EP1123586A1

EP1123586A1 - Method and apparatus for pseudonoise spreading in a cdma communication system

Info

Publication number: EP1123586A1
Application number: EP99956605A
Authority: EP
Inventors: Daisuke Terasawa; Avneesh Agrawal; Yu-Cheun Jou; Brian Harms; Brian K. Butler
Original assignee: Qualcomm Inc
Current assignee: Qualcomm Inc
Priority date: 1998-10-19
Filing date: 1999-10-18
Publication date: 2001-08-16
Also published as: HK1040332A1; CN1331866A; KR20010080248A; JP2002528951A; WO2000024136A1; AU1317600A

Abstract

Method for PN spreading a SCMA signal using a truncated PN sequence. Method for generating said truncated PN sequence. Method for generating a second truncated PN sequence by masking the first truncated sequence. A method for generating a phase shifted version of a truncated PN sequence.

Description

METHOD AND APPARATUS FOR PSEUDONOISE SPREADING IN A CDMA COMMUNICATION SYSTEM

BACKGROUND OF THE INVENTION

I. Field of the Invention

The present invention relates to spread spectrum communications.

More particularly, the present invention relates to a novel and improved method and apparatus for pseudorandom spreading in a direct sequence code division multiple access (CDMA) communication system.

II. Description of the Related Art

As wireless communication technology has advanced, the demand for high speed data services in a wireless environment has grown dramatically. The use of code division multiple access (CDMA) modulation is one of several techniques for providing digital wireless transmission that is well suited for the transmission of digital data. Other methods of digital wireless transmission include time division multiple access (TDMA) and frequency division multiple access (FDMA).

However, the spread spectrum modulation technique of CDMA has significant advantages over other digital modulation techniques. The use of CDMA techniques in a multiple access communication system is disclosed in U.S. Patent No. 4,901,307, entitled "SPREAD SPECTRUM MULTIPLE ACCESS COMMUNICATION SYSTEM USING SATELLITE OR TERRESTRIAL REPEATERS", assigned to the assignee of the present invention and incorporated by reference herein. The use of CDMA techniques in a multiple access communication system is further disclosed in U.S. Patent No. 5,103,459, entitled "SYSTEM AND METHOD FOR GENERATING SIGNAL WAVEFORMS IN A CDMA CELLULAR TELEPHONE SYSTEM", assigned to the assignee of the present invention and incorporated by reference herein. The method for providing digital wireless communications using CDMA modulation was standardized by the Telecommunications Industry Association (TIA) in TIA /EIA /IS-95-A Mobile Station-Base Station Compatibility Standard for Dual-Mode Wideband Spread Spectrum Cellular System (hereafter IS-95).

The multipath properties of the terrestrial channel produce, at the receiver, signals having traveled several distinct propagation paths. One characteristic of a multipath channel is the time spread introduced in a signal that is transmitted through the channel. The spread spectrum pseudonoise (PN) modulation used in a CDMA system allows different propagation paths of the same signal to be distinguished and combined, provided the difference in path delays exceeds the PN chip duration. If a PN chip rate of approximately 1 MHz is used in a CDMA system, the full spread spectrum processing gain, equal to the ratio of the spread bandwidth to the system data rate, can be employed against paths having delays that differ by more than one microsecond. A one microsecond path delay differential corresponds to a differential path distance of approximately 300 meters. The urban environment typically provides differential path delays in excess of one microsecond.

Another characteristic of the multipath channel is that each path through the channel may cause a different attenuation factor. For example, if an ideal impulse is transmitted over a multipath channel, each pulse of the received stream of pulses generally has a different signal strength than other received pulses.

Yet another characteristic of the multipath channel is that each path through the channel may cause a different phase on the signal. If, for example, an ideal impulse is transmitted over a multipath channel, each pulse of the received stream of pulses generally has a different phase than other received pulses. This can result in signal fading.

