TW200818755A - Detection of time-frequency hopping patterns - Google Patents

Abstract

The present invention relates to a method for detecting a pilot pattern, comprising a pilot signal sp(t), in a received signal r(t) implemented in an Orthogonal Frequency Division Multiplexing (OFDM) system. The method comprises the steps: computing a value of a log-likelihood function Λ (τ0, ν0) for a hypotheses space specified by the pilot signal sp(t) and an initial time frequency offset (τ0, ν0), and relating the computed value to a reference value to detect the pilot pattern.

Description

200818755 IX. Description of the Invention: [Technical Field] The present invention relates to a method for measuring a time-frequency skip mode in an OFDM (Orthogonal Frequency Division Multiple Jade) system. [Prior Art] Conventionally, a wireless communication channel within a single carrier transmission system is modeled as having a time varying impulse response g(u), which may be due to multipath

For any given time "system frequency selective. The frequency selectivity of the channel can be added by observing the known preamble signals emitted during the time of interest," and the day-to-day selectivity is usually observed by observation. The signals are periodically inserted into a number of known signals for tracking. However, in a terrestrial mobile communication environment, the channel selectivity is mainly due to the movement of the material end. As long as the transport degree is guaranteed, the channel can be modeled. For a non-time-varying delay Doppler response h (T, which indicates that a delay value τ and a Dubre offset U into the signal scatterer composite value channel increases in the cause of death] this fact has been mainly used as In the domain, the pre-introduction miscellaneous ΪΪ: design constraints to avoid frequency stacking. Dubrie information slightly more complex: 4 in the channel tracking related filter design 'it needs to estimate the channel's Doppler expansion., - time · frequency The skip mode is a signal whose frequency content changes periodically or aperiodically as a function of time in a specific manner. ==: Many communication and radar applications. Recently, since the second is to adapt the 'OFDM' future The line communication system technology has been widely used as a synchronization signal. 123138.doc 200818755 In the 0-day system, the radio resources are essentially divided into orthogonal time-frequency units, so the natural design is late. Synchronization of _frequency division. In the ~~〇 FDM system, the preamble symbols are periodically placed in the time-frequency plane, and the estimator is estimated. Figure i shows the regular interval preamble mode (indicative (4) and - Coase Tas (Cc) Sta_column mode ((d)--example, as is well known to those skilled in the art, the Costas-Dragon series is generated by shifting the horizontal scan line of the regular interval preamble pattern in a specific order Many

Can change one. The Costas array pattern is revealed in "media constraints on sonar design and performance" (IP c(10)as, easc〇nc〇nv.Rec., 1975, pp. 68A-68L). The ^ subcarriers within the _ 0FDM symbol have - centroids (10) ·, including the cyclic prefix Tep sec ·. Thus the 'subcarrier spacing is fs = 1 / (Ts _Tcp) Hz. For the original rule In the interval mode, in the time domain, a preamble symbol is inserted every ^^ 〇 FDM symbols, that is, Tp = NTS, and a preamble symbol is inserted every other subcarriers in the frequency domain, that is, fp = Mfs. Relative to the first subcarrier, Each mode may have a subcarrier offset index 〇^ h Μ. Any preamble mode may be specified by a one-dimensional time-frequency array, and its array element C[n, m] is in the nth OFDM symbol. The composite value of the preamble symbol is transmitted on the subcarriers of the melon. Unless otherwise stated, in the case where a preamble symbol exists, C[n, m] is η Γ ' in the absence of the presence of "〇,". The corresponding continuous time signal of the preamble mode on the q time domain period can be expressed as an OFDM symbol sequence as follows: Ο) • 卜 nTs) n=0 123I38.doc (2) 200818755

