AU4297799A - Method and system for determining the position of a mobile terminal in a cdma mobile communications system - Google Patents

Description

WO 99/52235 PCT/SE99/00577 -1 METHOD AND SYSTEM FOR DETERMINING THE POSITION OF A MOBILE TERMINAL IN A CDMA MOBILE COMMUNICATIONS SYSTEM 5 CROSS-REFERENCES TO RELATED APPLICATIONS This Application for Patent claims the benefit of priority from, and hereby incorporates by reference the entire disclosure of, co-pending U.S. Provisional Application for Patent Serial No. 60/081,117, filed April 8, 1998. 10 This Application also claims the benefit of priority from, and hereby incorporates by reference the entire disclosure of, co-pending U.S. Application for Patent Serial No. 08/989,491, filed December 12, 1997. BACKGROUND OF THE INVENTION 15 Technical Field of the Invention The present invention relates in general to the mobile communications field and, in particular, to a method and system for use in determining the position of a mobile radio terminal in a mobile radio system following a Code Division Multiple Access (CDMA) standard. 20 Description of Related Art In existing and future mobile communication systems, the demand for mobile positioning ("geolocation") capabilities has been increasing. Preferably, the mobile positioning functions will be performed, at least partly, within the cellular system 25 involved rather than relying completely on an external system (e.g., the Global Positioning System or GPS). A basic positioning problem encountered in cellular systems is in determining how to perform appropriate Time of Arrival (TOA), Time Difference of Arrival (TDOA), or similar propagation time measurements. Once any one of these measurement techniques is available, a number of algorithms exist that can 30 be used to calculate the geographical position of a mobile station (MS). CONFIfIMATION COPY WO 99/52235 PCT/SE99/00577 -2 The existing state of the art for positioning in cellular systems can be divided generally into the following categories: 1. Location area - The location of the current cell provides a rough indication of an MS's position. 5 2. Handover solutions - The simplest positioning concept (but providing reasonable accuracy) is based on a method in which handovers (including soft handovers) are made to a number of base stations (BSs). Each of these BSs measures the propagation time between itself and the MS involved. This method is relatively simple to implement because it involves very little change in the radio part. Also, the 10 BSs do not require the use of an absolute time-reference. However, designers believe that the "hearability" of this method is unsatisfactory (i.e., handover to two other BSs at different geographical locations is possible only in a small number of cases), especially if the system utilizes a 1-cell frequency re-use approach. 3. Antenna array solution - If a BS has an antenna array, an MS's position 15 can be calculated from the estimated direction and round-trip delay of the communication signal. 4. GPS solution in an MS - A GPS receiver can be included in a MS. However, this approach requires excessive computational and receiver complexity in the MS. 20 5. Downlink solutions - These solutions are based on measurements made by an MS. The MS measures signals transmitted by the BSs (e.g., in CDMA systems, the pilot channel data). The methods used for these solutions require an absolute time reference in, or synchronization of, the BSs. 6. Uplink solutions - These solutions are based on measurements made by 25 the BSs. The BSs measure a signal transmitted by the MS (e.g., a lengthy, known training sequence). The methods used for these solutions require an absolute time reference in, or synchronization of, the BSs. 7. Combined uplink/downlink solutions - Commonly-assigned U.S. Patent Application Serial No. 08/93 5,421 describes an approach that combines an uplink and 30 downlink solution, and thereby avoids the requirement for an absolute time reference in, or synchronization of, the BSs.

