the ASE of different network systems and the transmission power P2 of MBS described in the present application. DETAILED DESCRIPTION

[0055] The present application will be further described in detail below with reference to the accompanying drawings, so that those skilled in the art can implement the present application according to the description and drawings.

[0056] As shown in Figure 1 , the present application provides a RIS-assisted RSMA hyper dense network. The RIS-assisted RSMA hyper dense network specifically includes: a macro base station (MBS), a small base station (SBS), a user, and a reconfigurable intelligent surface (RIS).

[0057] The construction method of the RIS-assisted RSMA hyper dense network is: based on the random geometry theory, a hybrid spatial distribution model is constructed. The small base station (SBS) deployed in the hotspot area with high user density is modeled by Poisson cluster process (PCP), and the small base stations are independent and identically distributed around the parent point (cluster center). The Poisson cluster process is divided into TCP and MCP. Among them, the TCP parent point obeys the homogeneous Poisson point process (HPPP) with intensity λ p , and the child points are point processes of two-dimensional independent Gaussian distribution with variance σ 2 around the parent point. The MCP parent point obeys the HPPP with intensity λ, but the child points are uniformly distributed in a disc area with the parent point as the center and the radius R c . Users and base stations often appear in the same hotspot area, in order to reflect the spatial correlation, the users of the micro base station layer (users in the hotspot area) Both macro base stations (MBS) and macro base stations (MBS) utilize the Poisson Cluster Process (PCP) for modeling, with both base stations and users sharing the same parent PPP. For larger coverage areas, the Poisson Point Process (PPP) is used, where the spatial locations of macro base stations are independent and uniformly distributed. To accurately reflect the random arrival and dynamic mobility characteristics of users, the macro base station layer focuses on users (mobile users). Homogeneous Poisson modeling is employed. All base stations and users are equipped with a single antenna, and reconfigurable smart surfaces are deployed along the propagation path between the base station and user equipment to optimize the wireless channel environment and improve system transmission performance. Furthermore, the distance from the RIS to the base station is significantly shorter than the distance from the RIS to the user. Typical users select their serving base station based on a maximum average received power correlation strategy. RSMA technology is introduced to achieve proactive management of inter-user interference and spectrum resource allocation.

[0058] This invention also provides a method for analyzing the coverage performance of RIS-assisted RSMA ultra-dense networks, the specific process of which is as follows.

[0059] S1 system modeling

[0060] S1.1 Downlink System Model

[0061] This invention takes into account the dense base station deployment of UDN. Φ1 represents the SBS layer following a PCP distribution, reflecting its dense deployment in hotspot areas; Φ2 represents the MBS layer following a PPP distribution, reflecting its distribution over a wide area. To distinguish different types of BSs, an index set is introduced. in Indicates the SBS layer. This represents the MBS layer. In user distribution modeling, a homogeneous Poisson point characterization process is used for mobile users. Spatial distribution of users (at the macro base station layer); PCP process is used to analyze users (at the micro base station layer) within hotspot areas. Modeling is performed, and Φ1 and They share the same parent PPP. The RIS-assisted UDN downlink system model is as follows: Figure 1 As shown.

[0062] All base stations and users are equipped with a single antenna, and reconfigurable smart surfaces are deployed along the propagation path between the base station and user equipment to optimize the wireless channel environment and improve system transmission performance. Since random geometry is used for modeling, the location of a typical user does not affect the distribution characteristics of other users; therefore, it is assumed that the typical user is located at the origin. The serving base station is defined as the base station closest to the typical user, and a fixed distance exists between the serving base station and its nearest RIS. Where ||x c|| represents the distance from the serving base station to a typical user, while the distance from the RIS to a typical user is For the convenience of subsequent calculation, it is defined that the distance from the RIS to the serving base station satisfies Therefore, the distance from the RIS near the serving base station to a typical user is According to the user association to the maximum average received power strategy, the RIS position near the serving base station is defined as where P k represents the transmission power of the k-th layer serving base station, and a is the path loss coefficient. And When represents the distance from the serving BS to a typical user, and when the typical user is associated with the i-th layer BS, the signal SIR of the typical user can be expressed as:

[0063]

[0064] where P i is the transmission power of the serving BS, P j is the transmission power of the interfering BS. g x is the small-scale fading gain of the interfering base station. is the small-scale fading gain of the serving base station. i represents the serving base station, and j represents the interfering base station.

