Initial synchronization method for ultra-dense cell-free large-scale MIMO (Multiple Input Multiple Output) control plane

Abstract

The invention provides an initial synchronization method for an ultra-dense cell-free large-scale MIMO (Multiple Input Multiple Output) control plane, which relates to the technical field of # imgabs0 # networks, and comprises the following steps: firstly, allocating TRP IDs (Transmit Receiving Point Identifiers) for replacing traditional PCI (Physical Cell Identifiers) to different TRP; secondly, based on the TRP ID, converting a synchronization signal distribution problem of the cloud wireless access network into a TRP ID distribution problem for executing a preset constraint condition; and converting the TRP ID allocation problem into a minimum k partition optimization problem, and solving the minimum k partition optimization problem to obtain an optimal allocation scheme of the TRP ID meeting the preset constraint condition. According to the invention, the strong interference of the ultra-dense cell-free large-scale multiple-input multiple-output broadcast channel can be effectively reduced, and the initial synchronous access success rate is improved.

Description

Technical Field

[0001] The present invention relates to The technical field of networks, in particular, relates to an initial synchronization method for an ultra-dense cell-free large-scale multi-input multi-output control plane. Background Art

[0002] With the development of network mobile communication technology, it is expected that by 2029, the number of mobile user subscriptions will reach 9.3 billion, and global mobile data traffic is expected to triple to 450 exabytes per month. It is expected that 5G will become the mainstream mobile access technology in the near future. As an upgrade to 5G, it is expected to provide enhanced network capabilities in the future, including Peak data rate, 10 per square kilometer 6 -10 8 To achieve these ambitious goals, ultra-dense cell-free massive multiple-input multiple-output (MIMO) has become a key enabling technology.

[0003] Ultra-dense cell-less massive MIMO (Multi-Input Multiple-Output) is a network consisting of a large number of transmission reception points (TRPs) and users distributed over a large area, known as a cloud radio access network (CF-MM). While the control plane is crucial for CF-MM, most research on CF-MM focuses on initial access and handover, neglecting initial synchronization. To address this issue, existing technologies have proposed using synchronization signals compliant with Long Term Evolution (LTE) and New Radio (NR) standards. LTE, the core technology standard for fourth-generation mobile communications, was developed by the Third Generation Partnership Project (3GPP) to improve the speed and efficiency of 3G networks. NR, the radio access technology standard for fifth-generation mobile communications, is also led by 3GPP and supports more flexible frame structures and higher frequency bands, optimized for scenarios such as enhanced mobile broadband and ultra-reliable low-latency communications. However, regardless of whether LTE or NR synchronization signals are used, the broadcast channels of ultra-dense cell-less massive MIMO suffer from strong interference, resulting in a high rate of initial synchronization failure. Summary of the Invention

[0004] In order to solve the problem that the broadcast channel of ultra-dense and cell-free large-scale multi-input and multi-output (MIMO) has strong interference, resulting in a high initial synchronization access failure rate, the present invention proposes an ultra-dense and cell-free large-scale multi-input and multi-output (MIMO) control plane initial synchronization method, which can effectively reduce the strong interference of the broadcast channel of ultra-dense and cell-free large-scale multi-input and multi-output (MIMO) and improve the initial synchronization access success rate.

[0005] In order to achieve the above technical effects, the technical solutions of the present invention are as follows: A method for initial synchronization of a large-scale multi-input multi-output control plane in an ultra-dense cell-free environment comprises the following steps: S1 to different transmission reception points TRP assigned to replace the traditional physical cell identifier PCI transmission reception point identifier TRP ID; S2 based on the transmission reception point identifier TRP ID, the cloud radio access network synchronization signal allocation problem is converted into the transmission reception point identifier TRP ID allocation problem of executing preset constraints; S3. Convert the transmission reception point identifier TRP ID allocation problem into a minimum k partition optimization problem, solve the minimum k partition optimization problem, and obtain an optimal transmission reception point identifier TRP ID allocation solution that meets the preset constraints.

[0006] Preferably, the cloud wireless access network includes a central processing unit and a plurality of transmission reception points TRP, and the number of the transmission reception points TRP is greater than the number of transmission reception point identifiers TRP ID.

[0007] Preferably, converting the synchronization signal allocation problem of the cloud radio access network into a transmission reception point identifier TRP ID allocation problem that executes preset constraints includes: S21. Mapping the transmission reception point identifier TRP ID to the synchronization signal of the cloud radio access network, the synchronization signal including the primary synchronization signal PSS and the secondary synchronization signal, and encapsulating the synchronization signal and the physical broadcast channel in several consecutive symbols to obtain a synchronization signal block; S22. The calculation expression for the transmission reception point identifier TRP ID is restored by the user equipment UE using the time position of the synchronization signal block as follows:

[0008] in, A value representing the transmission reception point identifier TRP ID, Indicates the number of groups into which the transmission reception point TRP is divided, Indicates the first parameter in the transmission reception point identification group, Indicates the second parameter in the transmission reception point identification group.

[0009] Preferably, the first parameter Transmitted by a physical broadcast channel, the location of the demodulation reference signal of the physical broadcast channel Determined by the transmission reception point identifier TRP ID.

[0010] Preferably, the preset constraints include: adjacent transmission reception points TRPs shall not use the same transmission reception point identifier TRP ID; two adjacent transmission reception points TRPs as switching targets of the current transmission reception point TRP shall not use the same transmission reception point identifier TRP ID; all transmission reception points TRPs participating in joint transmission in the coordinated multi-point transmission mode shall not have the same ID.

[0011] Preferably, converting the transmission reception point identifier TRP ID allocation problem into a minimum k partition optimization problem includes: S31. Constructing a weighted graph for cloud wireless access networks ,in represents the set of vertices to which the transmission reception point identifier TRPID is assigned, Represents the relationship between transmission receiving points TRP, setting a weighted graph The The vertex is the Transmission reception point TRP; S32. Determine the position of the demodulation reference signal according to the transmission reception point identifier TRP ID as follows:

[0012] in, Indicates remainder calculation; S33. Let the position of the demodulation reference signal This is equivalent to determining the time position of the primary synchronization signal PSS or the secondary synchronization signal SSS, and converting the transmission reception point identifier TRP ID allocation problem into a minimum k partition optimization problem as follows:

[0013] in, represents the cost function, Indicates taking the minimum value, Representing a weighted graph The edge weights of Represents a collection of partitions, Represents the first partitions, Indicates the The transmission reception point TRP is assigned to the A binary variable with partitions, Indicates the location of a transmission receiving point Assigned to A binary variable with 1 partition.