WO 99/52235 PCT/SE99/00577 -3 As such, the second, third and fourth solutions have some serious drawbacks as mentioned above. The stand-alone downlink solution (5) has an inherent problem in not "hearing" enough of the BSs involved. This is the so-called "near-far problem", which can be an especially serious 5 problem in a CDMA system. The combined uplink/downlink solution-(7) also has the following drawbacks: 1) The positioning process can take a considerable amount of time, because measurements have to be performed on both the uplink and downlink; 2) the reliability of the positioning process is reduced, because "hearability" is limited to the link (up or down) with the poorest performance; and 3) the positioning process 10 uses up more information bandwidth than normal. The stand-alone uplink solution (6) has some significant problems. For example, the near-far problem can be resolved on the uplink by increasing the transmitter power of the MS, and transmitting a known signal for a relatively long period of time. However, this method produces some serious effects on system 15 performance. For example, in order to solve the hearability problem, the MS's signal has to be transmitted at a relatively high power. This high power transmission causes serious capacity reductions in the serving and surrounding cells. Also, if an MS transmits a known sequence, it has to either replace the normal speech channel with this signal (probably causing speech interruption), or send a signal in parallel. The latter 20 approach increases the MS's complexity, battery drainage, and usage of information bandwidth. Of course, these problems are exacerbated if the signal is transmitted for a relatively long period of time. However, as described in detail below, the present invention successfully resolves the above-described problems. 25 SUMMARY OF THE INVENTION In accordance with a preferred embodiment of the present invention, a method and system are provided for determining the position of MSs in a CDMA cellular system, in which each symbol to be transmitted by a MS is first spread by a short code (SC), and the resulting signal is further spread by a long code (LC). The spreading 30 code is divided into N-chip sections. Consequently, even if the transmitted symbols are unknown, the N-chip sections of the resulting signal are known (at least to within an WO 99/52235 PCT/SE99/00577 -4 unknown phase difference). The MS then transmits the resulting spread signal. At the receiving BS, the received spread signal is correlated (and despread) with the known codes, and the original data is reconstructed. As such, the timing of the received signal can be determined with an accuracy of at least the time interval of the unknown phase 5 difference. An important technical advantage of the present invention is that the position of mobile stations can be determined with minimal interference to the system. Another important technical advantage of the present invention is that the position of mobile stations can be determined without utilizing additional 10 communication resources. Still another important technical advantage of the present invention is that the position of mobile stations can be determined based on a normal traffic signal, which significantly decreases the transmission power required in comparison with existing mobile station positioning approaches. 15 BRIEF DESCRIPTION OF THE DRAWINGS A more complete understanding of the method and apparatus of the present invention may be had by reference to the following detailed description when taken in conjunction with the accompanying drawings wherein: 20 FIGURE 1 is a diagram of an exemplary frame format that can be used to implement a preferred embodiment of the present invention; FIGURE 2 is a schematic diagram of an exemplary MS positioning scenario that illustrates how to implement the preferred embodiment of the present invention; FIGURE 3 is a flow diagram of an exemplary method that can be used to 25 implement the preferred embodiment of the present invention; FIGURE 4 is a diagram that illustrates an exemplary bank of sliding correlators that can be used to implement the preferred embodiment of the present invention; and FIGURE 5 is a diagram that illustrates an exemplary bank of sliding correlators that can be used to implement a second embodiment of the present invention. 30 WO 99/52235 PCT/SE99/00577 -5 DETAILED DESCRIPTION OF THE DRAWINGS The preferred embodiment of the present invention and its advantages are best understood by referring to FIGUREs 1-5 of the drawings, like numerals being used for like and corresponding parts of the various drawings. 5 Essentially, in accordance with a preferred embodiment of the present invention, a method and system are provided for determining the position of MSs in a CDMA cellular system, in which each symbol to be transmitted by a MS is first spread by a SC, and the resulting signal is further spread by a LC. The spreading code is divided into N-chip sections. Consequently, even if the transmitted symbols are unknown, the N 10 chip sections of the resulting signal are known (at least to within an unknown phase difference). The MS then transmits the resulting spread signal. At the receiving BS, the received spread signal is correlated (and despread) with the known codes, and the original data is reconstructed. As such, the timing of the received signal can be determined with an accuracy of at least the time interval of the unknown phase 15 difference. Specifically, FIGURE 1 is a diagram of an exemplary frame format that can be used to implement a preferred embodiment of the present invention. As mentioned above, in most existing and planned CDMA cellular systems (and under normal operating conditions), each symbol to be transmitted by a MS is first spread by a short 20 code (SC). The time duration of the SC is typically equal to the symbol period. The resulting signal is then scrambled by at least one LC. The time duration of the LC is typically substantially longer than the symbol period, and it can even be up to several weeks or months long. As such, in the following description, a combination of LCs is referred to as the "LC," and a combination of all codes are referred to simply as the 25 "code". The coded signal is then modulated and transmitted over the air interface. At the receiver (e.g., a BS), the received signal is correlated with the known code (despread), and the original data is reconstructed. One existing MS positioning approach is for an MS to transmit a known signal (so-called "positioning pulse") to facilitate the position determination process. In 30 contrast, in accordance with the present invention, the MS whose position is to be determined is not required to change its transmitted signal to facilitate the positioning WO 99/52235 PCT/SE99/00577 -6 process, although a small change can be advantageous in some cases, as described below. In some instances, parts of the MS's transmitted signal are known. In these cases, this additional information can be utilized to improve the positioning performance. A second exemplary embodiment that utilizes such additional 5 information is described in detail below. Referring to FIGURE 1, the symbols to be transmitted by a MS are denoted as SI, S2,..., and the spreading codes are denoted as C 1 , C 2 ,... The coded symbols to be transmitted are denoted as SIC 1 , S 2

