TWI232043B - Method and apparatus for cell search for W-CDMA with non-ideal sampling - Google Patents

Abstract

The invention discloses a cell search method for a CDMA system for a three-stage cell search procedure. The cell search method has the steps of matching of incoming signals by a chip matched filter; over-sampling the incoming signals N times against a chip rate by a sampling device; transmitting the N over-samplers into a serial test unit, selecting a best sample among N over-samples by the serial test unit controlled by a first stage detector; and transmitting the best sample of the N over-samples into a second stage detector and a third stage detector for a trial. The best sample of the N over-samples has a maximum sampling value of the N over-samples. The invention significantly reduces the effects of non-ideal sampling so as to accomplish fast cell search.

Description

|| r: Case No. 911gli _η Revision V. Description of the invention (1) [Technical Field to which the Invention belongs] The present invention relates to a system for cod, division multiple access (CDMA) Cell search method and device, and more particularly, it relates to a cell search method and device for a wideband CDMA multiplexing access system. The method and device reduce the effect of non-ideal sampling in the system by using a cell search algorithm. . [Previous technology] The code division multiplexing access cellular system using a direct sequence spread spectrum multiplexing access technology greatly increases the channel capacity. This system has attracted considerable attention in recent research on mobile communication systems. Generally speaking, ‘due to the nature of frequency reuse’, the bandwidth efficiency of a code division multiplexed access system is better than other multiplexed access systems (such as frequency division multiplexed access and time division multiplexed access). In addition, cell planning in the code division multiplexing access system is fairly simple. Therefore, the code division multiplexed access system will be the mainstream in the future. The 3rd Generation Partnership Project (3 G P P) Wideband-Division Multiplexing Access / Division Duplex (W-CDMA / FDD) system has been adopted as one of the standards for the third-generation system of IMT-2000. In a cellular system of one-code multiplexed access, the method used by a user device (UE) to search for the best cell is called π-cell search. "A fast cell search system is very important. Switch-on delay (initial search), increase standby time (idle state search) and handover (operation state aCtiVe mode search) to maintain good communication

Page 7 ΐ 2 ι): Case No. 91105519_year month day__ 5. Description of the invention (2) Quality of information link. See U.S. Patent No. 6, 0 38, 250, issued to Shou et al., Entitled π Initialization Synchronization Method and Direct-Sequence Code Division Multiplexing Access Receiver Receiver for Asynchronous Cell System π (I nitia 1 Synchronization Method And Receiver for DS-CDMA Inter Base Station Asynchronous Cellular System) Reveals an initialization synchronization method for high-speed searched cells and a receiver for direct sequence type code division multiplexing access repeater Synchronized Cell System. A baseband acceptance signal is input to a matched filter and is associated with a spreading code provided by a spreading code generator. A signal electric power calculator calculates the electric power of the relevant output of the matched filter, and outputs the result to a long code synchronization timing determiner, a threshold value calculator, and a long code verifier ( long code identifier). During the initial cell search, the spread-spectrum code generator outputs a short code which is common to the control channel of each base station. After the long code synchronization schedule has been determined, each of the N chips that make up a part of the synthesized spreading code sequence is sequentially replaced and output.

See U.S. Patent No. 6, 1 8 5, 244 issued to Nystrom et al., Whose title is "Cell Search in Code Division Multiplexing Access Communication System", revealed in Cell Search in Code Division Multiplexing Access Communication During the system, a more efficient encoding plan requires a long code and a code frame time history. A code set with a sub-length of M_ (MQ — subscription), where the codeword contains a set of Q short codes, The code set is defined to a specific characteristic. The initial characteristics that are satisfied are that the codeword has no cyclic drift and thus generates a valid codeword; the others are

