KR100260457B1 - A fast search algorithm for inter-cell asynchronous w-cdma system using code hopping method - Google Patents

Abstract

PURPOSE: A base station asynchronous method using an M-notation hopping code and a method for acquiring an initial synchronization are provided to easily carry out initial synchronization acquisition and handoffs by shortening a pilot acquisition time in a W-CDMA cellular system. CONSTITUTION: If power is turned on, a mobile station detects pilot PN sequence start timing using an interface filter. If pilot PN sequence start timing is detected, the mobile station detects a hopping code using a non-coherent correlator. If the detection of the hopping code is completed, the mobile station generates a pilot PN sequence from the detected hopping code and M+1 numbers of binary code pairs. Using the generated pilot PN sequence, the mobile station confirms whether the hopping code was correctly detected. If the hopping code was correctly detected, the mobile station executes the demodulation of synchronous channel data. The base station acquires a 10 msec frame clock through the synchronous channel and enters an idle state. The mobile station acquires all the information of a system in the idle state.

Description

Base Station Asynchronous Method and Initial Synchronization Acquisition Method Using Mgin Hop Code

The present invention relates to a base station asynchronous method and a pilot search method for initial synchronization acquisition for a broadband direct sequence code division multiple access (hereinafter referred to as "W-CDMA") scheme, which is a potential candidate multiple access technology for next generation mobile communication.

The W-CDMA method, which is a ground-based wireless multiple access technology of the next generation mobile communication system (IMT-2000), which is currently being researched and developed in various countries in the world, is mainly based on North American base station synchronous W-CDMA and Japan and It is roughly classified into the base station asynchronous W-CDMA method which is being developed in Europe, etc., and both of the above two methods are being developed in Korea.

The base station synchronization method aligns the base station time of all base stations to an arbitrary reference time (absolute time), making it easier for the mobile station to acquire system time and handoff. At this time, since all base stations must receive a reference time from an external system such as a self-locating system (Golobal Positioning System: "GPS"), the base station synchronization method has a disadvantage that it is difficult to install various types of base stations.

1 shows a base station controller / switch 10 having a GPS receiver 11 and a base station A 20 having a GPS receiver 21 and a base station B 30 having a GPS receiver 31 and the base stations A and B. An example of a synchronous cellular system represented by a mobile station 40 located at is shown.

In contrast, FIG. 2 illustrates an asynchronous method in which the mobile station 40 is located between the base station controller / switch 15 and the base stations A and B 25 and 35 and the base stations A and B 25 and 35. In this method, it is easy to install various types of base stations because no system reference time is required. However, the pilot acquisition procedure and the handoff procedure of the mobile station can be more complicated than the synchronous method.

The base station asynchronous schemes proposed at home and abroad to reduce the pilot acquisition time are three types: the FMA2 (Frame Multiple Access 2) scheme proposed by the European Union, the scheme proposed by SK Telecom, and the scheme proposed by NTT DoCoMo in Japan.

In general, the base station synchronization method distinguishes each base station by an offset of the same code with respect to the absolute time, whereas the base station asynchronous method distinguishes each base station using a different code.

There are two common features of the above three base station asynchronous methods. One is to distinguish base stations by using different binary codes, and the other is that the pilot search time and the number of pilots used in the system. Is proportional.

Among the three asynchronous schemes proposed above, the FMA2 scheme uses different gold codes with short periods as pilot channel codes for each base station to distinguish each base station. The reason for using a short code here is to reduce the initial synchronization acquisition time of the mobile station. The disadvantages of this system are that the number of different pilot codes available in the system is small and for all codes that are used in the system during initial pilot search of the mobile station. Because of the need to search, it may take a long time to search for a pilot despite its short length.

3 shows an example of a code sequence structure transmitted through a pilot channel of the FMA2 scheme.

The base station asynchronous scheme proposed by SK Telecom basically uses two pilot channels for each base station, one of which transmits a very long binary code sequence to distinguish each base station, and the other is a long code. A channel for transmitting a short period binary code sequence including grouping and phase information.

In this method, the number of available pilot codes is relatively higher than that of the FMA2 method, and the reason for using additional pilot channels is to reduce pilot search time which is increased due to long periods of pilot codes transmitted in the main pilot channel.