[0065] S1.2 Transmission model

[0066] In order to study the performance of the UDN using RSMA technology in the framework of stochastic geometry, a typical two-user system is introduced in this paper. Considering the complexity of successive interference cancellation receivers, the application of two-user RSMA systems is usually more feasible in practical environments. The two-user system is used to divide the users in the coverage area of the serving BS into two regions, instead of two types of user point processes in u Φ . Regarding each serving BS, without loss of generality, we consider that each serving BS is associated with a user in the last round of user association process, which is called the fixed user. For simplicity, we assume that the distance between the fixed user and the connected serving BS is the same, denoted as the fixed distance r i (i = {1, 2}). According to the fixed distance r i , the coverage area of the serving BS is divided into the near region and the far region. The users in the near region are defined as "near users", and the users in the far region are defined as "far users". Since the distance ||x * || between the typical user and the serving BS is a random value, it is not possible to determine in advance whether the typical user is a near user or a far user. It is assumed that when the distance ||x * ‖ between the typical user and the serving BS is less than or equal to the fixed distance r iWhen the distance ||x * ||r i , the typical user is a far user and the fixed user is a near user.

[0067] S1.3 Channel Model

[0068] In the present application, a configuration scheme of passive RIS is adopted, that is, the RIS does not have the active signal transmission capability and can only realize signal transmission by reflecting the electromagnetic wave transmitted by the base station. The RIS-base station independent association mode is adopted, which is mainly due to the coupling relationship between the modeling of SBS and the distribution of users in the hot spot area. A single RIS can meet the coverage requirement, and the performance improvement brought by multi-RIS cooperation is limited. Based on this system architecture, the downlink communication link between the typical user and the serving base station can be divided into two parts: one is the direct link, and the other is the cascade link associated with the RIS.

[0069] Assuming that each RIS is a passive reflection element and has N reflection units, therefore, when considering large-scale fading, only the first reflection signal of the RIS is considered. In the cascade link of the RIS-assisted base station and the typical user, the large-scale fading from the base station to the RIS and from the RIS to the typical user is respectively:

[0070] η x,r =C||x-x r || -α

[0071] η r =C||x r || -α

[0072] wherein, is the intercept factor, f c is the carrier frequency, v c is the speed of light, and a is the path loss coefficient.

[0073] For small-scale fading, it is assumed that the channels between all antennas are modeled as i.i.d. and the channel variation is quasi-static. The baseband equivalent channels of the k-th layer base station to the RIS and the RIS to the typical user are respectively represented as:

[0074]

[0075] The reflection matrix coefficient of the RIS is set to to intelligently adjust the phase, wherein φ N ∈[0,2π) represents the phase shift coefficient, and it is assumed that each reflection unit can reach the maximum amplitude coefficient 1.

[0076] The received signal of the typical user can be represented as:

[0077]

[0078] where, denotes the large-scale fading from the serving base station to the typical user, denotes the large-scale fading from the serving base station to the RIS of the assisting serving base station, denotes the large-scale fading from the RIS of the assisting serving base station to the typical user, g0denotes the small-scale fading coefficient from the base station to the user, denotes the baseband equivalent channel from the serving base station to the RIS, denotes the baseband equivalent channel from the serving RIS to the user, denotes the baseband equivalent channel from the interfering base station to the RIS, denotes the baseband equivalent channel from the interfering RIS to the user, P tx denotes the base station transmit power, s i and s j denote the unit-power transmit symbols of the serving base station and the interfering base station, respectively, n0denotes the additive white noise.

[0079] S1.4useful signal power

[0080] Although it is possible that the RIS at other locations also passively reflects the signal to the typical user, the farther two-hop path will incur larger path loss and thus can be ignored. In addition, the present invention considers that the network is interference-limited, and the noise can be ignored, thus the useful signal power can be expressed as:

[0081]

[0082] where, denotes the phase offset caused by the serving base station to user link, denotes the channel phase of the nthreflection unit of the RIS (the RIS corresponding to the serving base station) to the serving base station, denotes the channel phase of the nthreflection unit of the RIS to the user, φ n denotes the controllable phase shift of the nthreflection unit of the RIS, denotes the channel coefficient of the nthreflection unit of the RIS to the serving base station, denotes the channel coefficient of the nthreflection unit of the RIS to the typical user.

