Satellite-ground two-step random access enhancement method based on irregular repeated time slot ALOHA

Abstract

The invention provides a satellite-ground two-step random access enhancement method based on irregular repeated time slot ALOHA. The satellite-ground two-step random access enhancement method comprises the following steps: step 1, designing a random access process method; 2, preamble design and timing detection are carried out; step 3, a receiving end demodulation and interference elimination process; and 4, performing simulation verification and performance evaluation. According to the invention, an irregular repeated time slot ALOHA strategy, a double-preamble (2-rooted PRACH preamble) detection technology, an orthogonal pilot channel estimation technology, a continuous interference elimination technology and the like are introduced, so that a satellite-ground two-step random access system level method is designed. Simulation results show that the method can improve the throughput of satellite-ground random access, significantly reduce the access delay, and improve the access speed and the access capacity.

Description

[Technical field]

[0001] Aiming at the multi-user scenario of satellite-to-ground two-step random access, the present invention designs a satellite-to-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA. In the case of a high probability of multi-user data collision, the method significantly improves the single access success rate of the user by adopting interference elimination technology, thereby improving the capacity and speed of satellite-to-ground two-step random access. The present invention belongs to the field of communication and signal processing technology. [Background technology]

[0002] The satellite-to-ground communication system is a key link in realizing the vision of 6G space-air-ground integration. Among them, low-orbit satellites (LEO) and geostationary orbit satellites (GEO) have shown great potential application value in the field of satellite-to-ground communications. Among all types of orbital satellites, LEO satellites stand out with their advantages of low launch cost and short link transmission delay; while GEO satellites are less affected by the Doppler effect due to their stationary characteristics relative to the ground. The satellite-to-ground communication system can not only provide same-frequency supplementary coverage for areas such as urban-rural fringe areas and urban edges where ground cellular network coverage is insufficient, thereby improving communication quality, but also provide different-frequency coverage services for areas such as remote mountain villages and polar regions where ground networks are difficult to cover continuously, ensuring that communication needs in coverage blind spots are met. [1] .

[0003] Random access (RA) is the core step to achieve uplink and downlink time and frequency synchronization and establish a reliable communication link between user equipment (UE) and base stations or satellites. It usually occurs when the user equipment is turned on for the first time or re-searches and accesses a new base station after disconnecting from the original base station. For the communication application scenarios of low earth orbit (LEO) and geostationary orbit satellites (GEO), the satellite-to-ground random access mechanism has more stringent requirements on the success rate of a single access, aiming to minimize the time cost introduced by the random access process.

[0004] In the traditional random access mechanism, data packet transmission generally follows the Slotted ALOHA (SA) strategy. Within the coverage of the base station or satellite beam, when the user terminal has a communication demand, it will immediately initiate an access request to the base station without pre-establishing a link or allocating time and frequency resources. If the receiver cannot successfully demodulate due to a data packet collision, the system will retransmit according to the hybrid automatic repeat request (HARQ) mechanism. However, since the SA strategy fails to effectively utilize the collision data packets, its access throughput (Throughput) limit is only 36.8% [2] .

[0005] To solve this problem, the irregular repetition slottedALOHA strategy (IRSA) was proposed. This strategy allows the transmitter to replicate and send data packets multiple times in multiple time slots, while the receiver uses successive interference cancellation (SIC) to utilize the colliding data packets, thereby achieving up to 80% asymptotic throughput. Applying this method to random access scenarios of LEO (low earth orbit satellite) or GEO (geosynchronous orbit satellite) communications with large link delay can effectively reduce the number of retransmissions of user data packets and significantly reduce access delay.

[0006] Compared with terrestrial random access scenarios, GEO or LEO communication scenarios are often characterized by high altitude, high dynamics, and large coverage. Especially in LEO scenarios, random access preamble detection needs to overcome the error caused by large Doppler frequency shift and accurately estimate the large round-trip delay difference. At the same time, if the IRSA method is introduced for random access, more accurate channel estimation and compensation are required to ensure the success rate of demodulation and decoding after interference elimination. [3][4] However, the current random access technology cannot fully meet the special needs of satellite-to-ground random access scenarios. Therefore, how to appropriately improve the existing technology to ensure the success rate and robustness of satellite-to-ground random access still faces huge technical challenges.

[0007] References:

[0008] [1] Yan Fuli. Research on 5G+ satellite converged communication applications[J]. Yangtze River Information and Communication, 2021, 34(06): 22-24. 3GPP.

[0009] [2] Shi Wenxiao. Communication Network Theory and Application[M]. Peking University Press, 2009.

[0010] [3]Casini E.,De Gaudenzi R.,Herrero ODRContention resolutiondiversity slotted ALOHA

[0011] (CRDSA): An enhanced random access scheme for satellite access packetnetworks[J]. IEEE transactions on wireless communications, 2007, 6(4): 1408-1419.