Cn(t)=%CnWiit-iTc) /=0 is the nth OFDM symbol, which is further composed of a sequence, the sequence modulation transmission filter pulse function μ (〇. ignores the cyclic prefix, the preamble mode The time-frequency array represents C[n,m] and is related to the discrete time series cn[i] by J2mni C[n9m]= J]cn[i^ ^ = .] 卜-0 (3) To modulate the data symbols in an OFDM system, the receiver needs to know the time-frequency response H(t,f) of the channel, which is the two-dimensional Fourier transform of the delayed Doppler response ^^^). If a sufficient number of substrate preambles are observed in time and frequency, then the output of the delayed Doppler correlator is a preferred approximation of the delayed Dubley response. The channel is zipped to have a delayed Doppler response, which represents the composite value channel gain of the sigma and the scatterer of the Doppler shift to the incoming signal. Assuming that the radio environment consists of a continuous scatterer (or "target,), each scatterer will introduce a specific delay and a Doppler shift to the signal propagating through it, then the received signal corresponding to the preamble is determined by It is concluded that: ΓΤ, 讳p — z*) ey27nVz^v + z(,) , ( 4 ) where z(t) is additive Gaussian white noise (AWGN), τ〇 and the initial timing and The frequency offset, and the maximum delay of the < Μ < Λ Ws 丄 = 丄 n system and the Doppler spread, which is less than or equal to the value supported by the leading density in the case of no stacking of $123138.doc 200818755. The target detector typically uses a correlator that matches the hypothesis signal, then finds the peak and compares it to the -specific limit value to determine the presence of the preamble. This related procedure may be computationally too complex, especially when it comes to potential assumptions. A known system using a time-frequency skip mode as a synchronization signal is disclosed in US 6,961,364 B1 to Lar〇ia et al. Different base stations use a non-slope mode and the detection algorithm is a maximum energy detector. SUMMARY OF THE INVENTION One object of the present invention is to provide a device and a method for comparing a preamble pattern in a 〇F D Μ system with less complexity in the prior art. An additional goal is to provide a method for synchronizing and detecting a foreign object in a communication network. ' This issue asks for the detection - before a certain standard is met in the best sense. • The problem of the time-frequency skip mode of (4). By applying the principle of generalized likelihood ratio 疋 (GLRT), the present invention provides an optimal approximate measurement for a given hypothesis on the preamble mode and its time: frequency offset. Based on this approximate measurement, a detection and synchronization of a preamble mode can be obtained. One advantage of the present invention is that the present invention can be used for initial synchronization of one of the FDM systems using different time-frequency hopping patterns to identify different devices. Another advantage is that although the best approximate measure is the energy integral at the output of a two-dimensional delayed Dubler correlator, the two-dimensional delayed Doppler correlation is not necessarily performed. 123138.doc 200818755 One of the advantages of one of the preferred embodiments of the present invention is that the current conduction mode is used by many devices that share a common structure (such as a per-slope mode). % ' can be stepped in to simplify its approximate degree 瞀. [Embodiment] # The object of the present invention is the existence of a measure, and then obtain an estimate by estimating an unknown parameter (:...). Rough initial time-frequency synchronization. Assume that the initial synchronization has been obtained, because N^ is known (τ〇, v〇), for a fan

The appropriate material signal such as the above-mentioned preamble signals defined on the circumference (V_v~Vmax), the most approximate channel of the channel is estimated, V) is below and the two-dimensional delayed Doppler image Ι (τ , ν)Related: I(r,v)= J is called ρ (Bu φ-如冶^Ki{rv)^X5p{r,v)^EshML{r,v) ^, (6) where A = ikKf dt The energy of the leading signal at the unspecified observation interval and

Xsp^^v)^\sPi})s\(t-T)e~2mdt W / The ambiguity function of the preamble signal Sp(t). In the initial synchronization phase, the initial time-frequency offset (τ〇, v〇) is unknown and therefore needs to be assumed. Therefore, detecting the presence of a preamble pattern and then identifying the device transmitting the pattern involves a search on the hypothesis space specified by the preamble Sp(1) and the initial time_frequency offset (τ〇, ν〇). First, starting with the assumption given by a Sp(1) and then defining a general 4乂/ then for the hypothesis, ie the delayed Dubler response of the channel begins in the delayed Doppler plane (τ〇, vG) And extended in its maximum delay Doppler extension range 123l38.doc 200818755. Since the channel response h(u5 V) is unknown, it is a redundant variable that needs to be removed. This can be done by applying the same method for the generalized approximate ratio test (GLrt). Assuming that the hypothesis (τ05 is) is correct, the first step estimates the excess variable 1ι(τ, ν). According to equation (6), this is derived from: (8) where and vGSv$vG+vmax. The next step is to replace the real channel response with its estimate in the logarithmic approximation function:

° 0 (9) After reconfiguring and removing extraneous terms, the logarithmic approximation function for this hypothesis becomes: A(~v0)=d:~|I(L

=Γ, J:1((1).,... and V (ια> In the broadest sense of the most approximate sense, the best approximate measure for this hypothesis is thus derived from equation (10), where Ι(τ, ν) is the delayed Doppler image observed by the hypothetical signal Sp(t). ” Estimating I ratio k in the possible hypothesis that the detection guidance mode is in Sp(t) and 〇0, This logarithm approximates the equivalent of the function or compares it to a specific defined row. Specialized (10) direct evaluation of eight (τ〇, v〇) is performed only by a delay du correlation, followed by AW 1 Integrate energy for all hypotheses that satisfy the Nyquist criteria (and therefore channel estimates) and the mode is valid, such a straight 123138.doc -10- 200818755 Many devices may assume that a large number of possible assumptions are too large. Fortunately, as can be seen in the following sections, there is a function for calculating the log-like approximation function without explicitly evaluating the delayed Doppler image ( Alternative method of τ, v). Even larger if these potential preamble patterns share a particular shared structure A reduction in complexity is possible. The discrete implementation of the detector is derived from the equation (ίο), and the generalized logarithm approximation function Λ(τ〇, v〇) is only in the channel's delayed Doppler image Ι (τ , v) is the integral of the assumed energy range, and the discrete approximation of the /P image on 1&() can be pushed. X The time-frequency offset version of the Doppler image can be expressed as ···+r0, μ + v〇) = Jr(r>; (Bu Bu (12 > Since only at the detection stage ^(〇=r(f 4-T〇)e-^2^ is one of the received signals) Offset version: = the amplitude of the Buhler image, so in the equation:::: delay = the item moves to the other side of the equation and defines - a new function for delaying the Doppler image of the channel: ^

Wtv) "handsome +, (Γ + Γ〇, ν + ν.) vj°pv ^J at where the subscripts τ〇, V〇 indicate the private μ main ^ ^ initial time * frequency offset hypothesis dependence in the abundance The integral in (13) can be expressed by the discrete sum of 苴^: J23l38.doc (14) 200818755 厶 QN-l : 2 heart! r0,J]e QN \r{t^nTs)cn {t-kTc)dt Q^p -1 ,2>- /=0 [nj^k)cn[i] where HU is a chip rate of 1 /T Hz in the delay domain and a code in the Doppler domain Slice rate QNTs sec. Downsampled channel delay Doppler image, and

|r(/ + w7; ^τ〇)μ(ί^ιΤ€)ώ (15) is a receiver-filtered, wave-like output that has a chip rate of 1/min. The time offset τ 〇 and the frequency offset ν 〇 ' taken under ηζ are shown in Fig. 2. It should also be clear that the delay assumption τ ❶ should preferably be chosen to be a multiple of the chip duration. Figure 2 shows a device 30 for generating a sample signal % from a received signal r(t). The received signal is input to a filter 3 1 having a wave function μ*(-t). The filtered signal is sampled using a sampler 32 at a chip rate} / Hz and a negative delay is added to a delay 33. The sampled and delayed signal is mixed in a mixer 35 with a signal from a source 34 to produce a sampled signal Ά, ζ·]. Discrete Frequency Domain Implementation As stated in equation (5), the maximum delay of the channel, the Doppler spread, vmax, should not be supported by the value of the other metrics. For the case where the integral in the equation (1 〇) is performed on the maximum value supported by the post-conduction degree, the ^r integral can be approximated to be in the range ^^ to ^^/from.../^ The discrete version of the sampled version 123138.doc •12-200818755 is derived from the following: · stomach η ♦ · κ ΐ ΐ ΐ 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在 在However, it is obvious that it is not necessary, as described below. (4) The detector can then reduce (4) complexity and compromise between performance and complexity. According to (4) requirements and equipment can: =: T window, when, _) the last (four) domain to be evaluated and can use DFT (discrete Fourier transform) in the frequency of 'small, tons of labor (17) where Γ *1 ν诹一' and r0, v. Art _j2nmi (18) corresponds to the brother η 〇 p j)