[0083] When the useful signal is superimposed by the signal vectors of the direct link and the RIS reflection link, the power of the synthesized signal is closely related to the phase relationship of the two vectors. According to the signal processing theory, the modulus square of the superposition of two complex vectors reaches the maximum value when their phases are the same, that is, the signal is beamformed. Therefore, in order to maximize the received power, it is necessary to compensate for the phase difference between the RIS reflection path and the direct path. Therefore, the optimal value of the phase offset is:

[0084]

[0085] By precisely regulating the phase offset of the RIS unit, the reflected signal is phase-aligned with the direct link at the receiving end, and the useful signal power is maximized, which is expressed as:

[0086]

[0087] Because the Nakagami-m fading is a more general fading model and contains the special case of Rayleigh fading, it is assumed that the small-scale fading of all channels is Nakagami-m fading. Specifically, the direct link may produce multipath propagation effects due to the reflection, scattering or diffraction of objects in the environment, so ||g0‖ ~ Nakagami(1, 1). In the RIS-assisted cascaded link, the signal is reflected to the typical user through the RIS, so the small-scale fading of the base station to the RIS where m h represents the shape parameter of the small-scale fading of the base station to the RIS, and the small-scale fading of the RIS to the typical user where m r represents the shape parameter of the small-scale fading of the RIS to the typical user. Due to the aggregation effect of random variables, the exact distribution of S is difficult to solve analytically. Therefore, the present application uses an approximate method: it is observed that the useful signal power can be expressed as the combination of two gamma distribution variables, and based on the moment matching method, the gamma distribution can be used to approximate the aggregated variable S ~ Γ(k s ,θ s ).

[0088] In the RIS-assisted cascaded link, which can be approximated as a Gaussian distribution according to the CLT:

[0089]

[0090] The cascaded link channel gain in the above formula is expressed as S r which can be approximated by a gamma random variable with a shape parameter and a scale parameter:

[0091]

[0092] Since Its q-th moment is denoted as:

[0093]

[0094] The cascaded link channel gain S r has been denoted by a gamma random variable with q-th moment:

[0095]

[0096] Using the moment matching method, the aggregate variable S is approximated as a gamma random variable with shape parameter and scale parameter:

[0097]

[0098] where the first and second moments of the aggregate variable S are:

[0099]

[0100] where,

[0101] S1.5interference signal power

[0102] Since the RIS-assisted reflection effect still exists for the interfering base station, the interference link is also divided into a direct link and a RIS-assisted cascaded link, where the small-scale fading of the direct link is Rayleigh fading, and the interference signal power is:

[0103]

[0104] where, ψ n denotes the phase and, denotes the channel phase from the interfering base station to the nth reflecting element of the RIS corresponding to the serving base station, denotes the channel phase from the nth reflecting element of the RIS to the user, φ n denotes the controllable phase shift of the nth reflecting element of the RIS, denotes the channel coefficient from the interfering base station to the kth RIS element, denotes the channel coefficient from the kth RIS element to the typical user, g j denotes the small-scale fading coefficient from the interfering base station to the user.

[0105] When the interference signal is reflected to the user through the RIS, φ k is uniformly distributed on [-π, π) to weaken the interference effect. Therefore, according to the CLT, when the number of reflecting elements N is large enough, obeys the circularly symmetric complex Gaussian distribution:

[0106]

[0107] As Thus g j Also subject to the CSCG distribution, the interference signal power S I Can be expressed as:

[0108]

[0109] The interference signal power S I The Laplace transform of:

[0110]

[0111] where, s denotes the ratio of the threshold to the scale parameter and the power, denotes the ratio of the interference power to the serving power, τ i denotes the SIR threshold value at which the typical user successfully demodulates and decodes the received signal when associated with the i-th tier BS; θ s denotes the scale parameter of the Gamma distribution of the equivalent approximation of the useful signal power; P i denotes the serving BS transmit power; r denotes the serving base station to typical user distance; Φ j denotes the set of BS locations of the tier in which the interfering base station is located; N denotes the number of reflecting elements per RIS; a denotes the path loss coefficient. η x,r denotes the large-scale fading from the base station to the RIS in the cascaded link of the RIS-aided base station and the typical user. η x denotes the large-scale fading of the direct link between the base station and the typical user, denotes the average number of sub-points of the cluster of interfering base stations of the SBS tier, P j denotes the interfering BS transmit power, denotes the transmit power of the interfering base stations of the SBS tier, denotes the transmit power of the interfering base stations of the MBS tier, denotes the base station density of the interfering base stations of the MBS tier, denotes the conditional probability density function of the distance y of the interfering base station to the typical user.