[0012] [4]Liva G.Graph-based analysis and optimization of contentionresolution diversity slotted ALOHA

[0013] [J].IEEE Transactions on Communications,2010,59(2):477-487. [Summary of the invention]

[0014] The purpose of the present invention is to design a satellite-to-ground two-step random access system-level method for the random access process in geosynchronous orbit satellite or low-orbit satellite communication scenarios by introducing irregular repetitive time slot ALOHA strategy, dual preamble sequence (2-rooted PRACH preamble) detection technology, orthogonal pilot channel estimation technology and continuous interference elimination technology. Simulation results show that this method can improve the throughput of satellite-to-ground random access, significantly reduce access delay, and improve access speed and access capacity.

[0015] The present invention proposes a satellite-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA, comprising the following steps:

[0016] Step 1: Design of random access process method

[0017] In the existing two-step random access process, the user terminal and the base station need to perform at least two signaling interactions to complete the access, namely, MsgA containing the random access preamble and control information and MsgB containing the random access response and contention resolution information. Consider that in an area covered by a single beam, there are m users who choose to initiate random access in the same cycle after demodulating the downlink broadcast signal. If the random access cycle time is divided into n s timeslots, each of which contains one MsgA signaling, and the normalized user load of the system is defined as:

[0018]

[0019] Assuming that the number of users successfully accessing in one round is q, the system throughput is defined as:

[0020]

[0021] Among them, each user in this n s The control information in MsgA needs to record the time slot position pointers of these MsgAs, and the receiver needs to store these n sThe receiver can detect all the information of each time slot, including the time slot data that collided because some users selected the same preamble code. The receiver determines which users have successfully accessed and which users need to re-access based on the demodulation results of this data.

[0022] MsgA in each time slot contains random access preamble and control information, which are transmitted on the physical random access channel (PRACH) and the physical uplink shared channel (PUSCH). Its time domain structure is as follows: Figure 2 As shown in Figure 2. Since the MsgA of each user cannot reach the satellite receiver synchronously and there is a large Doppler frequency shift (such as Figure 3 As shown in the figure), different users have different distances from the receiving base station, which causes the starting point of the receiving window to not coincide with the signaling frame header. Therefore, the data packet cannot be demodulated directly, and preamble detection is required for time-frequency compensation.

[0023] Step 2: Preamble design and timing testing

[0024] To solve the uplink synchronization problem of user data, random access uses the access preamble to estimate the timing advance (TA). In four-step random access, the TA estimate is fed back to each user through the random access response (RAR), and then the timing adjustment is performed to make the subsequent PUSCH data arriving at the receiver basically synchronized. In two-step random access, the TA value can be used to locate the position of PUSCH in MsgA, thereby omitting the feedback and timing adjustment steps, such as Figure 4 shown.

[0025] The Zadoff-Chu (ZC) sequence has become a widely used leading sequence due to its good autocorrelation and cross-correlation characteristics. The ZC sequence is a constant amplitude complex sequence. zc , a ZC sequence x with u as the physical root u (n), its general expression is:

[0026]

[0027] Where n is the sequence symbol index, j is the imaginary part sign, and equation (3) is generally called the root sequence, which is known to both the transmitter and the receiver. For a round-trip delay of τ i The i-th user receives the leading sequence under an ideal noiseless channel and can be expressed as y pre (n) = x u (n-τ i), the received sequence is cross-correlated with the root sequence, and the delay power spectrum (Power Delay Profile, PDP) can be expressed as:

[0028]

[0029] in, represents the complex conjugate of the root sequence with u as the physical root, and k is the time index of the delayed power spectrum. The user delay τ can be detected by detecting the peak position of the PDP spectrum. i , thereby achieving uplink synchronization.

[0030] In the satellite-to-ground communication scenario, the high-speed movement of the satellite will cause a significant Doppler frequency shift effect, which seriously interferes with the accuracy of the receiver's preamble detection algorithm and has a significant impact on the synchronization performance and demodulation performance. To meet this challenge, the present invention introduces a dual-root preamble sequence method to achieve robust TA estimation and Doppler frequency shift estimation (see the subsequent implementation section for specific steps), which can significantly improve the access success rate of the system.

[0031] Step 3: Receiver demodulation and interference elimination process

[0032] 1. Receiver time-frequency compensation

[0033] Without loss of generality, consider the transmission of coded modulation sequences in the frequency domain under noisy channels. The transmitter and receiver structures are as follows: Figure 5 and Figure 6 As shown. The transmitter uses discrete Fourier transform extended orthogonal frequency division multiplexing to send data, that is, the data bits to be transmitted are first processed by serial-to-parallel conversion and then digitally modulated. Next, the modulated symbols undergo discrete Fourier transform precoding and are mapped to different subcarriers to achieve frequency domain modulation. Finally, these frequency domain signals are converted back to the time domain through inverse Fourier transform, and a cyclic prefix is ​​added to complete the preparation for data transmission. The receiver is the inverse process of the above steps, which will not be repeated here. Let x[n] be the time domain baseband transmission signal after OFDM modulation. If N subcarriers are used for transmission, the normalized Doppler frequency shift of the transmission link is ε (see equation (12) for details), then the baseband sampling signal r[n] of the receiver can be expressed as:

[0034]

[0035] Where h[n] is the link channel coefficient and z[n] is Gaussian white noise.

