Multi-layer blind detection method in complex degree layer by layer reduced license-free SCMA

Abstract

The application belongs to the technical field of Internet of Things communication, and discloses a multi-layer blind detection method with gradually reduced complexity in license-free SCMA, constructs a system model, determines a user set and a codebook structure, adopts a license-free access mode between a base station and a user, regards a spectrum resource block as a decoding layer, starts matching of the user and a codebook code word from the first decoding layer, and gradually eliminates the determined interference of the matched user in the decoded layer by using the idea of successive interference cancellation, so that the decoding complexity is gradually reduced. Meanwhile, each decoding layer has a strict error threshold, which guarantees the decoding performance of each layer.

Description

Technical Field

[0001] This invention belongs to the field of Internet of Things (IoT) communication technology, and particularly relates to a multi-layer blind detection method in unlicensed SCMA with progressively decreasing complexity. Background Technology

[0002] The information society is evolving from the Internet era to the Internet of Things (IoT) era. If the Internet connects and serves people, then the IoT extends the reach of information networks from people to things, achieving "interconnectivity of everything." IoT applications place higher demands on air interface technology, requiring it to meet the exponential growth of user terminal devices and mobile data traffic. Traditional orthogonal multiple access technologies, such as TDMA, FDMA, and CDMA, all employ orthogonal physical resource allocation methods, which struggle to meet the low latency and ultra-high system capacity requirements of massive user connections.

[0003] Compared to orthogonal multiple access (OMA), non-orthogonal multiple access (NOA) allows more users to access the same resource block, significantly increasing the number of users and the system's transmission capacity. Sparse code multiple access (SCMA) is a code-domain NOA technology that combines multidimensional modulation with low-density signature multiple access, exhibiting good performance in spectral efficiency and throughput. The massive number of users in the Internet of Things (IoT) can utilize SCMA technology for large-scale access, while introducing an unlicensed random access mechanism to reduce user service transmission latency, save signaling overhead, and is suitable for large-scale, dynamically changing user numbers. Therefore, unlicensed SCMA technology plays a crucial role in the deployment and implementation of the IoT in future intelligent communications.

[0004] Traditionally, the receiver of an SCMA system employs the Message Passing Algorithm (MPA) for multi-user detection, which has lower complexity than the optimal Maximum A posteriori (MAP) algorithm while offering similar detection performance. However, the MPA algorithm requires prior knowledge of the SCMA codebook used by each user for user data detection. In the unlicensed SCMA system mentioned in this invention, since each user does not handshake with the base station, and users randomly acquire their codebooks and transmit data upon accessing the system, the base station needs to perform multi-user detection without knowing the matching of users and codebooks. This makes the traditional MPA algorithm unsuitable. Therefore, designing a simplified unlicensed SCMA blind detection algorithm is of practical significance.

[0005] Based on the above analysis, the problems and shortcomings of the existing technology are as follows:

[0006] (1) Power domain non-orthogonal multiple access (NOAMI) technology is the most researched and mature NOAMI technology, with significant research achievements in both access design and detection methods. It requires a significant channel gain difference between users to achieve good access performance. However, in IoT scenarios with a large number of devices, the devices are often densely deployed, making it difficult to achieve a significant channel gain difference between devices. Therefore, power domain NOAMI is not an effective solution for IoT user access.

[0007] (2) The MPA algorithm is usually used for multi-user data detection in the code domain SCMA system. There are also many improved MPA algorithms that have achieved lower complexity and higher detection performance. However, the MPA algorithm needs to know the SCMA codebook used by each user in advance during detection, so it is not suitable for unlicensed SCMA systems.

[0008] (3) There are also a few research results on unlicensed SCMA blind detection algorithms, but they are mainly for the case where the user and codebook match is known. Therefore, they cannot solve the blind detection problem when the base station does not know the user and codebook match under the unlicensed mechanism. Summary of the Invention

[0009] To address the problems existing in the prior art, this invention provides a multi-layer blind detection method in license-free SCMA with progressively decreasing complexity.

[0010] This invention is implemented as follows: a multi-layer blind detection method in unlicensed SCMA with progressively decreasing complexity includes: constructing a system model, determining the user set and codebook structure respectively; and connecting users to the base station using an unlicensed SCMA access method. Since users and the base station do not pre-handshake, users directly transmit information to the base station after randomly selecting a codebook, while the base station needs to perform blind reception detection when the matching of users and codebooks is unknown. In blind detection, firstly, a spectrum resource block is considered as a decoding layer. Starting from the first decoding layer, the base station attempts to match users and codebooks based on the sum of user information received at this layer and the indicator matrix in the codebook structure. Matching at this layer stops when the matching error threshold is reached, and the matching result is recorded. Secondly, the matching of users and codebooks is expanded layer by layer. The key is to utilize the idea of ​​continuous interference cancellation to eliminate deterministic interference from previously decoded layers at the current decoding layer, reducing the number of users and codebooks that need to be matched in the current decoding layer. Thirdly, as the number of decoded layers increases, the matching space of the current decoding layer also decreases, achieving a progressively decreasing complexity blind detection effect. Meanwhile, since each layer of decoding has an error threshold limitation, the accuracy of matching between the user and the codebook can be guaranteed for each layer of decoding.