[0014] Preferably, solving the minimum k partition optimization problem includes: S41. The minimum k partition optimization problem is equivalent to the maximum k cut problem; S42. Converting the maximum k-cut problem into a quadratic unconstrained binary optimization problem; S43. Use a multi-operator heuristic algorithm and a parameterized local search method to solve the quadratic unconstrained binary optimization problem to obtain the optimal allocation solution for the transmission reception point identifier TRP ID.

[0015] Preferably, the mathematical expression of the maximum k-cut problem is as follows:

[0016] in, Indicates taking the maximum value.

[0017] Preferably, the mathematical expression of the quadratic unconstrained binary optimization problem is as follows:

[0018]

[0019]

[0020] in, P (.) represents the penalty term, Indicates the The transmission reception point TRP is assigned to the A binary variable with 1 partition.

[0021] The present invention also proposes an ultra-dense cell-free large-scale multi-input multi-output control plane initial synchronization system based on the method, comprising: A transmission reception point identifier allocation module, configured to allocate a transmission reception point identifier TRP ID to different transmission reception points TRP, which is used to replace a traditional physical cell identifier PCI; a problem conversion module, configured to convert the cloud radio access network synchronization signal allocation problem into a transmission reception point identifier TRP ID allocation problem that executes preset constraints according to the transmission reception point identifier TRP ID; A solution module is used to convert the transmission reception point identifier TRP ID allocation problem into a minimum k partition optimization problem, solve the minimum k partition optimization problem, and obtain an optimal transmission reception point identifier TRP ID allocation solution that meets the preset constraints.

[0022] Compared with the prior art, the beneficial effects of the technical solution of the present invention are: The present invention proposes a method for initial synchronization of an ultra-dense, cell-free, large-scale, multi-input, multi-output control plane. First, a transmission receiving point identifier (TRP ID) is allocated to different transmission receiving points (TRPs) to replace the traditional physical cell identifier (PCI), and the synchronization signal allocation problem of the cloud wireless access network is converted into a TRP ID allocation problem that executes preset constraints, thereby realizing a new allocation of synchronization signals. The transmission receiving point identifier (TRP ID) allocation problem is then converted into a k-th partition optimization problem to obtain an optimal allocation solution. This effectively solves the problem of high initial synchronization access failure rate caused by strong interference in ultra-dense, cell-free, large-scale, multi-input, multi-output broadcast channels, can significantly improve the success rate of initial synchronization access, and optimize system communication performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A schematic diagram of an initial access process defined by 3GPP proposed in an embodiment of the present invention is shown; Figure 2 A diagram showing a time and frequency mapping of a primary synchronization signal (PSS) / secondary synchronization signal (SSS) in long term evolution proposed in an embodiment of the present invention; Figure 3 A diagram showing the time and frequency mapping of synchronization signal blocks in NR proposed in an embodiment of the present invention; Figure 4 FIG2 is a diagram showing a synchronization signal block burst set configuration in a new radio (NR) interface proposed in an embodiment of the present invention; Figure 5 A diagram showing the frequency position of SSB in the new air interface proposed in an embodiment of the present invention; Figure 6 A schematic diagram showing a flow chart of an ultra-dense cell-free large-scale MIMO control plane initial synchronization method proposed in an embodiment of the present invention; Figure 7 A diagram showing a TRP ID conflict principle proposed in an embodiment of the present invention; Figure 8 A diagram showing the principle of TRP ID confusion during UE handover proposed in an embodiment of the present invention; Figure 9 A diagram showing the principle of coordinated TRP confusion coordinated multi-point transmission in a cloud radio access network proposed in an embodiment of the present invention; Figure 10 A block diagram of an ultra-dense cell-free large-scale MIMO control plane initial synchronization system proposed in an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0024] The accompanying drawings are for illustrative purposes only and are not to be construed as limiting this patent; It is understandable to those skilled in the art that some well-known contents may be omitted in the drawings; To facilitate understanding of this embodiment, first, the prior art information of this embodiment is introduced as follows: CF-MM has recently attracted widespread attention from researchers. While not a completely new technology, it develops from the convergence of ultra-dense networks (UDN), distributed antenna systems (DAS), massive MIMO, and coordinated multipoint (CoMP), inheriting their advantages, such as improved coverage, high spectral efficiency (SE), and enhanced energy efficiency (EE). Previous studies have demonstrated the superiority of CF-MM in these areas. For example, HQ Ngo et al. demonstrated that CF-MM can increase single-user throughput by nearly five times at the 95th percentile compared to small cell solutions. Furthermore, the study found that CF-MM significantly improves single-user throughput while more than doubling radiated energy efficiency. Considering user quality of service (QoS) requirements, CF-MM can improve energy efficiency by an order of magnitude compared to co-located massive MIMO.

[0025] Previous literature has primarily focused on the data plane of ultra-dense CF-MM networks, focusing on key aspects such as spectral efficiency, energy efficiency, and throughput. In contrast, research on the control plane, such as initial access, synchronization, and calibration, has been somewhat underdeveloped. Indeed, statistics from the past three years confirm this concern. According to the Institute of Electrical and Electronics Engineers (IEEE) database, over 750 journal articles published between 2023 and 2005 investigated CF-MM, demonstrating that CF-MM has attracted significant research attention. Regarding the ratio of existing research on the data and control planes, approximately 76% of these papers focus on the data plane, while only 24% on the control plane, indicating that data plane research dominates the relevant field. However, according to the Third Generation Partnership Project (3GPP) standards, modern wireless networks, such as 4G and 5G / 5G-A, consist of both a data and control plane. For the future 6G, this architecture will inherit previous generations of networks. Furthermore, CF-MM is a promising technology for 6G, making research on the CF-MM control plane absolutely essential for next-generation wireless networks.

