CN111642006A

CN111642006A - Satellite random access timing detection method

Info

Publication number: CN111642006A
Application number: CN202010458043.XA
Authority: CN
Inventors: 甄立; 孙腾; 梁思远; 卢阳; 卢光跃
Original assignee: Xian University of Posts and Telecommunications
Current assignee: Xian University of Posts and Telecommunications
Priority date: 2020-05-27
Filing date: 2020-05-27
Publication date: 2020-09-08

Abstract

The invention provides a satellite random access timing detection method.A sending end cascades ZC sequences of K different root serial numbers in a time domain to construct a long leader sequence meeting the requirement of satellite beam coverage performance, and sends the long leader sequence on a specified physical random access channel resource; the receiving end directly obtains an accurate TA value through one-step timing detection and feeds back the TA value to the user for timing advance adjustment. The invention not only can effectively inhibit noise by flexibly adjusting the number of ZC sequences involved in correlation operation, but also can realize robustness to large carrier frequency offset, thereby improving access efficiency.

Description

Satellite random access timing detection method

Technical Field

The invention belongs to the technical field of satellite communication, and relates to a timing detection method which can be used for low-orbit satellite communication.

Background

In recent years, with the progress of society and the development of technology, the requirements for transmission rate and quality of service of mobile communication have become higher and higher, and the bandwidth and quality of service that can be provided by the existing mobile communication system have become less than satisfactory. With the rapid development of land mobile communication systems, the random access preamble design has gained wide attention of scholars at home and abroad, and has gained certain research and development in the fields of 5G millimeter waves, narrow-band internet of things, MIMO-OFDM and the like. However, since the satellite communication system has a wide beam coverage and a large satellite-to-ground propagation delay, the existing terrestrial random access preamble design method cannot be directly applied to the satellite communication system. On the other hand, the base station needs to obtain a Timing Advance (Timing Advance) TA value by detecting a preamble sent by the user and feed the TA value back to the user for Timing Advance adjustment. Considering the characteristics of Low operating signal-to-noise ratio, high dynamic link, Low complexity and the like in a Low Earth Orbit (LEO) satellite communication system, rapidly and accurately detecting a transmission preamble and estimating a TA value of the transmission preamble will face a greater challenge than a terrestrial scene. Because the satellite mobile communication system has larger satellite-ground link propagation delay, if the satellite-borne receiver cannot efficiently complete the timing detection process, the user can frequently initiate an access request, which greatly reduces the access efficiency. Therefore, the design of a high-efficiency and high-reliability receiving end timing detection algorithm is the key point for realizing uplink synchronization of the LEO satellite communication system.

At present, the random access preamble design still considers that the ground 5G preamble design principle is adopted as a preamble design research starting point, and the preamble has universality and good correlation performance under an ideal channel. However, the satellite beam has a large round-trip delay difference, which makes the duration of the general preamble sequence large and the subcarrier interval small, not only increases the complexity of the preamble detection process, but also is very sensitive to a large carrier frequency offset. Secondly, a common lte (long term evolution) random access timing detection method is mainly based on time domain correlation operation, specifically, a power delay spectrum is generated by using time domain period correlation operation of a receiving sequence and a local root sequence, then a correlation peak value larger than a preset threshold is extracted from a detection window, and a TA value of a user is further obtained by using a peak value position. However, the conventional methods are all directed to a low mobility scenario, and are not applicable to a high mobility scenario with a large carrier frequency offset, and are generally applied to a high signal-to-noise ratio environment and have high implementation complexity, so that the practical application requirements of the uplink synchronization technology on the LEO high-dynamic satellite-ground link cannot be met.

Disclosure of Invention

In order to overcome the defects of the prior art, the invention provides a satellite random access timing detection method, which is based on multi-sequence joint correlation, can be used for carrying out robust access preamble with low-complexity detection on a satellite communication environment, and estimating the value of Timing Advance (TA) in one step, not only can effectively inhibit noise by flexibly adjusting the number of Zadoff-Chu (ZC) sequences involved in correlation operation, but also can realize the robustness on large carrier frequency offset.