A fade occurs when multipath vectors are added destructively, yielding a received signal that is smaller than either individual vector. For example, if a sine wave is transmitted through a multipath channel having two paths where the first path has an attenuation factor of X dB, a time delay of d with a phase shift of Q radians, and the second path has an attenuation factor of X dB, a time delay of d with a phase shift of Q + _ radians, no signal would be received at the output of the channel. As described above, in traditional CDMA demodulator structures, the

PN chip interval defines the minimum separation two paths must have in order to be combined. Before the distinct paths can be demodulated, the relative arrival times (or offsets) of the paths in the received signal must first be determined. The demodulator performs this function by "searching" through a sequence of offsets and measuring the energy received at each offset. If the energy associated with a potential offset exceeds a certain threshold, a demodulation element, or "finger" may be assigned to that offset. The signal present at that path offset can then be summed with the contributions of other fingers at their respective offsets.

A method and apparatus of finger assignment based on searcher and finger energy levels is disclosed in U.S. Patent No. 5,490,165, entitled "FINGER ASSIGNMENT IN A SYSTEM CAPABLE OF RECEIVING MULTIPLE SIGNALS", filed October 28, 1993 and assigned to the assignee of the present invention and incorporated by reference herein. In the exemplary embodiment, the CDMA signals are transmitted in accordance with the Telecommunications Industry Association TIA /EIA/ IS-95- A entitled "MOBILE STATION-BASE STATION COMPATIBILITY STANDARD FOR DUAL-MODE WIDEBAND SPREAD SPECTRUM CELLULAR SYSTEM". The signals transmitted from a base station to a mobile station are referred to herein as forward link signals and the signals transmitted from a mobile station to a base station are referred to as reverse link signals. An exemplary embodiment of the circuitry capable of demodulating IS-

95 forward link signals is described in detail in U.S. Patent No. 5,764,592, entitled "MOBILE DEMODULATOR ARCHITECTURE FOR A SPREAD SPECTRUM MULTIPLE ACCESS SYSTEM", assigned to the assignee of the present invention and incorporated by reference herein. An exemplary embodiment of the circuitry capable of demodulating IS-95 reverse link signals is described in detail in U.S. Patent No.5,654,979, entitled "CELL SITE DEMODULATOR ARCHITECTURE FOR A SPREAD SPECTRUM MULTIPLE ACCESS COMMUNICATION SYSTEM," assigned to the assignee of the present invention and incorporated by reference herein. FIG. 1 shows an exemplary set of signals from a base station arriving at the mobile station. It will be understood by one skilled in the art that FIG. 1 is equally applicable to the signals from a mobile station arriving at the base station. The vertical axis represents the power received on a decibel (dB) scale. The horizontal axis represents the delay in the arrival time of a signal due to multipath delays. The axis (not shown) going into the page represents a segment of time. The signals in the common plane traveled along different paths arriving at the receiver at the same time, but having been transmitted at different times.

In a common plane, peaks to the right were transmitted at an earlier time by the base station than peaks to the left. For example, the left-most peak spike 2 corresponds to the most recently transmitted signal. Each signal spike 2 - 7 has traveled a different path and therefore exhibits a different time delay and a different amplitude response.

The six different signal spikes represented by spikes 2 - 7 are representative of a severe multipath environment. Typical urban environments produce fewer usable paths. The noise floor of the system is represented by the peaks and dips having lower energy levels.

The task of the searcher is to identify the delay as measured by the horizontal axis of signal spikes 2 - 7 for potential finger assignment. The task of the finger is to demodulate one of a set of the multipath peaks for combination into a single output. It is also the task of a finger, once assigned to a multipath peak, to track that peak as it may move in time.

The horizontal axis can also be thought of as having units of PN offset. At any given time, the mobile station receives a variety of signals from a base station, each of which has traveled a different path and may have a different delay than the others. The base station's signal is modulated by a PN sequence. A local copy of the PN sequence is also generated at the mobile station. Also at the mobile station, each multipath signal is individually demodulated with a PN sequence code aligned to its received time offset. The horizontal axis coordinates can be thought of as corresponding to the PN sequence code offset that would be used to demodulate a signal at that coordinate.

