Description of the related Art
The use of Code Division Multiple Access (CDMA) modulation techniques is but one of several methods for facilitating communications in which a large number of system users are present. While other techniques are known, such as Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA), and AM modulation schemes such as Amplitude Companded Single Sideband (ACSSB), CDMA has significant advantages over these other modulation techniques. The use of CDMA techniques in multiple access communication systems is disclosed in U.S. patent No. 4,901,307, entitled "spread spectrum multiple access communication system using satellites or terrestrial repeaters," and U.S. patent No. 5,103,459, entitled "system and method for generating signal waveforms in a CDMA cellular telephone system," both of which are assigned to the assignee of the present invention and are incorporated herein by reference. The method for providing CDMA mobile communications, referred to herein as IS-95, IS standardized by the telecommunications industry association in the united states under the designation "mobile station-base station compatibility standard for dual mode wideband spread spectrum cellular systems" by TIA/EIA/IS-95-a.
In the above-mentioned patents, multiple access techniques are disclosed in which a plurality of mobile station users communicate via satellite relay stations or terrestrial base stations (known as cell sites or cell sites) using Code Division Multiple Access (CDMA) spread spectrum communication signals, with each mobile station user having a transceiver. By employing CDMA communications, the spectrum may be reused multiple times, allowing for increased system user capacity. The use of CDMA techniques results in higher spectral efficiency than can be achieved using other multiple access techniques.
A method for synchronously demodulating data that has propagated (travel) from one base station along different propagation paths and for synchronously demodulating data redundantly provided at more than one base station is disclosed in U.S. patent No. 5,109,390 (' 390 patent) (entitled "diversity receiver in CDMA cellular communication system", assigned to the assignee of the present invention and incorporated herein by reference). In the' 390 patent, separately demodulated signals are combined to provide an estimate of the transmitted data with greater reliability than the demodulated data through either path or from either base station.
In general, handover can be divided into two types-hard handover and soft handover. In the hard handover procedure, when a mobile station leaves an initial (origin) base station and enters a destination base station, the mobile station disconnects a communication link with the initial base station and thereafter establishes a new communication link with the destination base station. During soft handoff, the mobile station completes the communication link with the destination base station before breaking the communication link with the initial base station. Thus, during soft handoff, the mobile station redundantly communicates with the initial base station and the destination base station for some period of time.
Soft handoff is less likely to drop (drop) calls than hard handoff. Further, when the mobile station travels near the coverage boundary of the base station, it makes repeated handover requests in response to small changes in the environment. This problem, known as ping-pong (ping-pong), is also greatly reduced by soft handover. The process for performing a soft handoff is described in detail in U.S. patent No. 5,101,501 (entitled "method and system for providing a soft handoff in communications in a CDAM cellular telephone system," assigned to the assignee of the present invention and incorporated herein by reference).
An improved soft handoff technique is disclosed in U.S. patent No. 5,267,261, entitled "mobile assisted soft handoff in a CDMA cellular communication system," assigned to the assignee of the present invention and incorporated herein by reference. In the system of patent 261, the soft handoff process is improved by measuring the strength of the "pilot" signal transmitted by each base station at the mobile station. These pilot strength measurements assist in the soft handoff process by facilitating the identification of viable base station handoff candidates.
The base station candidates may be divided into 4 groups. The first group, referred to as the active group, includes the base stations with which the mobile station is communicating. The second group, referred to as the candidate group, includes base stations whose signals are determined to be of sufficient strength to be useful to the mobile station but not currently in use. Base stations are added to the candidate set when their measured pilot energy exceeds a predetermined threshold TADD. The third group is the set of base stations that are in the vicinity of the mobile station (and not included in the active or candidate groups). And the fourth group is the remaining group containing all other base stations.
In IS-95, a base station candidate IS characterized by the phase offset of the pseudo-noise (PN) sequence of its pilot channel. When the mobile station searches to determine the strength of the pilot signal from the candidate base station, it performs a correlation operation in which the filtered received signal is correlated with a set of PN offset hypotheses. Methods and apparatus for performing correlation operations are described in detail in co-pending U.S. patent application serial No. 08/687,694 (filed 7/26 1996 entitled "method and apparatus for performing search acquisition in a CDMA communication system," assigned to the assignee of the present invention and incorporated herein by reference).