The technical scheme adopted by the invention for solving the technical problem comprises the following steps:

1) a transmitting end constructs a transmitting sequence, and concatenates ZC sequences with K different serial numbers in a time domain to construct a long leader sequence meeting the satellite beam coverage performance requirement, which is called a multi-root concatenated long leader sequence S ═ S₁,...,S_K]Wherein the length of each ZC sequence is to be robust to the carrier frequency offset; the user selects the multi-root cascade long leader sequence according to the random access configuration message in the system message SIB2 broadcasted periodically by the satellite, and sends the multi-root cascade long leader sequence on the specified physical random access channel resource;

2) the receiving end carries out cyclic shift m times N on a plurality of cascade long leader sequences sent by users_ZCAfter a length, the sequence S is obtained_m＝[S_m,1,...,S_m,K]Wherein m ∈ { 1., K-1}, and further converting the two sequences S ═ S [ -S [ ]₁,...,S_K]And S_m＝[S_m,1,...,S_m,K]Obtaining a new local sequence a after conjugate multiplication_m＝[a_m,1,...,a_m,K](ii) a Similarly, in the detection window at the receiving end, the received sequence at any timing index d and its cyclic shift sequence are correspondingly represented as

Wherein d ∈ { 1.,. KN_ZC}; the two sequences are subjected to conjugate multiplication to obtain a new receiving sequence b_m＝[b_m,1,...,b_m,K](ii) a Then, respectively take a_m＝[a_m,1,...,a_m,K]And b_m＝[b_m,1,...,b_m,K]The first subsequence a of_m,lAnd b_m,lAfter the sliding correlation operation is carried out, a correlation operation result can be obtained

L ∈ { 1.. multidot.K }, finally accumulating all correlation results and multiplying the correlation results by corresponding normalization coefficients to obtain a normalized timing measurement function

M ∈ { 1.,. K-1} represents the number of times of cyclic shift, L ∈ { 1.,. K } is the total number of subsequences used in correlation operation, a power delay spectrum is further obtained according to the obtained timing measurement function, a receiving end sets a threshold according to the power delay spectrum to obtain a correct position index corresponding to a unique peak value, and therefore a TA value is further obtained and fed back to the user for timing advance adjustment.

In the step 1), the length N of each ZC sequence_ZC839, the subcarrier interval corresponding to each ZC sequence is more than or equal to 1.25 KHz.

The invention has the beneficial effects that:

firstly, the invention uses a method of cascading a plurality of different root sequences to construct a plurality of new cascading long leader sequences, and only a unique correlation peak appears in a detection window when any ZC sequence in the constructed long leader sequences is used for timing detection in consideration of the minimum cross-correlation characteristic of the ZC sequences. That is to say, by using the proposed multiple concatenated long preamble sequences, an accurate TA value can be directly obtained through one-step timing detection, which is of great significance for improving access efficiency.

Secondly, the invention provides a corresponding timing detection method aiming at one-step TA estimation based on the built multiple cascade long leader sequences, the algorithm directly eliminates the influence of frequency deviation on the detection performance through correlation operation on the basis of building a new local and sending sequence, and can further inhibit the adverse influence of noise on the detection performance through flexibly selecting the number of sequences participating in the correlation operation, thereby improving the access efficiency.

Drawings

FIG. 1 is a schematic diagram of a configuration of a plurality of concatenated long preamble sequences;

FIG. 2 is a flow chart of a timing detection algorithm implementation of the present invention;

FIG. 3 is a graph of a power delay spectrum of a timing metric function of the present invention;

FIG. 4 is a graph comparing the detection performance of the present invention with the prior art method under different normalized frequency offsets;

fig. 5 is a graph comparing the detection performance of various random access methods under different signal-to-noise ratios.

Detailed Description

The present invention will be further described with reference to the following drawings and examples, which include, but are not limited to, the following examples.

Referring to fig. 1, the first partial sequence configuration of the present invention is implemented as follows:

and constructing a sending sequence meeting the requirements of the LEO satellite communication system.

The invention only considers the single user condition, firstly, in order to meet the requirement of a larger beam coverage area of the LEO satellite communication system, a mode of cascading a plurality of sequences is selected to generate a transmittable preamble sequence. Aiming at one-step TA estimation, K different root sequences are selected to be cascaded to generate a transmittable preamble sequence, namely a plurality of cascaded long preamble sequences, which are expressed as S ═ S [ S ]₁,...,S_K]. Wherein S_l＝ZC_lL ∈ { 1.,. K }, which indicates that the l-th length is N_ZCAnd any one of ZCs_lAre robust to carrier frequency offset and the value of K depends on the satellite beam coverage. As shown in fig. 1. So that the user can periodically broadcast the system message SIB2 (Sys) according to the satellitetem Information Broadcast) selects the multiple concatenated long preambles for transmission on the specified physical random access channel resource.