Note that each of the multipath peaks varies in amplitude as a function of time, as shown by the uneven ridge of each multipath peak. In the limited time shown, there are no major changes in the multipath peaks. Over a more extended time range, multipath peaks disappear and new paths are created as time progresses. The peaks can also slide to earlier or later offsets as the path distances change when the mobile station moves relative to the base station. Each finger tracks these small variations in the signal assigned to it.

In narrowband systems, the existence of multipath in the radio channel can result in severe fading across the narrow frequency band being used. Such systems are capacity constrained by the extra transmit power needed to overcome a deep fade. As noted above, CDMA signal paths may be discriminated and diversity combined in the demodulation process.

Three major types of diversity exist: time diversity, frequency diversity, and space /path diversity. Time diversity can best be obtained by the use of repetition, time interleaving, and error correction and detection coding that introduce redundancy. A system may employ each of these techniques as a form of time diversity.

CDMA, by its inherent wideband nature, offers a form of frequency diversity by spreading the signal energy over a wide bandwidth. The frequency selective fading that can cause a deep fade across a narrowband system's frequency bandwidth usually only affects a fraction of the frequency band employed by the CDMA spread spectrum signal.

The rake receiver provides path diversity through its ability to combine multipath delayed signals; all paths that have a finger assigned to them must fade together before the combined signal is degraded. Additional path diversity is obtained through a process known as "soft hand-off" in which multiple simultaneous, redundant links from two or more base stations can be established with the mobile station. This supports a robust link in the challenging environment at the cell boundary region. Examples of path diversity are illustrated in U.S. Patent No. 5,101,501 entitled "SOFT HAND- OFF IN A CDMA CELLULAR TELEPHONE SYSTEM", issued March 21, 1992 and U.S. Patent No. 5,109,390 entitled "DIVERSITY RECEIVER IN A CDMA CELLULAR TELEPHONE SYSTEM", issued April 28, 1992, both assigned to the assignee of the present invention. Both the cross-correlation between different PN sequences and the autocorrelation of a PN sequence, for all time shifts other than zero, have a nearly zero average value. This allows the different user signals to be discriminated upon reception. Autocorrelation and cross-correlation require that logical "0" take on a value of "1" and logical "1" take on a value of "-1", or a similar mapping, in order that a zero average value be obtained.

However, such PN signals are not orthogonal. Although the cross- correlation essentially averages to zero over the entire sequence length for a short time interval, such as an information bit time, the cross-correlation is a random variable with a binomial distribution. As such, the signals interfere with each other in much the same manner as if they were wide bandwidth Gaussian noise at the same power spectral density. It is well known in the art that a set of n orthogonal binary sequences, each of length n, for n any power of 2 can be constructed (see Digital Communications with Space Applications, S.W. Golomb et al., Prentice-Hall, Inc., 1964, pp. 45-64). In fact, orthogonal binary sequence sets are also known for most lengths that are multiples of four and less than two hundred. One class of such sequences that is easy to generate is called the Walsh function; a Walsh function of order n can be defined recursively as follows:

where W denotes the logical complement of W, and W(1) = |0| .

A Walsh sequence or code is one of the rows of a Walsh function matrix. A Walsh function matrix of order n contains n sequences, each of length n Walsh chips. A Walsh function matrix of order n (as well as other orthogonal functions of length n) has the property that over the interval of n bits, the cross-correlation between all the different sequences within the set is zero. Every sequence in the set differs from every other sequence in exactly half of its bits. It should also be noted that there is always one sequence containing all zeroes and that all the other sequences contain half ones and half zeroes.