The propagation delay between the base station and the mobile station is unknown. The unknown delay produces an unknown shift in the PN code. The search process attempts to determine the unknown shift in the PN code. To do so, the mobile station shifts the output of its searcher PN code generator in time. The range of search shifts is referred to as the search window. The search window is centered on the PN shift hypothesis. The base station transmits a message to the mobile station that physically represents the PN offset of the base station pilot. The mobile station will center its search window around the PN offset hypothesis.
The appropriate size of the search window depends on several factors, including the priority of the pilot, the speed of the search processor, and the expected delay spread (delay spread) of the multipath arrival. The CDMA standard (IS-95) defines three search window parameters. The search for pilots in the current and candidate sets is governed by a search window "a". The neighbor set of pilots is searched for in the range of window "N" and the remaining set of pilots is searched for in the range of window "R". The searcher window size is provided in Table 1 below, where the chips are
| SRCH_WIN_ASRCH_WIN_NSRCH_WIN_R | Window size (PN code slice) | SRCH_WIN_ASRCH_WIN_NSRCH_WIN_R | Window size (PN code slice) |
| 0 | 4 | 8 | 60 |
| 1 | 6 | 9 | 80 |
| 2 | 8 | 10 | 100 |
| 3 | 10 | 11 | 130 |
| 4 | 14 | 12 | 160 |
| 5 | 20 | 13 | 226 |
| 6 | 28 | 14 | 320 |
| 7 | 40 | 15 | 452 |
TABLE 1
Window size positioning (sizing) is a trade-off between search speed and the probability of losing a strong path outside the search window.
The base station sends a message to the mobile station specifying the PN hypothesis that the mobile station should search against its own PN offset. For example, the initial base station may instruct the mobile station to search for pilot 128PN chips prior to its PN offset. In response, the mobile station places its searcher demodulator 128 chips before the output chip cycle and searches for pilots using a search window centered at a particular offset. Once the mobile station is commanded to search for PN hypotheses to determine resources that can be captured to perform a handoff, it is critical that the PN offset of the destination base station pilot be very close in time to the directional offset. The speed of the search near the base station boundary is important because the delay in completing the required search results in a lost call.
In CDMA systems in the united states, synchronization of base stations is acquired by having each base station provided with a Global Positioning Satellite (GPS) receiver. However, there are situations where the base station may not be able to receive GPS signals. For example, in subways and tunnels, GPS signals are attenuated to the point where their use for timing synchronization base stations or micro base stations is prohibited. The present invention provides a method and system for providing timing synchronization in situations where a small portion of the network is able to receive and acquire timing from a centralized timing signal and a portion of the base stations are unable to receive the centralized timing signal.
Detailed description of the preferred embodiments
I. Overview of timing error calculation
Referring to fig. 1, a mobile station 60 is in communication with a base station 62 while it is approximately within a coverage area delineated by a base station coverage boundary 61. The base station 62 is synchronized with the rest of the network by using a central timing system such as the Global Positioning System (GPS). In contrast, the base station 64 is not synchronized with the central timing system. The base station controller 66 routes calls from the PSTN to the base station 62 or 64 by using a T1 line or other means. In addition, frequency synchronization is provided to the base station 64 via the T1 line.
The frequency synchronization can be made with acceptable accuracy for short time periods over the T1 line using methods known in the art. However, in these schemes, glitches (glitches) are common for providing frequency information. These glitches result in timing errors that are correctable by using the present invention. The phase intermittent correction of the present invention allows the use of less accurate frequency sources when needed due to the relationship between phase and frequency.
Referring to fig. 2, there is shown the transmission and corresponding time intervals used to synchronize the timing of the slave base station 64 and the synchronization timing of the reference base station 62. Signal path 500 illustrates transmission of the forward link from reference base station 62 to mobile station 60. The time interval during which such transmission takes place is designated τ1. At the mobile station 60, time aligns the start time of a frame transmission on the reverse link with the start time of a frame arrival on the forward link. The time alignment IS standardized in IS-95 and hardware conforming to its design IS added, so that methods and apparatus for performing the alignment are known in the art.