Referring to fig. 2, the second part of the timing detection algorithm of the present invention is implemented as follows:

step 1: and constructing a new local sequence used in a detection algorithm based on the transmitted multiple concatenated long preamble sequences.

In order to effectively eliminate the influence of carrier frequency offset on the detection performance, the algorithm constructs a new local sequence which can be used for correlation detection by a method of conjugate multiplication of a user transmission sequence and a cyclic shift sequence thereof. The concrete implementation is as follows: firstly, a plurality of cascade long leader sequences S ═ S transmitted by a user₁,...,S_K]Cyclically shifted by m times N_ZCAfter the length, the corresponding cyclic shift sequence S is obtained_m＝[S_m,1,...,S_m,K]Wherein m ∈ { 1., K-1}, N_ZCIs the length 839 of one ZC sequence; then, the transmitting sequence and the cyclic shift sequence are subjected to conjugate multiplication to obtain a new local sequence a_m＝[a_m,1,...,a_m,K]The sequence a_mIt will be used for subsequent correlation steps in the algorithm.

Step 2: a new received sequence for use in the detection algorithm is constructed based on the construction of the new local sequence.

When the receiving end receives the multiple concatenated long preambles transmitted by the user, correspondingly, at the timing position d, d ∈ { 1., KN_ZCThe received sequence and its cyclically shifted sequence are denoted in the same manner as

And

further carrying out conjugate multiplication on the two sequences to obtain a new receiving sequence b_m＝[b_m,1,...,b_m,K]. This sequence b_mWill also be used in the subsequent correlation step in the algorithm.

And step 3: and performing sliding correlation operation on the newly constructed local sequence and the receiving sequence.

For the convenience of the following discussion, the newly constructed local sequence and the l sub-sequence of the received sequence, i.e. a, are taken separately_m,lAnd b_m,lAre respectively represented as

Wherein

Representing the multiplication of two sequence elements. Using two subsequences a_m,lAnd b_m,lThe sliding correlation between them yields a correlation result corr_m,l(d) Is specifically shown as follows

And 4, step 4: the normalization results in a timing metric function c (d).

The timing metric function c (d), i.e. the function of the timing metric c (d), can be finally determined by accumulating all correlation results and multiplying by the corresponding normalization coefficients

Wherein M is in the range of { 1.,. K-1} to represent the number of cyclic shifts, and L is in the range of { 1.,. K } to be the total number of subsequences used for participating in the correlation operation, so that the influence of noise on the detection performance can be further suppressed by flexibly selecting L. And further obtaining a power time delay spectrum according to the obtained timing metric function. The only peak value can be detected by setting a proper threshold at the receiving end according to the power time delay spectrum, the TA value can be further obtained by using the correct position index corresponding to the peak value, and the TA value is fed back to the user for timing advance adjustment.

The effect of the present invention is further described by simulation below.

1. Simulation conditions are as follows:

the experiment was simulated using MatlabR2018b on a WINDOWS10 system with a CPU of core (TM) i 5-82503.40 GHz and a memory of 8.00 GB.

Selecting experimental parameters: in all the following experiments, the value of K was taken to be 8, and the corresponding root sequence number u₁,...,u₈Taking 1, 28, 101, 127, 179, 211, 337, 397, M as 7, ZC sequence length N_ZCAt 839, the channel condition is an AWGN channel.

2. Simulation content:

simulation experiment 1: a power time delay profile of the timing metric function is made and the experimental results are shown in fig. 3, where the horizontal axis represents the timing index and the vertical axis represents the associated calculated value.

Simulation experiment 2: the method of the invention is compared with the prior method in performance under different normalized frequency offsets, the experimental result is shown in figure 4, the horizontal axis represents the normalized frequency offset, and the vertical axis represents the timing mean square error of each method.

Wherein the method for He is selected from the group consisting of He Y, Cui G, Pengxu L I, et al, timing Advance timing Algorithm of Low Complexity Based on DFT Spectrum Analysis for software System [ J ]. China Communications 2015,12(4):140-150.