[0011] Furthermore, the construction of the system model includes:

[0012] The Internet of Things (IoT) includes one base station and J user devices. The user equipment uses SCMA uplink access to the base station, and the system has K available spectrum resources, where K = {f1, ..., f2}. K The system can support a maximum number of users. N represents the number of resource blocks a user can use simultaneously. At the transmitting end, the user equipment uses an SCMA encoder to directly map the B-bit information data stream into K-dimensional codewords according to its own dedicated codebook. Each codebook contains M codewords, where M = 2^32. B User u j codebook structure As shown below:

[0013]

[0014] Here, the number of rows represents the total number of spectrum resources in the system, with a total of K rows; when a user uses a certain frequency resource, the corresponding row is a non-zero row; c ij The 'c' is a complex number, representing the constellation point mapped to the user information; when a user does not use a certain frequency resource, the corresponding line is zero, c. ij =0; the number of columns represents the number of codewords M. If the codewords are transmitted in groups of 2 bits, M=4. Then, the user's resource usage is represented by the indicator matrix F, as shown in the following formula:

[0015]

[0016] Where the number of rows represents the number of usable resource blocks K, and the number of columns represents the total number of users J that the system can serve; ij =0 indicates user u j Unused resource block f i ;f ij =1 indicates user u j Use resource block f i The number of non-zero elements in each row represents the number of users using the spectrum resource, represented by row repetition d. v This indicates that the number of non-zero elements in each column represents the amount of spectrum resources used by the same user, indicated by column weight d. f This means that the non-zero elements in each row and each column are equal, and d f <K.

[0017] Furthermore, the multi-layer blind detection method in the license-free SCMA with progressively decreasing complexity includes the following steps:

[0018] Step 1: Randomly sort all users and select the top d users in ascending order. v For each user, randomly match the codebook associated with the first decoding layer to find d. vEach user selects the optimal combination of a codeword from the codebook and attempts to replace users in ascending order until the minimum error threshold condition of this layer is met.

[0019] Step two: Divide the system user set into pure users and impure users, and select d from the two user sets according to the mapping matrix F. v Each user forms a reused resource block f i user set And try replacing with a clean user. For uncertain users, continue until the minimum error threshold condition of this layer is met;

[0020] Step 3: Iterate through Step 2, matching resources, codebook codewords, and users for each resource layer until all users are matched with their codebooks and corresponding codewords, at which point the algorithm ends.

[0021] Furthermore, in step one, for the first resource block f1—the first decoding layer, there are a total of d v Unknown users are reused. All users are randomly sorted, and the top d users are selected in ascending order. v Each user randomly selects a codebook associated with resource block f1 and randomly chooses a codeword from that codebook as their transmission codeword. Since each codebook contains M codewords, d... v A total of users calculated Next, find d according to the following formula v Each user selects the optimal combination of a specific codeword from the codebook:

[0022]

[0023] Among them, y f1 This represents the signal received on resource block f1; This represents the set of users reused on resource block f1; σ represents the sum of multiple user transmitted signals multiplexed on resource block f1; 2 The noise power represents additive white Gaussian noise; δ f1 The error threshold represents the mean square error between the multi-user transmitted signal multiplexed on resource block f1 and the signal received by the base station on resource block f1.

[0024] After the first match, the remaining user set is represented as Sort the remaining users by user, and then select one user in ascending order each time to attempt a replacement. For a given user, calculate the error threshold δ' after the replacement. f1 And compare it with the error threshold before replacement: when δ' f1 ≥δ f1 If not, then no replacement will be performed. Keep it unchanged, while placing the user at the end of the queue of remaining users; when δ' f1 <δ f1 , Updated to u j Represents the replaced user, u j' This represents the replacement user, and the user being replaced is u. j They were placed at the end of the remaining user queue.

[0025] During the replacement and update process Users in the process are constantly being filtered, and each successful replacement... This will more closely resemble real-world multi-user reuse on resource block f1. The minimum threshold for determining a stopping condition on resource block f1 is set to... when At that point, the iterative search on the first decoding layer—resource block f1—is stopped, and the set of end users multiplexed on resource block f1 is determined. And record d v The codebook used by each user and the transmitted codewords in the codebook.

[0026] Furthermore, in step two, for the i-th resource block f i --The i-th decoding layer also has d v Each user reuses the resource block. At this point, the system user set is divided into two categories: the first category consists of users who have not yet been matched with any resource block, called pure users; the second category consists of users who have been matched with resource blocks, called impure users.

[0027] Based on the position of the non-zero element in the i-th row of the mapping matrix F, d can be selected from both pure users and semi-pure users. v Each user forms a reused resource block f i user set The selection method is as follows: based on the codebook information already matched by the semi-pure user, determine whether the semi-pure user uses resource block f. i If so, add the semi-pure user. Furthermore, semi-pure users do not participate in resource block f i The subsequent replacement process is for user sets The user identification part in the text is represented as Randomly select from clean users individual users The non-fixed user portion is selected, and codebooks not used by fixed users are randomly matched to the non-fixed users. The error threshold is calculated according to the following formula:

[0028]

[0029] in, It is in resource block f i The sum of multiplexed signals received from multiple users. Represents resource block f i The transmitted signal is mapped to the user's matched codebook and corresponding codeword; Representative resource block f i The transmitted signal mapped by the temporarily matched codebook and corresponding codeword for non-fixed users; σ 2 This represents the variance of additive white Gaussian noise; non-fixed users are replaced by other remaining pure users in subsequent iterations until a minimum threshold is reached.