Transmission 502 illustrates the transmission of a reverse link frame from the mobile station 60 to the reference base station 62. Signal 500 goes from base station 62 to mobile station 60(τ)1) Is equal to the time of signal 502 from base station 62 to mobile station 60 (or τ)1) Time of (d). Since base station 62 knows when it sent signal 500 and knows when it received signal 502, base station 62 can calculate a round trip delay time (RTD) time1) It is the calculation of the time error (tau)0′-τ0) A desired first value.
Signal path 504 is the reverse link signal transmission of the reverse link signal from the mobile station 60 to the slave base station 64 along a different propagation path. The time required for the signal 504 to propagate from the mobile station 60 to the slave base station 64 is designated τ2. The time of arrival of the reverse link signal 504 at the base station 64 is designated as T2. The time required for forward link signal 506 to travel from base station 64 to mobile station 60 is equal to tau2. In addition, the slave base station 64 may measure the time difference between the time it receives the reverse link signal from the mobile station 60 and the time it transmits its forward link signal to the mobile station 60. Will be provided withThe time difference is designated as RTD2. Knowing these times allows the calculation of the time error (τ)0′-τ0). The following description is given for calculating the time error τ0The method of' above.
First, it follows from fig. 2:
T2＝τ1+τ2and (1)
τ1+ΔT＝T0′+T2 (2)
By manipulating the terms of equations (1) and (2), the following equations are derived:
T2+ΔT＝T0′+2·τ2 (3)
2·τ2＝T2T0′+ΔT (4)
reduced sign, new variable RTD2Is defined as RTD2＝T2-T0′ (5)
RTD2＝T2-T0′ (5)
As can be seen from fig. 2:
T2＝T0+τ1+τ2 (7)
therefore, the temperature of the molten metal is controlled,
T2-T0＝τ1+τ2and (8)
RTD2＝2·τ2-ΔT
By substitution, time error (T)0′-T0) Equal to:
T0′-T0＝τ1-τ2+ΔT (9)
once the base station 64 knows its timing error amount (T)0′-T0) It adjusts its timing so that it is synchronized with the timing of the base station 62. These measurements may be erroneous, so in a preferred embodiment, multiple measurements are made redundantly to ensure timing correction accuracy.
Now, a method and apparatus for measuring each required time value in equation (12) will be described.
Measuring Round Trip Delay (RTD)1)
Fig. 3 is a flow chart illustrating the method of the present invention for synchronizing the timing of the slave base station 64 with the reference base station 62. In step 300, the synchronization method is initiated and the rover station 60 communicates with the reference base station 62 and falls within range to direct communication with the slave base station 64. In step 302, the Round Trip Delay (RTD) of the signal from the reference base station 62 to the mobile station 60 and from the mobile station 60 to the reference base station 62 is measured1) Time. This may be done by aligning the frame boundaries of frames received by the mobile station 60 with the frame boundaries of frames transmitted by the mobile station 60. Methods and apparatus for providing such alignment are known in the art. Thus, the Round Trip Delay (RTD) is measured1) As the time difference between the start time of the frame transmitted by the reference base station 62 and the start time of the frame received by the reference base station 62 from the mobile station 60.
Referring to fig. 4, forward link frames of data from a reference base station 62 are received at antenna 2 and provided through duplexer 3 to a receiver (RCVR) 4. Receiver 4 down-converts, filters, and amplifies the received signal and provides it to searcher 50 and TRAFFIC demodulator (TRAFFIC DEMODS) 54. The searcher 50 searches for pilot channels based on the neighbor set provided by the reference base station 62. The neighbor set list is provided as signaling data from the reference base station 62 on the traffic channel. A signal is provided to the control processor 55 indicating the start of a received frame from the reference base station 62. Control processor 55 generates and provides a time alignment signal to traffic modulator 58 that aligns the start time of a frame transmitted from mobile station 60 with the start time of a frame received at mobile station 60.
Data frames from a user of mobile station 60 are provided to traffic modulator 58, where traffic modulator 58 time aligns frames transmitted by transmitter (TMTR)56 with frames received by mobile station 60 from reference base station 62 in response to timing signals from control pre-process 55. The reverse link frames are upconverted, filtered and amplified by a transmitter 56 and then provided through a duplexer to be transmitted through an antenna 2.