The satisfying characteristic is that there is a one-to-one 1 (mapping) between the long code message and the valid code word, and the encoder will find that it appears to interfere with the movement of the messenger (thus the time frame of the code frame) and The transmitted codeword (ie, long code representation message) has a certain degree of accuracy and reasonable complexity. No. 6, in U.S. Patent No. 6, 28, 0, 0, 7 issued to Klm et al., "2," is the acquisition of a mobile communication system using asynchronous code division multiplexing access.肊 Base station method, revealing that a group of codes and cell codes are multiplexed = and used as a guide code. For different base stations, asynchronous communication is used for multiplexed access to the communication system; using multiplexed codes, interference Reduced due to the use of two pilot codes. One method for obtaining a cell base station in asynchronous multiplexing access is to include a base station control: a plurality of mobile stations and base stations, and the use of different For different base stations in sequence, the method of searching for a cell base station includes (a) specifying the cell, and the group code as the guide code (d) of the in-phase (丨 nphase) channel of the base station, and specifying the group code of κ and 'field cell The pilot code of the quadrature channel of the base station multiplexes the pilot code of the inphase channel and the quadrature: channel and generates the pilot code of the same phase and the quadrature channel. Please refer to the first figure, The figure will help to understand the simplified code frame structure of a 3rd Generation Cooperative Computing (3GPP) broadband frequency division multiplexing access / division multiplexing system. First of all, in the 3rd Generation Cooperative Computing (3Gpp) wideband frequency division In the code multiplexed access / frequency division multiplexed system, the cell search system typically uses three stages to form the 'edge two stage system, which includes two uniquely set synchronization channels and a common guidance channel. In the first stage, In 1 丨 〇, the primary (pr i mary) synchronization channel (PSCH) 1 1 1 is used for time slot (siot) synchronization. The primary synchronization channel

a 丨 1 (= ± 1) system 111 contains the primary synchronization ant (PSC) defined as acp, which depends on the existence of the diversity transmission (diversi ty 311 ^ 4. 〇11) transmitted at the base station. In the second stage 120, the secondary channel (SSCH) 121 is used for code frame / code group (code gr0up) identification. The secondary synchronization channel 121 contains secondary synchronization codes (SSCs) defined as acs, where the coefficient cs is similar to that of the primary synchronization channel. In the third stage} 3, 0, the common guidance channel (CPICH) 131 is used for the decision of downloading the scrambling code. As shown in the figure, 15 slots are included in the 10 (millisecond code box), and because this system uses 3.84 million chips per second (elbow (: 1 ^ 03/36 (:) Speed. So, there is 384〇〇

There are chips in each frame and 256 chips in each time slot. In addition, the primary synchronization channel and the secondary synchronization channel are 256 chips often and only transmitted at the beginning of the / time slot boundary. / The traditional cell search of the 3rd Generation Partnership Project (3GPP) wideband code division multiplexed access / frequency division multiplexed system can be divided into two general categories: serial search and catheter Pipelined search. The sequential search needs to go through three synchronization stages one after the other before the next new search starts. A sequential search (s e r i a 1) search usually takes more time than a pi pel ined search. The three stages of the ducted (p i p e 1 i n e d) search can be operated simultaneously, so the search time can be reduced. Of course, a high cost of complexity and computational power must be paid. 'Please refer to the second figure' This figure is a simplified diagram for a traditional three-stage sequential cell search algorithm. For simplicity, a complete three-stage search will be referred to as a trial (t r i a 1). In sequential cell search,

Page 10 \ 4imm, i I ή 丨 91 91519519_ 年月 日 _ «_ V. Description of the invention (5) The experiments do not overlap until the search is successful, that is, there can be only one at a time. Phases (ie, blocks 2 1 1, 2 2, and 2 1 3) are in operation, and therefore consume lower power loss, but take longer search time. On the other hand, referring to the third figure, this figure is a simplified diagram for a conventional three-stage catheterized cell search algorithm. In catheterized search, detection at different stages is performed simultaneously, and tests are overlapped. That is, boxes 3 1 1, 3 2 1 and 3 3 1 are in the same test. Obviously, in the guided search, there are more possible tests in a fixed time, and therefore a faster search is obtained. Of course, the price paid is greater power consumption. In the past, only this ducted search was more interesting. It should be noted that for catheterized search, there is no additional hardware device required than sequential cell search. In fact, the overall process time of an experiment does not require 30 milliseconds (ms), as shown in the figure. The cell search time of (K +2) χ 10 (ms) is required for each search if the cell search succeeds in the Kth test. However, the general basic assumptions are often used in the previous technology of the 3rd Generation Partnership Project (3GPP) wideband CDMA multiplexed access / frequency multiplexed cell search. That is, the sampling at the chip-matched filter output is perfect (ideal sampling). In fact, the sampling system at the chip-matched filter output is not perfect (ie, non-ideal sampling). The non-ideal sampling will reduce the signal-to-noise ratio and increase the cell search time. Due to the non-ideal sampling, the operational characteristics of the cell search were severely disrupted. [Summary of the Invention]