[0036] The preamble sequence arrives at the receiver after a certain delay τ. Due to the existence of the cyclic prefix (CP), the output sequence after removing the cyclic prefix will be the cyclic shift of the original transmitted sequence:

[0037] y[n]=r[(n-τ)modN](6)

[0038] Where N is the sampling sequence length of an OFDM symbol after removing the cyclic prefix, that is, the number of subcarriers, and τ is the delay. Before decoding, the time domain sequence needs to be converted to the frequency domain, so the time domain signal y[n] needs to be compensated for time and frequency to make it as close as possible to the ideal received sequence y without frequency deviation and delay. cmp [n]:

[0039]

[0040] To ensure the success rate of frequency domain demodulation and decoding. The preamble detection performed in step 2 can estimate the time delay and normalized Doppler frequency shift ε at the same time, thereby realizing the compensation process.

[0041] 2. Interference elimination based on irregularly repeated time slot ALOHA

[0042] The receiver needs to receive and store the data information of all time slots. During demodulation, the receiver first detects the preamble code of each time slot, determines whether it has collided, and records the estimated delay value of each time slot, and then starts to traverse demodulation and interference elimination: demodulate and decode the data of the non-colliding time slots, obtain the time slot position of the remaining data copies through the decoding results, and eliminate the interference of the copies in these time slots. When traversing to the last time slot, if there are still data in the time slot that has not been successfully demodulated, it is necessary to repeat the traversal. Repeat the above process many times until all time slots are demodulated, or the remaining undemodulated time slots are all collision time slots, at which time the demodulation process is completed. All users who failed to access need to re-initiate access in the next round of access. If each user repeats the data packet different times, different system throughputs can be obtained; if dynamic adjustment is performed during the access process, the overall access speed can be further improved. For details, please refer to the implementation plan.

[0043] 3. Channel estimation of time slots to be eliminated

[0044] To ensure that interference cancellation can be performed efficiently, the receiver needs to perform channel estimation for all time slots where each user's MsgA is located. Since different users may choose the same time slot, the pilots used for channel estimation must be orthogonal, and the orthogonal pilots select a row or column of the Hadamard matrix. There is a time delay and Doppler frequency shift in the channel, so the receiver needs to perform corresponding time-frequency compensation on the pilot sequence before channel estimation can be performed. For details of the specific channel estimation calculation process, please refer to the implementation plan.

[0045] Step 4: Simulation verification and performance evaluation

[0046] Consider LEO communication scenarios and GEO communication scenarios. When the elevation angle is large, the transmission link is mainly dominated by direct components, and the reflected components are not obvious. The channel of the present invention is set as a Rice channel with a large direct component. In order to ensure the demodulation and decoding performance, the present invention adopts low-density parity check code (Low Density Parity Check, LDPC) as the encoding and decoding method, and adopts multi-carrier modulation and demodulation technology; in order to compare the access performance, the present invention simultaneously simulates and verifies the improvement of the access speed and capacity of the present invention and the existing two-step random access method, and statistically calculates the access speed, access capacity, access success rate, etc. in the simulation process. For details of the simulation process, please refer to the implementation steps.

[0047] The advantages and beneficial effects of the present invention are:

[0048] The present invention designs a satellite-to-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA for the satellite-to-ground two-step random access scenario. The core of the method is to repeatedly transmit MsgA data in the transmitter of the satellite-to-ground link, and integrate advanced interference elimination technology in the receiver and perform efficient demodulation, so as to achieve effective utilization of conflicting time slot data. This strategy can significantly improve the success rate of a user's single access, thereby enhancing the overall capacity and response speed of the satellite-to-ground two-step random access. In addition, the present invention also cleverly introduces a dual-root preamble sequence mechanism, which is designed to improve the detection success rate of the preamble code when facing a large Doppler frequency shift channel. At the same time, combined with the time-frequency compensated orthogonal pilot, the present invention realizes accurate estimation of multi-time slot channels, ensures high accuracy of each interference elimination and demodulation process, and thus stabilizes the reliability of access performance.

[0049] In summary, the present invention can significantly improve the success rate of two-step random access between satellite and ground in both low earth orbit and geosynchronous orbit systems, and effectively reduce the frequency of signaling interaction between users and satellites. This innovative achievement has great application potential and value in the field of satellite communications.

Brief Description of the Drawings

[0050] Figure 1 The present invention is implemented in an overall flow chart.

[0051] Figure 2 It is a two-step random access MsgA time domain structure.

[0052] Figure 3 Schematic diagram of uplink frame reception in the beam coverage area.

[0053] Figure 4 Schematic diagram of the receiver detection window.

[0054] Figure 5 , Figure 6Block diagram of DFT-s-OFDM transmitter and receiver.