[0026] Despite the critical importance of the control plane for cloud radio access networks (CF-MM), there is a paucity of literature in this area. Most studies focus on initial access, handover, and other aspects. By leveraging the clustering characteristics of user equipment (UE), Chen et al. studied the uplink initial access and interference suppression issues of CF-MM. Recent papers in the Journal of Wireless Communications explored the previously overlooked handover characteristics of CF-MM. Specifically, the former aims to extend the traditional handover concept to the more complex transmission reception point (TRP) and UE association scenarios in CF-MM, while the latter aims to control the number of handovers in user-centric CF-MM. All of these studies have not considered the initial synchronization issue of CF-MM.

[0027] The Cloud Radio Access Network (CF-MM) was proposed to overcome the shortcomings of ultra-dense networks (UDNs), such as high handover rates and uneven coverage. CF-MM advocates the use of architectures similar to distributed antenna systems (DAS) and coordinated multi-point transmission (CoMP), aiming to mitigate the edge effects inherent in cellular-based wireless networks. Generally speaking, CF-MM refers to a network consisting of a large number of transmission reception points (TRPs) and users distributed over a large area. Wu Hanqiang et al. introduced a typical CF-MM, which consists of a central processing unit and a large number of distributed TRPs. All TRPs share the same cell. For each user, the TRP transmits coherent data to that user. This is similar to Scenario 4 of CoMP, except that CF-MM has a larger number of distributed TRPs.

[0028] Similar to DAS and CoMP, all TRPs in CF-MM belong to the same physical cell. The primary goal is to eliminate cell edges, thereby providing more uniform coverage. These characteristics are based on the fact that all TRPs share the same cell identifier, the physical cell identifier (PCI). When all TRPs in CF-MM share the same PCI, this poses significant challenges for user access to the network. This is even more pronounced in ultra-dense CF-MM, where TRPs are densely and widely deployed. To understand this issue, we first describe the function of PCI in current wireless networks.

[0029] Whether it is Long Term Evolution (LTE) or New Radio (NR), PCI is a crucial system parameter. In current wireless networks, cells broadcast PCI to all UEs. Only after the UE receives this parameter can it access the network. standard, Closely related to the initial synchronization signal, these signals include the primary synchronization signal (PSS) and the secondary synchronization signal (SSS). Based on the above analysis, TRPs in CF-MM share the same PCI, and therefore, the PSS / SSS are also identical. Consequently, all TRPs broadcasting the same synchronization signal will cause strong interference in the broadcast channel. When the interference exceeds the threshold, the UE's synchronization success rate will drop to zero, and the UE will be unable to access the network.

[0030] The technical solution of the present invention is further described below with reference to the accompanying drawings and embodiments.

[0031] Example 1 To further illustrate the ultra-dense cell-free massive MIMO control plane initial synchronization method proposed in this embodiment, we first analyze the incompatibility challenge between NR / LTE initial synchronization and CF-MM. To understand the incompatibility between the LTE / NR standard and CF-MM and why this issue is worth studying, we first briefly describe the initial access process. We then delve deeper into the details of the LTE / NR initial synchronization process to explain the issue. This will help us identify the root cause of the problem and provide a suitable solution.

[0032] According to the 3GPP standards, a UE must perform certain steps before it can receive and transmit data. These steps include cell search / reselection, receiving system information, and random access. The complete process is called LTE / NR initial access. Figure 1 As shown, a brief description is as follows: Cell search and selection: User equipment (UE) detects physical signals and channels to select a cell.

[0033] System information reception: User Equipment (UE) configuration channel and maps it to to receive the Master Information Block (MIB).

[0034] Random access: User equipment (UE) establishes uplink synchronization and obtains a specific identifier for wireless access communication.

[0035] In the first step, the user equipment (UE) attempts to find an acceptable cell on any public land mobile network (PLMN) by searching all supported frequencies. This step includes a series of synchronization phases, also known as initial synchronization, during which the UE determines time and frequency parameters. These parameters are crucial for the UE to demodulate downlink signals and transmit uplink signals with correct timing.

[0036] In Long Term Evolution (LTE) or New Radio (NR), the initial synchronization signal is strictly associated with a physical cell, which is represented by a physical cell identity (PCI), which is defined as follows

[0037] in They are the physical layer cell identification group and the physical layer identification within the physical layer cell identification group. In fact, It is a sector ,for and New Wireless (NR), The value of is defined as follows:

[0038] From formula (1), we can see that if we want to obtain the physical cell identifier (PCI) of the base station, we must obtain and For Long Term Evolution (LTE) / New Radio (NR), these two parameters are closely related to synchronization signals, namely the Primary Synchronization Signal (PSS) and the Secondary Synchronization Signal (SSS). Since the generation methods of the Primary Synchronization Signal (PSS) and Secondary Synchronization Signal (SSS) in Long Term Evolution (LTE) and New Radio (NR) are different, we will explain them in detail respectively.

[0039] For the Primary Synchronization Signal (PSS) and Secondary Synchronization Signal (SSS) in Long Term Evolution (LTE), the Primary Synchronization Signal (PSS) is defined as , which is generated by the frequency domain Zadoff-Chu (ZC) sequence as follows:

[0040] in yes The root sequence index of is given in Table I.

[0041] Table I Root sequence index of primary synchronization signal

[0042] The secondary synchronization signal (SSS) of length 62 is the interleaved concatenation of two M sequences of length 31, denoted here as SSS1 and SSS2, which differ between subframes 0 and 5.

[0043] (4) in , superscript and The following rules are used to identify the physical cell identity (PCI): The first parameter is derived from

[0044]

[0045]

[0046]

[0047] In formula (3) and formula (4), and Defined as M sequence There are two different cyclic shifts of

[0048]

[0049] in ,here Defined as

[0050] The initial conditions are .

[0051] In frequency division duplex (FDD) mode, the primary synchronization signal (PSS) is transmitted in the last symbol of time slot 0 and time slot 10, while the secondary synchronization signal (SSS) is transmitted one symbol in advance.