When powered on, the mobile station first searches for additional pilot channels to obtain long code phase information and grouping information. That is, by introducing the concept of grouping, the mobile station can obtain a pilot without searching all kinds of long codes and search regions.

4 shows an example of a code sequence structure transmitted through a pilot channel of the scheme proposed by SK Telecom.

The main drawback of this approach is the reduction of the channel capacity of the forward link due to the use of two pilot channels, and the introduction of the concept of grouping, which can complicate the design of cellular networks and the long code used by the system even when grouping. The larger the number, the shorter the pilot search time.

The method proposed by NTT DoCoMo assigns different long codes to each base station and introduces the concept of grouping like SK Telecom's method, but uses only one pilot channel.

A feature of this approach is to periodically mask long codes and then insert group identification codes with short periods in that portion. That is, the masked part is a composite component of the short code (all base stations common) and the group identification code used to distinguish the pilot channel from other channels, and the mobile station obtains the long code phase information and grouping information from this part.

FIG. 5 shows an example of a code sequence structure transmitted through a pilot channel of the NTT proposed scheme. The system receives a pilot I-specific M-code hopping code, a common binary code pair, and a binary orthogonal code pair. Q sequence generator 50, multipliers 52 and 53 for multiplying the outputs of the pilot I and Q sequence generators 50 and the outputs of the pilot Walsh code generator 51, and the respective multipliers 52, respectively. ) And FIR filters 54, 55 and gains 56, 57 that filter and gain the output of 53.

The mobile station acquires the timing information of the long code transmitted by the current base station by using a matching filter for the common short code, and obtains the group information of the long code by checking the group identification code.

In this way, like the SK telecom method, the mobile station can acquire a pilot without searching all kinds of long codes and all search sections.

The method is less serious than the SK telecom method using two pilots, but in the masking part, the orthogonality of the pilot channel and the traffic channel disappears, so that the capacity of the forward link can be reduced.

In addition, as SK Telecom introduces the concept of grouping, cellular network design can be complicated, and the larger the number of long codes used, the larger the pilot search time.

The purpose of the present invention is to achieve a pilot search time regardless of the number of pilots used in the system and the number of pilots used in the system, even if the initial pilot acquisition time of the mobile station is very short, even if the operation is asynchronous between base stations. Since no grouping concept is required to reduce, the cellular network is simple to design and does not require redundancy for grouping, enabling high-speed data transmission without attenuation of the channel capacity of the forward link, and thus a potential candidate multiple for next-generation mobile communications. The present invention provides a base station asynchronous method and an initial synchronous acquisition method using an M-based hopping code that can easily reduce initial pilot acquisition time and handoff in a broadband direct sequence code division multiple access cellular system.

In order to achieve the above object, the present invention distinguishes each base station by using different M-ary hopping codes when the base stations are asynchronous for the wideband code division multiple access scheme, and each base station has the unique M-ary hopping code and M + for each base station. Combining one binary code pair to generate a pilot sequence, one of the M + 1 pairs of binary codes is used as a common code to find the starting point of the pilot sequence, and the remaining M codes of the M code are Characterized by the use of the alphabet.

FIG. 1 is a schematic diagram of a synchronous system that maintains synchronization between base stations using a GPS and distinguishes base stations using a phase difference of the same code with respect to absolute time.

2 is a schematic diagram of a system for distinguishing base stations with different codes and operating asynchronously between base stations.

3 is a pilot code sequence time chart of a base station asynchronous scheme (FMA2) proposed by the European Union.

Figure 4 is a pilot code sequence time chart of the base station asynchronous method proposed by SK Telecom.

Figure 5 is a base station asynchronous pilot code sequence time chart proposed by NTT DoCoMo.

FIG. 6 is a pilot channel diagram illustrating a pilot PN sequence by combining base stations with different M binary hop codes according to the present invention, but combining a unique M binary hop code and M + 1 binary codes for each base station.

7 is a time chart of a pilot PN sequence generated by combining a binary code and an M binary jump code in accordance with the present invention.

8 is a block diagram of an embodiment of a pilot sequence start point detector using a baseband matched filter of I and Q common codes according to the present invention.

9 is a block diagram of an embodiment of an M-ary non-coherent parallel correlator for code element detection of an M-based hop code according to the present invention.

10 is a flowchart illustrating an initial system synchronization acquisition procedure of a mobile station according to the present invention.