Acquisition of mobile station by subordinate base station
Fig. 6 shows a traffic channel modulator 58 of a mobile station 60. The data frame is provided to the frame formatter 200. In the exemplary embodiment, frame formatter 200 generates and appends a set of Cyclic Redundancy (CRC) check bits and generates a set of tail bits. In an exemplary embodiment, the frame formatter 200 follows the frame format protocol standardized in IS-95 and IS described in detail in U.S. Pat. No. 5,600,754 entitled "method and System for arranging vocoder data to mask transmission channel induced errors," assigned to the assignee of the present invention and incorporated herein by reference.
The formatted data frames are provided to an encoder 202 that encodes data for error correction and detection. In an example embodiment, encoder 202 is a convolutional encoder. The encoded data symbols are provided to an interleaver 204 which reorders the symbols according to a predetermined interleaving format. The reordered symbols are provided to Walsh mapper 206. In the exemplary embodiment, Walsh mapper 206 receives eight coded symbols and maps the set of symbols to 64 chip Walsh sequences. The Walsh symbols are provided to spreading means 208 which spread the Walsh symbols according to the long spreading codes. The long PN code generator 210 generates a pseudo-random (PN) sequence that spreads and distinguishes the data from reverse link transmission data from other mobile stations in the vicinity.
In an exemplary embodiment, according to quadrature phase shift keying (Q)PSK) modulation format, in which the I and Q channels are spread according to a short PN sequence. According to respective PN generators (PN)1And PNQ) Spreading means 214 and 216, which perform a second spreading operation on the data with the short PN sequences provided by 212 and 218, provide spread data.
In step 304, the slave base station 64 acquires the reverse link signal transmitted by the mobile station 60. The base station controller 66 sends a signal to the slave base station 64 indicating the PN code offset that the mobile station 62 uses to spread its reverse link signal. In response to this signal from the base station controller 66, the slave base station 64 searches for the mobile station 60 centered on the PN offset represented by the signal from the base station controller 66.
In the exemplary embodiment, the slave base station 64 group loads (bank load) its searcher long code PN generator 106 and its short code PN generators 108 and 110 (shown in fig. 9) based on signals from the base station controller 66. The searcher process of the slave base station 64 is also described in detail herein.
Fig. 7 shows the arrangement of the slave base station 64. In the slave base station 64, a signal representing the PN of the mobile station 60 is received from the base station controller 60. The message is provided to the control processor 100. In response, control processor 100 calculates a window search range centered at a particular PN offset. The control processor 100 provides the search parameters to the searcher 101 and, in response to those parameters, the slave base station 64 directs a search for signals transmitted by the mobile station 60. The signal received by the antenna 102 of the slave base station 64 is provided to a receiver 104, wherein the receiver 104 downconverts, filters and amplifies the received signal and provides it to the searcher 101. In addition, the received signal is provided to traffic demodulator 105, which demodulates reverse link traffic data and provides the data to base station controller 60. The base station controller 66 in turn provides it to the public switched telephone network.
Fig. 9 shows the searcher 101 in detail. Demodulation of reverse link signals is described in detail in co-pending U.S. patent application No. 08/372,632 (filed on 3.1.1995 entitled "cell-site demodulator architecture for spread spectrum multiple access communication systems") and co-pending U.S. patent application No. 08/316,177 (filed on 30.9.1994 entitled "multipath search processor for spread spectrum multiple access communication systems"), both of which are assigned to the assignee of the present invention and incorporated herein by reference. An estimate of the PN offset of the mobile station 60 is provided from the base station controller 66 to the control processor 100. In response to the PN offset estimates provided by the base station controller 60, the control processor 100 generates initial long PN sequence hypotheses as well as initial short PN sequence hypotheses for performing searches by the slave base stations 64. In an example embodiment, control processor 100 sets shift registers that load PN generators 106, 108, and 110.