Nfft DFT and N know J2mii φ, β= ~[φ i, /=0 is the length of the received signal of the 5th cycle of the tiger - p. DFTlv η (19) fft 〇, 1,...,Ν at the nth Discrete frequency domain representation of the preamble signal within the 0 job symbol. Using FIG. 1 to visually illustrate that the rows within the array correspond to a given DF din of the received samples on a continuous 〇 FDM symbol for a given (τ., V0), this particular embodiment is by way of The energy sum at these locations of the model is assumed to approximate the approximate function. It should be noted that c[n,m] is also non-zero only when a preamble symbol is transmitted on the 1" subcarriers within the nth 〇FDM symbol. It should also be noted that if 123J38.doc 200818755 |C(n,m)丨 is constant for all (non-zero) preamble symbols and if %, (4) forms a plurality of lines of strange slopes in the time-frequency plane, then this embodiment Reduction to the prior art 0 provided in US 6,961,364 β1 mentioned in the background section. In addition, all possible assumptions for the initial frequency offset ~ should be noted, with regular intervals being selected at appropriate intervals. Next, the coffee can make

Calculate with pad zero DFT. For example, for a particular τ. Assume that a length with a pad-LNfft sequence (20) will produce a length-LNfft sequence, and a subsequence of 0 + L(Nfft -1) corresponds to the above-mentioned regular interval & or its hopping variable, The following is an evaluation of the logarithmic approximate function of the offset hypothesis for the given initial in the discrete frequency domain: iron 』'] 丨 = gg | ', JHc'hm] 丨 2 (21) which is only in discrete time-frequency The sum of the received signal energies at these locations of the hypothetical preamble pattern on the plane. Discrete Time Domain Implementations In some cases, it is possible to calculate the log likelihood measure directly in the time domain. For the preamble modes described in conjunction with the map, the time-frequency map C[n, m] is only non-zero at the subcarrier indices m+m, y+2M, ..., and is used as the symbol index n. A hopping sequence of a function and M is a unique insertion period within the number (not necessarily before the original regular interval pattern 123138.doc -14·200818755 insertion period). Then equation (21) becomes: o//^i /r-i =Σ ΣΚο,νο^ί^ +mAf] 2 rt=0 m=*0 f (2 2) where κ = Nfft / M. Use m = to replace the equation from equation (10). Obtained in equation (22): ν-'γ-1|τ r η (2 Nffi-, ,, *«〇Sl =^Σ Σ〇[ E^. J«5/K,〇k/-^]je"y^ (23) It relates to the cyclic self-correlation of these backward received signals, which is K times and several shorter length DFTs. Detecting multiple hopping modes can be greatly simplified if it has a specific shared structure. For example, 秸^ straw is assigned to different devices by the cyclic offset mode. The ❹] can then be implemented by using a two-dimensional cyclic correlator that matches the time-frequency mapping of the base mode. Figure 3 shows the cyclic offset mode of the Cosbus array. - The second mode