Yan No. 91105519 January __Revision V. Description of the invention (6) The main purpose of the present invention is to provide a cell search method for a code division multiplexing access system, and more particularly for a wideband code division multiplexing access. System to reduce non-ideal sampling effects and quickly complete cell searches. A secondary object of the present invention is to provide a cell search device for a code division multiplexing access system, and more particularly for a wideband code division multiplexing access system 'to realize cell search under non-ideal sampling effects and Does not increase the complexity of the hardware. In order to achieve the above-mentioned main purpose, the present invention provides a cell search method for a code division multiplexing access system, and more particularly for a wideband code division multiplexing access system to reduce non-ideal sampling effects. A new cell search algorithm that uses a three-stage channel system in the 3rd Generation Partnership Project (3GPP) standard. The first phase has a primary synchronization channel (PSCH) for time slot synchronization; the second phase has a secondary synchronization channel (SSC Η) for code frame / code group identification after time slot synchronization; and the third The stage has a common guidance channel (CPICH) for the decision to transmit the scrambled code. The cell search method includes the following steps: the received signal first passes through a chip matched filter (ch ip matched fi Iter), and the signal 7 > The medium sampling frequency is N times the chip rate. In this case, the N consecutive sampling points are randomly changed at regular intervals to determine whether the signal detector is used as the input of subsequent signal detectors, and the signal detectors for signal inspection are divided into three stages. In the first phase, the second phase is mainly to detect the timing of the primary synchronization channel (PSCH). Because the primary synchronization channel is synchronized with the time slot, the blindness of the synchronization channel is reported at the beginning. If the sequence is broken, the time slot synchronization is completed at the same time.

Description of the invention (7) The two-stage signal detector is based on the primary, pre-identified group co-detection: dagger-time secondary synchronization channel detection. Because the sequence of the second level: track's step-by-step passes through the bat code, the signal carried by the 8-step channel after the encoding is predicted. Therefore, when the second order is completed, 俨 = can be provided as code frame and code group synchronization. In addition, the third phase of the message = ^ detection, that is, the code frame code step can be completed, and the second phase of the code frame to use the time slot test of the Jieyi ’, the younger brother. Among them, the 3 code group selected in the third stage is synchronized to perform chaotic code comparison. If the selected scramble code is lower than this: the scramble code will be further detected with a threshold and three stages. Otherwise, if this is the threshold value, another further confirmation is started. When the mixed IL code enters a synchronous verification unit through the garbled code of J, the scrambled code taken is accepted. Otherwise, after V verifies the earlier detection, it is selected because the synchronous verification unit processes jealousy = restarts another three-stage detection. Late time τ produces seconds (ms)). Therefore, 1 long (it is assumed in this description that the delay cannot pass the synchronization verification unit. The garbled code passes the first threshold but at the same time, it will greatly reduce the time spent in the synchronization verification unit to obtain two copies: Λ method One of the features, where the best sampling point where no interference occurs is the best sampling point in the _ sampling point, which has the standard ΐ :: Ϊ supply-a cell search device for code division multiplexing Take the number ρ on the wideband code division multiplexed access system. The device processes the first stage with a primary synchronization channel PSCH for time slot synchronization; the second stage has a secondary synchronization channel / SCH at time After slot synchronization, it is used for code frame / code group identification; and in the third stage, the common channel (CPICH) is used for the decision of transmitting scrambled codes; 谠 The cell search device includes: a chip matched filter for Match one advance ^ Case No. 91105519! 423i (W3 i two-: one fifth, invention description (8) ^ ^ --- feL _ input signal; a sampling element, connected to the stone horse piece _ A chip rate is oversampled N times. samp 1 er), connected to the sampling element, used to pass ~ sampler (d 〇w η-points; a sequential test unit, connected to the descending ladder f is _ oversampling oversampling points to choose a maximum sampling Point; a complex — for the N sequential testing units for obtaining the first order === test family, connecting a two-stage detector, connecting to the sequential testing unit for obtaining the code group f Code frame synchronization; a third-stage detector connected to the sequential test unit to determine a scrambled code; and an authentication unit connected to the third-stage detector 'to determine whether the test was successful. The The authentication unit further includes: a comparator connected to the third-stage detector to test the output value of the third stage and a threshold value "this comparison; a first determining element connected to the comparator to determine Is the scramble code correct? If the threshold is exceeded, the selected scramble code enters a synchronous verification unit, otherwise the test is considered a failure and a new test will be restarted with the delay; the synchronous verification Unit, linked to that first decision A determining element for verifying the scramble code; and a second determining element connected to the synchronous verification unit for determining whether the scramble code is accepted; wherein if the f-mixed gas code passes the second determining element, the test is successful Otherwise, a new f-test is restarted after a delay time T 麾 seconds (ms), and the sampling points of the matched filter are assumed to be uniformly distributed. A feature of the device according to the present invention, in which no interference occurs 'The optimal sampling point in the N oversampling point has the largest sampling value in the N oversampling point. Other features and advantages of the present invention will be described in detail below and attached.