[0052] In time division duplex (TDD) mode, PSS is sent in the third symbol of time slot 2 and time slot 12, while SSS is sent three symbols in advance. In the frequency domain, the base station maps the synchronization signal to 63 subcarriers arranged symmetrically around the DC carrier and punctures the middle elements, such as Figure 2 shown.

[0053] Based on the above description, initial synchronization helps the user equipment (UE) obtain basic orthogonal frequency division multiplexing (OFDM) system parameters, namely symbol and frame timing, carrier frequency and sampling clock, which are obtained by demodulating the synchronization signals (ie, PSS and SSS).

[0054] First, the UE blindly demodulates the PSS. At this point, we can obtain , however, we cannot determine the frame timing because PSS is sent in both the first and second half of a frame. Secondly, it can demodulate SSS from a specific position, and then according to the sequence of SSS, the current subframe moment can be determined because the SSS sequence of subframe 0 and subframe 5 is different. This also helps us obtain After demodulating PSS and SSS, we can calculate PCI according to equation (1).

[0055] For Cloud Radio Access Network (CF-MM) networks that use synchronization signals similar to those used in Long Term Evolution (LTE), due to equations (2), (3), and (4), different transmission reception points (TRPs) have the same PCI, which results in the same PSS and SSS sequences. Furthermore, for LTE, these TRPs transmit the PSS and SSS on the same radio resource blocks (RBs), so their synchronization signals interfere with each other. This interference is stronger and more pronounced in CF-MM networks due to the large number of TRPs.

[0056] Conclusion: Since TRPs in CF-MM have the same PCI, if the same synchronization process as LTE is used in CF-MM, the synchronization signal interference between different TRPs will inevitably be strong.

[0057] For the PSS and SSS in New Radio (NR), we will first briefly introduce the synchronization signals in NR and then analyze the challenges faced when using them in cell-free massive multiple-input multiple-output (MIMO) systems. It is generated based on an M sequence of length 127, which is defined as follows:

[0058]

[0059]

[0060] in , the initial conditions are:

[0061] Secondary Synchronization Signal (SSS) for New Radio (NR) Defined as:

[0062]

[0063]

[0064]

[0065]

[0066] in

[0067]

[0068] The initial conditions are:

[0069]

[0070] From equations (8) and (9), we can see that the primary synchronization signal (PSS) sequence is The sector ID is determined by the physical cell identifier (PCI), while the secondary synchronization signal (SSS) is determined by the PCI group number. and sector ID Although the formula for New Radio (NR) differs from that for Long Term Evolution (LTE), the PSS / SSS is mapped one-to-one to the PCI. That is, when the PCIs of two cells are the same, the PSS / SSS sequences are also the same.

[0071] Although the synchronization signals of different Transmit Reception Points (TRPs) are identical, they do not interfere with each other if they are transmitted in different time-frequency domains. The problem is that the time-frequency of the synchronization signal is highly correlated with the PCI, so different TRPs have no choice but to send the PSS / SSS on the same time-frequency resources. We will explain this process in detail below.

[0072] To improve the efficiency of the user equipment (UE) synchronization process, in the new radio (NR), the PSS / SSS and physical broadcast channel (PBCH) are encapsulated into 4 consecutive symbols, called synchronization signal blocks (SSBs), such as Figure 3 As shown. In the frequency domain, PSS and SSS are mapped to 127 consecutive subcarriers. For SSS, there are 8 unused subcarriers below it and 9 unused subcarriers above it. In the time domain, PSS and SSS are transmitted in the first and third symbols respectively. PBCH occupies two complete orthogonal frequency division multiplexing (OFDM) symbols, namely the second and fourth symbols, with a span of 240 subcarriers. In the third OFDM symbol, SSS spans 48 subcarriers above and below. The demodulation reference signal (DMRS) of PBCH occupies 144 resource elements (REs), and the rest is used for PBCH payload. The frequency position of PBCH DMRS depends on PCI, that is, DMRS The location through PCI mode 4 is obtained.

[0073] Base station (BS) with periodic Most broadcast A collection of SSBs. Here It is related to the frequency and its value can be 4, 6, or 8. Can be 5, 10 (default), 20, 40, 80 and The transmission of SSB in the SS Burst Set is limited to one In the window, such as Figure 4 shown.

[0074] In summary, if different TRPs use the same synchronization signal and choose to send the synchronization signal in the maximum number, a synchronization signal conflict will occur in the time domain.

[0075] Regarding the frequency domain, such as Figure 5 As shown in Figure 2, the frequency position of the SSB is determined by the parameter absoluteFrequencySSB, which represents the center frequency of the SSB block. According to the literature

[22] , this parameter is used for the serving cell, which means that when the serving cell does not change, the parameter remains unchanged. For co-sited multipoint (CF-MM), all TRPs belong to the same cell, that is, the same serving cell with a unique PCI. The absolute frequency SSB of all TRPs is the same, so different TRPs send their synchronization signals on the same frequency resources.

[0076] Conclusion: Similarly, if NR-like synchronization signals are used in CF-MM (Cooperative Full Spectrum Multi-Connectivity), different TRPs (Transmission Reception Points) will use the same synchronization sequence in the same time-frequency domain, i.e., the PSS (Primary Synchronization Signal) / SSS (Secondary Synchronization Signal) from different TRPs will interfere with each other.

[0077] Our goal is to design a novel synchronization scheme, namely, to propose an initial synchronization method for ultra-dense cell-free large-scale MIMO control plane. Figure 6 , including the following steps: S1 to different transmission reception points TRP assigned to replace the traditional physical cell identifier PCI transmission reception point identifier TRP ID; S2. Based on the transmission reception point identifier TRP ID, converting the synchronization signal allocation problem of the cloud radio access network into a transmission reception point identifier TRP ID allocation problem that executes preset constraints; the cloud radio access network includes a central processing unit and a plurality of transmission reception points TRP, the number of the transmission reception points TRP being greater than the number of transmission reception point identifiers TRP ID; S3. Convert the transmission reception point identifier TRP ID allocation problem into a minimum k partition optimization problem, solve the minimum k partition optimization problem, and obtain an optimal transmission reception point identifier TRP ID allocation solution that meets the preset constraints.