C

2 ,... As shown, the spreading codes are divided into a plurality of N-chip sections. The "length" in time of the N-chip section 10 sequences is equal to the symbol interval. As such, the spread signal (e.g., SIC 1 ) can be viewed as a repeated known signal, but with an unknown phase. As illustrated by the coded symbols (e.g., SC 1 , etc.), and in accordance with the present invention, even if the symbols are unknown, N-chip sections of the resulting signal are known (to within an unknown phase difference). 15 At a receiver (e.g., at a BS), the received spread signal is time-dispersed (due to channel effects)and contains a great deal of interference (primarily from other users). However, for illustrative purposes, it can be assumed at this point that there no time dispersion effects or interference. It can also be assumed that the timing of the signal is known to within an M-chip uncertainty. The timing of the received spread signal can 20 be determined by correlating the code (e.g., C 1 ) with M different N-chip sections of the signal, which corresponds to M different time-shifts. A finer resolution can be obtained if the signal is sampled at a higher rate than the chip-rate and more than M correlations are performed. The results (e.g., output of a correlator) of these correlations are set forth in a 25 vector. The timing of the received signal is determined from the location of the largest peak value in this vector (which corresponds to a certain time-shift). In a similar manner, the timing of the received signal can be determined by correlating with other codes (e.g., C 2 , C 3 , etc.) instead of C 1 , but at time-shifts delayed by one or more symbol intervals. 30 In the presence of interference, it is preferable for correlation to be performed with more than one N-chip code sequence, because the above-described peak value will WO 99/52235 PCT/SE99/00577 -7 not be discernible. In order to resolve this problem, the vectors obtained from the correlations with C 1 , C 2 , etc. can be combined. The transmitted signal is modulated by the unknown symbols. Consequently, the correlations are preferably combined non coherently (i.e., the absolute value of the correlations are taken before the combination 5 is derived). The preferred embodiment described below with respect to FIGURE 2 illustrates an exemplary implementation of combining and filtering the correlation output vectors. As such, after combining a sufficient number of correlation vectors, the resulting M-vector will have a discernible peak value, which will provide the timing of the desired received signal. 10 In the presence of time-dispersion, the above-described M-vector derived from the received signal can include several peak values due to multipath propagation. In this case, the first peak that is higher than a predefined threshold value is likely to be the line-of-sight component, and it is preferably selected to provide the timing of the received signal. By selecting the first peak value, the non-active BSs can relatively 15 easily determine the timing of the received signal from the MS by using a bank of sliding correlators (e.g., as described with respect to the preferred embodiment below). Notably, the timing of the received signal from the MS whose position is to be determined can be ascertained even while the MS continues transmitting its normal data traffic. Consequently, in accordance with the present invention, the system will 20 experience no information loss or loss in capacity due to a new, non-information carrying signal being transmitted. Therefore, there is no need to radically increase the MS's transmission power, which would cause an increase in interference and thus a loss in capacity (and possibly RF blocking of the signal to the active BS). If the data transmitted by the MS of interest has a relatively high data-rate, or 25 equivalently, a low spreading factor, the symbol time interval will be quite short. In this case, the gain from the non-coherent combining process is decreased, because the coherent combining process (i.e., the correlations) is performed for fewer chips (N is smaller). In order to alleviate this problem, the MS can be ordered by the network to alter the data rate so that the length of the transmitted symbol is increased. This effect 30 is possible without lowering the data-rate, by using rate matching (e.g., code puncturing) and an increase in transmitted power to make up for the decreased coding WO 99/52235 PCT/SE99/00577 -8 gain. This solution is a convenient one, because it is likely that the transmission power will be incremented upward in any event, as described earlier. On another note, if the MS of interest is using a packet channel (i.e., the MS is not transmitting continuously), the cellular system preferably selects a channel on 5 which the MS transmits continuously during the MS positioning process (e.g., a physical control channel). FIGURE 2 is a schematic diagram of an exemplary MS positioning scenario that illustrates how to implement the preferred embodiment of the present invention. In the description that follows, it can be assumed that the position of the MS 120 is 10 being determined by an uplink method that utilizes an unknown transmitted signal. However, note that the invention is not intended to be so limited and can also include a downlink method, which is a similar application to that of the uplink method. The downlink case is omitted for the sake of brevity. Also, note that a combined uplink/downlink MS positioning methodology can be applied. 