Case No. 91105Fi1Q Month, Amendment-V. Description of Invention (9) The illustrations and defined patent scope are obvious. [Embodiment] The main idea of the first-stage sequential testing method is that the selected best sampling point has the maximum sampling value of the N oversampling point. The optimal sampling point selected in the first stage can increase the operating characteristics when non-ideal sampling is considered. Please refer to the fourth figure, which shows a cell searching method according to the present invention, which uses the first stage sequential testing method. According to the present invention, a total of 5 1 2 scramble codes are used in different cells at the time of downloading and are used repeatedly in the entire system. The code is subdivided into 64 groups, each group has 8 codes. Each code is 3 840 chips long and is therefore extended to the entire code frame. Since the cell positions are not synchronized, the code always starts its new cycle at the border of the code frame. As shown in the fourth figure, this method deals with three stages. The first stage (4) has a primary synchronization channel (fPSCH) for time slot synchronization. By using the same primary synchronization channel at each seek and by sending only the primary synchronization channel at the edge of the time slot, time slot synchronization can be easily achieved by synchronizing to the primary synchronization channel. Furthermore, the generalized hierarchical Golay sequence (generalized hierarchical Golay seqUence) is used as the primary synchronization code because it is easy to implement. The second stage 4 2 0 has a secondary synchronization channel (SSCH) which is used for code frame / code group identification after time slot synchronization. Synchronous group identification can be achieved by detecting the secondary synchronization channel, where 哕, ... are spread by one of the 16 orthogonal spreading codes, which is called secondary iso-g code. 'Carved face 3: R! Car 丨! η cable number ^, one ____________; 5. Description of the invention '(10) In order to reduce mutual interference, the secondary synchronization code is orthogonal to the primary synchronization code. In addition, in order to achieve fast frame / code group identification, the secondary synchronization channel is further encoded into a group of 64 code words, using a (5, 3) uninterrupted Reed-Solomon code (comma —f ree Reed s〇1⑽㈣c〇de, CFRS) 'where each codeword of the group is represented as a code group. Because of the good characteristics of the non-interval segment, once the code group is specified, the code frame synchronization system is completed. The third stage, 430 has a common guide channel (cpiCH), which is used to pass down the chaos of chaos. After the child code group is specified, by selecting one of 8 codes in the group, the chaotic code system can be simply determined by using a common guidance channel. The fifth diagram is a private diagram of a cell search method according to an embodiment of the present invention. The method uses a first-stage sequential test (s e) a ^ / n, STS1 method. To describe the process, the sixth diagram also shows the cell search architecture according to the present invention, which uses a serial test in stage—

Page 16 1'STSl) method. As shown in the figure, the architecture uses the first stage sequential test method (STS1) method. A cell search device is used in a code division multiplexing access system, and more particularly is used in a wideband code division system. The device processes three stages of channels. A chip matching filter 2 1 0, a sampling element 2 2 0 is connected to the chip matching filter ^ 1 ′, a down-sampler 2 3 0 is connected to the sampling element 2 0 ′, a The sequential test unit 2 4 5 is connected to the sampler 2 3 0, a first stage, the tester 2 40 is connected to the sequential test unit 2 45, a second-stage detector surname 50 is connected to the sequential Test unit 245, a third-stage detector with 260, ,,, and. Up to the sequential test unit 245, an authentication unit 28o is connected to the third stage. ¾ han · [; cable number 91105519 month day 月 π: 5. Description of the invention (11) Segment detector 260. The authentication unit 2 0 0 further includes a comparator 2 7 0 connected to the third stage detector 2 60, a first determining element 2 7 5 is connected to the comparator 2 7 0, and a synchronous verification unit 2 90 is connected to the first determining element 〆 27.5, and a second determining element 295 is connected to the synchronous verification unit 2 900. Please refer to the fifth and sixth figures. In step 600, the matched filter 210 is located at the front end of the cell search device 2000, and is used to detect a waveform entering the number 190 of h. In step 6 10, the sampling element 2 2 0 oversamples the incoming signal 1 90 at a chip rate N times, and the signal after the oversampling is expressed as Yi (k), Y2 (k),,, , YN (k), and send it to a down-sampler 230. In step 615, the down-sampler 230 transmits the N oversampling # numbers to the sequential test unit 2 4 5 and is controlled by the first stage detector 2 4 0 The sequential test unit 2 4 5 selects the best sampling point among the N oversampled signals. In step 6 2 0, the sequential testing unit 2 4 5 ′ transmits the optimal sampling point to the second-stage detector 250 and the third-stage detector 26 0. Box 4 1 1 in the first stage, box 4 2 1 in the second stage, and box 4 3 丨 in the third stage (see the fourth figure) are tested for the same best selected sampling point 'where the best The sampling point has the largest sampling value among the N oversampling points, so the effect of non-ideal sampling can be reduced. Once the test is wrong, the next new test will test the randomly selected sampling points again. In the first phase of the box, the second phase of the box and the third phase of the box will also test the best sampling point for the same selection. Such tests will continue until the scramble code is accepted by the synchronous verification process. What's more, the selection of the sampling points is realized sequentially, and each test requires (N + 2) χ 10 milliseconds (m s) ’. Therefore, if the whole cell search is successful in the κth test, the required