11 is a waveform diagram of a binary code signal having a length of four.

12 is a configuration diagram of a longest length sequence generator having a period of 63.

13 is a graph showing autocorrelation values of an extended longest length code.

14 is a graph showing the cross-correlation value of the Dean Gold Code.

15 shows a hopping code set generated from an R-S code set.

16 and 17 are graphs showing cross-correlation values of two arbitrary jump codes.

18 is a timing diagram of the forward channel.

※ Explanation of symbols about main part of drawing ※

10,15: Base station controller / switch 11,21,31: GPS receiver

20,25,30,35: base station 40: mobile station

50: pilot sequence generator 51: pilot code generator

54,55: Filters

Hereinafter, the present invention will be described with reference to the accompanying drawings.

The pilot channel of the forward link in a CDMA cellular system is used by the mobile station as a reference for initial system synchronization for the current base station and for time, frequency, and phase tracking for signals transmitted at the current base station.

First, in the base station asynchronous method of the present invention, each base station is assigned a hopping code unique to the base station having a length L and a size of a value each code element can have (hereinafter, referred to as an "alphabet size"). Receive.

Each base station combines the base station-specific hopping code and M + 1 binary code pairs (QPSK PN spreading) to generate base station-specific pilot PN (binary) sequence pairs. The length of the binary code used here is N, one of the M + 1 binary code pairs is used as a common code pair to find the starting point of the pilot sequence, and the other M pairs are the base station's unique jump code alphabets. It is used in common at all base stations. 6 shows a pilot channel structure of any one base station of the present invention.

In the present invention, the first α binary code pairs of the pilot sequence are used by all base stations. Therefore, the period of the pilot sequence is (α + L) × N chips.

FIG. 7 shows M-based jump codes {4, 1, 0, 5, 6, 3} and {6, 0, 2, 3, 4 when M = 7, L = 6, N = 16 and α = 1. , 5} shows a time chart for a pilot PN sequence transmitted from two base stations.

In the present invention, the mobile station finds a starting point of a pilot sequence transmitted from a base station having the largest received signal size using a matched filter for a common binary code. By comparing the sampling values of the ary parallel correlator, L code elements constituting the unique hopping code used by the base station can be derived.

FIG. 8 shows an example of a pilot sequence starting point detector using a matched filter of I and Q common codes, and FIG. 9 shows an M-coin noncoherent parallel correlator as an example for detecting code elements of a hopping code. .

In the base station asynchronous method of the present invention, the initial pilot search and frame clock acquisition procedure of the mobile station will be described according to the flowchart of FIG.

First, when the power is turned on, the mobile station detects the starting point of the pilot sequence transmitted by the current base station by using a matching filter for the common code of FIG. Upon detecting the starting point of the pilot sequence according to the execution of the above steps, the mobile station detects the code element of the leap code every N chips using the correlator of FIG.

After the detection of the L code elements constituting the hopping code by the above code element detection process, the mobile station generates a pilot PN sequence to be used by the mobile station from the detected hopping code and M + 1 binary code pairs.

Thereafter, the mobile station uses the generated pilot PN sequence to check whether the hopping code is properly detected through correlation with the received signal.

If the hopping code is not properly detected in this step, the process returns to the initial state and if it is properly detected, demodulation of the synchronization channel data is performed.

The sync channel basically includes information about an interleaver frame clock and a hopping code used by the current base station.

The mobile station acquires information on the frame clock through the synchronization channel, compares the hopping code currently used by the mobile station with the hopping code obtained from the synchronization channel through the synchronization channel data demodulation, and corrects any errors.

The sync channel frame has a length of 10 msec and the start point of the frame aligns with the base station time reference.

The mobile station, which acquires 10msec base station time through the synchronization channel, enters an idle state, where the mobile station acquires all the information of the system.

The main feature of the proposed method is that the initial synchronization acquisition time is very short and the initial synchronization acquisition time is different from the number of different pilot sequences used in the system. It doesn't matter.

Of the 2 (M + 1) binary codes used for the present invention, two (I, Q) common codes have good autocorrelation characteristics and the remaining 2M binary codes are orthogonal.

nality) and the crosscorrelation property defined as Equation 1 between 2 (M + 1) whole codes should be good.

In Equation 1, Tc is one chip length of a binary code, and x (t), y (t), and z (t) each represent one of 2 (M + 1) binary codes and have a value of ± 1. .