15 The exemplary MS positioning method 250 is illustrated by the flow diagram shown in FIGURE 3. The positioning method commences at step 252 by the serving BS (e.g., 100) sending an order (control message via the air interface) to the MS 120 to use a predefined spreading factor and transmission power level. A power increase can, for example, be accomplished with a typical fast power control feedback approach 20 used in a known CDMA system. In many instances, a rate change does not have to be performed. Also, a different procedure for handling voice/data inactivity may be needed, since it is desired that the MS 120 transmit even during the voice/data inactivity period. At step 254, a set (e.g., at least one) of nearby non-active (MS 120 not 25 connected to) BSs (e.g., 1 10a and/or 1 10b) is/are ordered by the network to start searching for and correlating with the code used by the MS 120. This code search is preferably performed with a bank of sliding correlators. An exemplary bank of sliding correlators (200) that can be used to implement the preferred embodiment of the present invention is illustrated in FIGURE 4. In the correlator bank 200, a baseband 30 signal, x(t), enters the sliding correlators (1 -M). Each block "D" denotes a chip delay. At step 256, the BSs correlate the code with that in the baseband signal with a WO 99/52235 PCT/SE99/00577 -9 multiplier 2 0 2 1-m and a register 2 0 6 -m. The register output is added (2041.M) for feedback to the output of the multiplier 2 0 2 1-. and coupled to the register input. In this exemplary embodiment, every N chips the register 2 0 6 1-m is reset. Then, the current value of the correlated signal, y(t), in the respective register is sampled and 5 passed through an absolute value function (e.g., filter) 2 0 8 1m, and then added 2 10 1-m to a second register 2 12 1-m. As shown, there are M of these correlators and non coherent combiners. The output values, z(1)-z(M), from these combiners constitute the correlation function from which the TOA value of the received signal can be estimated (step 258). At step 260, the BS estimates the MS's position from the 10 estimated TOA of the received signal. At step 262, the network can then order (via a control message over the air interface) the MS 120 to resume its normal operation. In operation, the length of the transmitted symbols can be defined as N chips long. Also, it can be assumed (for simplicity) that the received signal is sampled at the chip rate. In the case of a faster sampling rate, more sliding correlators can be used. 15 If there are no interfering users, the output from the bank (200) sliding correlators could have one or more magnitude peaks for every transmitted signal. These peaks correspond to the different paths that constitute the communication channel. The first of these peaks is most likely the direct-path, and the arrival time of this path can be used as a time-stamp for the received signal. A set of these time-stamps, which can be 20 obtained from different BSs, can be used to estimate the position of the MS 120. One of the known location estimation algorithms can then be used at the network side to calculate the MS's position. If the cell size associated with a particular BS is relatively large, more correlators can be implemented in the bank 200, since the uncertainty of the timing in 25 the other BSs becomes larger. For example, if the timing uncertainty with respect to another BS is 32is (equal to about 9 km), approximately 128 correlators would be needed in the band if the chip-interval is 0.24 [ts. In the presence of interference, the plurality of peak values in the correlator (as described earlier) most likely will not be discernible at the more remote BSs. However, 30 this problem can be resolved by combining a large number of M-vectors. This combining process is preferably performed non-coherently, because the transmitted WO 99/52235 PCT/SE99/00577 -10 data is unknown. This combining function can be performed by taking the absolute value of the output of the filter banks, and then adding the resulting sequences together. In other words, a vector of data values that is M entries long can be so obtained. The m* entry of this vector is equal to z(m)=|y(m,1) + ly(m,2)| +... ly(m,P)1, (1) 5 where y(m,n) is the n* output of the m' sliding correlator (in the bank 200), and P is the total number of combined vectors. Alternatively, the vector combining can be performed by summing the squared modulus of the output of the filter banks. In this case, the mth entry would take the form z(m)=|y(m,1)| 2 +|y(m,2) 1+... ly(m,P)1 2 . (2) 10 As such, the present invention is not intended to be limited to a specific method of performing the non-coherent vector combining. The exemplary embodiment shown in FIGURE 4 illustrates how such sliding correlator and non-coherent combining functions can be implemented. The first peak in the above-described vector, which is preferably higher than a 15 predefined threshold value, indicates the arrival time of the direct-path signal from the MS 120. However, an estimate of the TOA value for this direct-path signal can be refined by using a known signal processing technique. For example, the location of this peak value can be determined by using one of a number of known interpolation/smoothing techniques. 