In the system described in the '459 patent, the call signal begins as a 9600 bit per second information source which is then converted by a rate 1/2 forward error correction encoder to a 19,200 symbols per second output stream. Each call signal broadcast from a cell is covered with one of sixty-four orthogonal Walsh sequences, each sixty-four Walsh chips, or one symbol, in duration. Regardless of the symbol being covered, the orthogonality of all Walsh sequences ensures that all interference from other user signals in that cell are canceled out during symbol integration. The non-orthogonal interference from other cells as well as mutlipath limits capacity on the forward link.

In IS-95, all user signals transmitted by a base station are quadrature phase shift key (QPSK) spread using the same in-phase (I) channel PN sequence and quadrature (Q) channel PN sequence. Each base station in a CDMA system transmits in the same frequency band using the same PN sequence, but with a unique offset relative to an unshifted PN sequence aligned to a universal time reference. The PN spreading rate is the same as the Walsh cover rate, 1.2288 MHz, or 64 PN chips per symbol. In the preferred embodiment, each base station transmits a pilot reference. In the description of the present invention different information is transmitted on the I and Q channels which substantially increases the capacity of the system.

The pilot channel is a "beacon" transmitting a constant zero symbol and spread with the same I and Q PN sequences used by the traffic bearing signals. In the preferred embodiment, the pilot channel is covered with the all zero Walsh sequence 0. During initial system acquisition the mobile searches all possible shifts of the PN sequence and once it has found a base station's pilot, it can then synchronize itself to system time. As detailed below, the pilot plays a fundamental role in the mobile demodulator rake receiver architecture well beyond its use in initial synchronization.

FIG. 2 depicts a radio's generic rake receiver demodulator 10 for receiving and demodulating the forward link signal 20 arriving at the antenna 18. The analog transmitter and receiver 16 contain a QPSK downconverter chain that outputs digitized I and Q channel samples 32 at baseband. The sampling clock, CHLPX8 40, used to digitize the receive waveform, is derived from a voltage controlled temperature compensated local oscillator (TCXO).

The demodulator 10 is supervised by a microprocessor 30 through the databus 34. Within the demodulator, the I and Q samples 32 are provided to a plurality of fingers 12a-c and a searcher 14. The searcher 14 searches out windows of offsets likely to contain multipath signal peaks suitable for assignment of fingers 12a-c. For each offset in the search window, the searcher 14 reports the pilot energy it found at that offset to the microprocessor. The fingers 12a-c are then surveyed, and those unassigned or tracking weaker paths are assigned by the microprocessor 30 to offsets containing stronger paths identified by searcher 14. Once a finger 12a-c has locked onto the multipath signal at its assigned offset it then tracks that path on its own until the path fades away or until it is reassigned using its internal time tracking loop. This finger time tracking loop measures energy on either side of the peak at the offset at which the finger is currently demodulating. The difference between these energies forms a metric which is then filtered and integrated.

The output of the integrator controls a decimator that selects one of the input samples over a chip interval to use in demodulation. If a peak moves, the finger adjusts its decimator position to move with it. The decimated sample stream is then despread with the PN sequence consistent with the offset to which the finger is assigned. The despread I and Q samples are summed over a symbol to produce a pilot vector (PT, PQ). These same despread I and Q samples are Walsh uncovered using the Walsh code assignment unique to the mobile user and the uncovered, despread I and Q samples are summed over a symbol to produce a symbol data vector (DT, DQ). The dot product operator is defined as

P(n) • D(n) = Pτ(n)Dτ(n) + P_Q(n)D_Q(n), (2)