[0030] Furthermore, the replacement process and rules are as follows:

[0031] (1) Sort the remaining clean users, and each time select a user from the remaining clean users in ascending order to randomly replace them. A non-fixed user in the system; if Then no replacement will be made, and clean users will be placed at the end of the queue of remaining clean users, maintaining...

[0032] (2) If Then replace and update. The replaced users will be placed at the end of the queue of remaining clean users;

[0033] (3) If Then replace and update. And stop the search. Represents resource block f i The minimum threshold for determining the stopping condition; will be reused in resource block f i The set of end users on is denoted as And record d v The codebook and corresponding codewords matched for each user.

[0034] Another object of the present invention is to provide a multi-layer blind detection system in license-free SCMA with progressively decreasing complexity, which implements a multi-layer blind detection method in license-free SCMA. The multi-layer blind detection system in license-free SCMA with progressively decreasing complexity includes:

[0035] The system model building module is used to construct the system model and determine the user set and codebook structure.

[0036] The search and matching module is used to randomly sort all users and determine their transmitted codewords, finding d. v Each user selects the optimal combination of a certain codeword from the codebook and performs a search and matching.

[0037] The user set determination module is used to divide the system user set into clean users and non-clean users, selecting d. v Each user forms a reused resource block f i user set And calculate the error threshold;

[0038] The blind detection module is used to iteratively execute the user set determination module, matching resources with users, users with codebooks and codewords for each resource layer, until all users are matched with codebooks and corresponding codewords.

[0039] Another object of the present invention is to provide a computer device including a memory and a processor, the memory storing a computer program, which, when executed by the processor, causes the processor to perform the steps of the multilayer blind detection method in the license-free SCMA with progressively decreasing complexity.

[0040] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the multilayer blind detection method in the license-free SCMA with progressively decreasing complexity.

[0041] Another objective of this invention is to provide an information data processing terminal for implementing the multi-layer blind detection system in the license-free SCMA with progressively decreasing complexity.

[0042] Based on the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solution to be protected by this invention are as follows:

[0043] First, addressing the technical problems existing in the prior art and the difficulty of solving them, this paper closely analyzes, in conjunction with the technical solution to be protected by this invention and the results and data obtained during the research and development process, how the technical solution of this invention solves the technical problems, and the inventive technical effects brought about by solving these problems. The specific description is as follows:

[0044] This invention's multi-layer blind detection method in license-free SCMA, with progressively decreasing complexity, sets a high error threshold for subsequent users during initialization, resulting in highly reliable optimal matching users and significantly improving decoding performance. The resulting set of impure users provides some effective matching information to the second decoding layer, reducing the search space between users and resource blocks and lowering computational complexity to some extent. In subsequent layer-by-layer decoding, the increasing number of decoding layers and the growing number of impure users mean more certainty when searching for users at each layer, simplifying the user search process. As decoding progresses, more and more impure users complete the matching process, further reducing the number of users requiring matching and shrinking the search space. Later decoding layers have lower computational complexity, achieving a progressively decreasing complexity blind detection effect. Simultaneously, strict user replacement thresholds at each layer ensure decoding performance.

[0045] Second, considering the technical solution as a whole or from a product perspective, the technical effects and advantages of the technical solution to be protected by this invention are specifically described as follows:

[0046] This invention proposes a blind detection method for unlicensed SCMA systems, addressing the unknown user-codebook matching situation at the base station. This method utilizes the concept of continuous interference cancellation, eliminating deterministic interference obtained from each decoded layer layer by layer, thereby continuously reducing the matching space between the number of users and the number of codebooks at each decoding layer. The more decoded layers, the more deterministic interference can be eliminated, and the fewer users and codebooks need to be matched, resulting in a progressively lower computational complexity. This is particularly important for receiver detection in massive IoT connection scenarios. Otherwise, the receiver detection of a large number of users would place a significant burden on the base station's computation, hindering its implementation. Furthermore, the limiting effect of the detection threshold at each decoding layer also ensures the accuracy of multi-user decoding.

[0047] Third, as supplementary evidence of the inventive step of the claims of this invention, it is also reflected in the following important aspects:

[0048] (1) The expected benefits and commercial value of the technical solution of this invention after transformation are as follows:

[0049] With the accelerated implementation of the national new infrastructure strategy, the pace of 5G network construction in China has also significantly accelerated, bringing major development opportunities to industries such as the Internet of Things (IoT) and ushering in a period of explosive growth. The IoT era means that "everything in the world can be interconnected," and the surge in the number of smart devices is a prominent feature of this era. Accessing a large number of smart devices to the network through license-free SCMA technology is crucial, and low-complexity receiving and detection methods are one of the key issues for successful IoT deployment. This invention's technical solution is applicable to the following aspects: first, smart living applications in areas such as smart homes, connected vehicles, and smart wearables; second, industrial digitalization applications in areas such as smart agriculture, smart factories, and smart healthcare; and third, intelligent governance applications in areas such as smart fire protection, environmental monitoring, and smart metering. This invention's technical solution is not limited to the above applications; it is applicable to related fields such as intelligence, smart technology, monitoring, and sensing, further leveraging the role of IoT in empowering industrial upgrading, improving governance capabilities, and enriching social life, promoting the integrated development of IoT with various industries and the integration of the digital economy with the real economy. The Internet of Things has become an important driving force for a new round of global technological revolution and industrial transformation. Therefore, the technical solution of this invention has high expected returns and commercial value after transformation.