Grid enzyme 5Γ, signed two said H12) # No. 91105519 year correction The cell search time is (K + N + 1) x 10 milliseconds (m s). In step 6 30, the first stage detector 240 is used to obtain the time slot synchronization of the first stage. A non-coherent matched filter, where the matched filter is divided into a plurality of sections, and the output of each section is merged with its absolute quality, divided into four blocks and used as a It is used as a detector for time slot synchronization in the first stage 3 10. The non-uniform cumulative slot system is properly found as additive white Gaussian noise beyond 15 hours. For better characteristics, the boundary of more than one time slot in the first stage is selected to enter the next stage.

In step 64, after the time slot synchronization, the code group and code frame synchronization can be completed in the second stage. The second-stage detector 25 uses 6 matching filters and wave detectors to detect the secondary synchronization code. The coherent accumulation in the second stage is possible by using the channel estimates from the first stage. After collecting 15 secondary synchronization codes, they are related to the 6 4 C F R S codeword, and each codeword has 15 possible cyclic shift positions. In this way, 960 correlation values are obtained. Therefore, in the end, the code group and the cyclic position movement system with respect to the maximum value are determined as reasonable code group and code frame boundaries, respectively. In step 650, in the third stage, a strategy is used to detect the scrambled code by selecting one of eight codes in a group, wherein the code group has been identified in the second phase. Basically, in the third stage, the third stage detector 260 selects a mixed | L code at this stage. The third-stage detector 260 is composed of eight active correlation detection probable confusion codes every two slices. The third-stage detector makes a judgment and selects the maximum value among the eight active correlation detectors. The corresponding chaotic code is recorded once and finally, after 150 records (one code frame length), the value of the maximum number of votes is sent

;; Case No. 91105519 A_η Amendment 5. Invention description (13) to a certification unit to determine whether the test is correct. The value of the maximum number of votes is compared with a threshold value of 77. The threshold value 77 (/ is determined according to a constant false alarm rate. In step 660, the output 2 6 1 from the third-stage detector 2 6 0 is in accordance with a threshold value 0 2 7 1 for test comparison. The first decision element 275 is used after the comparator 270 to determine whether the scramble code is correct. If the threshold is not exceeded, the test is considered a failure and a new test It will restart without delay. If the threshold is exceeded, the scramble code enters the synchronization verification unit 290. In step 6 70, the second decision element 2 95 verifies whether the scramble code is accepted , If the scramble code passes the second decision element 2 95, the test is successful, otherwise a new test is restarted after a delay time T p milliseconds (ms), and the initial sampling point of one of the matched filters is It is assumed to be uniformly distributed. Extensive computer simulations are used to explore the comparison of the operating characteristics of the new cell search algorithm with traditional methods. The effect of non-ideal sampling is emphasized here. All numerical results are based on N = 2. Doppler drift system is 185.2 Hz (100-km / hr), each stage of the detection system is 10 milliseconds (ms), Ί > 2 50 milliseconds (ms) and 7 / 〇 is a 10-error alarm In addition, the transmission power of the actual channel is expressed as follows. First, the primary synchronization channel and the secondary synchronization channel have the same power, and the common guidance channel and the synchronization channel (primary synchronization channel + secondary synchronization channel) The power ratio is fixed. Second, the power of the common guidance channel is 10% of the total transmission power. In other words, during the cell search process, 80% of the total transmission power contributes to intracellular interference. Finally, a 1. geometric factor