where Pι(n) and P_Q(IT) are respectively the I and Q components of the pilot vector P for symbol n and Dι(n) and Dg(n) are respectively the I and Q components of the data vector D for symbol n. Since the pilot signal vector is much stronger than the data signal vector it can be used as an accurate phase reference for coherent demodulation; the dot product computes the magnitude of the data vector component in phase with the pilot vector. As described in co-pending U.S. Patent No. 5,506,865, entitled "PILOT CARRIER DOT PRODUCT CIRCUIT" and assigned to the assignee of the present invention, the dot product weights the finger contributions for efficient combining, in effect scaling each finger symbol output 42a-c by the relative strength of the pilot being received by that finger. Thus the dot product performs the dual role of both phase projection and finger symbol weighting needed in a coherent rake receiver demodulator. Each finger has a lock detector circuit that masks the symbol output to the combiner 42 if its long term average energy does not exceed a minimum threshold. This ensures that only fingers tracking a reliable path will contribute to the combined output, thus enhancing demodulator performance. Due to the relative difference in arrival times of the paths to which each finger 12a-c is assigned, each finger 12a-c has a deskew buffer that aligns the finger symbol streams 42a-c so that the symbol combiner 22 can sum them together to produce a "soft decision" demodulated symbol. This symbol is weighted by the confidence that it correctly identifies the originally transmitted symbol. The symbols are sent to a deinterleaver/decoder circuit 28 that first frame deinterleaves and then forward error correction decodes the symbol stream using the maximum likelihood Viterbi algorithm. The decoded data is then made available to the microprocessor 30 or to other components, such as a speech vocoder, for further processing.

To demodulate correctly, a mechanism is needed to align the local oscillator frequency with the clock used at the cell to modulate the data. Each finger makes an estimate of the frequency error by measuring the rotation rate of the pilot vector in QPSK I, Q space using the cross product vector operator:

P(n) X P(n-l) = P,(n)P_Q(n-l) - P,(n-l)P_Q(n) (3)

The frequency error estimates from each finger 44a-c are combined and integrated in frequency error combiner 26. The integrator output, LO_ADJ 36, is then fed to the voltage control of the TCXO in the analog transmitter and receiver 16 to adjust the clock frequency of the CHIPX8 clock 40, thus providing a closed loop mechanism for compensating for the frequency error of the local oscillator.

If the transmission power of signals transmitted by the base station to a mobile station is too high, it can create problems such as interfering with other mobile stations. Alternatively, if the transmission power of signals transmitted by the base station is too low, then the mobile station can receive multiple erroneous frames. Terrestrial channel fading and other known factors can affect the received power of signals transmitted by the base station. As a result, each base station must rapidly and accurately adjust the transmission power of the signals which it transmits to the mobile stations.

In a useful method for controlling the transmission power of signals transmitted by a base station, the mobile station transmits a signal or message to the base station when the power of a received frame of data deviates from a threshold or is received in error. In response to this message, the base station increases its transmission power of signals transmitted by the base station. A method and apparatus for controlling transmission power is disclosed in U.S. Patent No. 5,056,109, entitled "METHOD AND APPARATUS FOR CONTROLLING TRANSMISSION POWER IN A CDMA CELLULAR TELEPHONE SYSTEM," assigned to the assignee of the present invention and incorporated by reference herein.

Recently there has been a great increase in interest in providing high speed digital data over wireless communication links. In providing for these high speed data links, systems have been developed that use greater bandwidths. One method for providing high speed data involves the use of CDMA techniques based closely on the developments made in the deployment of IS-95 systems. A system proposed by the Telecommunications Industry Association to provide high speed data over wireless communication systems is entitled "cdma2000".

In the transmission of CDMA signals, it is desirable to reduce the peak to average ratio of transmission power through a power amplifier of a wireless communication device. A method for reducing the peak to average ratio is through the use of complex PN spread as described in U.S. Patent Application Serial No. 08/856,428, entitled "REDUCED PEAK TO AVERAGE TRANSMIT POWER HIGH DATA RATE IN A CDMA WIRELESS COMMUNICATION SYSTEM," filed April 9, 1996, assigned to the assignee of the present invention and incorporated by reference herein. In complex PN, the information signals T and Q' are PN spread in accordance with the following equations:

I = I' PN, - Q' PN_Q, (4)

Q = I' PN_Q + Q' PN,,. (5)

where PN_t and PN_Q are distinct PN spreading codes. This complex PN spreading is very useful in cases where the amount of information may be substantially higher on of the information signal channels (T or Q') than the other. The complex PN spreading serves to balance the load, thus reducing the peak to average ratio. The aforementioned U.S. Patent Application Serial No. 08/856,428 also describes a method of despreading the complex PN spread data.