[0050] (2) The technical solution of this invention fills a technical gap in the industry both domestically and internationally:

[0051] Existing unlicensed SCMA blind detection technologies mainly focus on two aspects: first, the blind detection problem at the receiver when the number of active users is unknown; and second, the blind detection problem when multiple users select codebooks and collisions occur. These studies address receiver detection solutions based on known user-codebook matching at the base station, but cannot solve the blind detection problem when the base station has unknown user-codebook matching. The technical solution of this invention proposes a low-complexity solution to this problem. Attached Figure Description

[0052] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0053] Figure 1 This is a flowchart of a multi-layer blind detection method in license-free SCMA with progressively decreasing complexity, provided in an embodiment of the present invention.

[0054] Figure 2 This is a schematic diagram of an IoT unlicensed SCMA uplink provided in an embodiment of the present invention;

[0055] Figure 3 This is an interaction diagram of a multi-layer blind detection system in license-free SCMA with progressively decreasing complexity, provided in this embodiment of the invention.

[0056] Figure 4 This is a schematic diagram of user and codebook matching provided in an embodiment of the present invention;

[0057] Figure 5 This is a matching diagram of the user and the codebook codewords after the first layer of decoding provided in this embodiment of the invention;

[0058] Figure 6 This is a schematic diagram of the multi-user matching situation on resource block f2 in the F matrix provided in an embodiment of the present invention;

[0059] Figure 7 This is a matching diagram of the user and the codebook codewords after the second layer of decoding provided in an embodiment of the present invention;

[0060] Figure 8 This is the final matching diagram between the user and the codebook codewords after the third layer decoding provided in this embodiment of the invention. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0062] To address the problems existing in the prior art, this invention provides a multi-layer blind detection method in license-free SCMA with progressively decreasing complexity. The invention will be described in detail below with reference to the accompanying drawings.

[0063] like Figure 1 As shown, the multi-layer blind detection method in license-free SCMA with progressively decreasing complexity provided by this embodiment of the invention includes the following steps:

[0064] S101, Randomly sort all users and select the top d in ascending order. v For each user, randomly match the codebook associated with the first decoding layer to find d. v Each user selects the optimal combination of a codeword from the codebook and attempts to replace users in ascending order until the minimum error threshold condition of this layer is met.

[0065] S102, divide the system user set into pure users and impure users, and select d from the two user sets according to the mapping matrix F. v Each user forms a reused resource block f i user set And try to replace with the remaining clean users. For uncertain users, continue until the minimum error threshold condition of this layer is met;

[0066] S103, iteratively execute step S102, matching resources, codebook codewords and users for each resource layer until all users are matched with codebooks and corresponding codewords, at which point the algorithm ends.

[0067] As a preferred embodiment, the multi-layer blind detection method in license-free SCMA with progressively decreasing complexity provided by this invention specifically includes the following steps:

[0068] 1. System Model

[0069] like Figure 2 As shown, assume there is 1 base station and J user devices in an Internet of Things (IoT) network. The user equipment uses SCMA to access the base station via uplink, and the system has K available frequency resources. The maximum number of users that this system can support is N represents the number of resource blocks a user can use simultaneously. At the transmitting end, the user equipment uses an SCMA encoder to directly map the B-bit information data stream into K-dimensional codewords according to its own dedicated codebook. Each codebook contains M codewords, where M = 2^32. B User u j codebook structure As shown in formula (1):

[0070]

[0071] Here, the number of rows represents the total number of spectrum resources in the system, with a total of K rows; when a user uses a certain frequency resource, the corresponding row is a non-zero row, c ij The 'c' is a complex number, representing the constellation point mapped to the user information; when a user does not use a certain frequency resource, the corresponding line is zero, c. ij =0. The number of columns represents the number of codewords M. For example, if the codewords are transmitted in groups of 2 bits, M = 4. The user's resource usage can be represented by an indicator matrix F, as shown in formula (2):

[0072]

[0073] Where the number of rows represents the number of usable resource blocks K, and the number of columns represents the total number of users J that the system can serve. ij =0 indicates user u j Unused resource block f i ;f ij =1 indicates user u j Resource block f was used i The number of non-zero elements in each row represents the number of users simultaneously using the spectrum resource, expressed as row repetition d.v To represent; the number of non-zero elements in each column represents the amount of spectrum resources used by the same user, expressed by column weight d. f This indicates that the non-zero elements in each row and each column are equal, and d f <K.

[0074] 2. Algorithm Flow

[0075] Assume the channel information h for each user equipment j It is known that all codebooks have been stored at the base station. However, due to the unlicensed access method used between the base station and users, the base station is unaware of the matching status between users and the codebooks, necessitating blind detection of uplink information for multiple users. The blind detection process treats each spectrum resource block as a decoding layer, proceeding layer by layer from the first decoding layer to match user codewords with the codebook. Continuous interference cancellation is employed to eliminate definite interference from already matched users at higher layers, thereby continuously reducing the matching space between the number of users and the number of codebooks, thus achieving a gradual reduction in decoding complexity. The specific blind detection process is described below:

[0076] Step 1: For the first resource block f1 (i.e., the first decoding layer), there are a total of d... v An unknown user reuses this resource block. Randomly sort all users and select the top d users in ascending order. v Each user randomly selects a codebook associated with resource block f1, and then randomly chooses a codeword from that codebook as their transmission codeword. Since each codebook contains M codewords, d... v The total calculation required by each user Next. Find d according to formula (3). v Each user selects the optimal combination of a certain codeword from the codebook. The advantage of doing this is that it raises the threshold from the beginning, enabling highly accurate search matching in future iterations.

