KR101414610B1 - Method For Generating 2 or more Sequence Set, And Method For Generating Sequence For The Same - Google Patents

Abstract

A method for generating two or more sequence sets and a sequence generation method therefor are disclosed.

Specifically, in generating two or more sequence sets, one or more sequences are selected from each of a first type sequence and a second type sequence generated based on two or more different lengths, thereby generating two or more sequence sets A method is disclosed. Alternatively, in generating two or more sequence sets, it is possible to select one or more sequences in each of the first type sequence and the second type sequence in which the first type sequence is modulated using the heterogeneous sequence, thereby generating two or more sequence sets A method for generating a signal is disclosed.

In addition, in generating the sequence based on the length of the available subcarriers excluding the DC subcarriers, the sequence is not allocated to the DC subcarriers but is continuously allocated in the region excluding the DC subcarriers using the frequency domain circular motion. A method for generating a sequence to be placed is disclosed.

3 sequence sets, correlation properties, PAPR

Description

[0001] The present invention relates to a method for generating two or more sequence sets and a method for generating a sequence for the same,

The following description relates to a method for generating two or more sequence sets that can be applied to a specific channel and a method for generating a sequence therefor.

Hereinafter, a sequence used in the OFDM, OFDMA, and 3GPP LTE systems to which the above-described methods are applied will be described.

In recent years, demand for high-speed data transmission has been increasing, and OFDM has been adopted as a transmission method of various high-speed communication systems as a method favoring such high-speed transmission. Hereinafter, orthogonal frequency division multiplexing (OFDM) will be described. The basic principle of OFDM is to divide a data stream having a high-rate into a large number of data streams having a slow-rate, and to transmit them simultaneously using a plurality of carriers. Each of the plurality of carriers is referred to as a subcarrier. Since orthogonality exists between a plurality of carriers of the OFDM, even if the frequency components of the carriers overlap each other, detection on the receiving side is possible. The data stream having the high-speed data rate is converted into a data stream having a plurality of low data rates through a serial-to-parallel (S / P) converter, and a plurality of sub- And then the respective data strings are summed and transmitted to the receiving side. OFDMA is a multiple access method in which subcarriers are allocated according to a transmission rate required by multiple users in the entire band in such OFDM.

The OFDM scheme is disadvantageous in that the PAPR (Peak-to-Average-Power Ratio) or CM (Cubic Metric) of the transmission signal is very large. Since the data in the frequency domain is transmitted using the multi-carriers through the IFFT, the magnitude of the OFDM signal amplitude can be expressed by the sum of the sizes of the multi-carriers. However, if the phases of the multi-carriers coincide with each other, a signal having a high maximum value such as an impulse is generated in the OFDM signal to have a very high PAPR or CM. Such a transmission signal according to the OFDM scheme lowers the efficiency of the high output linear amplifier and operates in the nonlinear region of the high output amplifier, causing signal distortion.

Meanwhile, a channel and a sequence used in a recently proposed 3GPP (Third Generation Partnership Project) LTE (Long Term Evolution) system will be described below.

In general, a terminal first performs communication with a base station by performing synchronization with a base station in a synchronization channel (hereinafter referred to as "SCH") and performing cell search. Such an SCH can have a hierarchical structure, and thus can be divided into a main synchronization channel (hereinafter referred to as a "P-SCH") and an auxiliary synchronization channel (hereinafter referred to as "S-SCH").

A series of processes for performing synchronization with a base station and acquiring a cell ID to which the mobile station belongs is called a cell search. In general, a cell search is performed by initial cell search performed when the initial terminal is powered on, and neighbor cell search when a terminal in a connection or idle mode searches for an adjacent base station. (Neighbor cell search).

The P-SCH used in a communication system using multiple orthogonal subcarriers such as OFDM or SC-FDMA preferably satisfies the following conditions.

First, in order to detect good performance on the receiving side, the time-domain auto-correlation characteristic for the sequence constituting the SCH should be good.

Second, the complexity of synchronization detection should be low.

Third, the Peak-to-Average Power Ratio (PAPR) (or CM) should be low.

Fourth, if the SCH can be utilized for channel estimation, it is desirable that the frequency response has a constant value. That is, it is known that a flat response in the frequency domain is the best channel estimation performance in terms of channel estimation.

A Zadoff-Chu sequence is discussed as a CAZAC (Sequential Amplitude Constant Auto-Correlation) sequence as a sequence used for a channel used in LTE including the SCH as described above.

Two types of CAZAC sequences are used: GCL (Generalized Chirp-Like) CAZAC and Zadoff-Chu CAZAC. These are in conjugate complex numbers, and the GCL CAZAC can be obtained by taking the conjugate complex number of Zadoff-Chu. Zadoff-Chu CAZAC is given as follows.

Here, n denotes a sequence index, L denotes a length of a CAZAC sequence to be generated, and M denotes a sequence ID. At this time, M can be expressed as a set of natural numbers that are mutually adjacent to L.

When a Zadoff-Chu CAZAC sequence (hereinafter referred to as a "ZC" sequence) given by Equation (1) is denoted by c (k; N, M), all three characteristics are as follows.

In Equation (2), the CAZAC sequence always has a size of 1, and Equation (3) shows that the auto-correlation function of the CAZAC sequence is represented by a delta function. Here, autocorrelation is based on circular correlation. Equation 4 also shows that the cross-correlation function is always constant when N is a prime number.

As described above, in the case of the CAZAC sequence, a total of L-1 sequences can be generated when the length (L) of the required sequence is a prime number. However, if the length of the sequence is not a prime number, . As described above, when the sequence length L required in the communication system due to the length of the resource block or the like is not a decimal length, the following method can be proposed as a solution to this problem.

First, there is a truncated sequence generating method.

1 is a diagram for explaining a method of generating a sequence according to a cut type sequence generation method.

In this method, when the required length L of the system is not a prime number, a sequence is generated by setting a prime number (X) larger than L to L in Equation (1). Then, among the generated sequences, a sequence of a length longer than L is truncated to an L length.

Since the sequence generated by the above-described method cuts out a part of a sequence, it is possible to improve the autocorrelation and the cross- , The characteristic having a value of 0 in other cases and the characteristic in which the cross-correlation value always has a constant as shown in Equation (4) deteriorates. In addition, it can not be guaranteed that the number of sequences corresponding to L-1 is removed when a sequence having poor correlation characteristics is removed. In addition, by cutting out a part of the generated CAZAC sequence, degradation may also be caused in the characteristics of the CAZAC sequence having a low PAPR characteristic.

Meanwhile, in order to solve the above-mentioned problems, a technique of selecting a maximum decimal length (X) less than a length (L) required in a communication system to generate a CAZAC sequence and inserting a padding portion in a portion having a length of LX Has been proposed.

2 is a diagram for explaining a method of generating a sequence according to a padding-type sequence generation method.

According to the padding-type sequence generation method, when the length L required by the system is not a prime number, a sequence is generated by setting the largest prime number (X) among prime numbers smaller than L to L in Equation (1). Thereafter, a sequence having an L length is generated by padding a generated sequence C1 with 0 for a length C2 corresponding to L-X.

In the padding-type sequence generation method, a part of the generated sequence is divided as shown in FIG. 1 by dividing the sequence by setting the correlation calculation portion of the sequence as the C1 portion in FIG. 2 There is an advantage that the deterioration of the autocorrelation and the cross-correlation property does not occur. However, when the entire length of the applied sequence is considered, the correlation characteristic and the PAPR characteristic may be degraded due to the padding portion C2 of 0 as well.

On the other hand, there is a cyclic extension scheme in a manner similar to that described above with reference to FIG. According to the cyclic extension method, when the length L required by the system is not a prime number, the sequence is generated by setting the largest prime number (X) among prime numbers smaller than L to L in Equation (1). Thereafter, a sequence having an L length is generated by copying and inserting a length (C2) corresponding to L-X in the generated sequence (C1) in a part of the generated sequence.

According to this method, it is possible to reduce the non-correlation and the deterioration of the PAPR characteristic in the 0-padding scheme described above.

Hereinafter, a case of selecting two or more sequence sets to be applied to a specific channel will be described.

Considering the above considerations, when designing a sequence in a sequence-based arbitrary transmission channel, important considerations include the above-mentioned Peak-to-Average Power Ratio (PAPR) or CM (Cubic Metric) (Periodic and aperiodic) auto / cross-correlation property.

However, when two or more sequence sets are generated using only a sequence generated based on the same length, or using only a sequence of the same kind and applied to a specific channel, the above-described PAPR or CM characteristic, There is a problem that it is difficult to satisfy all of the above.