is represented by the "(4) Μ symbol The original mode of the cyclic offset of the disc (10) + W subcarriers. As in the case of the specification interval, the modes with different subcarrier offsets are completely orthogonal. For a particular class of Costas arrays, there is a different identity. The melon mode has a maximum of _ coincidence per cycle. See the example in Figure 6. It should be noted that for the length L - Costas order There are a total of L χ μ χ n different cycles (4) Shifts are used to identify different cells 'if they are time synchronized. On the other hand, for the unsynchronized network, there are LxM different cyclic frequency offsets. 123I38.doc • 15-200818755 In addition to artificially induced cyclic shifts, relative to other references such as patterns from different base stations or sampling points of a particular terminal, for various reasons, the 'preamble mode can also have a local scale, Offset. This initial offset is essentially the same as and represented by (τ〇, v〇) above, except that it is in the range of <7; and is within the constraints of </5. Anything outside this range 'The offset will be folded into the exponents η, μ and more. Detecting a cyclic offset preamble mode depends on its presence in a hypothetical time-frequency coordinate: τρ (24) νο = (shore + 彡)Λ +ν0 is shown on an enlarged scale in Figure 3. To demonstrate how a two-dimensional correlator can be implemented, we consider an example that is simplified by setting the frequency domain leading insertion period to 丨 from the one shown in Figure 3. For each of the initial time-frequency offsets (τΰ, VG) Assume that an LX two (6 X 7) input p train is formed by performing N length DFT on the symbol duration and X placing the frequency domain samples in N consecutive rows, as shown in the first matrix in FIG. The initial time-frequency offset hypothesis should preferably be selected as part of the symbol duration and sub-carrier width to avoid double counting. Once the input array is set, the pattern search begins. First, the Costas array is expressed. One of the frequency index hopping sequences {〇, 2, 1, 4, 5, 3, X}, where "χ,, indicates that _ does not have any + pay. This sequence is placed under the input array for visual illustration. In the -step, the rows of the array are cyclically rotated by a column (subcarrier) index by a number corresponding to the hopping sequence, as shown by the second matrix 52 in FIG. 4 and spanning the rows (symbols) In addition to the indication "χ", and 123138.doc -16- 200818755 produces a line vector of size L=6. The base of the array is placed in the image. 5 in the middle of the output system cycle to the right / n / in a second step 'the Costasis sequence, as shown in Figure 4, the third 1 ==, the loop of the ring (four) into the array, then The resulting vector is placed 2 as added in the first step. ... is placed in the second row 56 of the output array in Figure 5. Moved. V continues to exhaust all N possible Costas sequence cyclic biases 5〇 Will then contain the target - given