[0055] Figure 7 This is the random access preamble structure used in the present invention.

[0056] Figure 8 It is a schematic diagram of eliminating interference of data packets of a receiver according to the present invention.

[0057] Fig. 9 This is the channel estimation method adopted by the present invention.

[0058] Fig.10 The program of the present invention comprises a simulation module.

[0059] Fig.11 The figure is a comparison of the first access throughput curves of the present invention and the baseline solution based on the SA strategy as a function of user load.

[0060] Fig.12 The figure is a comparison of the first access success rate curves of the present invention and the baseline solution based on the SA strategy as a function of user load.

[0061] Fig.13 The figure is a comparison of the curves of average signaling interaction times versus user load between the present invention and the baseline solution based on the SA strategy.

[0062] Fig.14 The figure is a comparison of the total rounds required for all users to access the present invention and the baseline solution based on the SA strategy with the user load variation curve.

[0063] Fig.15 The figure is a comparison of the access speed variation curves of the present invention and the baseline solution based on the SA strategy with the user load. [Specific implementation method]

[0064] In order to more clearly illustrate the purpose, technical solutions and advantages of the present invention, the following Figure 1-15 The present invention is further elaborated in detail. The described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, but should not be understood as limiting the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0065] The present invention proposes a satellite-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA, comprising the following steps:

[0066] Step 1: Design of random access process method

[0067] Consider a low-orbit satellite single-beam coverage area. Assume that there are m users in the area who demodulate the beam downlink broadcast signal and choose to initiate random access in the same period. The time domain length of MsgA is defined as a time slot length. The time period from when all users initiate random access to when they complete reception is divided into n s time slots, the normalized user load of the system is defined as:

[0068]

[0069] Assuming that the number of users successfully accessed in the first round is q, the system throughput is defined as:

[0070]

[0071] Figure 1 The implementation block diagram of the present invention is shown. Based on the basic framework of two-step random access, after the user obtains the cell timing information and the preamble format information, a preamble pool for use is generated. First, the user randomly selects the preamble root sequence in the preamble pool, and adds a cyclic prefix, a protection interval, etc. according to the current system parameters to form an access preamble. At the same time, the user generates a data packet of a certain code length to form MsgA signaling, in which the preamble (on the PRACH channel) and the data packet (on the PUSCH channel) of the same user are frequency-division multiplexed, and the selection of the user's time slot is not related to the selection of the preamble; in the time domain, the preamble and the data packet are time-division multiplexed, and there is a certain interval between the two to ensure that adjacent time slots do not interfere with each other, such as Figure 2 shown.

[0072] Then the user randomly selects several time slots to send the exact same MsgA, where the data part of MsgA contains the pointer of the time slot selected by the user. The number of repetitions of MsgA for each user is different, and different repetition situations are expressed as polynomials:

[0073]

[0074] The exponent β of each term of the polynomial represents the number of repetitions, and the coefficient λ β Represents the proportion of users who repeat this number of times to the total number of users, satisfying:

[0075]

[0076] Different repetition situations will bring different system throughput performance. The number of repetitions can be dynamically adjusted according to different user loads to ensure that the access efficiency is continuously optimized.

[0077] The satellite receives and stores the MsgA information of all accessed users, and introduces interference elimination technology for demodulation and decoding. For users who successfully access, the satellite receiver returns MsgB to the user terminal on the corresponding Physical Downlink Shared Channel (PDSCH) resource, which contains the information of the successfully accessed user and its TA estimation value.

[0078] Finally, the user terminal receives MsgB, detects and determines whether re-access is required. If the access is successful, the timing advance information TA can be used to adjust the timing for subsequent data transmission.

[0079] Step 2: Preamble design and uplink synchronization

[0080] There is a large Doppler shift in the satellite-to-ground communication link. The normalized Doppler shift ε is defined as the absolute Doppler shift f d Ratio to subcarrier spacing Δf:

[0081]

[0082] The decimal part of ε will lead to multiple peaks, causing false detection; the integer part will cause the peak of the delayed power spectrum to shift, causing false detection. Therefore, it is necessary to design the preamble for high-speed mobile scenarios. For the preamble code of multi-carrier transmission, the impact of Doppler frequency shift is also related to the absolute Doppler frequency shift and the subcarrier spacing.