In order to solve the problems described above, the present invention provides a communication system in which a correlation set and a PAPR (or CM) characteristic are both excellent in selecting a set of two or more sequences used for a specific channel, And a method for generating two or more sets of sequences that can be performed.

It is also possible to generate a sequence such that a sequence is not assigned to a DC subcarrier in the frequency domain to generate a configuration sequence used to generate such a set of two or more sequences, or to generate a sequence to be applied to a particular channel independently thereof Method is used, the generated sequence maintains the correlation property and the PAPR (or CM) property in both the time domain and the frequency domain.

According to an aspect of the present invention, there is provided a method of generating two or more sequence sets, the method comprising: generating a first type sequence and a second type sequence based on two or more different lengths, , And generating the set of two or more sequences using the selected sequences.

The first type sequence may be a sequence generated by performing DC puncturing based on an available subcarrier length including a DC subcarrier, and the second type sequence may be a sequence generated by performing a DC puncturing on the DC subcarrier among the available subcarriers A sequence generated based on the excluded length, and a sequence generated in such a manner that the sequence is not allocated to the DC subcarrier.

More specifically, the two or more sequence sets may be three sequence sets, and one of these three sequence sets may be selected from the first type sequence, two from the second type sequence, or alternatively from the first type sequence 2, and a second type sequence, and if two sequences out of the first type sequence are selected, the two sequences selected are two sequences constituting a conjugate symmetry pair desirable.

Preferably, the sequence selected in the sequence selection step of the second type sequence is a sequence generated so as to be continuously arranged in a region excluding the DC subcarriers using a frequency domain circular motion. The second type sequence is generated using a Zadoff-chu (ZC) sequence, and the available subcarrier length including the DC subcarrier may be 64. [

Also, in the sequence selection step, the sequences selected from the first type sequence and the second type sequence may be selected in consideration of at least one of a cross correlation property and a PAPR characteristic.

Meanwhile, if the first type sequence is a sequence generated based on a length less than an available subcarrier length, and the second type sequence is a sequence generated based on a length greater than the available subcarrier length, 1 type sequence may be extended by cyclic copy or zero insertion by the available subcarrier length and the generated second type sequence may be truncated by the available subcarrier length.

According to another aspect of the present invention, there is provided a method for generating two or more sequence sets, the method comprising: generating a first type sequence and a second type sequence by modulating the first type sequence using a heterogeneous sequence, Selecting one or more sequences in the sequence, and generating the set of two or more sequences using the selected sequences.

Here, the first type sequence may be a Zadoff-chu (ZC) sequence, and the second type sequence may be a sequence that modulates the first type sequence using a Hadamard sequence.

Also, a method of generating two or more sequence sets according to another embodiment of the present invention includes selecting one or more sequences in each of a first type sequence and a second type sequence of two or more different types, And generating the set of two or more sequences using the selected sequences.

According to another aspect of the present invention, there is provided a method of generating a sequence based on a length of an available subcarrier excluding DC subcarriers, the method comprising: allocating the sequence to the DC subcarriers, And a sequence is generated so as to be continuously arranged in an area excluding the DC subcarriers.

In this case, when the generated sequence is an even length, the frequency domain circular motion may be a circular motion that is shifted by half the length of the sequence. More specifically, the frequency domain circular motion may shift the sequence to the left or right The length of the sequence is shifted by half the length of the sequence length.

According to the embodiment of the present invention as described above, it is possible to generate two or more sequence sets that are both excellent in the correlation property and the PAPR (or CM) characteristic and can reduce the calculation amount in the receiver.

Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The following detailed description, together with the accompanying drawings, is intended to illustrate exemplary embodiments of the invention and is not intended to represent the only embodiments in which the invention may be practiced.

The following detailed description includes specific details in order to provide a thorough understanding of the present invention. However, those skilled in the art will appreciate that the present invention may be practiced without these specific details. In some instances, well-known structures and devices are omitted or shown in block diagram form around the core functions of each structure and device in order to avoid obscuring the concepts of the present invention. In the following description, the same components are denoted by the same reference numerals throughout the specification.

As described above, the present invention provides a method for generating a set of three sequences having superior correlation characteristics and PAPR (or CM) characteristics in selecting a set of two or more sequences used for a specific channel in a communication system do. To this end, a method of generating two or more sequence sets by a method of generating two or more sequences based on different lengths according to each embodiment of the present invention and a method of generating two or more sequences using a sequence modulated by a heterogeneous sequence And a method of generating a set.

Ⅰ. First Embodiment-A method of producing a semiconductor device having two or more sequence  According to the generation method, two or more sequence  How to create a set

In order to describe the method according to an embodiment of the present invention, a method of generating each configuration sequence that can be used in the above-described two or more sequence sets will be described. Meanwhile, although the Zadoff-Chu (ZC) sequence is described as an example as an example of the sequence used in the description of each method, the described scheme can be applied to any other sequence. In the following description, the case where three sequence sets are generated as two or more sequence sets is mainly described, but the sequence included in the sequence set may be any number of two or more.

≪ First Method >

First, a method for generating a sequence in the time domain without DC puncturing as a method that can be used in a single subcarrier system such as WCDMA as a first method of generating each constitution sequence will be described.

In this method, a sequence generated in the time domain is defined in the time or frequency domain and transmitted. This method is effective when the region where processing of the transmitted channel is performed is the time domain. For example, in the case of a sync channel of a WCDMA system, the P-SCH signal generated / defined in the time domain can be directly transmitted in the time domain and correlation can be performed at the receiving end to perform symbol synchronization. However, this method causes a DC offset problem due to the signal being transmitted to the DC carrier portion.

≪ Second Method >

Next, a method for generating a sequence in the time domain using DC puncturing as a second method of generating each configuration sequence will be described.

In this method, a sequence generated in a time domain is defined and transmitted in a time or frequency domain, and puncturing (or "suppression" or "suppression") of a frequency domain sequence element (ie, a chip) nullification "). This method differs from the first method in that it performs puncturing in the DC region. This method is also more effective in the case of a channel processing in the time domain as in the first method. This DC puncturing has the advantage of solving the DC offset problem described above.

In addition, this method is similar in terms of correlation performance to the third method, which will be described below, when a CAZAC sequence such as a Zadoff-Chu sequence is used. This is because the CAZAC sequence maintains CAZAC characteristics in both time / frequency domains. The only difference in this case is the difference in whether the sequence is ZC or Generalized Chirp-Like (GCL), and both ZC and GCL maintain CAZAC characteristics.

According to the second method, the sequence looking at the time domain is ZC and the sequence looking at the frequency domain is GCL. In addition, it is possible to optimize the signal characteristics of the time / frequency domain of all the sequences generated through the use of ZC equally with the CAZAC characteristics. Also, in order to maintain the CAZAC characteristic in both the time and frequency regions for all the root sequence index numbers, it is preferable that the sequence becomes continuous at the time of mapping the sequence. In the case of the second method, Continuity can be maintained when mapping in all frequency ranges.

<Third Method>

Next, a method of generating a sequence in the frequency domain using DC puncturing as a third method of generating each configuration sequence will be described.

This method is a method of performing mapping in the frequency domain after generating a sequence, and then performing DC puncturing. This method is more effective for channels processing in the frequency domain and can also solve the DC offset problem by DC puncturing.

Also, this method is similar in terms of correlation performance when a CAZAC sequence such as a Zadoff-Chu sequence is used like the second method described above. According to the second method described above, the sequence viewed in the time domain is ZC and the sequence observed in the frequency domain is GCL, but according to the present method, the sequence observed in the frequency domain becomes ZC. In addition, it is possible to optimize the signal characteristics of the time / frequency domain of all sequences generated through the use of ZC to have CAZAC characteristics. Also, in order to maintain the CAZAC characteristic in both the frequency and the time domain with respect to all the root sequence index numbers, it is preferable that the CAZAC characteristic is continuous in the mapping of the sequences. According to this method, continuity can be maintained in all areas of time / frequency.

&Lt; Fourth method >

Finally, as a fourth method of generating each configuration sequence, a method of generating a sequence in the frequency domain without using DC puncturing will be described.

This method is a method of properly mapping the DC region and the guard subcarrier in consideration of the position of the DC sub-region and the guard sub-carrier when the mapping is performed in the frequency domain after generating the sequence. Method. This method is more effective for a channel processing in the frequency domain, and can solve the DC offset problem like the second and third methods described above.