[0078] This method should ensure that CF-MM is compatible with the current NR standard, while also considering the future Based on the above analysis, if NR synchronization signals are to be used in CF-MM, a feasible approach is to have different TRPs broadcast their respective PSS / SSS at different times, i.e. one TRP broadcasts each PSS / SSS in each cycle. Send PSS / SSS once. The challenge lies in determining the optimal time slot for each TRP within each cycle to minimize interference. Solving this problem requires considering actual deployment scenarios and ensuring that the timing and allocation strategies are adaptable to different network configurations while effectively reducing interference.

[0079] exist In the network, physical cell identities (PCIs) are assigned based on a predefined set provided by Operations and Maintenance (OAM). In practice, PCI assignment follows a specific scenario, and once assigned, each cell broadcasts its synchronization signal using the assigned PCI. However, in the CF-MM architecture, where there are no traditional cells, the TRP ID will serve the equivalent of the PCI. Therefore, we recommend using the TRP ID instead of the PCI and assigning the PSS / SSS to each TRP based on its TRP ID. It's important to note that this isn't just a simple replacement; the distinction will become clear shortly.

[0080] A key question arises: how to transmit the TRP ID to the UE (user equipment)? Given that the PCI is communicated to the UE via the PSS / SSS, we transform the cloud radio access network synchronization signal allocation problem into a transmission reception point identifier (TRP ID) allocation problem that enforces predefined constraints, including: S21. Mapping the transmission reception point identifier TRP ID to the synchronization signal of the cloud radio access network, the synchronization signal including the primary synchronization signal PSS and the secondary synchronization signal, and encapsulating the synchronization signal and the physical broadcast channel in several consecutive symbols to obtain a synchronization signal block; S22. The calculation expression for the transmission reception point identifier TRP ID is restored by the user equipment UE using the time position of the synchronization signal block as follows: (11) in, A value representing the transmission reception point identifier TRP ID, Indicates the number of groups into which the transmission reception point TRP is divided, Indicates the first parameter in the transmission reception point identification group, Indicates the second parameter in the transmission receiving point identification group. and the second parameter , we designed the following signaling to carry them: for : It can be clearly determined by the SSB time position (i.e., SSB index). Therefore, if the UE decodes the SSB index, then this parameter is determined.

[0081] Another parameter :It can be transmitted by PBCH (Physical Broadcast Channel). For the current , the input bit sequence length of the PBCH payload is 32, and the output coded bit sequence length is 864. The PBCH payload uses polar coding. According to the polar coding process, we can add up to 9 bits to the input bit sequence without extending the total output coded bit sequence length. Therefore, the coded bit sequence is also 864, and It can be equivalent to 512.

[0082] Since the first parameter Transmitted by the physical broadcast channel, we also recommend using TRP ID instead of PCI to determine the location of DMRS (Demodulation Reference Signal) of PBCH in CF-MM ,Location Determined by PCI, that is, PCI mode 4, but in CF-MM, all TRPs share the same PCI, then interference between different TRPs is inevitable, so we recommend using TRP ID to determine the location of DMRS It is a suitable alternative to CF-MM. By leveraging the above ideas, we ensure that both PCI and TRP ID can be transmitted to the UE in a compatible manner using traditional NR procedures.

[0083] Based on the above analysis, the number of Transmission Reception Point (TRP) IDs (TRP IDs) is limited to approximately 4096. However, the number of TRPs deployed in the Cloud Radio Access Network (CF-MM) is much greater. Therefore, it is inevitable that different TRPs will reuse the same TRP ID. In practice, we need to consider some preset constraints.

[0084] The preset constraints include: adjacent transmission reception points TRPs shall not use the same transmission reception point identifier TRP ID; two adjacent transmission reception points TRPs as switching targets of the current transmission reception point TRP shall not use the same transmission reception point identifier TRP ID; all transmission reception points TRPs participating in joint transmission in the coordinated multi-point transmission mode shall not have the same ID.

[0085] First, we should avoid using the same ID for adjacent TRPs, which will cause TRP conflict problems, such as Figure 7When this happens, the user equipment (UE) cannot access the network because it cannot distinguish from which access point its initial access starts. Secondly, we should also avoid the current TRP's two adjacent handover target TRPs using the same ID or having Figure 8 The same modulus value as shown in . When this happens, the mobile UE will experience TRP switching problems simply because it cannot decide its switching target during the switching process, which will lead to TRP switching failures and call drops in the overlapping area. We realize that TRP conflict and confusion are similar to the physical cell identity (PCI) allocation problem, however, TRP ID and PCI have different functions and constraints. Finally, we should consider the constraints of the coordinated multi-point transmission (CoMP) mode [6]. In CoMP mode, all TRPs participating in the joint transmission should have different IDs, otherwise the UE cannot distinguish between different measurement reports, such as channel state information (CSI) measurements. These TRPs are called CoMP sets. Figure 9 As shown in the figure, there are two TRPs participating in the joint transmission. However, for traditional CF-MM, all TRPs can participate in the cooperative transmission. However, in practice, due to the bandwidth limitation of the fronthaul link, it is very challenging to let all TRPs participate in the cooperative joint transmission. Therefore, we consider that the number of TRPs in a CoMP set is limited, such as 3 or 4.

[0086] In summary, the synchronization signal design for the Cloud Radio Access Network (CF-MM) translates into a TRP ID assignment problem. Our goal is to minimize synchronization signal interference. Furthermore, we must avoid physical broadcast channel (PBCH) demodulation reference signal (DMRS) collisions, TRP ID conflicts, handover confusion, and CoMP confusion. Considering all these constraints, we first establish a mathematical model and then design an efficient solution to assign IDs to all TRPs in the CF-MM.

[0087] We design a mathematical model for the synchronization signal of the Cloud Radio Access Network (CF-MM). The entire network consists of a central processing unit (CPU) and TRPs, the maximum number of available TRP IDs is , we believe that the number of TRPs is greater than the number of available IDs, i.e. It is worth noting that this is consistent with actual deployments. In practice, a typical wireless network service area has about 10,000 base stations. When the same area is served by CF-MM, the number of TRPs exceeds 10,000, which is much larger than the maximum number of available TRP IDs.