; 7 All the Lx N of the cyclic offset may be assumed to be the generalized logarithm: 円 可 可 可 概 概 概 概 概 概 概 概 概 概 概 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否 是否Clearly reveal two prominent peaks, one at 01 〇 'μ = 〇) 'represented as the first in row 55 of output array 50 and the other one at (η = 2, μ = 2), expressed as an output array 5 of the 5th, 59 of the 57th. The vector placed in line 57 is derived from the fourth matrix 54 of the clock, as described above. The related program shown in Figures 4 and 5 occurs offline in a memory buffer containing previously captured data. It can be implemented alternately in a sliding window: new in the 'sequence' = the material continues to arrive and fill up Array, while overflowing the old two materials in the ^ ^ brother, the memory buffer loading and vector tracking {疋 可 可王口 P by modifying the address indicator without physically moving the buffer content to complete. Finally, even this example shows that the extension of a single ^^ to a cycle of the Costas array is relatively straightforward. A computationally efficient implementation adds energy over multiple cycles to form the input array prior to association. The following is a brief description of the specific embodiment of the sliding window for this particular situation, 123138.doc -17- 200818755. The offline specific embodiment described above does not change 'except that the data has been loaded into the buffer, so the data acquisition step can be omitted.丨·Execute the division on the receiving sample for the - symbol duration. The length of the bribe depends on the initial frequency (the assumption that 7 W = the absolute square of the rain output is placed in the first row of the wheeled array - the row is - 歹 "subcarrier" index the wheel into the array Rotate by the soil (the physical content or the number of a finger _ column. 胄 should add the input _ in the Costas order = row (symbol) index and place the (4) orientation on the first line of the output array Resets the wheeled array metrics displaced by step 3 and rotates the input/output (rotating the rows to the right) a position. Obtaining fragments under the received samples over the - symbol duration may overlap The former depends on the assumption of τ〇. 7· Advance to step 1. The simulation result is: evaluate the performance of the detector' to consider the array leading mode 60 described in Figure 6. The system is based on the production ^ ^ preamble mode 61"Two access points have -base =, sheep on the % offset nfp and cycle offset in time (10) 62 ' as for (m,n) = (2,; between 忒4 two modes 2. 3· 4, 5. 6. 123138.doc -18 - 200818755 The number of coincidences in period 63 is in this example - Because of the (four) GF(7)-ideal periodic Costas sequence system is also used for the Coase preamble nucleus 61, 6.2. Maximum delay spread ------ ————--- Array 1 Array 2 - -N____ Μ Γ ------ — .. 16 16 —-===== _30_ 8 Nfft 1024 512 __ncd 64 64 Costas sequence length Z 16 (GF17) 3〇 (GF31) Assume LX Μ X TV The number of 4096 7440 capsule I 丨 is used for the parameters of the two Costas array. The parameters of the two specific arrays that are not simulated are shown in Table 1. In order to make a fair comparison, both have - about & 256 Preamble density. The FFT size of the OFDM symbol with the length -16 array is 1〇24. Since M=!6 in this case, the maximum delay extension that can be supported is 1〇24/16=64 chips, which is set. In the length of the cyclic prefix, the camphor symbol FFT size system 512 for the length_3〇 array, that is, the 〇FDM symbol interval is half of the first array. However, the prefix cycle length remains the same, so it can accommodate The same is true for each channel implementation in the simulation, introducing an initial random time_frequency offset (τ〇, ν〇)' which is evenly distributed Interval [〇, Ts] and [0, to 2]. Correlation • Make a single hypothesis (τ〇, Vo) when forming the input array = (〇, 〇, this point corresponds to one in the time domain Ts Sec. Roughly search for the preamble modes with the fs Hzi spacing in the frequency domain. If the real target is within a pay and subcarrier on either side of the detected location, a successful detection is declared. After obtaining cell identification and coarse synchronization, a further fine search can be performed to establish the boundary of the delayed Doppler response of the channel. In the case of a preferred SNR (signal-to-noise ratio) or another 123138.doc -19. 200818755, this can involve some simple interpolation. Table 2 lists some of the other parameters shared by all simulations. The simulated two power delays the Dubler wheel gallery. The "straight" 屯^ Κ is only implemented as a zero-mean Gaussian variable and the "full" channel has a ^ ^ v ^ Case3 X Bessei profile, which is spread over the entire maximum delay Dub &; 于杜布勒£ domain. It represents the two extremes of the channel-human sequence. The actual performance in practice should be roughly between.