The signal is received by the antenna 102, down-converted, filtered and amplified and passed to the correlator 116. Correlator 116 correlates the received signal with combined long and short PN sequence hypotheses. In the exemplary embodiment, the short PN hypotheses are generated by multiplying the short PN hypotheses generated by PN generators 108 and 110 by the long PN sequence generated by PN generator 106. One of the combined PN sequence hypotheses is used to despread the I channel and the other is used to despread the Q channel receiving the QPSK signal.
Two PN despread signals are provided to Fast Hadamard Transform (FHT) processors 118 and 120. In co-pending U.S. patent application No. 08/173,460 (filed on 12/22/1993 entitled "method and apparatus for performing a fast hadamard transform", assigned to the assignee of the present invention and incorporated herein by reference.) FHT processors 118 and 120 correlate despread signals with all feasible Walsh symbols to provide energy computation means (I)2+Q2)122 provide the resulting amplitude matrix. The energy calculation means 122 calculates the energy of the elements of the amplitude matrix and provides the energy value to the maximum detector 124 which selects the largest energy dependency. The maximum correlation energy is provided to accumulator 126, which accumulates the energy for a plurality of Walsh symbols, and based on these accumulated energies, a determination is made as to whether mobile station 60 can be acquired at that PN offset.
Adjusting initial timing by a slave base station
Once the mobile station 60 is acquired, the slave base station 64 adjusts its timing so that the mobile station 60 can successfully acquire its forward link transmission in block 306. The slave base station 64 operates by determining the difference between the PN offset used to acquire the reverse link signal from the mobile station 60 and the PN offset used by the reference base station 62 to receive the reverse link signal from the mobile station 60. Using this PN offset difference, the slave base station 64 adjusts the timing of its pilot signal in such a way that it will fall within the search window of the mobile station 60 when the mobile station 60 searches for its pilot signal.
Acquisition of subordinate base stations by mobile stations
The slave base station 64 needs to have some indication of time in the search for the mobile station signal. In the preferred embodiment, the time error of the slave base station 64 is maintained at or below 1 millisecond by another synchronization scheme. There is a scheme of keeping the slave base station 64, which cannot receive the GPS signal, at a time of low accuracy. One possible way to achieve the degree of initial synchronization is to manually set the time of the slave base station 64 at a certain interval. A second method is to use a WWV receiver to set the time, which is known in the art. Unlike GPS signals, WWV centralized timing signals are transmitted at very low frequencies and are able to penetrate (vertical inter) tunnels and subways. However, WWV receivers cannot provide the degree of time synchronization required to conduct CDMA communications.
In the exemplary embodiment, the slave base station 64 adjusts its timing based on the assumption that the mobile station 60 is directly in the vicinity of the slave base station 64. Initial timing adjustments are then made without any propagation delays between the slave base station 64 and the mobile station 60. Thereafter, the slave base station 64 adjusts its PN sequence generators 72 and 74 in advance (forward in time), which accounts for the increasing propagation delay time between the slave base station 64 and the mobile station 60. Once the mobile station 60 acquires the pilot channel of the slave base station 64, the final timing adjustment for the slave base station 64 can be performed based on the calculations described above using the normal procedure.
The pilot channels of different base stations are distinguished from each other by the phase of their PN generators, as IS known in the art and standardized in standard IS-95. The reference base station 62 instructs the mobile station 60 to search for the slave base station 64 through the neighbor list the reference base station 62 indicates the signaling data that can be acquired at the PN phase offset using the pilots of the slave base station 64, where the PN phase offset is described relative to the received PN offset of the reference base station 62. The message is demodulated and decoded by a traffic demodulator 54 and provided to the searcher 50. In response, the searcher 50 performs a search focused on a PN phase offset around a PN phase that is specified in the signal from the reference base station 62.
The pilot signal is typically generated by a linear feedback shift register, the implementation of which is described in detail in the above-mentioned patent. In order to acquire the pilot signal from the slave base station 64, the mobile station 60 must synchronize to the received signal from the slave base station 64 at phase φ and frequency ω. The purpose of the searcher operation is to find the phase phi of the received signal. As previously described, the slave base station 64 may be provided with relatively accurate frequency synchronization using a T1 link from the base station controller 66, as is known in the art. The mobile station finds the phase of the received signal by testing a set of phase hypotheses, called search windows, and determining whether one of the offset hypotheses is correct.