However, even if a CAZAC sequence such as ZC is used, this method can not maintain a perfect CAZAC characteristic in both regions for all sequences because discontinuity occurs in a DC subcarrier region when a sequence is mapped to a subcarrier . In other words, since all the generated sequences can not be optimized to maintain the CAZAC characteristics in both the time and frequency regions, they can be optimized only for specific frequency-domain mapped positions.

In the case of generating three sequence sets using a sequence generated using the above-described four methods, the first method itself has a DC offset problem, and thus, according to the second to fourth methods Generating a set of three sequences using the generated sequence may have the following problems. That is, when a sequence set is determined for a channel to be used for a specific purpose by using any one of the second to fourth methods, when one generation method for one length is applied for a region to be processed, It may be difficult to construct a sequence set that satisfies all desirable characteristics (e.g., correlation property, PAPR or CM characteristics) of the channel.

To be more specific, a concrete example of the synchronization channel of the current 3GPP LTE system will be described with reference to the P-SCH. Of course, the principle to be described below is basically based on the fact that the process of transmitting and detecting a sequence is performed on any sequence-based channel (e.g., uplink / downlink reference signal (RS), ACK / NACK, control channel, Applicable. In addition, even if detection is performed in a manner other than correlation detection as a detection method in an arbitrary sequence-based channel, it is very important to select a sequence set satisfying the above-mentioned basic characteristics, So that the sequence can be selected.

The P-SCH considered in the 3GPP LTE system up to now has a center band (DC) regardless of the transmission band (e.g., 1.25 MHz, 2.5 MHz, 5 MHz, 10 MHz, 15 MHz, 20 MHz, (1.095 MHz) including subcarriers (including subcarriers). Therefore, considering the DC subcarriers, the number of practical usable subcarriers is 72. Therefore, the number of subcarriers occupied by the P-SCH may be equal to or smaller than 72.

In this case, considering the sequence length or the number of subcarriers used, it is preferable to use all 72 subcarriers simply in terms of resource utilization. However, the use of 72 subcarriers means that a total of 73 subcarriers (i.e., 1.095 MHz) including the DC subcarriers are used. In this case, when the symbol synchronization is performed, correlation is performed with good performance and low calculation amount A re-sampling operation corresponding to a multiple of 1.095 MHz (i.e., 1.095 x MHz) is required. However, since the sampling clock used for OFDM symbol transmission in LTE is a multiple of 1.92 MHz, the complexity of using an additional interpolator or a decimator for resampling at 1.095 x MHz is required. However, if 1.92 MHz is divided by a natural number, the corresponding sampling clock (for example, 0.96, 0.48 MHz) can be used to reuse the conventional sampling clock, You only need to perform decimation.

Therefore, in the following description, the case where 0.96 MHz is used, that is, 64 subcarriers including the DC subcarriers are used as the usable subcarriers is mainly described, but the number of usable subcarriers is 72 subcarriers (1.08 MHz) , 70 subcarriers (1.05 MHz), 71 subcarriers (1.065 MHz), 73 subcarriers (1.095 MHz), and the like.

Also, since the sequence used in the P-SCH of the current 3GPP LTE is a ZC-based sequence, the following description is based on the ZC sequence. The ZC sequence has excellent correlation / PAPR (CM) / retention characteristics in the time and frequency domain.

In the case of a sequence-based channel, the correlation property should be good as described above, and the above-mentioned PAPR problem is also important due to the distortion problem in the non-linear region of the amplifier when amplification is considered in transmission. This PAPR is directly related to the flatness of the signal in the time domain. Finally, the frequency-flatness is a requirement when the corresponding sequence is used for channel estimation in the frequency domain. Considering the correlation estimation of the S-SCH by performing channel estimation in the P-SCH in the current LTE, consideration is also given .

Hereinafter, a case where three sequence sets are generated using each of the first to fourth methods will be described in detail, and problems occurring in such a case will be described. Here, the second method and the third method focus on the case of generating a sequence based on the length 64, but a sequence having a different length will also be examined in some cases. However, the second, third, and fourth methods have different lengths corresponding to the DC subcarriers, and the sequences generated by these methods have different lengths.

Considering these points, a case will be described below in which three sequence sets are generated considering only each method.

&Lt; 3 &gt; sequence  For set creation>

When generating three sequence sets using each sequence generated according to the present method, each sequence itself is not suitable because it has the DC offset problem as described above.

&Lt; 3 by the second method sequence  For set creation>

The sequence (P ^M (n)) defined in the time domain according to the present method can be expressed as follows.

Here, n represents a time domain index.

On the other hand, the sequence P ^M (k) defined in the frequency domain according to the present method can be expressed as follows.

Here, k represents a frequency domain index.

Comparing Equation (6) with Equation (5), it can be seen that the time-domain sequence and the frequency-domain sequence of the sequence generated according to the present method are essentially the same signal only in the transmitted region.

In this case, the sequence generated according to the present method is a ZC sequence in the time domain, and it can be confirmed that the GCL becomes a GCL in the frequency domain as described above.

Here, GCL is defined by an equation generated by multiplying a DFT coefficient of a ZC sequence.

As an example according to the present method, a case will be described in which a sequence is generated with N = 64 and L = 64, and the sequence is mapped to a frequency domain for transmission.

The possible types of ZC sequences of length 64 are 32, and the periodic characteristics of these 32 sequences all retain CAZAC characteristics (although DC is punctured, self / cross correlation, from PAPR point of view, . However, from an aperiodic point of view, it does not consistently maintain its characteristics for PAPR / CM (in the case of 4x oversampling) with self / cross correlation. In the case of other channels, in particular, in the case of a synchronous channel that performs symbol synchronization, symbol synchronization must be performed. Therefore, not only the above-described periodic correlation property but also the above-mentioned aperiodic characteristic is also very important.

As a concrete example of generating three sequence sets using the method of generating a sequence with N = 64 and L = 64 as described above and mapping to the frequency domain and transmitting the sequence, M = 1, 63 is selected to generate three sequence sets.

In the case of a ZC sequence, it is preferable to select a conjugate symmetry pair to support simultaneous correlation when selecting two or more sequence sets. For example, in the case of generating a sequence based on N = 64, selecting two of these sequences is M = (1,63), (3,61), (5,59), (7,57) ..., (31, 33), etc., and M = (1,70), (2,69), (3,68), ..., (35, 36, and when N = 63, M = (1,62), (2,61), (4,59), ..., (31,32) are selected. Therefore, in a preferred embodiment of the present invention, when selecting two or more sequence sets, two sequences out of the selected sequence sets are proposed to select the conjugate pair so that simultaneous correlation is possible.

FIGS. 3 and 4 illustrate the autocorrelation of three sequences of ZC sequence indexes M = 1, 3 and 63 as a specific example of generating three ZC sequence sets with N = 64 and L = 64 in the second method Characteristic and cross-correlation characteristics, respectively.

As can be seen in FIGS. 3 and 4, it can be seen that the selection of a sequence of M = 1, 3, 63 as described above does not satisfy the restriction to optimize periodic / aperiodic magnet / have.

Specifically, the aperiodic autocorrelation property can be satisfied by selecting another root sequence. In other words, a large side lobe of autocorrelation characteristic in the case of M = 3 in Fig. 3 can be overcome by selecting another sequence set. However, in the case of the cross-correlation, even if any of the three sequence combinations is generated by an even-numbered ZC, a correlation peak value of at least 24% or more occurs as shown in FIG. I can not.

For reference, when considering the above-mentioned requirements, two of the three sequence sets are desirably selected to select a sequence set constituting a pair of symmetric pairs satisfying the simultaneous correlation, and the other one is selected for other conditions (aperiodic correlation, PAPR, CM, etc.). It should be noted that the above-described M = 1,63 combination condition is applied only to the second method.

In the case of using only the second method, M = 1, 19, 63 can be selected as an optimal combination when N = 64 and PAPR or CM are not considered. However, in this case, the CM at M = 19 is 4.2103 dB, which is not suitable (it is assumed that a root sequence index having a CM of 2.6 dB or less is selected in this combination). Also in this case, the peak value of the cross correlation has a large value of about 25%.

Therefore, the optimal combination that can be selected in consideration of the correlation property, PAPR, CM can be M = 1, 51, 63. In this case, each non-oversampled PAPR is 1.0904dB, 1.0821dB, and 1.0904dB, respectively, and the 4x oversampled CM is 1.3713dB, 2.521dB, and 1.3713dB, respectively.

FIG. 5 is a graph showing aperiodic autocorrelation and cross-correlation characteristics for three sequence sets of M = 1, 51, and 63 among the sequences generated according to the second method.