[0112] S1.6 RSMA transmission protocol

[0113] The RSMA of the application adopts a two-user model to analyze network performance. After adopting the RSMA technology, decoding failure of one user will not affect the decoding process of other users. The information transmitted to the user by the serving base station is divided into a "private" part and a "public" part. Among them, the public stream refers to information that multiple users hope to receive, and the private stream refers to information for a specific user. Typically, the typical user first decodes the public stream. When the user receives the public stream signal, the private stream signals of all users in the same frequency band will interfere. Therefore, the SIR of the typical user decoding the public stream is defined as follows:

[0114]

[0115] Where β' is the private stream allocation coefficient, β t ' and β f ' are the private stream allocation coefficients of the typical user and the fixed user respectively, P i represents the transmission power of the serving base station, P j represents the transmission power of the interfering base station. When the typical user successfully decodes the public stream, the interference of the public stream can be eliminated when decoding the private stream. Therefore, the SIR of the typical user decoding the private stream is defined as follows:

[0116]

[0117] In the two-user RSMA system, the typical user and the fixed user share the same public stream information, whether it is a near user or a far user. The difference in channel conditions will be reflected in the decoding process of the private stream. When the typical user is a near user, its private stream is usually allocated with lower power, while the private stream of the fixed user is allocated with higher power. Conversely, when the typical user is a far user, its private stream will be allocated with higher power, while the private stream of the fixed user will be allocated with lower power. In order to facilitate subsequent calculation and formula expression, we preset β t ' as a lower private stream allocation coefficient, and β f ' as a higher private stream allocation coefficient. When the typical user is a near user, the SIR of the typical user decoding the private stream is defined as follows:

[0118]

[0119] When the typical user is a far user, the SIR of decoding the private stream is defined as follows:

[0120]

[0121] If the common stream or the private stream decoding fails, the decoding process will be interrupted. Unlike the two-NOMA system, in the two-user RSMA system, the decoding interruption of the far user will not affect the decoding of the near user. Therefore, when the typical user is the near user, the decoding probability of the near user can be defined as:

[0122]

[0123] When the typical user is the far user, the private stream power allocation coefficient will be exchanged accordingly, and the decoding probability of the far user is defined as follows:

[0124]

[0125] where τ c represents the signal-to-interference ratio threshold of the common stream decoding, and τ p represents the signal-to-interference ratio threshold of the private stream decoding.

[0126] Performance analysis of S2

[0127] S2.1 Association probability

[0128] According to the Slivnyak-Mecke theorem, the random selection of the typical user will not affect the distribution of the point process. Since x c represents the position of the BS closest to the typical user, therefore ||x c || represents the distance between the candidate serving BS and the typical user.

[0129] In the PCP, if the position of the cluster center is z, f(y|z) represents the PDF of the distance y of the point in the sub-process to the origin. Although the distribution of the sub-process does not depend on the position z of the cluster center, its conditional distance distribution is closely related to the size of z.

[0130] When ||x c || in the parent PPP under the condition, the cumulative distribution function (CDF) and the probability density function are represented as follows:

[0131] CDF:

[0132] PDF:

[0133] where represents the average value of the number of sub-points of each cluster in the SBS layer. represents the conditional probability density function of the distance y of the base station to the typical user. represents the cumulative distribution function of the distance y of the base station to the typical user.

[0134] In this modeling process, PCP adopts Thomas cluster process (TCP) whose child points are points process of two-dimensional independent Gaussian distribution with variance σ k 2 generated around the parent point. Therefore, the PDF of distance y is defined according to the following Marcum Q function:

[0135]

[0136] where I0(t) is the zeroth-order first kind modified Bessel function. If the cluster center position z is at the origin, the PDF of y is defined as follows:

[0137]

[0138] When , the location distribution of the candidate serving BS is homogeneous PPP (HPPP), and therefore the distance ||x c || follows Rayleigh distribution, whose CDF and PDF are expressed as follows:

[0139] CDF:

[0140] PDF:

[0141] where λ k represents the density of Φ k .

[0142] Under the parent PPP condition, the association probability of a typical user associated with the i layer BS in the UDN is defined as follows:

[0143] ① When , the definition when the typical user is associated with SBS is:

[0144]

[0145] ② When , the definition when the typical user is associated with MBS is:

[0146]

[0147] where represents the average number of child points of the cluster of the layer where the serving base station is located, represents the average number of child points of the cluster of the interfering base station of the SBS layer. represents the conditional probability density function of the distance from the serving base station to the typical user, represents the transmission power of the interfering base station of the SBS layer, denotes the transmit power of the interfering base station of the MBS layer, denotes the base station density of the interfering base station of the MBS layer. denotes the cumulative distribution function of the distance between the micro base station and the typical user when the micro base station is the interfering base station.