[0077]

[0078] in, This represents the sum of signals received on resource block f1; This represents the set of users reused on resource block f1; This represents the multiplexed signal transmitted on resource block f1; σ 2 The noise power represents additive white Gaussian noise; δ f1 The error threshold represents the mean square error between the multi-user transmitted signal multiplexed on resource block f1 and the signal received by the base station on resource block f1.

[0079] After the first matching, the remaining user set can be represented as Next, sort the remaining users by user, and select one user in ascending order each time to attempt a replacement. For a given user, calculate the error threshold δ' after the replacement. f1 And compare it with the error threshold before replacement. When δ' f1 ≥δ f1 If not, then no replacement will be performed. Keep it unchanged, and place the user at the end of the queue of remaining users; when δ' f1 <δ f1 , Updated to u j Represents the replaced user, u j' This represents the replacement user, and the user being replaced is u. j They were placed at the end of the remaining user queue.

[0080] During the replacement and update process Users in the process are constantly being filtered, and each successful replacement... This will more closely resemble the real user reused on resource block f1. The minimum threshold for determining a stopping condition on resource block f1 is set to... when At that point, the iterative search on the first decoding layer (resource block f1) is stopped, and the set of end users reused on resource block f1 is determined. And record d v The codebook used by each user and the transmission codewords in the codebook.

[0081] Step 2: For the i-th resource block f i (i.e., the i-th decoding layer), also has d v Each user reuses the resource block. At this point, the system user set can be divided into two categories: the first category consists of users who have not yet been matched with any resource block, called pure users; the second category consists of users who have been matched with resource blocks, called impure users.

[0082] Based on the position of the non-"0" element (i.e., "1" element) in the i-th row of the mapping matrix F, d can be selected from both pure and impure users. v Each user forms a reused resource block f i user set The selection method is as follows: First, based on the codebook information already matched by the non-pure user, it can be determined whether the non-pure user has used resource block f. i If so, then add the non-pure user. Furthermore, these non-pure users do not participate in resource block f. i The subsequent replacement process is for user sets The user identification part in the text can be represented as: Secondly, randomly select from clean users. Individual users, as The non-fixed user portion is selected, and codebooks not used by fixed users are randomly matched to the non-fixed users. The error threshold is calculated according to formula (4):

[0083]

[0084] in, It is in resource block f i The sum of signals received from multiple users. Represents resource block f i The transmitted signal is mapped by the user based on the matched codebook and its corresponding codeword; Representative resource block f i The transmission signals of non-fixed users based on their temporarily matched codebooks and their corresponding codeword mappings; σ 2 This represents the variance of the additive white Gaussian noise. Non-fixed users may be replaced by other remaining clean users in subsequent iterations until the minimum threshold is reached.

[0085] The specific replacement process and rules are as follows:

[0086] (1) Sort the remaining clean users, and each time select a user from the remaining clean users in ascending order to randomly replace them. A non-fixed user in the system. If... If no replacement is made, the clean user will be placed at the end of the queue of remaining clean users, and this will be maintained.

[0087] (2) If Then replace and update. The replaced users will be placed at the end of the queue of remaining clean users;

[0088] (3) If Then replace and update. And stop the search. Represents resource block f i The minimum threshold for determining the stopping condition. At this point, it will be reused in resource block f. i The set of end users on is denoted as And record d v The codebook matched by each user and its corresponding codewords.

[0089] Step 3: Iterate through Step 2, matching resources with users, users with codebooks and their codewords layer by layer, until all users and codebooks and their corresponding codewords are matched, at which point the algorithm ends.

[0090] like Figure 3 As shown, the multi-layer blind detection system in license-free SCMA with progressively decreasing complexity provided in this embodiment of the invention includes:

[0091] The system model building module is used to construct the system model and determine the user set and codebook structure.

[0092] The search and matching module is used to randomly sort all users, determine the user's transmitted codeword, and find d. v Each user selects the optimal combination of a certain codeword from the codebook and performs a search and matching.

[0093] The user set determination module is used to divide the system user set into clean users and non-clean users, selecting d. v Each user forms a reused resource block f i user set And calculate the error threshold;

[0094] The blind detection module is used to iteratively execute the user set determination module, matching resources with users, users with codebooks and codewords for each resource layer, until all users are matched with codebooks and corresponding codewords.

[0095] To demonstrate the inventiveness and technical value of the technical solution of this invention, this section provides specific product or related technology application examples of the technical solution claimed.