Page 19 SS9U05519 Years Amendment 5. Explanation of the invention (14) factor) GWPbPpsch + p ssch + Ppc) / p X is used to modularize the position of the user device in the cell. A higher G value indicates that the user equipment is closer to the base station. The cumulative distribution function (CDF) of the search time is used to explore the characteristics of different search algorithms. Please refer to the seventh figure, which shows a signal model according to the present invention under non-ideal sampling effects. Using a base-band representation, the received signal r (t) is given as: Γ (ί) = Σ] Α / ν ^ 〇 (ί) + ^ 550 (ί) + ^ ρς (ί) μ-iTc ~ τ) + Λ ^ ηι (ί)) < ^, 〇 + V ^ nx (t) (EQ -1) where Ppsc, Cpsc, pssc, Cssc and Ppc, cpc are respectively Primary synchronization channel 'The power and spreading code of the secondary synchronization channel and the common pilot channel. g (τ, t) is a complex value representing the Rayleigh attenuation gain, h (t) is a square root of a raised cosine function (SqUare root raised cosine shaping function) and has a rough factor of approximately 0.22. Tc is the chip period of the user device. r is an initial random delay and is modeled with a random variable. The variable has a uniform distribution across the entire (-0.5Tc, 0.5Tc). In addition, P and P are the powers of intracellular interference η I (t) and intercellular interference η χ (t), where η 丨 (t) and nx (t) are zero white with a single variable. Noise (zero mean additive white Gaussian noise with unity variance). According to the mathematical formula E Q-1, it can also be expressed as: ⑼ + \ ^ Sssc (t) + KSpc⑴j + this ―) 丨 of x, 0 + v ^ nx (t)

Case No. 91105519 Rev. V. Description of Invention (15) (EQ-2), where psc, (t) = c psc (i) x h (t-iTc-r) S ssc (t) — C ssc (

h (t-iTc-T (EQ-3) (EQ-4) s

\) y cpc (i) x h (t-iTc-r) (EQ-5). There are three observation systems of the above model that deserve to be proposed here. First, for simplicity, only flat fading channels are considered, and only those channels that are searched by the cell are dealt with clearly; all other channels are included in the interference term η! (T) or n x (t). Second, r is expressed as the effect of non-ideal sampling, which was ignored in the prior art, and third, the model assumes that the effect is from the same frequency source of instability of the oscillator. The eighth graph is a graph showing the comparison of different cell search methods under the non-ideal sampling effect. Where " ideal π is used to represent ideal sampling. As shown in the figure, non-ideal sampling causes severe damage to operating characteristics, especially in the case of lower signal-to-noise ratio (SNR) (a smaller G value), a more practical example. In this figure, for all algorithms and signal-to-noise ratio gain G of 6dB, a 90% chance (occas i ons) will be completed in 110 milliseconds. On the other hand, when the signal-to-noise ratio gain G is OdB, the method STS1 of the present invention completes a search rate of 90% in 300 milliseconds, but it takes 450 milliseconds (ms) for the traditional algorithm. Please note that in a very low percentage of searches

Page 21_face 3 Π Case No. 91105519 ± Amendment V. Description of Invention (16) The traditional algorithm has better characteristics under the standard CDF, which can be attributed to fewer opportunities (〇ccasi ο ns) Sampling in traditional algorithms is closer to the point of optimization. Although the present invention has been explained in terms of its preferred embodiments, from the above, we can see that most of the modifications and changes can be realized without departing from the spirit of the original purpose of the present invention and the concept of innovation. It should be understood that the specific embodiments shown are not intended or meant to limit the invention; the disclosed are intended to encompass the scope of the appended patent and all amendments will fall within the scope of the patent.

Page 22

Modification: Xixi :: Display and display a broadband code division for the 3rd Generation Partnership Project (3GPP). Ma Xigong receives the simplified code frame structure diagram of the Λ knife frequency multiplexing system. Xi Gong ^ Figure 2 f / One is used for the 3rd Generation Partnership Project (3GPP) broadband frequency division code. Gong takes the knife, the traditional sequential cell search method of the Xi Gong system. (In this example, the process time of the first stage is set to 10 milliseconds (^)) The third picture shows a multi-frequency access / frequency division for 3rd Generation Cooperative Computing (3GPP) wideband code division Traditional ducted cell search method for multiplexed systems. (In this example, the process time of each stage is set to 10 milliseconds (ms)) The fourth diagram shows a cell search method according to the present invention, which uses the first stage sequential test (seria 1 testinstage-1, STS1) method. The fifth diagram is a flowchart of a cell search method according to the present invention, which uses the first stage sequential testing method. The sixth diagram shows a cell search architecture according to the present invention. The architecture uses a first-stage sequential testing method. The seventh diagram shows a signal model according to the present invention. The eighth graph is a graph showing the comparison of different cell search methods under the non-ideal sampling effect. Explanation of drawing number: 111 primary synchronization channel 121 secondary synchronization channel 131 common guidance channel 1 9 0 signal

I Case No. 91105519 Amendment 245 250 260 270 271 275 280 290 295 Schematic description 2 0 0 Cell 2 10 Chip 2 2 0 Sampling 2 3 0 Drop 2 4 0 A certification together, a second search device, a matched filter element, a sampler, a phase detector, a test unit, a phase detector, and a detector value π 0.