As can be seen from FIG. 5, it can be seen that the optimal combination of M = 1, 51, and 63 described above also exhibits a cross-correlation peak value of about 25.4%.

In the case of using the second method, even if the PAPR or CM requirements are ignored, the optimum combination is M = 1, 51, 63 as the selected one.

When a signal is transmitted according to Equation (5), the signal may be transmitted through a DAC (Digital-Analog Convert) in a time domain, and a pulse shaping filter and an oversampling may be selectively performed Lt; / RTI &gt; Meanwhile, when a signal is transmitted according to Equation (6), a signal mapped as Equation (6) may be transmitted in the frequency domain through an IFFT OFDM modulation method considering data and a protection subcarrier. For both of these methods, the operation of the receiving end may be the same.

In the second method, it may be preferable to use a sequence of a length corresponding to the average number of subcarriers in the most optimal case, such as when the resampling rate does not matter. That is, this corresponds to the case of N = 73 in the above example, and in this case, the combination of M = 1, 67, 72 can be selected as the most optimal combination.

6 is a graph showing autocorrelation characteristics and cross-correlation characteristics when a combination of M = 1, 67, and 72 is selected as a combination of three sequence sets when N = 73 when generating a sequence according to the second method to be.

As can be seen from FIG. 6, the PAPRs that are not oversampled in this case are 1.0205dB, 1.0205dB, and 1.0205dB, respectively, and the 4x oversampled raw CMs for 4x oversampling are 1.183dB, 1.183dB , 2.1862 dB, and the maximum cross-correlation value is about 19%.

&Lt; 3 &gt; sequence  For set creation>

As an example of applying the third method described above, a sequence is generated with N = 64 and L = 64 as in the case of the second method described above, and the case of mapping to a frequency domain and transmitting is described below. This method is the same as the second method described above in terms of performance, and as mentioned in the above description, only the sequence in the time domain is GCL. Therefore, in the case of using only the second method described above, the same method can be applied to the case of using the present method. Accordingly, the third method has a problem that the cross-correlation characteristic can not be optimized as in the second method do.

Also, in the case of using this method, two of the three sequence sets are desirably selected to select a combination satisfying the simultaneous correlation, and the other sequence is selected in consideration of other conditions such as aperiodic correlation, PAPR, CM, It is possible to do.

When an optimal set is found for the case of generating three sequence sets using only the third method, M = 1, 59, 63 can be selected. At this time, the non-oversampled PAPR is 1.0904dB, 1.0821dB, and 1.0904dB, respectively, and the 4x oversampled CM is 1.3713dB, 2.521dB, and 1.3713dB, respectively.

7 is a graph showing the aperiodic magnet / cross correlation characteristics when M = 1, 59, 63 are selected as a set of three sequences out of the sequences generated according to the third method.

As can be seen from FIG. 7, it can be seen that the optimum combination also has a cross-correlation peak value of about 24.3%.

The optimal combination selected as described above is also selected in the case of ignoring the PAPR or CM requirement. Therefore, in the case of generating three sequence sets using only the third method, the three sequences It is difficult to satisfy both the periodic / aperiodic magnet / cross correlation property, PAPR or CM characteristics as in the case of generating the set.

&Lt; 3 by the fourth method sequence  For set creation>

As an example of generating three sequence sets using only the fourth method, an example is described in which a sequence is generated with N = 64 and L = 63 and is mapped to the frequency domain for transmission. This method is a method of generating a sequence of L = N-1 and mapping it to a frequency domain in order to generate an N chip sequence in the time domain. For example, when N = 64, a sequence of L = 63 is generated and is mapped by avoiding the DC position.

Basically, in order to maintain the CAZAC characteristic in both the time and frequency regions in the CAZAC sequence, continuous mapping must be maintained as in the second and third methods. However, when mapping in the frequency domain, the present method may have a problem that the discontinuous region occurs due to the DC interval, and the CAZAC characteristic is destroyed in the time domain. Due to this influence, when performing symbol synchronization in the time domain, it is not possible to perform simultaneous correlation, i.e., the property of the symmetric pair can be destroyed.

Therefore, in the embodiment of the present invention, when the sequence is generated by the fourth method, in order to maintain the continuity of the mapped sequence, circular motion is applied in the frequency domain as shown in FIG. 6 to remove the discontinuity in the DC domain Lt; / RTI &gt;

FIG. 8 is a diagram for explaining a method of generating a sequence by using a frequency domain circular motion according to an embodiment of the present invention when a sequence is generated by the fourth method so that the continuity of the sequence is maintained.

Specifically, FIG. 8 illustrates a method of cyclically shifting the sequence to the right by a length corresponding to half of the sequence length in order to solve the problem of sequence discontinuity in the DC region. However, FIG. 8 is an example of solving the discontinuity problem in the DC region through the circular movement in the frequency domain by half the length of the sequence when the length of the corresponding sequence is an even length. On the other hand, when the length of the sequence is an odd number length, the problem of the DC region discontinuity can be solved by a method similar to that shown in FIG. For example, after the position at which the mapping of the sequence starts is set to the second subcarrier, not the first subcarrier of the available subcarrier, the sequence may be continued except for the DC region within the subcarrier window set in this manner Mapping method can be used. However, in the following description, for convenience of description, a method of generating sequences such that the sequences are continuous in a portion excluding the DC region using cyclic shift is referred to as a "half-shift method" do.

Based on the above description, the case of generating three sequence sets according to the fourth method will be summarized as follows.

FIG. 9 is a diagram illustrating a case where a set of M = 1, 2, and 62 is selected as a concrete example of generating three sequence sets according to the fourth method, and a case in which a half shift according to an embodiment of the present invention is not applied Time domain constellation.

Based on this, the signal generated by the fourth method itself can not maintain the characteristic having a constant amplitude in the time domain which is the opposite region due to the discontinuity problem mentioned in the above description. Such a problem can be confirmed through FIG. 9 (a).

However, the method of inserting a sequence so as to maintain the continuity as possible can use the half moving method according to the embodiment of the present invention as described above. (For reference, in the case of the second method or the third method, the continuity condition is satisfied, so that all the same or similar characteristics are maintained for all cyclic shift combinations.) Among these combinations, for example, When the circular motion is applied and inserted as shown in FIG. 6, the constellation is as shown in FIG. 9 (b). (Of course, when the circulation movement is applied to the left side by 32 lengths, it may have the same characteristics because it is symmetrical with the above-described case where the circulation movement is applied to the right side).

If a sequence is generated using the fourth method, a root sequence forming a symmetric pair for simultaneous correlation in the time domain is selected only when a sequence is generated by applying a half moving method according to an embodiment of the present invention It can be seen that there is a number. That is, in FIG. 9 (b), when M = 1, 62, it can be seen that a characteristic having a constant amplitude similar to the case of generating the sequence by the second method and the third method is maintained. In other words, according to one embodiment of the present invention, it is possible to find a combination having a characteristic of maintaining a constant amplitude through proper cyclical movement in the frequency domain.

However, from the viewpoint of the method of generating three sequence sets, only two sequences satisfying the half shift method according to the embodiment of the present invention as described above can be selected, and the remaining one sequence satisfies these conditions A problem that can not be selected occurs.

When three sets of sequences are selected using only the present method, it is preferable to select a sequence satisfying the simultaneous correlation for two sequences out of the three sequences, and the other sequence may be selected from other sequences such as aperiodic correlation, PAPR or CM You can select using the conditions. However, in the case of generating three sequence sets using only the fourth method, since the sequence that can be generated using the half shift method according to the embodiment of the present invention is a maximum of two sequences, There is a problem that it is impossible to generate a sequence set that satisfies the characteristic of maintaining a constant amplitude.

If the requirements of PAPR or CM are ignored for the half shift mentioned in the above description, the combination of M = 1, 62 and 61 is optimal. In this case, the PAPRs which are not oversampled are 1.0298 dB, 1.0298 dB, 3.6715dB, and the 4x oversampled CM is 1.2881dB, 1.2881dB, and 4.3712dB, respectively.

FIG. 10 is a graph showing aperiodic autocorrelation characteristics and cross-correlation characteristics when M = 1, 62, and 61 are selected as the optimal combination when three sequence sets are generated using the fourth method.

However, when the optimal combination as described above is selected, it is also understood that the maximum peak value of the cross correlation has about 22%.

For reference, in the case of the sequences generated by the second method and the third method, the DC puncturing causes a slight distortion in the desired characteristics as described above for the combination of each pair of symmetrical pairs due to the signal in the time domain . However, in the case of the sequence generated using the cutting movement according to the embodiment of the present invention described above with reference to FIG. 6, the symmetry property is perfectly paired without generating distortion such as DC puncturing in the time domain .