[0096] Assume an Internet of Things (IoT) system has one base station and six user devices. User equipment accesses the base station via unlicensed SCMA uplink, and the system has 4 available frequency resources. The maximum number of users that this system can support is At the transmitting end, the user equipment uses an SCMA encoder to directly map the 2-bit information data stream into 4-dimensional codewords according to its own dedicated codebook. Each codebook contains 4 codewords, where 4 = 2. 2 User u j codebook structure As shown in formula (5):

[0097]

[0098] The row count represents the total number of spectrum resources in the system, and there are 4 rows in total. When a user uses a certain spectrum resource, the corresponding row is a non-zero row, c. ij Take a complex number to represent the constellation point to which the user information is mapped; when a user does not use a certain spectrum resource, the corresponding row is zero, c ij =0. The column count represents the number of codewords, which is 4.

[0099] User resource usage can be represented by an indicator matrix F, as shown in formula (6):

[0100]

[0101] Here, the number of rows represents the number of usable resource blocks (4), and the number of columns represents the total number of users the system can serve (6). Row weight d v =3 indicates; column weight d f =2 to represent d f <K.

[0102] The specific blind testing process is as follows:

[0103] Step 1: For the first resource block f1, there are a total of d v = 3 unknown users are using this resource. Randomly sort all users, and select the first 3 users in ascending order. Randomly assign each of these 3 users a codebook associated with resource block f1. The structure of these three codebooks conforms to the form of the first 3 columns of the indicator matrix F, because the first 3 columns of F are associated with resource block f1. The matching of codebooks and users is as follows: Figure 4 As shown.

[0104] exist Figure 4 In the codebook, users u1, u2, and u3 are matched respectively. and Users u4, u5, and u6 wait in the queue. Users u1, u2, and u3 randomly select a codeword from their respective codebooks as their transmission codeword. Since each codebook contains 4 codewords, the 3 users need to calculate 4 codes in total. 3 Next, according to formula (3), find the optimal combination of three users selecting a certain codeword from the codebook. The purpose is to raise the threshold from the beginning and achieve high-quality search matching in future iterations.

[0105] Assuming that, after calculation, user u1 matches the codebook The first codeword, user u2 matches the codebook. The second codeword, user u3 matches the codebook. The error threshold δ obtained during the fourth codeword k1 If the highest value is found, then the optimal combination is initialized. At this point, the remaining user set can be represented as... Next, among the remaining users, sort them by user, and select one user in ascending order each time, i.e., user u4 attempts to replace. For a given user, calculate the error threshold δ' after the replacement. f1 And compare with the error threshold before replacement. when Then no replacement will be performed. Keep it unchanged, and place user u4 at the end of the queue of remaining users. when Updated to j = 1, 2, or 3, u j Represents the user being replaced, and also the user u to be replaced j At the end of the remaining user queue

[0106] During the replacement and update process Users in the process are constantly being filtered, and each successful replacement... This will more closely resemble the user set that is actually reused on resource block f1. The minimum threshold for determining a stopping condition on resource block f1 is set to... when When the iteration search on the first resource block f1 stops, the set of end users reused on resource block f1 is obtained. And record The codebooks used by the three users and the transmitted codewords in the codebooks, assuming that at this time... The matched codebook and codeword are The superscript number represents the corresponding codeword in the codebook. After decoding is completed on the first resource block f1, the matching status between the user and the codewords in the codebook is as follows: Figure 5 As shown.

[0107] Step 2: For the second resource block f2 (the second decoding layer), there are also 3 users sharing the same memory. At this point, as follows... Figure 6 As shown, the positions of the "1" element in the second row of the mapping matrix F are 1, 4, and 5. The position "1" determines the relationship with the codebook. The matched user u1 reuses this resource block, and positions "4" and "5" indicate that two other unknown users reuse this resource block. In this layer's multi-user search, the matching results from the first layer are partially applied here, reducing the matching space. Compared to the first layer requiring a search of three unknown users, this layer only needs to search two, and the search space for unknown users is smaller, requiring only searching among clean users who haven't yet been matched with any codebook. The set of clean users in this layer is {u5, u6, u3}.

[0108] In the clean user set {u5, u6, u3}, sorted in ascending order, first select user u5 and user u6 to match the codebook associated with resource block 2. and (As can be seen from the second row of the F matrix, there are a total of 3 codebooks that match the second resource block, of which 1 codebook already exists.) There are two more associated codebooks remaining after matching with user u1. and Similarly, first calculate the error threshold according to formula (7):

[0109]

[0110] Among them, y f2 It is the sum of signals received on resource block 2. This indicates that user u1, based on the matched codebook, performed a match on resource block 2. The transmitted signal generated by the first codeword in the resource block represents the fixed user portion of resource block 2; σ represents the sum of the transmitted signals of the non-fixed users (i.e., users u5 and u6) matched on resource block 2; 2 This represents the variance of the additive white Gaussian noise. Non-fixed users may be replaced by other remaining clean users in subsequent iterations until the minimum threshold is reached.

[0111] The specific replacement process and rules are as follows:

[0112] (1) Sort the remaining clean users, and each time select a user from the remaining clean users in ascending order to randomly replace them. If a non-determined user in the system, If no replacement is made, the clean user will be placed at the end of the queue of remaining clean users, and this will be maintained.

[0113] (2) If Then replace and update. The replaced users will be placed at the end of the queue of remaining clean users;

[0114] (3) If Then replace and update. And stop iteratively searching on the second resource block f2. This represents the minimum threshold for determining a stopping condition on resource block f2. The set of end users reused on resource block f2 is then denoted as... And record The codebooks used by multiple users and the corresponding codewords in the codebooks, assuming that at this time... The matched codebook and codeword are The superscript number represents the corresponding codeword in the codebook. The matching result between the user and the codebook and its corresponding codewords after the second layer of decoding is as follows: Figure 7 As shown.