Page 24

Claims (1)

Ί ':: Γ-II. Ί 土. Ί'mm.w Mongolian No. 91105519 A_ Amendment VI. Patent Application Scope 1. A cell search method for a code division multiplexing access system. The cell search process at this stage includes the following steps: a chip matching filter, match (incoming signal); by a sampling element, the incoming signal is oversampled N times at a chip rate ; By a down-sampler (d 0w η-samp 1 er), the N oversampling signals are transmitted to a sequential test unit; by the sequential test unit controlled by a first-stage detector at A best sampling point is selected from the N oversampling points; and the selected best sampling point from the N oversampling points is transmitted to a second stage detector and a third stage detector to complete a test (tria 1 ). 2. The cell search method according to the scope of the patent application 1, wherein the test of the cell search further includes the following steps: With the detector of the first stage, the time slot synchronization of the first stage can be obtained; The detector detects that the code group is synchronized with the code frame; by a third-stage detector, a scramble code is selected; by a comparator, the scramble code is compared with a threshold value of 7? The first decision element behind the comparator determines the scramble code; wherein if the threshold value is exceeded, the selected scramble code enters Page 25 l _._ r Request No. 91105519 Amendment VI. Patent application scope and a synchronous verification unit, otherwise the test is deemed to fail and the new test will be restarted without delay; and by one after the synchronous verification unit The second decision element determines the scramble code; if the scramble code passes the second decision element, the test is successful; otherwise, a new test is restarted after a delay time T production second (ms), and the An initial sampling point is assumed to be uniformly distributed. Where the N oversampling is at the N oversampling, where the first is obtained using the generalization 3. The best sampling point among the cell search method points according to the scope of the patent application has the largest point in the absence of interference Sampling value. 4. The time slot synchronization according to the stage of the cell search method according to the scope of the patent application 2 further includes the following steps: The first detector has a genera 1 ized hierarchical Go 1 ay sequence in a primary synchronization channel as a Primary synchronization code. 5. A cell search method according to the scope of the patent application, wherein the second phase of code group and frame synchronization includes the following steps: The second detector is used in a secondary synchronization channel, which is orthogonal to the primary synchronization code. 1 of 16 secondary synchronization codes. 6. The cell search method according to the scope of the patent application 5, wherein the secondary synchronization channel is further encoded into a group of 64 code words by a (1,5,3) uninterrupted Reed-Solomon code (c 0 mma -free R eed S ο 1 〇m ο nc 〇de, Page 26 (Amendment VI. Patent Application Scope (CFRS)), where each codeword of the group is listed as a code group to identify a code frame edge and code group. 7. The cell search method according to the scope of patent application 2, wherein after the code group is identified, the chaotic code can be determined by using a common guidance channel and selecting one of 8 code words. 8. The cell search method according to the scope of the patent application, wherein the selection of the sampling points is realized sequentially, and (N + 2) x 10 milliseconds are required for each test, so if the entire cell search is performed at the Kth time The experiment was successful, and the required cell search time was (K + Ν +1) x 10 milliseconds (ms). 9. The cell search method according to claim 2 wherein the first stage detector uses a non-homogeneous synthetic matched filter to perform time slot synchronization in the first stage. 10. The cell search method according to patent application scope 4, wherein the first stage detector selects more than one time slot boundary in the primary synchronization code as a candidate for a better operation, although only one candidate will be considered . 11. The cell search method according to claim 5 wherein the second-stage detector uses 16 matched filters for detecting the secondary synchronization code. 12. The cell search method according to the scope of the application patent 6, wherein the boundary between the code group and the code frame in the second stage is the maximum value of the movement position of the code group and the cycle. 1 3. The cell search method according to the scope of patent application 2, wherein the threshold β 〇 is determined according to a constant false alarm rat e. 1 4. The cell search method according to the scope of application for patent 1, wherein the cell search Page 27 1 丨 Multiple, ¢ 1212043 ^ No. 91105519 Amendment VI. Patent Application Scope Fetch / Frequency Duplex System. A search method, in which the cell searches a cell search device of a personal digital assistant (PDA) system, and the cells of the second stage and the third stage are equipped with an entry signal; a matched filter is used to sample the N times; |, Connected to the sampling element, the _ finder method is used for wideband CDMA multiplexing1 5. The cell finder method according to the scope of patent application 1 is used for mobile devices and wireless systems. 