In the case of generating three sequence sets by any one of the first to fourth methods introduced in the above description, three sequence sets satisfying all the characteristics of periodic / aperiodic self / cross correlation, PAPR or CM It can be seen that it is difficult to generate. Therefore, in the embodiment of the present invention, in generating two or more sequence sets, the sequence generated according to at least two methods among the first to fourth methods described above is defined as a first type sequence and a second type sequence, respectively And generates one or more sequences from each of the first type sequence and the second type sequence to generate a sequence set capable of satisfying all desired characteristics required in any sequence-based channel by generating two or more sequence sets .

In order to explain one embodiment of the present invention through a more specific example, the following description will exemplify a set of three sequences to be used in a synchronous channel, but three sets of sequences according to an embodiment of the present invention may have the same May be applied to an arbitrary sequence-based channel such as an RACH, an ACK / NACK, an arbitrary control channel for transmitting control information such as CQI, etc., in the same manner, as well as a synchronous channel. In the case of generating two or more arbitrary number of sequence sets . &Lt; / RTI &gt; In the following description, the ZC sequence is used as a sequence to be used, but the same can be applied to any other sequence as long as the characteristic is maintained. In the following description, the generation of three sequence sets using each of the first to fourth methods described above and the generation of a time domain sequence of N = 64 chips are described for the sake of consistency. However, 72, 73, 63 and the like can be applied by the same principle.

In addition, the conditions mentioned for selecting the optimal combination in the case of generating the three sequence sets by each of the first to fourth methods described above are the same as those in the case where three sequences are generated by using the respective methods as the only method In the case of generating two or more sequence sets using two or more methods as in the embodiment of the present invention, the conditions mentioned in the respective methods can be relaxed and applied, and in accordance with an embodiment of the present invention In the case of generating two or more sequence sets using two or more methods, it is selectively applied in consideration of the characteristics of the above-mentioned conditions.

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, a specific embodiment of the present invention will be described.

<First Example - Two in the fourth method Sequence , One in the second method Sequence  How to choose>

In this embodiment, it is proposed to generate three sequence sets by combining the generated sequence using the half shift and the sequence generated according to the second method in the fourth method.

In the case of selecting the sequence in the fourth method (specifically, length (L) = 63), it is preferable to select one of M = 1 or M = 62 considering PAPR as described above.

Considering these considerations, according to the present embodiment, a sequence is generated with L = 63 to select a root sequence index M = 1, 62, a circular motion is applied to the sequence by 32 lengths, and the mapping is applied to the corresponding subcarrier (The optimal combination in the case of selecting two sequences in the fourth method), generates a sequence with L = 63, selects a root sequence index M = 51 (or 13) It is possible to generate the remaining one sequence according to the method.

The correlation characteristics of the selected combination will be described below. However, since the periodic correlation characteristic is the same for the selected combination in all the IDFT operations, a detailed description thereof will be omitted.

FIG. 11 is a diagram illustrating a method of selecting two sequences of M = 1 and 62 among the sequences generated according to the fourth method according to the first embodiment of the present invention, and selecting M = 51 Is a graph showing non-periodic autocorrelation characteristics and cross-correlation characteristics.

11, PAPRs that are not oversampled for each of the sequences M = 1 and 62 and the sequences M = 51 among the sequences generated according to the second method among the sequences generated according to the fourth method, 1.0298dB, 1.0298dB and 1.0821dB, respectively, and the 4x oversampled CM is 1.2881dB respectively. 1.2881 dB, and 2.521 dB, respectively. In the case of generating three sequence sets according to the present embodiment, compared to the case of generating three sequence sets using only the second method, superior performance is obtained for CM and at least one pair of cross correlation pairs . Therefore, it can be seen that the overall cross-correlation level shows better characteristics than the case of generating the three sequence sets using only the second method, although it is almost the same as the case of the second method in terms of the maximum peak value.

The sequence selected by the optimal combination among the sequences generated according to the second method is M = 1 or 63, but in this embodiment, the index of the sequence generated according to the second method based on the length 64 is selected as 51 The index of the lane of the sequences generated according to the second method is selected in order to avoid the high cross-correlation problem with these sequences because the sequence index selected as the optimal combination is M = 1, 62 according to the fourth method.

<2nd Example - Two in the fourth method Sequence , One in the third method Sequence  How to choose>

In this embodiment, it is proposed to generate three sequence sets by combining the generated sequence using the half shift and the sequence generated according to the third method in the fourth method.

When selecting a sequence for three sequence sets from the sequence generated using the fourth method (L = 63), it is preferable to select one of M = 1 or M = 62 considering PAPR.

Accordingly, a method of generating three sequence sets according to the present embodiment can be defined as follows. That is, as a method of generating a sequence according to the fourth method and generating two sequences by a half moving method, a sequence is generated with L = 63 to select a root sequence index M = 1, 62, And the remaining sequences are generated with L = 64 to select the root sequence index M = 5 (or 59) to select the sequence generated according to the second method.

The correlation characteristics of the selected combination will be described below. However, since the periodic correlation characteristic is the same for the selected combination in all the IDFT operations, a detailed description thereof will be omitted.

12 shows a second embodiment of the present invention, in which two sequences of M = 1 and 62 are selected from the sequences generated according to the fourth method, and M = 5 Is a graph showing the aperiodic autocorrelation characteristic and the cross-correlation characteristic in the case of selecting the non-periodic autocorrelation characteristic.

12, the PAPR that is not oversampled for each of the sequences M = 1, 62 and M = 5 among the sequences generated according to the fourth method and the sequence generated according to the third method, 1.0298dB, 1.0298dB and 1.0821dB, respectively, and the 4x oversampled CM is 1.2881dB respectively. 1.2881 dB, and 2.521 dB, respectively. In the case of generating three sequence sets according to the present embodiment, compared with the case of generating three sequence sets using only the third method, superior performance is obtained for CM and at least one pair of cross correlation pairs . Thus, it can be seen that the overall cross-correlation level shows better characteristics than the case of generating three sequence sets using only the third method, although it is almost the same as that of the third method in terms of the maximum peak value.

The sequence selected by the optimal combination among the sequences generated according to the third method is M = 1 or 63, but in this embodiment, the index of the sequence generated according to the third method based on the length 64 is selected as 5 The index of the lane of the sequences generated according to the third method is selected in order to avoid the high cross-correlation problem with these sequences, since the sequence index selected as the optimal combination is M = 1, 62 according to the fourth method.

<Third Example - one in the fourth method Sequence , Two in the second method Sequence  How to choose>

In this embodiment, it is proposed to generate three sequence sets by combining the generated sequence using the half shift and the sequence generated according to the second method in the fourth method.

When selecting a sequence for three sequence sets out of the sequence of the type generated by the fourth method (L = 63) as described above, it is preferable to select one of M = 1 or M = 62 in consideration of PAPR .

Specifically, in the present embodiment, a sequence with L = 64 is generated by the second method to select a root index M = 13, 51, a sequence with L = 63 is generated by the fourth method, and a root index M = 1 (or M = 62), and generates a set of three sequences using a sequence that is circularly shifted by 32 lengths.

Considering the PAPR, the root index that can be selected in the sequence through the fourth method is M = 1 or 62, so that the root index selected in the sequence through the second method is M = 1, 63 Is preferably avoided.

The correlation characteristics of the selected combination will be described below. However, since the periodic correlation characteristic is the same for the selected combination in all the IDFT operations, a detailed description thereof will be omitted.

13 is a diagram illustrating a third example of a method according to an embodiment of the present invention. In the second method, a sequence of M = 13 and 51 is selected from a sequence having a length of 64 and a sequence of M = 1 Is a graph showing the aperiodic autocorrelation characteristic and the cross correlation characteristic of the three sequence sets selected as the in-sequence.

13, PAPRs that are not oversampled for each of the sequences M = 13 and 51 and M = 1, respectively, generated according to the second method and M = 1 among the sequences generated according to the fourth method, 1.0821dB, 1.0821dB and 1.0298dB, respectively, and the 4x oversampled CM is 2.521dB respectively. 2.521 dB, and 1.2881 dB, respectively. Also, the cross-correlation peak value at this time is about 23.8%.

In the case of the second method, M = 1 or M = 63 is optimal. When 1 or 63 is selected, M = 1 (or 62), which is the length L 63 selected in the fourth method, The next lane index was selected because it could have a high cross-correlation problem with the sequence.