[0148] S2.2 Conditional coverage

[0149] Since the coverage is defined as the probability that the SIR experienced by the typical user is greater than the threshold value required for successful demodulation and decoding. The useful signal power S has been equivalently approximated by a gamma distribution, therefore, the coverage under RIS assistance can be derived by the upper incomplete gamma function as:

[0150]

[0151] According to the above formula, the conditional coverage of the typical user associated with the i-th layer BS under the PPP condition is:

[0152]

[0153] where, τ i denotes the SIR threshold value of the receiving end for successful demodulation and decoding of the received signal when the typical user is associated with the i-th layer BS; θ s denotes the scale parameter of the gamma distribution of the equivalent approximation of the useful signal power; P i denotes the transmit power of the serving BS; denotes the conditional probability density function of the serving distance.

[0154] To simplify the derivation, the following functions are defined:

[0155]

[0156] After deriving the conditional coverage of the UDN under RIS assistance, the conditional coverage of the RSMA system under RIS assistance needs to be obtained in combination with the RSMA transmission protocol. Unlike the two-user NOMA system, both users in the RSMA system need to decode their own public stream and private stream. The decoding interruption of the far user will not affect the decoding of the near user. Under the PPP condition, the conditional coverage of the typical user in the two-user RSMA system associated with the i-th layer BS can be expressed as:

[0157]

[0158] In the two-user RSMA system, the conditional coverage of the near user is defined as follows:

[0159]

[0160] where:

[0161] The conditional coverage of a remote user associated with the i-th tier BS is defined as:

[0162]

[0163] S2.3 Network coverage

[0164] When calculating the coverage, the relevant distance distribution of a typical user to the serving BS should be considered, so the parent PPP under the condition that the relevant distance of a typical user to the i-th tier serving BS in the UDN is denoted as follows:

[0165] ① When , the relevant distance of a typical user to the SBS is defined as:

[0166]

[0167] ② When , the relevant distance of a typical user to the MBS is defined as:

[0168]

[0169] where is the PDF of the relevant distance of a typical user to the i-th tier serving BS.

[0170] Under the condition of the parent PPP , the conditional coverage of a typical user associated with the i-th tier BS is expressed as follows:

[0171] ① When

[0172]

[0173] ② When , the following formula is used:

[0174]

[0175] In order to simplify the formula and analyze the coverage more intuitively, the following symbols are introduced:

[0176]

[0177] When a typical user is subject to the PCP, the location of the serving BS comes from the typical cluster or other clusters. In order to accurately describe the coverage, the PDF of the distance ||z0|| from the origin to the center z0 of the typical cluster should be considered, and the relevant distance of a typical user to the i-th The coverage of a typical user associated with a tier BS is defined as follows:

[0178] 1. When , the following formula is used:

[0179]

[0180] 2. When , the following formula is used:

[0181]

[0182] where denotes the density of parent PPP ; and denotes the probability density function of the distance from the origin to the center of a typical cluster z0.

[0183] When a typical user is subject to a PPP, the coverage of a typical user associated with the i th tier BS cannot take into account the PDF of the distance ||z0|| any more, and thus can be expressed as follows:

[0184] 1. When , the following formula is used:

[0185]

[0186] 2. When , the following formula is used:

[0187]

[0188] Since the users Φ u are distributed according to PPP and PCP respectively, the probability of each user being selected as a typical user is equal. By defining the density of and denoting the average number of users in each user cluster, the probability of a typical user selected belonging to or can be calculated as follows:

[0189]

[0190] According to the conditional coverage of RIS-RSMA systems , the RIS-aided RSMA hyperdense network coverage

[0191]

[0192] S2.4 Area spectral efficiency

[0193] RIS-assisted UDN, the i-th The ergodic rate of the i-th tier BS is defined as:

[0194]

[0195] In the RIS-RSMA system, the i-th tier The ergodic rate of the i-th tier BS can be calculated by the following expression:

[0196]

[0197] where, denotes the rate component of the near-zone common flow, denotes the rate component of the near-zone typical user's private flow as a near user, and the fixed user's private flow as a far user, respectively, denotes the rate component of the far-zone common flow, denotes the rate component of the far-zone typical user's private flow as a far user, and the fixed user's private flow as a near user, respectively.