[0115] Step 3: For resource block 3 (3rd decoding layer), only user 5 and the codebook remain. Matching has not yet been performed, but can be performed directly, saving the user the step of matching with the codebook. Let user 5 select a codebook. The four codewords in the matrix are used, and the error threshold is calculated based on the correspondence in the third row of the indicator matrix F, as shown in formula (8):

[0116]

[0117] in, For the error of resource block 3, This represents the sum of signals received on resource block 3. This represents the sum of the transmitted signals from users u3 and u4 as fixed users. Launched on behalf of user u3 The third codeword in the codebook, Launched on behalf of user U4 The third codeword in the codebook, User u5 transmits codebook The nth codeword in the codebook. Formula (8) shows that the end user u5 should match the codebook. The ability of The smallest codeword.

[0118] Assuming user u5 matches the codebook The second codeword in the middle If the value is the smallest, then the decoding for that layer is complete. The final matching result between the user and the codebook and its codewords is as follows: Figure 8 As shown.

[0119] In fact, the blind detection algorithm provided in this embodiment of the invention is a three-dimensional matching process of resource blocks, codebook codewords, and users. First, during the initialization of the first decoding layer (resource block f1), the randomly selected d... v Each user is randomly assigned to a d-level system associated with the first decoding layer. v In each codebook, there are M codewords. By analyzing d... v Each user calculates the error value for the combination of M codewords in their respective matching codebook to obtain the optimal combination result. Subsequent replacement matching processes are all based on this optimal combination initialization. Its important significance is to set a high error threshold for subsequent users, making the obtained matching users highly reliable and greatly improving decoding performance.

[0120] At the end of each decoding layer, the resulting set of impure users provides definite matching information for subsequent decoding layers, reducing the search space for matching users and resource blocks in the current decoding layer. As the number of decoded layers increases, the number of impure users also increases, providing more definite matching information for the ongoing decoding layer, thus simplifying the user search process. Simultaneously, with each layer of decoding, the more users among the impure users who complete the entire matching process, the lower the computational complexity of the later decoding layers, achieving a blind detection effect with decreasing complexity layer by layer. Furthermore, the strict user replacement threshold at each layer ensures decoding performance.

[0121] It should be noted that embodiments of the present invention can be implemented in hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the above-described devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuitry such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of the above-described hardware circuitry and software, such as firmware.

[0122] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A multi-layer blind detection method in license-free SCMA with progressively decreasing complexity, characterized in that, include: Step 1: Construct a system model and determine the user set and codebook structure. ; Step two: Randomly sort all users and determine their transmission codewords to find... Each user selects the optimal combination of a certain codeword from the codebook and performs a search and matching. Step 3: Divide the system user set into clean users and non-clean users, and select... Each user forms a reused resource block f i user set And calculate the error threshold; Step four, iteratively execute step three, matching resources with users, users with codebooks and codewords for each resource layer, until all users are matched with codebooks and corresponding codewords; The construction of the system model includes: When the Internet of Things includes 1 base station and J user devices The user equipment uses SCMA to access the base station via uplink, and the number of available frequency resources in the system is K. The system can support a maximum number of users. N represents the number of resource blocks a user can use simultaneously. At the sending end, the user equipment uses an SCMA encoder to directly map the B-bit information data stream into K-dimensional codewords according to its own dedicated codebook. Each codebook contains M codewords, where M=2. B ,user codebook structure As shown below: ； The number of rows represents the total number of spectrum resources in the system, which is K rows in total; when a user uses a certain frequency resource, the corresponding row is a non-zero row. The plural form represents the constellation point to which user information is mapped; when a user does not use a certain frequency resource, the corresponding line is zero. The column number represents the number of codewords M. If the codewords are transmitted in groups of 2 bits, M=4. Then, the user's resource usage is represented by the indicator matrix F, as shown in the following formula: ； Where the number of rows represents the number of usable resource blocks K, and the number of columns represents the total number of users J that the system can serve; Indicates user Unused resource block f i ; This indicates that user j uses resource block f. i The number of non-zero elements in each row represents the number of users simultaneously using the spectrum resource, expressed as row counts. This indicates that the number of non-zero elements in each column represents the amount of spectrum resources used by the same user, indicated by column weights. This means that the non-zero elements in each row and each column are equal, and .

2. The multi-layer blind detection method in license-free SCMA with progressively decreasing complexity as described in claim 1, characterized in that, The multi-layer blind detection method in the license-free SCMA with progressively decreasing complexity includes the following steps: Step 1: Randomly sort all users and select the top d users in ascending order. v For each user, randomly match the codebook associated with the first decoding layer to find d. v Each user selects the optimal combination of a codeword from the codebook and attempts to replace users in ascending order until the minimum error threshold condition of this layer is met. Step two: Divide the system user set into pure users and impure users, and select d from the two user sets according to the indicator matrix F. v Each user forms a reused resource block f i User set U(f i And try to replace U(f) with a clean user. i For uncertain users in the layer, continue until the minimum error threshold condition of this layer is met; Step 3: Iterate through Step 2, matching resources, codebook codewords, and users for each resource layer until all users are matched with their codebooks and corresponding codewords, at which point the algorithm ends.