<Fourth Example - one in the fourth method Sequence , Two in the third method Sequence  How to choose>

In this embodiment, it is proposed to generate three sequence sets by combining the generated sequence using the half shift and the sequence generated according to the third method in the fourth method.

When selecting a sequence for three sequence sets out of the sequence of the type generated by the fourth method (L = 63) as described above, it is preferable to select one of M = 1 or M = 62 in consideration of PAPR .

Specifically, in the present embodiment, the sequence of L = 64 is generated by the third method to select the root indexes M = 5 and 59, the sequence of L = 63 is generated by the fourth method, and the root index M = 1 (or M = 62), and generates a set of three sequences using a sequence that is circularly shifted by 32 lengths.

Considering the PAPR, since the root index that can be selected in the sequence through the fourth method is M = 1 or 62, the root index selected in the sequence through the third method is M = 1, 63 Is preferably avoided.

The correlation characteristic of the selected combination is similar to that of the third embodiment, and a detailed description thereof will be omitted.

In the case of the third method, M = 1 or M = 63 is optimal. When 1 or 63 is selected, M = 1 (or 62), which is the length L 63 selected in the fourth method, Sequence and a high cross-correlation problem, the next lane index is also selected as in the case of the third embodiment.

<Fifth Example - one in the fourth method Sequence , Two in the third method Sequence  Selection( PAPR , CM Method is not considered)

The second method and the third method are considered to be symmetrical to each other. Accordingly, when three sequence sets are generated using a sequence having a length of 64 and a sequence having a length of 63 without considering PAPR or CM, in the case of selecting three sequences out of the sequences according to the third method and the fourth method .

Specifically, according to a third method, a sequence having a length of 64 is generated, two sequences of M = 1 and 63 are selected, a sequence of length 63 is selected according to a fourth method, and M = 2 (or 61) is selected to select three sequence sets.

The correlation characteristics of the selected combination will be described below.

14 is a view of a fifth embodiment according to an embodiment of the present invention, in which a sequence of M = 1 and 63 is selected from a sequence having a length of 64 according to a third method and a sequence of M = 2 Is a graph showing the aperiodic autocorrelation characteristic and the cross correlation characteristic of the three sequence sets selected as the in-sequence.

14, PAPRs that are not oversampled for sequences of M = 1 and 63 and sequences of M = 2, respectively, generated according to the third method and those generated by the fourth method, 1.0904 dB, 1.0904 dB, and 3.6715 dB, respectively, and the 4x oversampled CM is 1.3713 dB. 1.3713 dB, and 4.3712 dB, respectively. At this time, the cross correlation peak value is about 20.8%.

In the three sequence set generation method according to an embodiment of the present invention as described above, only the sequence generated through only one of the first to fourth methods has a periodic / aperiodic self / cross correlation property and a PAPR or CM It is difficult to generate three sets of sequences satisfying all the characteristics. Therefore, a method of generating three sequence sets by selecting at least one of the sequences generated through at least two methods among the first to fourth methods is proposed will be.

Specifically, in the case of generating the three sequence sets using only the second method and the case of generating the three sequence sets using only the third method, the sequence of the sequence generated to have the even length as described above A sequence set that can be selected to satisfy the desired cross-correlation property can extract only a maximum of two sequences, specifically a sequence that forms a pair of symmetric pairs, and exhibits a high cross-correlation peak value of at least 24% even if any three sequence sets are selected . In addition, when three sets of sequences are generated by using the odd number of lengths, the sequence of the symmetric pair is not established in the odd number length sequence. Therefore, it is impossible to select the three sets of sequences maintaining the above- Able to know. In addition, in the case of generating a sequence having a length of 63 as described above with respect to the case of generating three sequence sets using only the fourth method, a half-moving method according to an embodiment of the present invention described above Since the number of maximum sequences that can be generated is two (i.e., left half shift and right half shift), it is also difficult to select the three sequence sets maintaining the above-mentioned desirable characteristics.

Accordingly, in one embodiment of the present invention, by selecting one or more sequences in each of two or more types of sequences, specifically two types of sequences having different lengths, it is possible to select three or more sequences that satisfy the correlation and PAPR (CM) You can select an open sequence set. With this method, three sequence sets with better performance can be selected without increasing the complexity.

In one embodiment of the present invention described above, a method of generating two or more sequences, specifically, a method of generating two or more sequence sets from two or more types of sequences having different lengths has been described. Hereinafter, another exemplary embodiment of the present invention will be described with reference to the accompanying drawings. In generating two or more sequence sets, a sequence of a specific type and a sequence of a type obtained by modulating the specific type sequence using different kinds of sequences are used to generate two or more sequence sets To create a new file.

Ⅱ. SECOND EMBODIMENT - sequence  Generated by modulation Sequence  Using more than 2 sequence  How to create a set

The above-mentioned ZC sequence is most effective as a sequence used in 3GPP LTE at present. However, due to the time-frequency duality characteristic of the ZC sequence, synchronization in an environment with a frequency offset may result in losing timing in the time domain.

Here, the time-frequency domain duality refers to a phenomenon in which a sequence shifted in one region is also shifted in another region.

For example, when timing synchronization is performed in a time domain in an environment having a frequency offset in a synchronization channel (even when frequency synchronization is performed in a frequency domain in an environment with a time offset) (partial correlation).

In the case of a CAZAC sequence such as a ZC sequence, ideal correlation characteristics as described above with respect to equations (3) and (4) are obtained. However, when partial correlation is performed due to frequency offset as described above, Is destroyed.

If there is a frequency offset, an ambiguity peak occurs in addition to the desired peak corresponding to the original timing.

15 is a graph for showing an example in which an ambiguous peak value occurs when time synchronization is performed in an environment having 5 ppm frequency offset.

Specifically, FIG. 15 shows an example in which an ambiguous peak value occurs at a position of about 75% when time synchronization is performed. Accordingly, if the timing is adjusted to the required peak value in FIG. 15, it does not affect the next step of frequency synchronization, but if the timing at the timing synchronization is adjusted to the ambiguity peak , The correct frequency synchronization can not be performed due to the above-described time-shift duality. That is, when timing synchronization is performed, the estimation performance of frequency synchronization depends on how low the second peak is compared to the peak value required.

Therefore, in one embodiment of the present invention, it is proposed that the ZC sequence be modulated and used by using not only the ZC sequence described above but also a heterogeneous sequence other than the ZC sequence. The sequence used for the modulation of the ZC sequence is preferably a Hadamard sequence, but any other random sequence may be used.

As a specific example according to this method, an example in which a ZC sequence is modulated using a Hadamard sequence will be described. In the following description, a method of generating the corresponding sequence in the time domain, that is, a method of generating a sequence in a time domain, converting the sequence into a frequency domain signal through DFT, and performing DC puncturing on the sequence , The corresponding sequence may be generated in the frequency domain and directly mapped to the corresponding subcarrier. Also, a region in which modulation is performed using the Hadamard sequence is also possible in the time domain and the frequency domain.

First, in the specific example according to the present embodiment, the ZC sequence is generated in the time domain. The ZC sequence generated in the time domain can be expressed by Equation (1). In this example, the case where N = 64 is assumed, but the concept of modulating two or more sequences is the same, so that the same method can be applied to other cases. In this example, the case where M = 1 is specifically described.

Here, a Hadamard sequence is selected as a sequence for performing modulation. The Hadamard sequence selected in this example selects the next sequence, which is the fifth column of the Hadamard sequence having a length of 64.

H = [1 1 1 1 -1 -1 -1 -1 1 1 1 -1 -1 -1 -1 1 1 1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 - 1 1 1 1 1 -1 -1 -1 -1 1 1 1 -1 -1 -1 -1 1 1 1 1 -1 -1 -1 -1 1 1 1 -1 -1 -1 -1]

)는 다음과 같이 나타낼 수 있다. 이때,

는 구성성분대 구성성분(element-by-element) 곱을 수행하는 연산을 의미한다.Accordingly, a Zahar modulated ZC sequence (

) Can be expressed as follows. At this time,

Means an operation that performs element-by-element multiplication.