[0083] The present invention uses dual preamble sequences as access preambles. The two preamble sequences have different physical roots, and both ZC sequences are root sequences without cyclic shift. In the actual transmission process, the two preamble sequences can be sent at different times and the same frequency (i.e., mode A), or at the same time and the same frequency (i.e., mode B). Compared with the traditional preamble, each sequence of the enhanced preamble uses two root sequences such as Figure 7 As shown, assuming that the two root sequences are:

[0084]

[0085] Among them, N zc Represents the length of the preamble sequence, and u1,u2 represent the two physical roots selected by the user. After the two preamble sequences pass through the Doppler frequency shift channel, the receiver uses the root sequences corresponding to the two preambles to calculate the PDP spectrum in a circular correlation manner, and obtains two peaks, whose positions are set to p1 and p2 respectively. After analysis and calculation, it can be seen that the positions of the two peaks are related to the normalized Doppler frequency shift ε, the physical roots u1,u2 and the actual normalized delay τ, satisfying:

[0086]

[0087] Where [ε] means the normalized Doppler shift is rounded to the nearest integer. represents the modular multiplicative inverse of u1,u2, that is, it satisfies:

[0088]

[0089] The above can estimate the normalized Doppler shift and delay They are:

[0090]

[0091] Equation (16) shows that the Doppler frequency shift can be roughly estimated by rounding and the normalized delay can be estimated more accurately by using two delayed power spectra. In order to make the normalized delay estimation accuracy of the above method as large as possible, u1 and u2 should be selected in pairs and satisfy the following relationship:

[0092]

[0093] The physical roots available for mode A satisfy:

[0094]

[0095] The physical roots available for mode B satisfy:

[0096]

[0097] In addition, when designing the frame structure, it is also necessary to consider the relationship between the user delay difference within the coverage of a single beam and the preamble length and preamble subcarrier spacing. Assume that the maximum user delay difference is RTD max , the subcarrier spacing is Δf, and the PDP spectrum peak position needs to be limited to the preamble length N zc To ensure that it can be detected, the following relationship should be satisfied between the three:

[0098]

[0099] Among them, the leading sequence length is N zc =839, the preamble subcarrier spacing is Δf = 1.25kHz, which can be used as the LEO system with a beam radius of R = 45km and the satellite-to-ground random access preamble with a radius of R = 60km. In addition, it is also necessary to max Design the cyclic prefix CP and guard time (GT) of corresponding length. The time domain lengths of CP and GT in the present invention are both equal to the maximum round-trip delay difference RTD. max .

[0100] Step 3: Receiver demodulation and interference elimination process

[0101] 1. Receiver time-frequency compensation

[0102] according to Figure 2 The receiver MsgA structure shown in the figure, the receiver needs to detect the preamble code of each time slot and record the delay of the detected user. For the colliding time slot, if the preamble code in the time slot is different, the delay of the colliding user can be estimated separately according to the properties of the ZC sequence. Therefore, if two or more preamble codes are detected in a single time slot, it means that a collision has occurred in the time slot. In the present invention, the receiver starts the traversal demodulation process after completing the preamble code detection of all time slots. The demodulation needs to skip the colliding time slot and demodulate the time slot with only one MsgA received. The sequence y[n] of the receiver after removing the cyclic prefix is ​​expressed as:

[0103] y[n]=r[(n-τ)modN](21)

[0104] Where τ represents the delay, N represents the time domain sampling sequence length of an OFDM symbol after removing the cyclic prefix, that is, the number of subcarriers, and r[n] represents the receiver baseband sampling signal, which is related to the transmitter baseband signal x[n] as follows:

[0105]

[0106] Where h[n] is the link channel coefficient and z[n] is Gaussian white noise. Substituting equation (22) into equation (21) yields:

[0107]

[0108] Timing advance detected using preamble To estimate the delay τ, use To estimate the actual normalized Doppler frequency shift ε, the purpose of time-frequency compensation can be achieved. The compensation result It can be expressed as:

[0109]

[0110] The compensation result can be used to perform channel estimation, and then realize demodulation and decoding. The channel estimation method is detailed in Section 3 of this section.

[0111] 2. Interference cancellation based on irregularly repeated time slot ALOHA

[0112] Consider Figure 8In the simple access scenario shown, user 1 and user 2 replicate their data frames twice in the current transmission, selecting time slots 2, 4 and time slots 2, 3 respectively; user 3 replicates its data frame three times, selecting time slots 1, 2, and 4. Therefore, only one data frame is transmitted in time slots 1 and 3, and no collision occurs. However, three data frames and two data frames are stacked in time slots 2 and 4, respectively. According to the definition of formula (10), this access number can be expressed by Λ(x) = 0.67x 2 +0.33x 3 This means that 2 / 3 of the user data is repeated twice, and 1 / 3 of the users data is repeated 3 times.

[0113] The first time slot contains only the data frame of user 3. After demodulating and decoding this time slot, the interference of user 3 in time slots 1, 2, and 4 can be eliminated at the same time using the time slot pointer information contained in the decoded data; the third time slot contains only the data frame of user 2. After demodulating and decoding this time slot, the interference of user 2 in time slots 2 and 3 can be eliminated at the same time; because the data interference of users 2 and 3 has been eliminated, the data of user 1 in time slots 2 and 4 is "easy to demodulate", so it can also be demodulated and decoded. At this point, all users have completed decoding.

[0114] From the above analysis, we can see that the user's data frames are randomly scattered in each time slot after multiple copies. As long as one of these data copies can be demodulated and decoded, all copies can be eliminated. Different from the time-slot ALOHA protocol, the improved ALOHA protocol utilizes conflicting time slots and uses continuous interference elimination technology to make the time slots that were originally undemodulated become "easy to demodulate", reducing the probability of retransmission and further improving throughput.