Fig. 5 shows the mobile station searcher 50 in detail. The spread spectrum signal is received at the antenna 2. The apparatus aims to achieve synchronization between a pseudo-random noise (PN) sequence generated by the PN sequence generator 20 and a received spread spectrum spread by the same PN sequence of unknown phase transmitted by the slave base station 64. In the exemplary embodiment, pilot signal generator 76 (fig. 7) and PN generator 20 are maximum length shift registers that generate PN code sequences that are used to spread and despread pilot signals, respectively. Thus, the operation of acquiring synchronization between the code for despreading the received pilot signal and the PN spreading code of the received pilot signal includes determining the time offset of the shift register.
The spread spectrum signal is provided by the antenna 2 to the receiver 4. The receiver 4 down-converts, filters and amplifies the signal and provides the signal to a despreading element 6. The despreading element 6 multiplies the received signal by the PN code generated by the PN generator 20. Due to the random noise-like nature of the PN code, the product of the PN code and the received signal should be substantially zero except for the moments that are close to synchronization.
The search controller 18 provides the offset hypotheses to the PN generator 20. The bias hypothesis is determined based on the signal transmitted by the reference base station 62 to the mobile station 60. In the exemplary embodiment, the received signal is modulated by Quadrature Phase Shift Keying (QPSK), so PN generator 20 provides the PN sequence for the I modulation component and the separate sequence for the Q modulation component to despreading element 6. The despreading element 6 multiplies the PN sequence with its corresponding modulation component and provides the product of the two output components to coherent accumulators 8 and 10.
Coherent accumulators 8 and 10 sum the products beyond the length of the product sequence. Coherent accumulators 8 and 10 are responsive to signals from searcher controller 18 for resetting, locking and setting the addition period. The sum of the products is supplied from adders 8 and 10 to squaring device 14. The squaring means 14 squares each sum and adds the squared values.
The squared sum is provided by squaring means 12 to a non-coherent combiner 14. The non-coherent combiner 14 determines the energy value of the output from the squaring device 12. The non-coherent accumulator 14 is used to cancel out the effect of frequency differences between the base station transmit clock and the mobile station receive clock and to facilitate detection statistics in fading environments. The non-coherent accumulator 14 provides an energy signal to a comparison device 16. The comparison means 16 compares the energy value with a predetermined threshold provided by the searcher controller means 18. The result of each comparison is then fed back to the searcher controller 18. The results fed back to the searcher controller 18 include the energy of the correlation and the PN offset caused in the measurement.
In the present invention, the searcher controller 18 outputs the PN phase at which it is synchronized with the base station 64. This offset is used to calculate the time error, as described further below.
In the exemplary embodiment, when the mobile station 60 acquires the slave base station 64, it calculates the difference between the time it receives the signal from the slave base station 64 and the time it receives the signal from the reference base station 62. The value is provided to a message generator 52 that generates a message representing the difference. On the reverse link, the message is sent as signaling data to the reference base station 62 and the slave base station 64, which return the message to the base station controller 66.
Measuring delay between transmission of forward link signal from slave base station and reception of reverse link signal at slave base station
In step 311, the slave base station 64 measures the time (T) it receives the reverse link signal from the mobile station 602) And the time (T) it transmits the forward link signal to the mobile station 601) The difference between them. The slave base station 64 stores the PN offset at the time it transmits its forward link signal and, upon detecting the reverse link signal from the mobile station 60, calculates the time difference RTD2. In the exemplary embodiment, the calculated time difference is provided by the slave base station 64 to the base station controller 66, and adjustments to the timing calculations are performed at the base station 66. Those skilled in the art will appreciate that the present invention readily extends to cases where the calculations are performed at either the base station or the mobile station.
Timing adjustment of slave base stations
The base station controller 66 in response performs the calculations described in equation (12) and sends an indication of the required timing adjustment to the slave base station 64. Referring to fig. 7, the timing adjustment signal is received by the slave base station 64 at the control processor 100. The control processor 100 generates and provides control signals to the timing adjustment processor 99. Timing adjustment processor 99 generates a signal that is measured by the amount indicated by the signal from base station controller 66. The time of the timing source 98 is changed.