The method of Gui is selected from the documents Cui G, He Y, Li P, et al, enhanced Timing With symmetry Zadoff-Chu Sequences for Satellite Systems [ J ]. IEEEcommunications Letters,2015,19(5): 747-.

Simulation experiment 3: the method of the invention is compared with the prior method in performance under different signal-to-noise ratios, the experimental result is shown in figure 5, the horizontal axis represents the signal-to-noise ratio, and the vertical axis represents the timing mean square error of each method.

The method for He is selected from The group consisting of He Y, Cui G, Li P, et al, random access preamplified on time pre-compensation for LTE-satellite system [ J ]. The journal of China Universities of Posts and telecommonications 2015,22(3):64-73.

The method of Li is selected from the documents Li C, Ba H, Duan H, et al.A two-step time delay differentiation estimation method for initial random access in satellite LTEs [ C ]. International Conference on Advanced Communication technology. IEEE,2014:10-13.

The method of ZHen is selected from the documents ZHEN L, Qin H, Song B, et al.

3. Simulation experiment result analysis:

as can be seen from fig. 3, the power delay profile has a distinct unique peak, which illustrates that the proposed method can effectively detect the preamble sequence.

Figure 4 depicts the timing mean square error curves for different methods at different normalized frequency offsets when the signal-to-noise ratio is-11 dB. It is worth noting that due to the conjugate symmetry of ZC sequences, both the Gui and He methods achieve relatively low timing mean square error in the presence of integer-times normalized frequency offset. However, if there is a decimal-times normalized frequency offset, its timing estimation performance is drastically degraded, and the worst case is obtained when the decimal-times normalized frequency offset is set to a multiple of 0.5. In contrast, the timing mean square error performance of the method of the present invention remains almost unchanged for different normalized frequency offset values, which indicates that the method of the present invention can completely eliminate the adverse effect of carrier frequency offset on timing estimation. In addition, the timing estimation performance of the method of the present invention can be further improved by increasing L, and it can be clearly seen that when L is 8, the method of the present invention can be significantly superior to the existing method.

Figure 5 shows the timing estimation mean square error for various detection methods at different signal to noise ratios with the carrier frequency offset set to 750 Hz. As is apparent from fig. 5, the method of He in which the subcarrier spacing is small has a large timing error at different signal-to-noise ratios because the frequency offset greatly affects its timing performance. The Li method uses a terrestrial preamble format, but when the frequency offset is greater than half the subcarrier spacing (i.e., 62.5Hz), the performance of the method will be degraded. Although the method of Zhen can effectively overcome the effect of frequency offset, its timing accuracy is susceptible to noise. Compared with the methods, the method can further reduce the influence of noise on timing detection by increasing L because the number of the subsequences used in the method is adjustable.

Claims

1. A satellite random access timing detection method is characterized by comprising the following steps:

1) a transmitting end constructs a transmitting sequence, and concatenates ZC sequences with K different root serial numbers in a time domain to construct a long leader sequence meeting the satellite beam coverage performance requirement, which is called a multi-root concatenated long leader sequence S ═ S₁,...,S_K]Wherein the length of each ZC sequence is to be robust to the carrier frequency offset; the user selects the multi-root cascade long leader sequence according to the random access configuration message in the system message SIB2 broadcasted periodically by the satellite, and sends the multi-root cascade long leader sequence on the specified physical random access channel resource;

2) the receiving end carries out cyclic shift m times N on a plurality of cascade long leader sequences sent by users_ZCAfter the length, the sequence S is obtained_m＝[S_m,1,...,S_m,K]Wherein m ∈ { 1., K-1}, and further converting the two sequences S ═ S [ -S [ ]₁,...,S_K]And S_m＝[S_m,1,...,S_m,K]Obtaining a new local sequence a after conjugate multiplication_m＝[a_m,1,...,a_m,K](ii) a Similarly, in the receiving-end detection window, the received sequence at any timing index d and its cyclic shift sequence are correspondingly represented as

And

2. The satellite random access timing detection method of claim 1, wherein: in the step 1), the length N of each ZC sequence_ZCAnd the sub-carrier interval corresponding to each ZC sequence is more than or equal to 1.25KHz for 839.