20 The above-described non-coherent combining function can be performed during a relatively long time duration without experiencing any detrimental effects on the communication system involved. The primary limitation on this function is the time stability of the MS 120. However, since the MS is accurately synchronized to the serving BS (e.g., 100), the MS's time stability does not pose a significant problem. In 25 any event, even if the non-coherent vector combining function can be performed during a relatively long time duration, this approach is likely to be attractive for many applications in that the position determination estimate can be made quite rapidly. In WO 99/52235 PCT/SE99/00577 -11 order to speed up the position determination process, and in some cases to make it possible to perform at all, the network can order the MS 120 to increase its transmit power. Typically, when such a situation occurs, the MS is quite close to the serving BS (i.e., a "near-far" hearing problem). As a result, the MS can cause increased 5 interference primarily in the serving cell. However, this increased interference potentially can be alleviated by a known interference cancellation technique. At this point it is important to note that, in accordance with the present invention, the typical increase in the MS's transmit power which is needed is much lower than that required by existing methods, whereby a special positioning signal is 10 transmitted from the MS. Using one of these existing methods, the positioning signal has to be transmitted at a relatively high power level (up to 25 dB above normal) in order to minimize the data loss. Consequently, these existing MS positioning methods can cause severe interference and possibly even RF blocking of signals in the serving cell. 15 FIGURE 5 is a diagram that illustrates an exemplary bank of sliding correlators (300) that can be used to implement a second embodiment of the present invention. As shown, in the bank 300, a baseband signal, x(t), enters the sliding correlators (1 -M). Each block "D" denotes a chip delay. The correlation of the code with that in the baseband signal is performed by a multiplier 3 0 2 1.m and a register 3 06 1-m. The register 20 output is added (3041-m) for feedback to the output of the multiplier 3 02 1m. and coupled to the register input. In this exemplary embodiment, every N chips the register 30 6 1.M is reset. Then, the current value of the correlated signal, y(t), in the respective register is sampled and passed through a switch 3 07 1-m. The position of the switch is determined by the state of the current symbols in the correlated signal. In other words, 25 if the current symbols in the correlated signal are known, then the switch is positioned to coherently combine the symbols (with a register 314 1-m and associated conjugate signal multiplier and adder as shown). Otherwise, if the symbols in the correlated signal are unknown, the signal is non-coherently combined (e.g., as described above with respect to FIGURE 4). As such, the switch 307 1-m couples the signal with the 30 unknown symbols to a norm function (e.g., filter) 3 0 8 1-m, and the resulting signal is then added 3 10 1-m to a register 3 12 1-.. As shown, there are M of these correlators, WO 99/52235 PCT/SE99/00577 -12 and M coherent and non-coherent combiners. The output values, z(1)-z(M), from these correlators constitute the correlation function from which the TOA value of the received signal can be determined. As shown, most of the elements shown in FIGUREs 4 and 5 are virtually 5 identical. However, in the exemplary embodiment shown in FIGURE 5, the basic difference from the embodiment shown in FIGURE 4 is that in FIGURE 5 the signals are coherently combined when the received data is known, and non-coherently combined when the received data is unknown. In FIGURE 5, when the received symbols are known, the correlations are multiplied with the complex conjugate of the 10 corresponding symbols and combined coherently. In this case, the combined vectors corresponding to the vectors (1) and (2) above, will take the respective forms z(m)= sl*y(m, 1)+s 2 *y(m,2)+.. .s*y(m,k) + y(m,k+1) ... + y(m,P) , (3) and z(m) =si*y(m, 1) +s 2 *y(m,2) +...s*y(m,k) 2 + y(m,k+ 1) 2... + y(m,P) 2 (4) where si* is the complex conjugate of the symbol, I, and k is the number of known 15 symbols. Although a preferred embodiment of the method and apparatus of the present invention has been illustrated in the accompanying Drawings and described in the foregoing Detailed Description, it will be understood that the invention is not limited to the embodiment disclosed, but is capable of numerous rearrangements, modifications 20 and substitutions without departing from the spirit of the invention as set forth and defined by the following claims.