In FIG. 11, as an example, x (t), y (t), and z (t) when N is 4 are shown by the time chart.

In addition, the jump cord used by the present invention should have good cross-correlation properties. When there are two hopping codes defined as in Equation 2, the cross-correlation property of the two hopping codes is represented as in Equation 3.

In Equation 3, δ (x, y) has a value of 1 when x and y are the same and 0 when they are different. In addition, (x) _L means the remainder of X divided by L.

In the present invention, an extended gold code is considered as an example of a binary code. The extended gold cord is obtained by adding one more zero to the end of the gold cord with a length of 2n-1. Therefore, the length N of the extended gold cord is 2 ⁿ .

Gold sequences of period 2 ⁿ −1 are generated from two maximum length sequences having the same period.

As an example, the present invention considers the case where n = 6. This value may be used in actual system design. 11 shows two longest length generators having a generator polinomial given by way of example.

In Equation 4, D is a delay operator, and the two longest length sequences having respective generation polynomials have a period of 63.

When the longest length sequences generated by the generated polynomials are expressed as m1 and m2, respectively, the number of gold sequences generated therefrom is 65 (= 2 ⁶ +1) as follows.

Each sequence of the gold sequence set given above has a period of 63. The extended gold code set used in the present invention is made by extracting only one period of each sequence constituting the gold sequence set and adding one zero at the end.

Equation 6 shows the extended gold code set thus created.

Each code of the extended gold code set given by the above equation is 64 in length and orthogonality is satisfied between the codes except for c ₂ .

In addition, each code in the above equation is a modified gold code, but the cross-correlation property between the codes is still good, and since c ₁ and c ₂ are elements of the extended gold code set and modified longest length codes, respectively, the auto-correlation property is very good.

FIG. 13 shows autocorrelation characteristics for the extended longest length code c ₁ . In addition, the cross-correlation property given by Equation 1 between the extended gold codes g ₇ , g ₂₃ , and g ₁₁ is shown in FIG. 14.

13 and 14, it can be seen that the extended gold code set satisfies a necessary and sufficient condition to be a binary code set to be used in the present invention.

The alphabet size M of the jump sign used in the embodiment of the present invention is 31. That is, the required binary code pair is 32 pairs including the common code pair.

Therefore, only 64 of the 65 gold codes given in Equation 6 are used. That is, two of the 64 gold codes are used as a common code pair and the remaining 62 codes are 31 pairs. This is shown in equation (7).

The hopping code set used as an example in the present invention uses a subset of the R-S code sets having alphabet size M of 31. 15 shows a relationship between the hopping code set and the R-S code set.

The R-S code word defined in GF (M) and implying k + 1 information digits is given as a low space of the following generator matrix. (Here GF stands for Galois Field and all operations are done modulo M.)

Β in Equation 8 is a primitive element of GF (M). The arbitrary code r generated from the generation matrix is given by Equation (9).

r = [no, n1, n2, ..., nk] G, where nk ∈ {0, 1, 2, ..., M-1}

The length L of the RS code is M-1 and it can be seen from Equation 9 that the number of RS codes constituting the RS code set is M ^{k + 1} .

A characteristic of the R-S code is that it does not exceed k when the cross-correlation value (number of overlaps) given by Equation 3 is j = 0. However, the R-S code set itself is not suitable as a jump code set to be used in the present invention. This is because the condition of the cross-correlation value being less than or equal to k does not consider the time shift of the code (j ≠ 0 in Equation 3).

A code set whose cross-correlation value is less than or equal to k at any time shift from an RS code set with M ^{k + 1} codes is fixed by fixing the second element of the input k + 1 tuple, the value of n1 to any nonzero value. It can be produced as a combination of values that the remaining k elements can have.

Therefore, it is possible to extract a set of leap codes having M ^K elements having less than k cross-correlation values between code words at any time shift from the RS code set having M ^{k + 1} elements.

Table 1 shows the number of different jump codes (different pilots) available in the system depending on the value of k when M is 31.

k (maximum cross-correlation value) Hop Code Count (pilot count) 2 31 ² = 961 pcs 3 31 ³ = 29791 pcs 4 31 ⁴ = 923521 pcs 5 31 ⁵ = 28629151 pcs

In the present invention, numbers are used for each of the M ^K jump codes extracted from the RS code set. The code number is generated as shown in Equation 10 by using the remainder except for n ₁ which is a fixed value among k + 1 input tuples.