T = [1 0.9988 - 0.049068i 0.98079 - 0.19509i 0.90399 - 0.42756i -0.70711 + 0.70711i -0.33689 + 0.94154i 0.19509 + 0.98079i 0.74095 + 0.67156i -1 -1.2246e-016i -0.67156 + 0.74095i 0.19509 + 0.98079i 0.94154 + 0.33689i -0.70711 + 0.70711i 0.42756 + 0.90399i 0.98079 -0.19509i -0.049068-0.9988i 1 + 4.8986e-016i -0.049068-0.9988i -0.98079 + 0.19509i 0.42756 + 0.90399i -0.70711 + 0.70711i 0.94154 + 0.33689 i -0.19509 - 0.98079i -0.67156 + 0.74095i -1 -1.1022e-015i 0.74095 + 0.67156i -0.19509-0.98079i -0.33689 + 0.94154i -0.70711 + 0.70711i 0.90399-0.4575i -0.98079 + 0.19509i 0.9988-0.049068i 1 + 1.9594e-015i -0.9988 + 0.049068i 0.98079 -0.19509i -0.90399 + 0.42756i -0.70711 + 0.70711i 0.33689-0.994154 0.19509 + 0.98079i -0.74095-0.67156i -1 + 4.911e-016i 0.67156-0.74095i 0.19509 + 0.98079i -0.94154 - 0.33689i -0.70711 + 0.70711i -0.42756 -0.90399i 0.98079-0.19509i 0.049068 + 0.9988i 1 + 4.4087e-015i 0.049068 + 0.9988i -0.98079 + 0.19509i -0.42756-0.90399i -0.70711 + 0.7071 1i -0.94154 - 0.33689i -0.19509-0.98079i 0.67156-0.74095i -1 -1.6659e-014i -0.74095-0.67156i -0.19509-0.98079i 0.33689-0.994154 -0.70711 + 0.70711i -0.90399 + 0.42756i -0.98079 + 0.19509 i -0.9988 + 0.049068i]

After generating the Hadamard-modulated ZC sequence in this manner, the conversion to the frequency domain using DFT, the DC puncturing, and (optionally) the conversion to the time domain signal can be applied in the same manner as in the existing sequence generation method .

16 is a graph showing a profile when a Hadamard modulated ZC sequence generated according to an embodiment of the present invention is time-synchronized in an environment with 5 ppm frequency offset.

The comparison of FIG. 16 with FIG. 15 shows that the level of the ambiguous peak value is significantly reduced. In addition, the sequence according to the present embodiment maintains the time / frequency flatness characteristics due to the existing CAZAC characteristics and does not affect the cross-correlation characteristics between the sequences at all (specifically, an ambiguous peak value occurs at 10% box).

On the other hand, when selecting the above-mentioned Hadamard sequence, it is possible to combine only those corresponding to the 16th column from the first column when N = 64 (in the above example, N = 64, the fifth column is selected).

To summarize the above-described embodiment, in order to solve the problem of the time / frequency ambiguity that may occur in the CAZAC sequence (ZC, Frank sequence, etc.), by using the modulated sequence using the Walsh Hadamard sequence, It is possible to solve the problem of the above-mentioned ambiguity due to the offset, and in particular, it is preferable to generate using only a part of the Hadamard sequence in order to maintain a flat characteristic in both time / frequency.

On the other hand, in the case of selecting three sequence sets using the generated sequence according to the example using the Z-modulated ZC, the following combinations can be selected. This corresponds to an example of generating a sequence in the time domain.

First sequence: A sequence in which the ZC sequence corresponding to M = 1 is modulated by the Hadamard sequence corresponding to the 11th Hadamard sequence.

Sequence 2: A sequence in which the ZC sequence corresponding to M = 63 is modulated by the Hadamard sequence corresponding to the 11th Hadamard sequence.

Sequence 3: A sequence in which the ZC sequence corresponding to M = 51 is modulated by the Hadamard sequence corresponding to the 8th Hadamard sequence.

Here, the Hadamard index is 0 to 63.

According to the embodiment of the present invention as described above, a sequence less susceptible to frequency offset than a case of using a pure ZC sequence can be generated and used.

According to this embodiment, the ZC sequence of 64 length is generated and applied as it is in the above example. However, in another example of the present invention, after generating the ZC sequence of length greater than 64, A method of generating a ZC sequence having a length smaller than 64, and a method of expanding the generated sequence as described above with reference to FIG. 2, or a combination of various methods Can be used.

In the meantime, the Hadamard sequence used in the above-described embodiment has a Hadamard sequence having a length of 8, specifically a sequence [1 1 1 1 -1 -1 -1 -1] having a length of the fifth Hadamard sequence of 8 It can be seen that it has a repeated form. In another embodiment of the present invention using this principle, a method of repeatedly using a small-length Hadamard sequence in using a Hadamard sequence is proposed.

For example, when a 72-length ZC sequence is modulated using the Hadamard sequence, there is no 72-length Hadamard sequence, so the 8-length Hadamard sequence is repeated 9 times to generate the 72-length sequence described above . The generated 8 × 9 binary sequence can be modulated in a 72-length ZC sequence in an element-by-element manner to generate a Hadamard-modulated ZC sequence according to this embodiment.

Here, when iterating, it is possible to perform simple repetition such as 1 1 1 1 -1 -1 -1 -1 1 1 1 -1 -1 -1 -1, but 1 1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 1 1 1 1, it is also possible to repeat the sign by inverting it at the next iteration.

To illustrate, let's look at an example of a ZC sequence with the simplest length of 12.

Let the ZC sequence of length 12 be represented as follows.

a (6) a (7) a (8) a (9) a (10) a (11)

On the other hand, the Hadamard sequence of length 4 can be expressed as follows.

[h (0) h (1) h (2) h (3)]

Using these sequences, the result modulated according to the present embodiment can be expressed as follows.

(3) a (4) h (0) a (5) h (1) a (2) a (2) a (6) h (2) a (7) h (3) a (8) h (0)

Of course, as described above, the Hadamard sequence having a length of 4 can be generated as follows using the inverted form for each iteration without simply repeating it.

a (4) (- h (0)) a (5) (-) (2) a (3) a (3) h (1)) a (6) (-h (2)) a (7) (-h (3) (10) (- h (2)) a (11) (- h (3))]

In the above example, only a method of generating a sequence in the time domain is shown. However, when a sequence is generated and applied in the frequency domain, the same method can be applied. In other words, the generation method of the sequence is the same as that described above, and only the area mapping the generated sequence needs to be in the frequency domain.

In addition, the following sequence generation method is also applicable.

For example, if a sequence of length 32 is symmetrically mapped (total length 64), a sequence of 32 lengths is symmetrically connected to generate a sequence of 64 lengths and then mapped to another sequence of 64 lengths Modulation may be performed. Of course, it may be used in combination with a method of repeatedly using a small-length Hadamard sequence as described above.

To help you understand, let's look at the following example.

A method of generating a ZC sequence of length 8 by symmetrically connecting ZC sequences of length 4 in generating a ZC sequence of length 8 can be represented as follows.

a (2) a (3) a (2) a (1) a (2)

Here, when modulation is performed by applying a Hadamard sequence of length 4, it can be expressed as follows.

a (3) h (3) a (2) h (2) a (3) a (3) (1) h (1) a (0) h (0)]

That is, in the above example, the Hadamard sequence used for the modulation of the ZC sequence is also symmetrically connected and used for modulation of the ZC sequence.

However, the Hadamard sequence should generally have a power of 2 (or n / 12, n / 20 is a power of 2), and accordingly the Hadamard-modulated ZC sequence should also have an exponent of 2, To generate three sets of sequences using only one of the above-mentioned four methods, the above-described problems can be the same as those in the case of generating three sets of sequences only in a sequence having an even length as mentioned above The same is true for the above-described example for the three sequence sets generated using the modulated ZC sequence). In other words, when three sequences are generated by selecting three sequences only in a sequence having an even length, cross-correlation peak values may represent more than 24%, which may make optimization difficult in terms of cross-correlation.

Therefore, in one embodiment of the present invention, it is proposed to generate two or more sequence sets using a ZC sequence as well as a ZC sequence itself, as well as a ZarD modulated ZC sequence as described above.

Specifically, in the case of generating three sequence sets as an example of two or more sequence sets, it is assumed that two sequences are selected from a Hadamard-modulated ZC sequence, one sequence is selected from the ZC sequence itself, It is possible to generate one sequence from the ZC sequence itself and two sequences from the ZC sequence itself to generate three sequence sets.

As another example, in selecting three sequence sets, one sequence may be selected from a random sequence (or a binary sequence), and two sequences may be selected from a ZC sequence to generate three sequence sets. In this example, when the length of the ZC sequence is N, the index M of the two selected ZC sequences is preferably selected as 1, N-1.

As another example, in selecting three sequence sets, it is also possible to select two sequences from a random sequence (or a binary sequence), and select one sequence from the ZC sequence to generate three sequence sets .

Selecting one or more sequences of the two types of sequences may be selected in consideration of the cyclic / aperiodic auto / cross correlation property and the PAPR (or CM) characteristic as described above.