[0198] According to the conditional coverage probability definition of the near user in the two-user RSMA system, the decoding rate of the near-zone user is obtained as:

[0199] The decoding rate of the near-zone user's common flow is as follows:

[0200]

[0201] The decoding rate of the near-zone typical user's private flow is as follows:

[0202]

[0203] The decoding rate of the near-zone fixed user's private flow is as follows:

[0204]

[0205] According to the conditional coverage probability definition of the far user in the two-user RSMA system, the decoding rate of the far-zone user is obtained as:

[0206] The decoding rate of the far-zone user's common flow is as follows:

[0207]

[0208] The decoding rate of the far-zone fixed user's private flow is as follows:

[0209]

[0210] The decoding rate of the far-zone typical user's private flow is as follows:

[0211]

[0212] In RSMA system, the user's ergodic rate is composed of public ergodic rate and private ergodic rate. The public ergodic rate must satisfy To ensure that all public flows can be decoded.

[0213] The ASE of RSMA system assisted by RIS can be obtained from the definition of ASE:

[0214]

[0215] Where, denote the user's ergodic rate of PPP distribution and PCP at the i-th layer BS, respectively.

[0216] S3 simulation analysis

[0217] Next, the simulation analysis of RIS-assisted RSMA ultra-dense network (RIS-RSMA) coverage and ASE impact. Since the coverage and ASE are closely related to the SIR threshold τ, in order to simplify the analysis, it is assumed that τ t = τ f = τ. The simulation parameters are: the density of parent PPP MBS layer base station The density of SBS layer base station λ2=1×10 -6 m -2 , the transmit power of SBS layer base station P1=1W, the transmit power of MBS layer base station P2=20W, the path loss coefficient α=4, the Nakagami-m fading coefficient m=1, the range radius of near user r1=10m, the range radius of far user r2=50m, the average number of users per user cluster The density of SBS layer user The average number of SBS per SBS cluster The number of RIS reflecting units N=150, the carrier frequency f c =3.5GHZ, the Rayleigh parameter σ=20.

[0218] The relationship between the coverage of different network systems and the transmit power P2 of MBS is shown in Figure 2 . Among them, the power allocation coefficient β of the near user of the NOMA system is set to 0.1, the power allocation coefficient β' of the private flow in the RSMA system is 0.5, and the SIR threshold value τ=-10dB. The results show that with the increase of P2, the coverage is improved at the beginning, but after reaching a certain threshold value P2 *, beyond which the interference to SBS will increase, resulting in the decline of UDN coverage. As can be seen from the figure, although RSMA loses more coverage performance than NOMA under the same P2, the RIS-RSMA system compensates for the defect of the decline in coverage performance, and the coverage rate of the RIS-RSMA system is greater than that of the RSMA system when only the fading coefficient m is changed, which shows that RIS not only enhances the useful signal power by reconstructing the channel conditions, but also reduces the interference signal power.

[0219] The relationship between the ASE of different network systems and the transmission power P2 of MBS is shown in Figure 3 . With the increase of P2, the ASE gradually decreases. This is because the high-power signal of MBS produces significant interference to the spectrum multiplexing users of adjacent SBS, especially in the environment of dense deployment of SBS, the superposition of interference significantly reduces the ergodic rate of SBS, and thus the overall ASE also decreases. The data also show that the ASE of the RSMA system is better than that of NOMA and OMA under any transmission power P2. Therefore, in the complex environment of dense deployment of SBS and large transmission power of MBS, RSMA is more suitable than NOMA as a technology to optimize the spectrum utilization of UDN, and reducing the power allocation coefficient β' of private flow in the RSMA system can improve the ASE of UDN, because the multi-user gain of sharing public flow can effectively compensate for the decrease in private flow performance. Combined with the analysis of Figure 3 , too high P2 will cause the rapid decline of UDN coverage and ASE, so there is indeed an optimal transmission power value P2 * , which makes the coverage rate of the RIS-RSMA system reach the maximum, and its ASE is better than that of NOMA and OMA. At the same time, RIS technology can improve the ASE of the network by maximizing the active signal power, so RIS-RSMA can comprehensively improve the transmission performance of UDN.