The cdma2000 system employs orthogonal code channels that are subsequently spread by pseudonoise sequences. The system is designed to provide spreading to a predetermined set of chip rates including 1.2288 Mcps, 3.6864 Mcps, 7.3728 Mcps and 11.0592 Mcps. Providing a different set of PN generators for each chip rate requires additional hardware and drives up the cost of equipment. In addition, it is desirable that equipment be able to function at each of the chip rates. Moreover, it is desirable that a method for searching for pilot signals that provides fast and effective acquisition. SUMMARY OF THE INVENTION

The present invention is a novel and improved method for PN spreading a CDMA signal. The present invention teaches of a method for generating a truncated PN sequence. In addition, the present invention teaches of method for generating a second truncated PN sequence by masking the first truncated sequence. A method for generating a phase shifted version of a truncated PN sequence.

BRIEF DESCRIPTION OF THE DRAWINGS

The features, objects, and advantages of the present invention will become more apparent from the detailed description set forth below when taken in conjunction with the drawings in which like reference characters identify correspondingly throughout and wherein:

FIG. 1 is a illustration of the multipath signals in a CDMA environment;

FIG. 2 is an illustration of a RAKE receiver for receiving CDMA signals;

FIG. 3 is a block diagram of the initial processing of a CDMA signal in a third generation CDMA communication system;

FIG. 4 is a block diagram of the final processing of a CDMA signal in a third generation CDMA communication system; FIG. 5 is a block diagram illustrating the generation of the masked PN sequence used in IS-95 systems;

FIG. 6 is an illustration of a generic PN generator;

FIG. 7 is an illustration of the states of a truncated PN sequence and the states of an offset truncated PN sequence; FIG. 8 is block diagram of an exemplary embodiment of a circuit used to generate a truncated PN sequence;

FIG. 9 is an exemplary embodiment of the registers used in the LFSR of FIG. 8;

FIG. 10 is a block diagram of an exemplary embodiment of a circuit used to generate a truncated PN sequence and second truncated PN sequence using a single PN generator; FIGS, lla-llb are illustrations of the problem incurred in attempting to phase shift a truncated PN sequence;

FIG. 12 is a block diagram of circuit for providing phase shifted versions of a truncated PN sequence; and FIG 13 illustrates the preferred embodiment of the present invention whereby the initial phase of the PN sequences is the same regardless of the resultant chip rate.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

Although spectrally efficient multiple access systems such as those specified in IS-95 (referred to as cdmaOne systems) are capable of handling today's capacity demands, manufacturers and operators anticipate increasing demands on their systems as the interest in wireless data communications and increasing popularity of wireless telephony increase. In anticipation of this increasing demand, the International Telecommunication Union (ITU) initiated a program to standardize third generation wireless communication systems. The Telecommunications Industry Association submitted to the ITU a Radio Transmission Technology (RTT) candidate, entitled "The cdma2000 ITU-R Candidate Submission (0.18)" (hereafter the cdma2000 submission). The cdma2000 submission provides for operation at a set of different chip rates that allows the system to grow to meet capacity needs. The chip rate of CDMA systems governs the amount of data that can be transmitted by a system through the relationship with the required spreading gain factor. In particular, the cmda2000 submission provides for operation at 1.2288 Mcps (the chip rate of current cdmaOne systems), 3.6864 Mcps, 7.3728 Mcps, 11.0592 Mcps and 14.7456 Mcps. In cdmaOne systems, as is envisioned in cdma2000 systems, each base station spreads its forward link transmission using a common PN spreading code that is offset by a predetermined amount from other base stations in its vicinity. The amount of offset required between base stations is a function of the maximum propagation path that is anticipated by system designers. As described in the aforementioned U.S. Patent No. 5,764,592, the