[0088] Converting the transmission reception point identifier TRP ID allocation problem into a minimum k partition optimization problem includes: S31. Constructing a weighted graph for cloud wireless access networks ,in represents the set of vertices to which the transmission reception point identifier TRPID is assigned, Represents the relationship between transmission receiving points TRP, setting a weighted graph The The vertex is the Transmission reception point TRP; In S31, the topology of the cloud radio access network (CF-MM) is modeled as a weighted graph using Represents that the vertex set Indicates the allocation of TRP ID, edge set Represents the relationship between TRPs. Since the total number of TRPs is ,but . Design Vertex in TRP in CF-MM Because we want to assign IDs to different TRPs according to (11), that is, TRPs are divided into We use the set represents each group, so there is . TRPs in different groups have different modulus values, so the synchronization interference between them can be ignored. However, for those TRPs in the same group, the primary synchronization signal (PSS) / secondary synchronization signal (SSS) of different TRPs are the same because they have the same TRP ID, so there is synchronization interference between these TRPs. In order to improve the synchronization success rate, we should minimize the synchronization interference between different TRPs, which is our design goal. By Defined as a partition set, Defined as Transmission Reception Point (TRP) and Transmission Reception Point (TRP) The synchronization interference between The edge weight of . For each vertex and each partition , define binary variables , if the transmission reception point (TRP) Assigned to a partition , then the variable is 1, otherwise 0.

[0089] S32. Determine the position of the demodulation reference signal according to the transmission reception point identifier TRP ID as follows: (12) in, Indicates remainder calculation; when When the demodulation reference signal (DMRS) position This is equivalent to determining the time position of the primary synchronization signal (PSS) / secondary synchronization signal (SSS). Therefore, the demodulation reference signal (DMRS) collision problem is equivalent to the synchronization interference problem. Then the transmission reception point (TRP) ID allocation problem under consideration can be transformed into a minimum k partition optimization problem as follows: (13a) in, represents the cost function, which is used to represent the total interference between different groups. Indicates taking the minimum value, Representing a weighted graph The edge weights of Represents a collection of partitions, Represents the first partitions, Indicates the The transmission reception point TRP is assigned to the A binary variable with partitions, Indicates the location of a transmission receiving point Assigned to A binary variable with 1 partition.

[0090] The solving of the minimum k partition optimization problem includes: S41. The minimum k partition optimization problem is equivalent to the maximum k cut problem; In S41, the physical meaning of the maximum k-cut problem is to maximize the interference between different transmission reception point (TRP) groups. Since the primary synchronization signal (PSS) / secondary synchronization signal (SSS) used in different transmission reception point (TRP) groups are orthogonal in the code domain, the interference is very low.

[0091] Since the Max-k-cut problem has a fast and efficient solution, we transform the problem under consideration into an equivalent Max-k-cut problem formulated using binary quadratic optimization (BQO). The Max-k-cut problem is transformed into a quadratic unconstrained binary optimization problem, specifically including:

[0092]

[0093]

[0094]

[0095] (14d)

[0096] Where (14b), (14c) and (14d) are the constraints to avoid transmission reception point (TRP) ID conflict, handover confusion and coordination confusion respectively. represents the set of adjacent cells, Indicates the handover cell set, Represents a set of coordinated multi-point cells.

[0097] When the number of partitions is greater than 2 (i.e. 2), it has been proven that the minimum k partition (MkP) problem and the maximum k cut (Max-k-cut) problem are both It is very useful to transform problem (14) into an unconstrained problem. The equivalent penalty term corresponding to (14b) or (14c) is:

[0098] in is a positive scalar, then problem (14) can be rewritten as

[0099]

[0100] (16b)

[0101] Similarly, the equivalent penalty term corresponding to (16b) is

[0102] Then question (16) is transformed into:

[0103]

[0104]

[0105]

[0106] Constraint (18b) can be viewed as a linear equation, and the corresponding penalty term is:

[0107] Finally, problem (18) is equivalent to the following unconstrained form:

[0108]

[0109]

[0110] Now, the original problem (14) with constraints is transformed into a quadratic unconstrained binary optimization (QUBO) problem. Although the QUBO problem can be solved using readily available software such as Gurobi or Matlab, we cannot easily use these commercial software to solve problem (20) because the problem under consideration is very complex with more than 10,000 vertices.

[0111] S43. Use a multi-operator heuristic algorithm and a parameterized local search method to solve the quadratic unconstrained binary optimization problem to obtain the optimal allocation solution for the transmission reception point identifier TRP ID.

[0112] The algorithm proposed in S43 consists of two parts, as shown in Table II. The first part consists of the steps in Table II The second part is the last step, which is step 5 in Table II, and is based on the parameter local search maximum k-cut (PLS) algorithm. It can be regarded as a post-processing method to obtain a better solution. Its task is to find a solution that changes the most. A better solution for vertex partitioning (actually ), if such a partition exists; otherwise, the given partition is c-optimal. In the following, we will find that the proposed method can optimally solve the considered problem.

[0113] Table II Proposed algorithms for solving the TRP ID partitioning problem in CF-MM (Collaborative Fog Computing-Mobile Edge Computing)

[0114] In this embodiment, first, by allocating transmission receiving point identifiers TRP IDs that replace traditional physical cell identifiers PCIs for different transmission receiving points TRPs, and converting the synchronization signal allocation problem of the cloud wireless access network into a TRP ID allocation problem that executes preset constraints, a new allocation of synchronization signals is achieved. Then, the transmission receiving point identifier TRP ID allocation problem is converted into a k-th partition optimization problem to solve the optimal allocation solution, which effectively solves the problem of high initial synchronization access failure rate caused by strong interference in ultra-dense cell-free large-scale multi-input and multi-output broadcast channels, can significantly improve the success rate of initial synchronization access, and optimize system communication performance.