[Table 2] Some additional parameters for these simulations

Figure 7 illustrates one of the simulations of one of the single-cell false detection probabilities. The 71 line array 1 (Costas 16), with the oyster shell line, is filled with "channels, and the line with the circle 72 is array 2 (c〇stas 3〇), "full "channel. Dotted 73 series array with an asterisk! (CGStas 16),,, Straight, 'Channel, and Circled Virtual Line Array 2 (C0stas 30), " Straight, Channel. Despite the seriousness of the intra-channel diversity-human sequence, the effectiveness is still very strong in these two extreme cases. The larger sidelobe peak energy ratio of the second array (Costas 3〇) does not produce more gain until a higher SNR range in the "filled" range. This is mainly due to the fact that it has more likely assumptions (7440) and therefore more opportunities to make mistakes. This second set of simulations involves two cells with the same subcarrier 0 = 0. The first cell is located at (η, μ) = (〇, 〇), an average SNR is 0 dB, while the second small 123138.doc -20· 200818755 region is located at (η, μ) = (6, 7)' relative to The first cell has a variable signal power. In addition to the offsets mentioned earlier at the receiver, the two cells have a similarly distributed random relative time-frequency offset. A successful detection is announced in the dual-cell simulation only upon success. This occurs when the locations of the two largest metrics within the array correspond to the locations of the two targets. - A typical array for a second-intensity cell is shown in Figure 8 for reference. Figure 9 shows a simulation of one of the two-cell mishaps in the "full" channel at 0 dB SNR, while Figure 1〇 shows the 〇dB snr in, in a flat" channel - The double-cell miscalculation probability - the simulation - the pattern. The solid lines in Figures 9 and 10 represent the operation cycle and the dashed lines in Figures 9 and 1 represent the two observation periods. In Figures 9 and 10 The lines with the asterisk indicate array 1 (Costas 16) and the circles with lines in Figures 9 and 1 indicate the array " (Costas 30) PPl and P2 in Figures 9 and 10 are received separately The power of the leading signal. Considering the hidden facts, even if the most (four) is missed, most of the time is still measured and the strongest of the two communities, the performance in the "full" channel (Figure 9) is due only - (4) The observation period should be better. In a channel with certain time selectivity, the performance of the benefit can be improved by & plus the number of accumulated preamble cycles to improve the level, such as at Q == 2, 10% error rate The gain of 3 dB is evidenced by the other side. For the "flat" channel case, the gain observed for q=2 is mainly noise suppression. System rather than diversity, since the slope of the curves remains the same. In addition to observing more of the former Mo, the two # < 4 ' can be further improved by some other measurements or under the conditions of 123U8.doc -21 - 200818755 疋Detecting performance. For example, the assumed number of initial time-frequency offsets (TG, V❹) can be increased. This effectively increases the search density and thus the chance of finding the peak of the log-like approximation function at the expense of computational complexity. The avoidance constraint can identify the index (η, μ, 〇 to reduce the false alarm rate. Finally, the initial time-frequency offset in the cells and terminals is most likely to be more than the assumed offset in the simulation. The "symbol alignment in equal cells" is a basic assumption in OFDM systems) that will significantly reduce interference. A detector that performs the method for detecting a preamble mode can be naturally implemented in a node of a communication system, such as a base station, a mobile phone, or any other type of wireless communication device. This method is implemented in the -memory unit and executed by a processing device. The above design principles have always adhered to an orthogonal time-frequency self-division format inherent in an FDM system, thereby causing a transition between time and frequency (or delay and Dubler) domains (preferably using DFT). Receiver algorithm. Due to the demodulation of the poor material symbol, it is also used to complete, but _ dedicated and flexible • thin T plus device can handle almost all calculations when receiving data bits from a data machine. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows an example of one of the leading modes used in connection with the present invention. Fig. 2 shows a device for generating a sampling signal from _ ☆ 彳 receiving 遽 5 遽 r (t). Figure 3 shows the cyclic shift mode of the - Costas array. 123138.doc -22- 200818755 Figure 4 shows one of the two-dimensional correlators in accordance with the present invention. π ~ Body Embodiments Figure 5 shows an correlator output array in conjunction with Figure 5. Figure 6 shows an example of a two-cycle offset Costas array. Figure 7 shows a single cell survey with one observation period. One of the solid gases Figure 8 shows a typical detection metric. Figure 9 shows the pattern of dual cell detection in a full channel. Figure 10 shows one of the two-cell detection in the flat channel. This surname is missing. 1 [Main component symbol description] 30 Device 31 Filter 32 Sampler 33 Delayer 34 Source 35 Mixer 50 Output array 51 First Matrix .52 Brother-Matrix 53 Third Matrix 54 Fourth Matrix 55 Output Array 5 〇56 56 Round-out Array 5〇57 57 Peak 58 Peak 60 Costas Array 123138.doc •23 - 200818755 61 * Leading Mode 62 Leading Mode 63 cycle 71 solid line 72 solid line 73 dotted line 74 dotted line

123138.doc -24-

Claims (1)