3. The multi-layer blind detection method in license-free SCMA with progressively decreasing complexity as described in claim 2, characterized in that, In step one, for the first resource block f1—the first decoding layer, there are a total of Unknown users reuse resource blocks; randomly sort all users and select the top-ranked users in ascending order. Each user randomly selects a codebook associated with the f1 resource block and randomly chooses a codeword from that codebook as their transmission codeword; since each codebook contains M codewords, then... A total of users calculated Next, find according to the following formula Each user selects the optimal combination of a specific codeword from the codebook: ； in, This represents the signal received on resource block f1; This represents the set of users reused on resource block f1; This represents a multi-user transmission signal multiplexed on resource block f1; The noise power representing additive white Gaussian noise; The error threshold represents the mean square error between the multi-user transmitted signal multiplexed on resource block f1 and the signal received by the base station on resource block f1. After the first matching, the remaining user set can be represented as Next, sort the remaining users by user, and select one user in ascending order each time to attempt a replacement. For a specific user, calculate the error threshold after the replacement. And compare it with the error threshold before replacement; when If not, then no replacement will be performed. Keep the current order unchanged, and place the user at the end of the queue of remaining users; when , Updated to , Represents the replaced user. This represents the replacement user, and also the user being replaced. At the end of the remaining user queue; During the replacement and update process Users in the process are constantly being filtered, and each successful replacement... This will more closely resemble the real users reused on resource block f1; the minimum threshold for determining the stopping condition on resource block f1 is set to... ,when At this point, the iterative search on the first decoding layer, i.e., resource block f1, is stopped, and the set of end users multiplexed on resource block f1 is determined. and record middle The codebook used by each user and the transmission codewords in the codebook.

4. The multi-layer blind detection method in license-free SCMA with progressively decreasing complexity as described in claim 2, characterized in that, In step two, for the i-th resource block f i --The i-th decoding layer also exists Users are reused on resource blocks; the system user set is divided into two categories: the first category is users who are not matched with any resource blocks, called pure users; the second category is users who have been matched with resource blocks, called impure users. Based on the position of the non-"0" element in the i-th row of the indicator matrix F, the non-"0" element is equivalent to the "1" element, which can be selected from both pure users and impure users. Each user forms a reused resource block f i user set The selection method is as follows: First, based on the codebook information already matched by the non-pure user, it can be determined whether the non-pure user has used resource block f. i If so, then add the non-pure user. Furthermore, these non-pure users do not participate in resource block f. i The subsequent replacement process is for user sets The user identification part in the text can be represented as: ; Secondly, randomly select from clean users. Individual users, as The non-fixed user portion is used, and codebooks that fixed users have not used are randomly matched to the non-fixed users. The error threshold is calculated according to formula (4): ； in, It is in resource block f i The sum of signals received from multiple users. Represents resource block f i The transmitted signal is mapped to the user's matched codebook and corresponding codeword; Representative resource block f i The transmission signal of non-fixed users is based on the temporarily matched codebook and the corresponding codeword mapping; This represents the variance of additive white Gaussian noise; non-fixed users are replaced by other remaining pure users in subsequent iterations until a minimum threshold is reached. .

5. The multi-layer blind detection method in license-free SCMA with progressively decreasing complexity as described in claim 4, characterized in that, The replacement process and rules are as follows: (1) Sort the remaining clean users, and each time select a user from the remaining clean users in ascending order to randomly replace them. A non-fixed user in the system; if If the user is not replaced, the clean user will be placed at the end of the queue of remaining clean users, and the process will continue. ; (2) If Then replace and update. The replaced users will be placed at the end of the queue of remaining clean users; (3) If Then replace and update. And stop searching; Represents resource block f i The minimum threshold for determining the stopping condition is used; at this point, it will be reused in resource block f. i The set of end users on is denoted as and record middle The codebook matched by each user and its corresponding codewords.

6. A multi-layer blind detection system in license-free SCMA with progressively decreasing complexity, applying the multi-layer blind detection method in license-free SCMA as described in any one of claims 1 to 5, characterized in that, Multilayer blind detection systems in the license-free SCMA with progressively decreasing complexity include: The system model building module is used to construct the system model and determine the user set and codebook structure. ; The search and matching module is used to randomly sort all users, determine the user's transmission codeword, and find... Each user selects the optimal combination of a certain codeword from the codebook and performs a search and matching. The user set determination module is used to divide the system user set into clean users and non-clean users, and select... Each user forms a reused resource block f i user set And calculate the error threshold; The blind detection module is used to iteratively execute the user set determination module, matching resources with users, users with codebooks and codewords for each resource layer, until all users are matched with codebooks and corresponding codewords.

7. A computer device, characterized in that, The computer device includes a memory and a processor. The memory stores a computer program, which, when executed by the processor, causes the processor to perform the steps of the multilayer blind detection method in the license-free SCMA with progressively decreasing complexity as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the multilayer blind detection method in the license-free SCMA with progressively decreasing complexity as described in any one of claims 1 to 5.

9. An information data processing terminal, characterized in that, The information data processing terminal includes a multi-layer blind detection system in the license-free SCMA with progressively decreasing complexity as described in claim 6.

Images