[0115] Example 2 This embodiment simulates and verifies the ultra-dense cell-free large-scale MIMO control plane initial synchronization method described in the above embodiment. In our simulation, we used the OpenCellID dataset. This dataset provides key information for each cell, such as longitude, latitude, cell ID, and mobile country code (MCC). Specifically, in our simulation, we used LTE and NR cell data in a certain port area, with an MCC of 454. The maximum transmit power of each transmission reception point (TRP) was set to 40dBm. The total number of cells in the port area was approximately 10,000, and the minimum distance between cells was approximately 100 meters. We adopted a multi-slope path loss model.

[0116] In our simulation, each cell represents a Transmitter-Receiver Point (TRP) in the CF-MM. We compared the performance of the proposed method with the Quadratic Unconstrained Binary Optimization (QUBO) method and the Tabu Search method. As mentioned above, the proposed method is The proposed method combines the parametric local search (PLS) method with the traditional one. To evaluate these methods, four typical cases representing different areas of a port were selected. The areas of these cases range from small to large, including a street, an administrative district, an island, and finally an entire city. Specifically, the smallest area is a street, which is served by approximately 200 transmitters and receivers, while the coverage of an administrative district requires approximately 1,000 transmitters and receivers. An island is served by approximately 2,000 transmitters and receivers, and the entire city, including four areas, is covered by 10,000 transmitters and receivers.

[0117] The overall performance of different methods is shown in Table III. For the case where the number of transmitters (TRP) is small, such as 200 or 1000, we can find that if the interference (i.e. dual The value of ) is the most concerned indicator, then the proposed method, MOH + PLS (Multi-objective Heuristic Algorithm + Path Reconnection Algorithm), can achieve the best results. As can be seen from Table III, the tabu search takes less time to find a suitable solution. However, when the number of TRPs is 1000, it also violates some constraints. As a compromise, the performance of MOH (Multi-objective Heuristic Algorithm) is a good choice because its results are good enough and the time consumption is not very high. The PLS (Path Reconnection Algorithm) method can improve Solution with a large number of TRPs. When the number of TRPs is large, the simulation time of the BUBO algorithm is very long, for example, more than one day. Therefore, when the number of TRPs is greater than 2000, it is not a suitable method for solving the considered problem. Therefore, detailed results about this method are meaningless and we do not provide them in this table.

[0118] Table III Performance comparison between different methods

[0119] Secondly, we give the convergence behavior of the proposed method and the previous method. From Table IV, we can see that when the number of vertices is small, the convergence speed is faster and the proposed method can obtain the optimal solution because The value of is the same as that obtained by integer linear programming (ILP)

[31] . This verifies the optimality of the MOH solution in the case of low TRP number.

[0120] Table IV Performance comparison of different methods when the number of temporary reference points (TRP) is 200

[0121] We then show the performance of different methods when the number of TRPs (transmission receiving points) is large (e.g., 10,000), as shown in Table V. For a large number of TRPs, it is infeasible to find a solution to the integer linear programming (ILP) in a finite time, so we use an upper bound to verify the effectiveness of the proposed method. As shown in Table III, the best solutions obtained by MOH and MOH + PLS are 1330493 and 1337145, respectively. The difference between these values ​​and the upper bound (1350516) is approximately This shows that The difference between the final solution and the optimal solution is less than 2%. For the sake of brevity, we do not show the results of other cases.

[0122] Table V Performance comparison of different methods when the number of TRPs (transmission reception points) is 2000

[0123] In our final analysis, we compared the synchronization success rates of various methods. Table VI shows that the synchronization success rate is very low, below 30%, resulting in poor initial access rates and a poor user experience. Therefore, it is clear that NR (New Radio)-style synchronization signals are not suitable for CF-MM (Cellular-Free Massive Multiple-Input Multiple-Output) networks without significant modifications. In the results for Tabu Search (Tabu) and MOH&PLS, the synchronization signal is based on the design proposed in Section 3. For a scenario involving 10,000 TRPs (Transmission Receive Points), Table VI highlights that our proposed method outperforms other methods, thus validating its effectiveness in enhancing initial synchronization in CF-MM networks.

[0124] Table VI Synchronization rates of different methods when the number of TRPs (transmission reception points) is 10,000

[0125] In this embodiment, we study the initial synchronization problem in the control plane of CF-MM (Cellular-Free Massive Multiple-Input Multiple-Output) networks. Existing Long Term Evolution (LTE) and New Radio (NR) synchronization signals are not suitable for CF-MM deployments. To address this problem, we first propose a new synchronization signal designed specifically for CF-MM. Then, we reformulate the synchronization problem as a TRP (Transmission Receive Point) ID allocation problem. Finally, we solve this problem by integrating MOH and PLS methods. Simulation results show that when the number of TRPs is small, the proposed method achieves near-optimal performance, and the deviation of the solution from the optimal value is less than In future work, we aim to improve the scalability of the method in scenarios where the number of TRPs exceeds 10,000. In addition, we plan to utilize the proposed method and Technological development for the future Network development an optimized synchronization program.

[0126] Example 3 See also Figure 10 This embodiment further proposes an ultra-dense cell-free large-scale multi-input multi-output control plane initial synchronization system implemented based on the method described in the above embodiment, including: A transmission reception point identifier allocation module, configured to allocate a transmission reception point identifier TRP ID to different transmission reception points TRP, which is used to replace a traditional physical cell identifier PCI; a problem conversion module, configured to convert the cloud radio access network synchronization signal allocation problem into a transmission reception point identifier TRP ID allocation problem that executes preset constraints according to the transmission reception point identifier TRP ID; A solution module is used to convert the transmission reception point identifier TRP ID allocation problem into a minimum k partition optimization problem, solve the minimum k partition optimization problem, and obtain an optimal transmission reception point identifier TRP ID allocation solution that meets the preset constraints.

[0127] In this embodiment, first, by allocating transmission receiving point identifiers TRP IDs that replace traditional physical cell identifiers PCIs for different transmission receiving points TRPs, and converting the synchronization signal allocation problem of the cloud wireless access network into a TRP ID allocation problem that executes preset constraints, a new allocation of synchronization signals is achieved. Then, the transmission receiving point identifier TRP ID allocation problem is converted into a k-th partition optimization problem to solve the optimal allocation solution, which effectively solves the problem of high initial synchronization access failure rate caused by strong interference in ultra-dense cell-free large-scale multi-input and multi-output broadcast channels, can significantly improve the success rate of initial synchronization access, and optimize system communication performance.