[0220] The above simulation results show that the ASE of the RSMA system is better than that of NOMA and OMA under any transmission power P2 of the SBS layer base station. Therefore, in the complex environment of dense deployment of SBS and large transmission power of MBS, RSMA is more suitable than NOMA as a technology to optimize the spectrum utilization of UDN. This shows that the use of RSMA technology solves the problem that the ASE improvement of NOMA technology is limited due to its dependence on user channel differences and fixed decoding order, and further improves the spectrum utilization.

[0221] The simulation results show that, although RSMA loses more coverage performance compared with NOMA under the condition of the same SBS layer base station transmission power P2, the RIS-RSMA system compensates for the defect of the decline in coverage performance, and the coverage rate of the RIS-RSMA system is greater than that of the RSMA system only changing the fading coefficient m. Therefore, the RIS-assisted RSMA ultra-dense network compensates for the coverage performance loss caused by multi-user decoding and user interference accumulation of the RSMA technology, is superior to the existing technologies such as OMA and NOMA, and further improves the coverage rate of the system.

[0222] The RIS-assisted RSMA ultra-dense network provided by the application not only enhances the useful signal power by reconstructing the channel condition through RIS, but also reduces the interference signal power. The RIS-assisted RSMA system not only compensates for the defect of the decline in coverage performance, but also has a coverage rate gain greater than that caused by only changing the channel condition, and RIS can improve the ASE of the system by maximizing the active signal power itself, so that the RIS-RSMA system realizes comprehensive improvement of transmission performance.

[0223] The application further maximizes the useful signal power by regulating the phase offset of the RIS reflection unit, and further improves the transmission performance of the ultra-dense network.

[0224] The coverage performance analysis method of the RIS-assisted RSMA ultra-dense network provided by the application is used to accurately analyze the system performance. The useful signal power of the service base station assisted by RIS transmission is approximated as a gamma distribution by using the moment matching method, and the cumulative interference signal power is equivalent to a circularly symmetric complex Gaussian (CSCG) distribution based on the Central Limit Theorem (CLT). Since the Central Limit Theorem is applicable to a large number of independent interference sources, the more interference terms, the more accurate the approximation, so the CLT theorem is used to analyze the interference in the ultra-dense network. The complex interference is approximated as a Gaussian distribution by using the CLT theorem, thereby simplifying the calculation of coverage rate and regional spectral efficiency. And by using the moment matching method and the CLT theorem, the expression of the Laplace transform of the useful signal power and the interference power is more accurate, and the interference caused by RIS is described more accurately, thereby improving the accuracy of the calculation of coverage rate and regional spectral efficiency.

[0225] Although the embodiments of the application have been disclosed as above, they are not limited to the application and implementation listed in the specification and embodiments, and can be fully applied to various fields suitable for the application. Those skilled in the art can easily make other modifications, and therefore the application is not limited to specific details and the figures shown and described herein, without departing from the general concept defined by the claims and the equivalent scope.

Claims

1. A method for analyzing coverage performance of a RIS-assisted RSMA ultra-dense network, comprising: The method comprises the following steps: constructing a calculation model of a useful signal power distribution and a calculation model of an interference signal power; wherein the calculation model of the useful signal power distribution is in the form of a gamma random variable, and the calculation model of the interference signal power is in the form of a Laplace transform expression; obtaining a conditional coverage rate of a RIS-assisted RSMA super-dense network according to the calculation model of the useful signal power distribution and the calculation model of the interference signal power; de-conditioning the conditional coverage rate to obtain a coverage rate of the RIS-assisted RSMA super-dense network; the coverage rate of the RIS-assisted RSMA super-dense network is: wherein, denotes the coverage probability of a typical user subject to Poisson Point Process, PPP, distribution associated with the i-th tier base station in a RIS-RSMA system, denotes the coverage probability of a typical user subject to Poisson Cluster Process, PCP, distribution associated with the i-th tier base station in a RIS-RSMA system, SIR i denotes the signal-to-interference ratio of a typical user associated with the i-th tier base station, τ denotes the SIR threshold value for successful demodulation and decoding of the received signal at the receiver, denotes the probability that a typical user belongs to denotes the probability that a typical user belongs to denotes the probability that a typical user belongs to denotes the set of all base stations;​ determining a user's ergodic rate at a macro base station layer and a micro base station layer according to the conditional coverage rate of the RIS-assisted RSMA super-dense network, and obtaining a regional spectrum efficiency of the RIS-assisted RSMA super-dense network according to the ergodic rate; the regional spectrum efficiency of the RIS-assisted RSMA super-dense network is: wherein, respectively denote the ergodic rate of users subject to a Poisson point process and users subject to a Poisson cluster process at the i-th tier base station; denotes the micro base station tier, denotes the macro base station tier, respectively denote the probability that a typical user belongs to denotes the macro base station user, denotes the micro base station user; λ i denotes the density of the i-th tier base station, denotes the set of all base stations;​ wherein the ergodic rate comprises a public ergodic rate and a private ergodic rate; calculating the ergodic rate of the user at each base station layer based on a two-user model of RSMA: wherein, denotes the user's average speed at the i-th tier base station; denotes the rate component of the near-zone common flow, denotes the rate component of the near-zone typical user's private flow as a near user, and denotes the rate component of the far-zone common flow, denotes the rate component of the far-zone typical user's private flow as a far user, and wherein each service base station is associated with a user in the last round of user association process, and the user is called a fixed user; it is assumed that the distance between the fixed user and the connected service base station is the same, which is called a fixed distance; according to the fixed distance, the coverage area of the service base station is divided into a near zone and a far zone; the user in the near zone is defined as a near user, and the user in the far zone is defined as a far user; wherein the RIS-assisted RSMA super-dense network comprises: a plurality of micro base stations deployed in a hotspot area with high user density; the spatial positions of the micro base stations are modeled by a Poisson cluster process, and the micro base stations are independently and identically distributed around a cluster center; a plurality of macro base stations, the spatial positions of which are modeled by a Poisson point process, and the spatial positions of the macro base stations are independent and uniformly distributed; wherein each of the micro base stations and the macro base stations is respectively equipped with an RIS.