[0115] 3. Channel estimation for interference cancellation slots

[0116] When performing interference elimination, due to the different channel environments in different time slots, after successful demodulation and decoding in a certain time slot, in order to ensure the accuracy of interference elimination and the success rate of subsequent demodulation and decoding, it is necessary to obtain the channel coefficient of the time slot to be eliminated. Assume that the channel coefficient of the jth time slot of the i-th user is h ij [n] = h ij , whose MsgA data sequence x i [n] nested pilot sequence x pi [d], the pilot interval is Δn, which satisfies:

[0117] x pi [d] = x i [(d-1)Δn+1] (25)

[0118] The pilot sequence is a row or column of the Hadamard matrix, and different pilot sequences are orthogonal to each other.

[0119]

[0120] Among them, N p is the pilot length. The pilot sequence is associated with the preamble code. Different preambles correspond to different pilot sequences. Therefore, if the preamble codes in the same time slot are different, the corresponding pilot sequence can be mapped through the preamble, whether it is a collision time slot or a non-collision time slot, and then the channel estimation can be performed; within a time slot, it can be considered that the Doppler frequency shift and delay of each user are only related to their spatial position and have nothing to do with time. Therefore, the Doppler frequency shift and delay estimated by the preamble code can be used to perform time-frequency compensation on the pilot. Fig. 9 As an example, assuming that the pilot detection and time-frequency compensation are performed, the pilot sequence output by the receiver can be expressed as equation (24). According to the properties of the Hadamard matrix, the receiver uses the local pilot x pi [n] calculation, if the delay is estimated correctly, that is Then the estimated channel coefficient It can be expressed as:

[0121]

[0122] in, It can be obtained from equation (24), from which we can see that the error of channel estimation mainly comes from the Doppler frequency shift estimation deviation and the Gaussian white noise z[n] of the channel. In the absence of Gaussian noise (z[n] = 0) and the Doppler shift estimate is completely accurate In this case, the channel estimate is a perfect estimate

[0123] Step 4: Simulation verification and performance evaluation

[0124] For the single-beam low earth orbit (LEO) system and geosynchronous orbit (GEO) system, the present invention designs and builds the following system based on the relevant parameters listed in Table 1: Fig.10 The simulation module shown.

[0125] Table 1

[0126]

[0127] In the simulation process of the present invention, a certain number of users are first randomly generated within the beam coverage. Subsequently, these users are simulated to perform a series of operations, including the selection of preamble codes, the encoding and modulation of data, the selection of time slots and the transmission of data. At the receiver, the present invention simulates the processing processes of preamble code detection, channel estimation, data demodulation and decoding, and interference elimination. At the same time, the system collects statistics on access live data in real time for subsequent analysis. In order to fully verify the performance of the present invention in terms of random access performance, the simulation results of the present invention are compared in detail with the existing two-step random access baseline scheme, which is used to intuitively evaluate the performance advantages of the present invention in key indicators such as access success rate and access delay.

[0128] Figure 11-Figure 15 The performance of the present invention is compared with the existing random access method from different perspectives. Considering the signal-to-noise ratio is 6.5dB, the Rice factor K=10dB, and the code rate is R c =0.5, and in the scenario of multi-carrier transmission, the preamble code adopts a double-root preamble sequence that is robust to Doppler frequency shift. Fig.11 The first access throughput shown indicates the number of users that can access when all users initiate access at the same time under the current load, representing the access capacity of the two solutions. In the LEO scenario and GEO scenario, the peak throughput proposed by the present invention is increased by about 70% and 75% respectively compared with the current baseline solution. The first access success rate s is defined as the ratio of the number of first successful access q to the total number of current access m, that is, s = q / m = T / G. The obtained curve is as follows: Fig.12 As the user load increases, the probability of MsgA data collision increases, so the first access success rate of the existing baseline solution decreases. However, thanks to the interference elimination demodulation technology, when the normalized user load is less than 0.6, the present invention can still guarantee a success rate of more than 90%.

[0129] Simulation results show that in LEO and GEO scenarios with a signal-to-noise ratio of 6.5dB and a Ricean factor K=10, the throughput of the present invention can be increased by up to about 70% and 75%, and the access delay can be reduced by up to about 56% and 60%, significantly improving the random access performance.

[0130] In terms of access speed, the average number of signaling interactions can directly reflect the size of the average access delay per capita, such as Fig.13 As shown, it can be seen that the average latency of the existing solutions basically increases linearly with the normalized user load, that is, the average access latency is completely limited by the user load. However, the present invention can ensure that when the normalized user load is lower than 0.6, the average access latency is about the time of two signaling interactions, thereby improving the access speed. Fig.14The number of rounds required for all users to complete access is displayed to intuitively show the overall access speed, and the total number of rounds C required for the last user in the beam to complete access is recorded. Fig.15 Reflects the average number of users accessing each round after all users have completed their use. c , that is, m c =m / C, directly reflects the access speed. In LEO and GEO scenarios, the access delay proposed by the present invention can be reduced by about 56% and 60% respectively compared with the existing baseline solution.