Jump code number = n ₀ × M ^k-1 + n ₂ × M ^k-2 +, ..., + n _k-1 × M + _nk

As a design example, considering that M = 31, k = 2, and β = 3, there are 961 different jump codes (number of pilots) available in the system. In this case, any jump code r may be generated as shown in Equation (11).

Equation 12 shows codes 39, 257, 425 and 634 generated using Equation 10.

r ₃₉ = (14 7 1 25 10 19 5 28 0 4 2 25 14 5 8 8 20 9 18 20 2 19 4 16 7 9 28 16 10)

r ₂₅₇ = (30 2 24 21 11 3 22 1 11 16 23 3 8 19 16 24 15 1 14 21 2 19 23 15 28 28 18 22 30 18)

r ₄₂₅ = (28 6 20 7 0 19 2 29 6 24 24 3 30 22 3 22 19 28 0 10 18 30 20 10 5 29 18 13 2 5)

r ₆₃₄ = (25 16 23 9 24 24 22 29 12 22 12 25 17 4 2 19 29 0 2 3 23 19 20 16 3 17 9 0 15 4)

16 and 17 show the cross-correlation values (number of overlaps) defined by Equation 3 for jump codes 39 and 257 and 425 and 634, respectively, with the maximum value of cross-correlation exceeding k (= 2) at any time shift. It can be seen that.

18 shows an example of a timing diagram of a forward channel of any base station to which hopping code 39 has been assigned. The chip rate is assumed to be 4.096 Mcps, the frame length is 10msec, and α is 2. In this case, one period of the pilot sequence is 500 μsec (2048 chips) and is repeated 20 times during the 10 msec frame time.

As described above, the present invention basically does not require synchronization between base stations, and thus does not require a GPS receiver, and since the base station is distinguished using an M binary hopping code, the number of pilots that can be used without pilot reuse (= hopping) The number of codes) becomes very large, and the existing base station asynchronous schemes distinguish base stations using unique base station-specific binary codes. Since M + 1 binary orthogonal codes, which are common and have very short lengths and have very good cross-correlation properties, are used in combination, the pilot acquisition time of the mobile station is very short.

In addition, the present invention has a great advantage that the more the number of pilots used in the system is irrelevant to the existing methods that require a lot of pilot acquisition time, and the existing method of introducing a grouping concept to shorten the pilot acquisition time. Because it does not need to introduce the concept of grouping, the design of the cellular network is easy and the complexity of the terminal can be relatively reduced, and since the redundancy for group identity coding is not required, the capacity reduction of the forward link can be eliminated. In addition, the initial synchronization acquisition algorithm of the terminal is simplified.

Therefore, the inter-base station asynchronous method and the early synchronization acquisition method of the mobile station of the present invention for the W-CDMA cellular system of the present invention have an advantage that the pilot acquisition time is very short, which facilitates initial synchronization acquisition and handoff.

In particular, in the conventional asynchronous method, the larger the number of pilots used in the system, the longer the pilot acquisition time, while the present invention shows the advantage that the pilot acquisition time is the same regardless of the number of pilots used in the system.

Claims (4)

In the wideband code division multiple access scheme, each base station is distinguished by using different M binary hopping codes, and each base station combines the M binary hopping code and M + 1 binary code pairs unique to each base station to generate a pilot sequence. M binary leap, wherein one pair of binary codes of the M + 1 pair is used as a common code for finding a starting point of a pilot sequence, and the remaining M are used as code alphabets of the M binary jump code. Base station asynchronous method and initial synchronization acquisition method using code. 2. The base station asynchronous method and initial synchronization acquisition method according to claim 1, wherein the base station asynchronous code uses a subset of the R-S code set as the binary code. In the asynchronous method of claim 1, the mobile station finds the starting point of the pilot sequence using a matched filter for a common binary code, and then uses an M-parallel correlator to infer the hopping code used by the current base station to synchronize the system. And a base station asynchronous method and initial synchronous acquisition method using M-based hopping code. The method of claim 1, wherein the base station asynchronous method and initial synchronization acquisition method using the M-code hopping code, characterized in that an extended gold code is used as a binary code.