Claims (17)

1. A method for facilitating the positioning of a mobile station in a CDMA based cellular system using one or more short codes and a long code in an unknown signal transmitted from the mobile station, comprising the steps of: 5 ordering the mobile station to increase at least a portion of a transmission power level by a predetermined amount; correlating in at least one base station the unknown signal with parts of said one or more short codes and said long code; and non-coherently combining an output resulting from the correlating step. 10

2. The method of Claim 1, wherein at least one of said one or more short codes and said long code are divided into a plurality of equal sections.

3. The method of Claim 1, wherein at least one of said one or more short 15 codes and said long code are divided into a plurality of N-chip sections.

4. A method for facilitating the positioning of a mobile station in a CDMA based cellular system using a predefined combination of codes to spread an unknown signal transmitted from the mobile station, comprising the steps of: 20 receiving said spread unknown signal in at least one base station in said CDMA based cellular system; correlating in said at least one base station said received spread unknown signal with parts of said predefined combination of codes; and non-coherently combining correlation outputs resulting from said correlating 25 step.

5. The method of Claim 4, further comprising the step of detecting a timing of a signal transmitted from the mobile station, at said non-coherently combining step. WO 99/52235 PCT/SE99/00577 -14

6. A method for facilitating the positioning of a mobile station in a CDMA based cellular system using a predefined combination of codes to spread a partially known signal transmitted from the mobile station, comprising the steps of receiving said spread partially known signal in at least one base station in said 5 CDMA-base cellular system; correlating in said at least one base station said received spread partially known signal with parts of said predefined combination of codes; coherently combining correlation outputs resulting from said correlating step, corresponding to a known part of said partially known transmitted signal; and 10 non-coherently combining correlation outputs resulting from said correlating step, corresponding to an unknown part of said partially known transmitted signal.

7. The method of Claim 6, further comprising the step of non-coherently combining the results of the coherently combining step and non-coherently combining 15 step.

8. A correlator bank for code searching in a CDMA mobile communications system, comprising: a plurality of multipliers, each multiplier of said plurality of multipliers connected in parallel so as to receive a coded signal; 20 a first plurality of registers, an input of each register of said first plurality of registers coupled to an output of a respective multiplier; a plurality of filters, an input of each filter of said plurality of filters coupled to an output of a respective register of said first plurality of registers; and a second plurality of registers, an input of each register of said second plurality 25 of registers coupled to an output of a respective filter of said plurality of filters.

9. The correlator bank of Claim 8, wherein each filter of said plurality of filters performs a norm function. WO 99/52235 PCT/SE99/00577 -15

10. The correlator bank of Claim 8, wherein a combined output of said second plurality of registers comprises a correlation function for estimating a time of arrival of a received signal. 5

11. The correlator bank of Claim 8, wherein said coded signal comprises a baseband signal.

12. The correlator bank of Claim 8, wherein at least one multiplier of said plurality of multipliers is coupled to a second multiplier of said plurality of multipliers 10 with a one-chip delay.

13. The correlator bank of Claim 8, wherein said output of said respective register of said first plurality of registers is added to a respective input of said first plurality of registers. 15

14. The correlator bank of Claim 8, wherein at least one register of said plurality of registers is reset at every N-chip interval.

15. The correlator bank of Claim 8, wherein said output of said respective 20 register of said second plurality of registers is added to a respective input of said second plurality of registers.

16. A correlator bank for code searching in a CDMA mobile communications system, comprising: 25 a plurality of multipliers, each multiplier of said plurality of multipliers connected in parallel so as to receive a coded signal; a first plurality of registers, an input of each register of said first plurality of registers coupled to an output of a respective multiplier; a switch responsive to a state of symbols in a correlated signal; 30 a plurality of filters, a first input of each filter of said plurality of filters coupled to a first output of said switch if said state is associated with unknown symbols; WO 99/52235 PCT/SE99/00577 -16 a second plurality of registers, an input of each register of said second plurality of registers coupled to a second output of said switch if said state is associated with known symbols, an output of each register of said second plurality of registers coupled to a second input of each filter of said plurality of filters; and 5 a third plurality of registers, an input of each register of said third plurality of registers coupled to an output of a respective filter of said plurality of filters.

17. The correlator bank of Claim 16, wherein each filter of said plurality of filters performs a norm function. 10