Claims

1. A satellite-to-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA, characterized in that: The steps include: Step 1: Design of random access process method In the two-step random access process, MsgA contains random access preamble and control information, and MsgB contains random access response and contention resolution information; consider that in an area covered by a single beam, there are m users who choose to initiate random access in the same cycle after demodulating the downlink broadcast signal; MsgA of each time slot contains random access preamble and control information, which are transmitted on the physical random access channel PRACH and the physical uplink shared channel PUSCH respectively; because the data packet cannot be demodulated directly, preamble detection is required for time-frequency compensation; Step 2: Preamble design and timing testing To solve the uplink synchronization problem of user data, random access uses the access preamble to estimate the time advance TA; the TA estimate is fed back to each user through the random access response RAR, and then the timing adjustment is performed to synchronize the subsequent PUSCH data arriving at the receiver; the TA value is used to locate the position of PUSCH in MsgA, omitting the feedback and timing adjustment steps; Step 3: Receiver demodulation and interference elimination process 3.1、Receiver time-frequency compensation; 3.2, Interference cancellation based on irregularly repeated time slot ALOHA; 3.

3. Channel estimation of the time slot to be eliminated; Step 4: Simulation verification and performance evaluation Consider LEO communication scenarios and GEO communication scenarios; in the case of large elevation angles, the transmission link is dominated by direct components, and the reflected components are not obvious. The channel is set as a Ricean channel with a large direct component. In order to ensure the demodulation and decoding performance, low-density parity check code LDPC is used as the encoding and decoding method, and multi-carrier modulation and demodulation are used.

2. The satellite-to-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA according to claim 1, characterized in that: In step 1, if the random access cycle time is divided into n s time slots, each of which contains one MsgA signaling, and the normalized user load of the system is defined as: Assume that the number of users successfully accessing in one round is q, and define the system throughput as: Among them, each user in this n s The control information in MsgA needs to record the time slot position pointers of these MsgAs, and the receiver needs to store these n s The receiver can detect all the information of each time slot, including the time slot data that collided because some users selected the same preamble code. The receiver determines which users have successfully accessed and which users need to re-access based on the demodulation results of this data.

3. The satellite-to-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA according to claim 1, characterized in that: In step 2, the Zadoff-Chu (ZC) sequence is a constant amplitude complex sequence. zc , let u be the ZC sequence of physical roots x u (n), its expression is: Where n is the sequence symbol index, j is the imaginary part sign, and equation (3) is called the root sequence, which is known to both the transmitter and the receiver. For a round-trip delay of τ i The i-th user receives the leading sequence y in an ideal noiseless channel. pre (n) = x u (n-τ i ), the received sequence is cross-correlated with the root sequence, and the delayed power spectrum PDP is obtained as: in, represents the complex conjugate of the root sequence with u as the physical root, and k is the time index of the delayed power spectrum; the user delay τ is detected by detecting the peak position of the PDP spectrum i , to achieve uplink synchronization.

4. The satellite-to-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA according to claim 1 or 2, characterized in that: In step 2, a dual-root preamble sequence method is introduced to achieve robust TA estimation and Doppler frequency shift estimation; specifically: Define the normalized Doppler shift ε as the absolute Doppler shift f d Ratio to subcarrier spacing Δf: A dual preamble sequence is used as the access preamble. The two preamble sequences have different physical roots, and both ZC sequences are root sequences without cyclic shift. Suppose the two root sequences are: Among them, N zc Represents the length of the preamble sequence, u1, u2 represent the two physical roots selected by the user; after the two preamble sequences pass through the Doppler frequency shift channel, the receiver uses the root sequences corresponding to the two preambles to calculate the PDP spectrum in a circular correlation manner, and obtains two peaks, whose positions are set to p1 and p2 respectively. The positions of the two peaks are related to the normalized Doppler frequency shift ε, the physical roots u1, u2 and the actual normalized delay τ, satisfying: Where [ε] means the normalized Doppler shift is rounded to the nearest integer. Represents the modular multiplication inverse of u1,u2, that is, it satisfies: Estimate the normalized Doppler shift and delay They are: Among them, u1 and u2 should be selected in pairs and satisfy the following relationship:

5. The satellite-to-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA according to claim 4, characterized in that: In step 2, when designing the frame structure, it is also necessary to consider the relationship between the user delay difference within the coverage of a single beam and the preamble length and preamble subcarrier spacing; assuming that the maximum user delay difference is RTD max , the subcarrier spacing is Δf, and the PDP spectrum peak position needs to be limited to the preamble length N zc To ensure that it can be detected, the three should satisfy the following relationship: Among them, the leading sequence length is N zc =839, the preamble subcarrier spacing is Δf=1.25kHz, which is used as the LEO system with a beam radius of R=45km and the satellite-to-ground random access preamble with a radius of R=60km.