[0128] Obviously, the above embodiments of the present invention are merely examples for the purpose of clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. A person skilled in the art would be able to make other variations or modifications based on the above description. It is not necessary and impossible to enumerate all embodiments here. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the scope of protection of the claims of the present invention.

Claims

1. A method for initial synchronization of ultra-dense cell-free large-scale multi-input multi-output control plane, characterized in that: The following steps are involved: S1 to different transmission reception points TRP assigned to replace the traditional physical cell identifier PCI transmission reception point identifier TRP ID; S2 based on the transmission reception point identifier TRP ID, the cloud radio access network synchronization signal allocation problem is converted into the transmission reception point identifier TRP ID allocation problem of executing preset constraints; S3. Convert the transmission reception point identifier TRP ID allocation problem into a minimum k partition optimization problem, solve the minimum k partition optimization problem, and obtain an optimal transmission reception point identifier TRP ID allocation solution that meets the preset constraints.

2. The method for initial synchronization of ultra-dense cell-free large-scale multi-input multi-output control plane according to claim 1, characterized in that: The cloud wireless access network includes a central processing unit and a plurality of transmission reception points TRP, where the number of the transmission reception points TRP is greater than the number of transmission reception point identifiers TRP IDs.

3. The method for initial synchronization of ultra-dense cell-free large-scale multi-input multi-output control plane according to claim 1, characterized in that: The method of converting the synchronization signal allocation problem of the cloud radio access network into a transmission reception point identifier TRP ID allocation problem that executes preset constraints includes: S21. Mapping the transmission reception point identifier TRP ID to the synchronization signal of the cloud radio access network, the synchronization signal including the primary synchronization signal PSS and the secondary synchronization signal, and encapsulating the synchronization signal and the physical broadcast channel in several consecutive symbols to obtain a synchronization signal block; S22. Using the time position of the synchronization signal block, the user equipment UE recovers the transmission reception point identifier TRPID by the following calculation expression: in, A value representing the transmission reception point identifier TRP ID, Indicates the number of groups into which the transmission reception point TRP is divided, Indicates the first parameter in the transmission reception point identification group, Indicates the second parameter in the transmission reception point identification group.

4. The method for initial synchronization of ultra-dense cell-free large-scale multi-input multi-output control plane according to claim 3, characterized in that: The first parameter Transmitted by a physical broadcast channel, the location of the demodulation reference signal of the physical broadcast channel Determined by the transmission reception point identifier TRP ID.

5. The method for initial synchronization of ultra-dense cell-free large-scale multi-input multi-output control plane according to claim 4, characterized in that: The preset constraints include: adjacent transmission reception points TRPs shall not use the same transmission reception point identifier TRP ID; two adjacent transmission reception points TRPs as switching targets of the current transmission reception point TRP shall not use the same transmission reception point identifier TRP ID; all transmission reception points TRPs participating in joint transmission in the coordinated multi-point transmission mode shall not have the same ID.

6. The method for initial synchronization of ultra-dense cell-free large-scale multi-input multi-output control plane according to claim 5, characterized in that: Converting the transmission reception point identifier TRP ID allocation problem into a minimum k partition optimization problem includes: S31. Constructing a weighted graph for cloud wireless access networks ,in represents the vertex set to which the transmission reception point identifier TRP ID is assigned, Represents the relationship between transmission receiving points TRP, setting a weighted graph The The vertex is the Transmission reception point TRP; S32. Determine the position of the demodulation reference signal according to the transmission reception point identifier TRP ID as follows: in, Indicates remainder calculation; S33. Let the position of the demodulation reference signal This is equivalent to determining the time position of the primary synchronization signal PSS or the secondary synchronization signal SSS, and converting the transmission reception point identifier TRP ID allocation problem into a minimum k partition optimization problem as follows: in, represents the cost function, Indicates taking the minimum value, Representing a weighted graph The edge weights of Represents a set of partitions, Represents the first partitions, Indicates the The transmission reception point TRP is assigned to the A binary variable with partitions, Indicates the location of a transmission receiving point Assigned to A binary variable with 1 partition.

7. The method for initial synchronization of ultra-dense cell-free large-scale multi-input multi-output control plane according to claim 6, characterized in that: The solving of the minimum k partition optimization problem includes: S41. The minimum k partition optimization problem is equivalent to the maximum k cut problem; S42. Converting the maximum k-cut problem into a quadratic unconstrained binary optimization problem; S43. Use a multi-operator heuristic algorithm and a parameterized local search method to solve the quadratic unconstrained binary optimization problem to obtain the optimal allocation solution for the transmission reception point identifier TRP ID.

8. The method for initial synchronization of ultra-dense cell-free large-scale multi-input multi-output control plane according to claim 7, characterized in that: The mathematical expression of the maximum k-cut problem is as follows: in, Indicates taking the maximum value.

9. The method for initial synchronization of ultra-dense cell-free large-scale multi-input multi-output control plane according to claim 7, characterized in that: The mathematical expression of the quadratic unconstrained binary optimization problem is as follows: in, P (.) represents the penalty term, Indicates the The transmission reception point TRP is assigned to the A binary variable with 1 partition.

10. An ultra-dense cell-free large-scale multi-input multi-output control plane initial synchronization system implemented based on the method according to any one of claims 1 to 9, characterized in that: include: A transmission reception point identifier allocation module, configured to allocate a transmission reception point identifier TRP ID to different transmission reception points TRP, which is used to replace a traditional physical cell identifier PCI; a problem conversion module, configured to convert the cloud radio access network synchronization signal allocation problem into a transmission reception point identifier TRP ID allocation problem that executes preset constraints according to the transmission reception point identifier TRP ID; A solution module is used to convert the transmission reception point identifier TRP ID allocation problem into a minimum k partition optimization problem, solve the minimum k partition optimization problem, and obtain an optimal transmission reception point identifier TRPID allocation solution that meets the preset constraints.

Images