1 6. — For the three stages of code division multiplexing access processing (first stage, second search, the cell search device includes: a chip matched filter for matching a sampling element, connected to the The chip-in signal passes a down-sampler at a chip rate (down-sampler: used to transmit the N oversampling points; a down-sampler is used at the N sampling points; the sequential test unit is used for A sequential test unit is connected to the over-sampling point to select a largest first-stage detector, which is connected to the time slot to obtain the first stage; a second-stage detector is connected to the sequential test unit, and To obtain the code group and frame synchronization; a third-stage detector connected to the sequential test unit to determine a scrambled code; and an authentication unit connected to the third-stage detector to determine the test Whether it is successful. 1 7. The cell search device according to the scope of application for patent 16, wherein the authentication unit further includes: Fen page 1ιΜθ # 3 ·: Le ^ ± t Case No. 91105519 Λ_η Revision 6. Patent application scope a comparator connected to the third stage detector to test the third stage output value and a threshold value 77 comparison; a first decision element connected to the comparator to determine whether the scramble code is correct; wherein if the threshold value is exceeded, the selected scramble code enters a synchronization verification unit, otherwise the test Failed, and a new test will be restarted without delay; the synchronous verification unit is connected to the first determining element to verify the scramble code; and a second determining element is connected to the synchronous verification unit, using To determine whether the scramble code is accepted; wherein if the scramble code passes the second decision element, the test is successful; otherwise, a new test is restarted after a delay time T production second (ms), and one of the matched filters is The initial sampling points are assumed to be uniformly distributed. 18. The cell search device according to the scope of application for patent 16, wherein the optimal sampling point among the N oversampling points (YK k), Y 2 (k),, YN (k)) appears without interference. In the following, the N sampling point has the largest sampling value. 19. The cell search device according to the scope of patent application 17, wherein the first-stage detector uses a generalized hierarchical Golay sequence (eq u e n c e) as a primary synchronization code in a primary synchronization channel. 20. The cell search device according to the scope of application patent 17, wherein the second-stage detector is in a secondary synchronization channel and uses the same synchronization with the primary synchronization code. Page 29 Sixth, the scope of patent application is over 16 secondary synchronization codes. 2 1. The cell search device according to the scope of patent application 17 in which the secondary ^ step channel is further encoded into a group of 64 code words, with one (1, 5, 3) ..., intermittent Reed-Solomon stone A horse (comma-free Reed Solomon code, CFRS), where each codeword of the group is represented as a code group to identify a code frame edge and code group. ^ 2 2 According to the cell search device of patent application scope 17, after the code 4 is identified, the chaotic code can be determined by selecting one of 8 code words by using a common guidance channel. ψ 2 3 · The cell search device according to the scope of application patent 16 where the sampling ΐ ^ selection is realized in sequence 'and each test requires (N + 2) x 1 〇 = I pumping, t if the entire cell The search was successful only in the κ trial, and the required time is (K + Ν + 1) x 10 milliseconds (ms). Phase: 4 test cell search device according to patent scope 17, where the -time slot is identical in the first stage with a non-homogeneous synthetic matching filter … ^ No-Jie as a ^ ί choose more than one time slot in the primary synchronization code. The first order = candidate for Tochigi, although only-the candidate will be considered when the time slot meets the limit = t tester in the Choose more than one in the primary synchronization code to take into account. Below is-candidate for better operation, although only one candidate will be staged mM & rn)! ί One ΐ; Case No. 91105519, Amendment 6. Scope of patent application. 27. The cell search device according to the scope of application patent 21, wherein the boundary between the code group and the code frame in the second stage is the maximum value of the movement position of the code group and the cycle. 2 8. The cell search device according to the scope of patent application 17, wherein the threshold value 77 is determined according to a constant false alarm rate. 29. The cell search device according to the scope of application for patent 16, wherein the cell search method is used in a wideband division code multiplexed access / frequency division duplex system. 30. The cell search device according to the scope of application for patent 16, wherein the cell search method is used in a mobile device and a wireless personal digital assistant PDA system. Page 31