6. The satellite-to-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA according to claim 1, characterized in that: In step three, The receiver time-frequency compensation is as follows: considering the transmission of the coded modulation sequence in the frequency domain under a noisy channel, the transmitter uses discrete Fourier transform extended orthogonal frequency division multiplexing to send data, that is, the data bits to be transmitted are first processed by serial-to-parallel conversion, and then digital modulation is performed; then, the modulated symbols will undergo discrete Fourier transform precoding and be mapped to different subcarriers to achieve frequency domain modulation; finally, these frequency domain signals are converted back to the time domain through inverse Fourier transform, and a cyclic prefix is ​​added to complete the preparation for data transmission; the receiver is the inverse process of the above steps, which will not be repeated; Interference elimination based on irregularly repeated time slot ALOHA is as follows: The receiver needs to receive and store the data information of all time slots; during demodulation, the receiver first performs preamble detection on each time slot to determine whether a collision occurs and record the estimated delay value of each time slot, and then starts traversing demodulation and interference elimination: demodulate and decode the data in the non-colliding time slots, obtain the time slot positions of the remaining data copies through the decoding results, and eliminate the interference of the copies in these time slots; when traversing to the last time slot, if there are still data in the time slot that has not been successfully demodulated, it is necessary to repeat the traversal; repeat the above process many times until all time slots are demodulated, or the remaining undemodulated time slots are all collision time slots, at which time the demodulation process is completed; All users who fail to access need to re-initiate access in the next round of access; if different users repeat data packets for different times, they will obtain different system throughputs; The channel estimation of the time slot to be eliminated is as follows: The receiver needs to perform channel estimation for all time slots where each user's MsgA is located; since different users may choose the same time slot, the pilots used for channel estimation must be orthogonal, and the orthogonal pilots select a row or column of the Hadamard matrix; there are time delays and Doppler frequency shifts in the channel, so the receiver needs to perform corresponding time-frequency compensation on the pilot sequence before performing channel estimation.

7. The satellite-to-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA according to claim 1 or 6, characterized in that: In step 3, let x[n] be the time domain baseband transmission signal after OFDM modulation. If N subcarriers are used for transmission and the normalized Doppler frequency shift of the transmission link is ε, then the baseband sampling signal r[n] of the receiver is expressed as: Where h[n] is the link channel coefficient, z[n] is Gaussian white noise; The preamble sequence reaches the receiver after a certain delay τ. Due to the existence of the cyclic prefix CP, the output sequence after removing the cyclic prefix will be the cyclic shift of the original transmitted sequence: y[n]=r[(n-τ)modN] (6) Where N is the sampling sequence length of an OFDM symbol after removing the cyclic prefix, that is, the number of subcarriers, and τ is the delay. Before decoding, the time domain sequence needs to be converted to the frequency domain, so the time domain signal y[n] needs to be compensated for time and frequency to be as close as possible to the ideal received sequence y without frequency deviation and delay. cmp [n]: Ensure the success rate of frequency domain demodulation and decoding; the preamble detection estimates the time delay and normalized Doppler frequency shift ε at the same time, thereby realizing the compensation process.

8. The satellite-to-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA according to claim 1 or 6, characterized in that: Timing advance detected using preamble To estimate the delay τ, use To estimate the actual normalized Doppler frequency shift ε, the compensation result is It is expressed as: The compensation result is used to perform channel estimation and then realize demodulation and decoding.

9. The satellite-to-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA according to claim 1 or 6, characterized in that: Assume that the channel coefficient of the jth time slot of the i-th user is h ij [n] = h ij , whose MsgA data sequence x i [n] nested pilot sequence x pi [d], the pilot interval is Δn, which satisfies: x pi [d]=x i [(d-1)Δn+1] (25) The pilot sequence is a row or column of the Hadamard matrix, and different pilots are mutually orthogonal; Among them, N p is the pilot length; the pilot sequence is associated with the preamble code, and different preambles correspond to different pilot sequences. Therefore, if the preamble codes in the same time slot are different, whether it is a collision time slot or a non-collision time slot, the corresponding pilot sequence is mapped by the preamble to perform channel estimation.

10. The satellite-to-ground two-step random access enhancement method based on irregularly repeated time slot ALOHA according to claim 9, characterized in that: In a time slot, the Doppler frequency shift and delay of each user are considered to be related only to their spatial position and not to time. Therefore, the Doppler frequency shift and delay estimated by the preamble are used to perform time-frequency compensation on the pilot. According to the properties of the Hadamard matrix, the receiver uses the local pilot x pi [n] calculation, if the delay is estimated correctly, that is Then the estimated channel coefficient It is expressed as: The error in channel estimation comes from the Doppler frequency shift estimation bias and the Gaussian white noise z[n] of the channel; in the absence of Gaussian noise, z[n] = 0 and the Doppler frequency shift estimation is completely accurate, In the case of

Images