When generating the three sequence sets according to the embodiments of the present invention described above, one or more of the sequences based on different schemes and / or different lengths are selected to constitute three sequence sets. In one preferred embodiment of the present invention, a length part corresponding to the difference between the generated length and the length to which the sequence is to be applied is generated by circular copying or 0 , And on the contrary, it is generated based on a length larger than the length to which the sequence is to be applied, and then a length portion corresponding to a difference between the generated length and the length to which the sequence is to be applied is cut It can also be applied in combination with a method of generating a sequence.

The above-described first and second embodiments of the present invention can be used in combination with each other. For example, in generating two or more sequence sets, when one or more sequences are selected in each of a first type sequence and a second type sequence generated based on two or more different lengths to generate two or more sequence sets , Any one of the above-described sequences may use a sequence modulated by another sequence.

In another embodiment of the present invention, in generating two or more sequence sets as described above, a sequence set having better correlation characteristics and PAPR (CM) characteristics is generated by using two or more heterogeneous sequences to generate can do. For example, one or more sequences of two or more sequence sets may be selected from the ZC sequence, and the other one or more sequences may be selected from any other type of sequence, such as a PN sequence, to generate two or more sequence sets.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The foregoing description of the preferred embodiments of the present invention has been presented for those skilled in the art to make and use the invention. It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the present invention as defined by the following claims It can be understood that Accordingly, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

According to the embodiment of the present invention as described above, it is possible to generate two or more sequence sets that are both excellent in the correlation property and the PAPR (or CM) characteristic and can reduce the calculation amount in the receiver.

1 is a diagram for explaining a method of generating a sequence according to a cut-type sequence generation method;

2 is a diagram for explaining a method of generating a sequence according to a padding-type sequence generation method;

FIGS. 3 and 4 are diagrams showing examples of the case where three sequences of ZC sequence indexes M = 1, 3 and 63 are selected as a specific example of generating three ZC sequence sets with N = 64 and L = 64 according to the second method Correlation characteristic and cross-correlation characteristic, respectively.

FIG. 5 is a graph showing aperiodic autocorrelation and cross-correlation characteristics for three sequence sets of M = 1, 51, and 63 among the sequences generated according to the second method; FIG.

6 is a graph showing autocorrelation characteristics and cross-correlation characteristics when a combination of M = 1, 67, 72 is selected as a combination of three sequence sets when N = 73 when generating a sequence according to the second method .

FIG. 7 is a graph showing the aperiodic magnet / cross correlation characteristics when M = 1, 59, 63 are selected as a set of three of the sequences generated according to the third method.

8 is a diagram for describing a method for generating a sequence by using a frequency domain circular motion according to an embodiment of the present invention when a sequence is generated by the fourth method so that the continuity of the sequence is maintained.

FIG. 9 is a diagram illustrating a case where a set of M = 1, 2, and 62 is selected as a concrete example of generating three sequence sets according to the fourth method, and a case in which a half shift according to an embodiment of the present invention is not applied And a time domain constellation of the time domain.

FIG. 10 is a graph showing aperiodic autocorrelation characteristics and cross-correlation characteristics when M = 1, 62, and 61 are selected as optimal combinations when three sequence sets are generated using the fourth method.

FIG. 11 is a diagram illustrating a method of selecting two sequences of M = 1 and 62 among the sequences generated according to the fourth method according to the first embodiment of the present invention, and selecting M = 51 A graph showing the aperiodic autocorrelation characteristics and the cross-correlation characteristics in the case of selecting the non-periodic autocorrelation characteristic.

12 shows a second embodiment of the present invention, in which two sequences of M = 1 and 62 are selected from the sequences generated according to the fourth method, and M = 5 Is a graph showing the aperiodic autocorrelation characteristic and the cross-correlation characteristic in the case of selecting the non-periodic autocorrelation characteristic.

13 is a diagram illustrating a third example of a method according to an embodiment of the present invention. In the second method, a sequence of M = 13 and 51 is selected from a sequence having a length of 64 and a sequence of M = 1 And a non-periodic autocorrelation characteristic and a cross-correlation characteristic of a set of three sequences selected as a non-periodic autocorrelation characteristic.

14 is a view of a fifth embodiment according to an embodiment of the present invention, in which a sequence of M = 1 and 63 is selected from a sequence having a length of 64 according to a third method and a sequence of M = 2 And a non-periodic autocorrelation characteristic and a cross-correlation characteristic of a set of three sequences selected as a non-periodic autocorrelation characteristic.

FIG. 15 is a graph showing an example in which an ambiguous peak value occurs when time synchronization is performed in an environment having a 5 ppm frequency offset. FIG.

16 is a graph showing a profile when a Hadamard modulated ZC sequence generated according to an embodiment of the present invention is time synchronized in an environment with 5 ppm frequency offset;

Claims (23)

A method for generating two or more sequence sets, Selecting one or more sequences in each of a first type sequence and a second type sequence generated based on two or more different lengths; And Generating the set of two or more sequences using the selected sequences, Wherein the first type sequence is a sequence generated by performing DC puncturing based on an available subcarrier length including a DC subcarrier, Wherein the second type sequence is a sequence generated based on a length of the available subcarriers excluding the DC subcarriers, the sequence being generated in such a manner that the sequence is not allocated to the DC subcarriers. delete The method according to claim 1, Wherein the two or more sequence sets are a set of three sequences, In the sequence selection step, Selecting one of the first type sequences and selecting two of the second type sequences. The method according to claim 1, Wherein the two or more sequence sets are a set of three sequences, In the sequence selection step, Selecting two sequences out of the first type sequence, and selecting one of the second type sequences. 5. The method of claim 4, Wherein the two sequences selected from the first type sequence are two sequences constituting a conjugate symmetry pair. The method according to claim 3 or 4, Wherein the sequence selected in the sequence selection step of the second type sequence is generated so as to be continuously arranged in an area excluding the DC subcarrier using frequency domain circular motion. The method according to claim 1, Wherein the first type sequence and the second type sequence are generated using a Zadoff-chu (ZC) sequence, And the available subcarrier length including the DC subcarrier is 64. The method of claim 1, The method according to claim 3 or 4, Wherein the sequences selected in the first type sequence and the second type sequence are selected in consideration of at least one of a cross correlation property and a PAPR characteristic in the sequence selection step. The method according to claim 1, Wherein the second type sequence is a sequence generated based on a length less than an available subcarrier length, Wherein the first type sequence is a sequence generated based on a length greater than the available subcarrier length. 10. The method of claim 9, The generated second type sequence is extended by a cyclic copy or zero insertion by the available subcarrier length, Wherein the generated first type sequence is truncated by the available subcarrier length. The method according to claim 1, Wherein the second type sequence is a sequence of modulating the first type sequence using a heterogeneous sequence. A method for generating two or more sequence sets, Selecting one or more sequences in each of a first type sequence and a second type sequence in which the first type sequence is modulated using a heterogeneous sequence; And Generating the set of two or more sequences using the selected sequences, Wherein the two or more sequence sets are a set of three sequences, In the sequence selection step, Selecting one of the first type sequences and selecting two of the second type sequences. 13. The method of claim 12, The first type sequence is a Zadoff-chu (ZC) sequence, Wherein the second type sequence is a sequence of modulating the first type sequence using a Hadamard sequence. delete delete A method for generating two or more sequence sets, Selecting one or more sequences in each of a first type sequence and a second type sequence of two or more different types; And Generating the set of two or more sequences using the selected sequences, Wherein the two or more sequence sets are a set of three sequences, Wherein the first type sequence is a random sequence, the second type sequence is a Zadoff-chu (ZC) sequence, Selecting two sequences from the random sequence and one sequence from the Zadoff-Chesequence to generate the three sequence sets. delete delete delete 17. The method of claim 16, If the length of the Zadoff-Chech sequence is N, Wherein the indices of the two sequences selected from the Zadoff-Cheju sequences are 1, N-1, respectively. A method for generating a sequence based on a length of an available subcarrier excluding a DC subcarrier, Wherein the sequence is generated such that the sequence is not allocated to the DC subcarriers but is continuously arranged in an area excluding the DC subcarriers using a frequency domain cyclic shift. 22. The method of claim 21, If the generated sequence is an even length, Wherein the frequency domain circular motion is a circular motion that is shifted by a half length of the sequence length. 23. The method of claim 22, Wherein the frequency domain cyclic shift is a cyclic shift that moves the sequence by a half length of the sequence length to the left or right in the frequency domain.

Images