200818755 X. Patent Application Range: 1 · A method for detecting a pre-, ^ s sp (t) in a received signal r(t) implemented in an orthogonal division-edge OFDM system The method of the preamble mode is characterized in that the method comprises the steps of: calculating a one-to-segment approximation function for the hypothesis space specified by the preamble Sp(1) and the initial time-frequency offset, v〇) The (τ., v〇) value causes the calculated value to be correlated with the - reference value to the preamble mode. 2. As requested! The method 'where the reference value is a threshold value, and the step of correlating the calculated value includes comparing the calculated value to the threshold value. 3. In the method of claim i, the step of correlating the calculated value includes evaluating and comparing the calculated value between the preamble signal Sp(1) and the initial time_frequency offset (4) 4. The method of any one of clauses 1 to 3, wherein the step of calculating a value directly evaluates the log-like approximation function Λ(Τ(), Vq), the evaluation comprising the steps of: - Perform a delayed Doppler correlation, and - Integrate an energy into this hypothesis space. The method of any one of claims 1 to 4, wherein the logarithm approximation function is defined as: Λ(ν〇)= C bamboo. Where I(r,v)= jr(i)s9p(ir)e^j2^dt where Ι(τ,ν) is a delayed Doppler image, and r(t) is the received signal of the preamble mode, And <(t) is a hypothetical preamble signal. 123138-OK 200818755 6. The method of claim 5, wherein the signal r(t) obtains a sample, the method of 4 further comprises sampling the receiving step, the sample having a ^Jn, i] to calculate the logarithmic approximate function The step frequency offset is, ^=chip rate 1/7^ sample time offset τ〇 and 且 and the method i隹& becomes a multiple of the chip period I: the step includes selecting the delay hypothesis T . In the case of r• as in any of claims 5 to 6, A(r., ~) is the sample of the logarithm of the ^^0~((〇, v〇) kW^m \ am The discrete form of the form is approximated by ·(Υλ·V"β tr , f2 where., plus 他 〇 ^ 系 ^ to - in the delay domain W HZ and Doppler domain 8. Rate sampling - Sampling delay Doppler image. 方法 \7 任 任 任 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 。 9. The method of claim 8, wherein the logarithmic approximation function is evaluated by the following frequency domain: n-0 m*0 ^-10-1 ~, k, 22:, ==heart, / =〇J \ ———, Ί m»0 is the position of the hypothetical preamble pattern on a discrete time-frequency plane. #收信号能量和, where I* N贝-\ jlrtmi R^^mh X rr^ [nj} e ^ i^n 123138-OK 200818755 is one of the lengths of the sampled received signal corresponding to the rWS〇FDM symbol and the one-time-frequency map-Nfft DFT: I· "~M jlroni C Bu, + ΣΊ ~, W = 0,1,...,N—\ /a 离散 is a discrete frequency domain representation of one of the preamble signals in the 12th 〇FDM symbol. 1〇_如W, in any one of the methods 6 or 7, wherein The logarithmic approximation function eight (1), V〇) is evaluated in the time domain using the discrete Fourier transform (DFT). η. The method of claim ίο, where the logarithm approximation function is used by Intra-domain evaluation: r τ *2 Njp-\ , ON^\ A>r_i/ Nm-\ . , 2^w where K = Nfft/M. 12: The method of any of the claims ^" Used for pre-testing The method is for synchronizing and/or a device in a communication network. The method of claim 12, wherein each of the devices transmits a one-time-hop hopping preamble signal with a-cycle offset mode assigned to Each device. 4.4 A detection of crying in a OFDM system, receiving at least one preamble generated in the OFDM system is characterized as the detector comprising a memory unit and a Processing the device and configured to perform the method of any one of claims 1 to 13. 123138-OK 1 5 - an OFDM system comprising at least _ its violent platform and at least one mobile communication 2 The device is characterized by a 筮J 一一郎 point system adapted to transmit a preamble signal in 200818755, and a second node in the system is adapted to receive the preamble signal, the The two nodes are crying together, and include a detection port. The system is adapted to perform the request item I to the method. Turn into the software program' memory unit Within, the characteristics of the horse system can be used / any of the methods. 123138-ΟΚ -4-