2. The method of analyzing the coverage performance of RIS-assisted RSMA ultra-dense networks according to claim 1, characterized in that, The calculation model of the useful signal power distribution is: where κ s and θ s denote the shape and scale parameters of the gamma distribution equivalent to the useful signal power, respectively, E[S] and E[S 2 ] denote the first and second moments of the gamma distribution equivalent to the useful signal power, respectively, and S denotes the gamma distribution equivalent to the useful signal power.

3. The method of analyzing the coverage performance of RIS-assisted RSMA ultra-dense networks according to claim 2, characterized in that, The calculation model of the interference signal power is: wherein, s denotes the ratio of the threshold to the scale parameter and the power; denotes the ratio of the interference power to the serving power; τ i denotes the SIR threshold value for the typical user to successfully demodulate and decode the received signal when associated with the i-th tier base station; θ s denotes the scale parameter of the gamma distribution equivalent to the useful signal power; P i denotes the serving base station transmit power; r denotes the serving base station to typical user distance; Φj denotes the set of base station locations of the tier where the interfering base station is located; N denotes the number of reflecting elements per RIS; a denotes the path loss coefficient; denotes the set of all base stations; η x,r denotes the large scale fading from the base station to the RIS in the cascaded link of the RIS assisted base station and the typical user; η x denotes the large scale fading of the direct link between the base station and the typical user; denotes the average number of sub-points of the cluster of interfering base stations of the micro base station tier; P j denotes the interfering base station transmit power; denotes the transmit power of the interfering base stations of the micro base station tier; denotes the transmit power of the interfering base stations of the macro base station tier; denotes the base station density of the interfering base stations of the macro base station tier; denotes the conditional probability density function of the distance y of the interfering base station to the typical user; denotes the micro base station tier, denotes the macro base station tier.

4. A reconfigurable intelligent surface, RIS, assisted RSMA ultra-dense network, characterized in that, The method for analyzing the coverage performance of the RIS-assisted RSMA super-dense network according to any one of claims 1-3, wherein the RIS-assisted RSMA super-dense network comprises: a plurality of micro base stations deployed in a hotspot area with high user density; the spatial positions of the micro base stations are modeled by a Poisson cluster process, and the micro base stations are independently and identically distributed around a cluster center; a plurality of macro base stations, the spatial positions of which are modeled by a Poisson point process, and the spatial positions of the macro base stations are independent and uniformly distributed; wherein each of the micro base stations and the macro base stations is respectively equipped with an RIS.

5. The RIS-assisted RSMA ultra-dense network of claim 4, wherein, the phase offset of the reflecting unit of the RIS is set as: wherein φ n denotes the phase offset of the n-th reflecting element of the RIS, denotes the channel phase of the n-th reflecting element of the RIS to the user, denotes the channel phase of the n-th reflecting element of the RIS to the user, denotes the phase offset caused by the serving base station to user link.

Images