CDMA receiver generates a local version of the PN sequence in order to despread the received signal. The PN sequence generated in the receiver will be offset from the PN sequence used to perform the spreading at the base station becatise of the time it took the signal to propagate from the base station to the mobile station receiver. If PN offsets at the base stations in a given area are too close to one another, the base station PN spreading will not appear unique to the mobile and will prevent the mobile station from being able to discriminate between the received signals from the different base stations. Thus, the amount PN offset between base stations is a function of the maximum anticipate propagation time for a signal to reach the mobile station. Because the required PN offset is a function of propagation time, the minimum amount of PN chip offsets between base stations is the product of the maximum propagation time and the PN spreading rate. Moreover, the number of states required to be generated by a PN spreading is equal to the minimum amount of PN chip offset between base stations and the number of base stations that may be communicating with a mobile station or interfering with one another at a mobile station. Thus, a shorter PN code would be required of a system operating at 1.2288 Mcps than would be required of a system operating at 3.6864 Mcps.

FIG. 3 illustrates the downlink (forward link) transmission scheme proposed in the cdma2000 submission. Frames of data are provided to CRC and tail bit generator 102 which generates a set of parity bits for the frame, referred to as cyclic redundancy bits. The method for generating of cyclic redundancy check bits is well known in the art and a method for generating CRC bits is described in detail in U.S. Patent No. 5,504,773, entitled "MATHOD AND APPARATUS FOR THE FORMATTING OF DATA FOR TRANSMISSION", which is assigned to the assignee of the present invention and incorporated by reference herein. CRC and tail bit generator 102 then appends a set of tail bits to the frame which are used to clear the memory of the decoder and the receiver. The packet of data is then provided to encoder 104. Encoder 104 can either be a convolutional encoder or a turbo encoder. Convolutional encoders are well known in the art. Convolutional encoders allow for the use of trellis decoders at the receiver which greatly reducing the amount of energy need to correctly transmit the data to a remote station. Alternatively, encoder 4 can be a turbo encoder the design of which are well known in the art and an example of which is described in U.S. Patent No. 5,466,747, entitled "ERROR CORRECTION CODING METHOD WITH AT LEAST TWO SYSTEMATIC CONVOLUTIONAL CODINGS IN PARALLEL, CORRESPONDING ITERATIVE DECODING METHOD, DECODING MODULE AND DECODER", which is incorporated by reference herein.

The encoded symbols are provided by encoder 104 to interleaver 106. Interleaver 106 reorders the symbols to provide for time diversity which protects against burst type errors that are common in the wireless environment. In the exemplary embodiment, interleaver 106 is a block interleaver in which the data is read into a memory element in rows and read out in columns. The reordered symbols are provided to scrambling element 112.

Scrambling element 112 scrambles the interleaved data in accordance with a decimated long PN sequence. The scrambling is accomplished by performing a modulo-2 addition of the interleaver output symbol with the binary value of the long PN chip. The long code is generated by a linear feedback shift register (LFSR) which is passed through a masking function. The masking function is a function of the identity of the user, typically based on the user's electronic serial number (ESN). The long code generated by the LFSR is masked and provided to bit selector 110 which decimates the sequence to an appropriate rate. The decimated sequence is provided to scrambling element 112 which provides an additional level of call security to the user.

The scrambled symbols are provided to multiplexer and signal point mapping element 114. In the exemplary embodiment, each set of two bits is provided to multiplexer and signal point mapping element 114 which maps the set of binary bits a constellation consisting of the points (1,1), (1,-1), (-1,1) and (-1,-1). One of the points in this constellation mapping is provided on a first output to data channel gain element 116 and the second point in this constellation is provided on a second output to data channel gain element 118. The gain adjusted packets are provided to puncturing elements 122 and 28 which adjust the transmission gain of the packets. The gain adjusted packets are then provided to puncturing elements 122 and 124. Power control bits to control the transmission power of the remote station (not shown) are provided to power control channel gain element 120. Power control gain element 120 adjusts the gain of the power control bits and provides the gain adjusted power control bits to puncturing elements 122 and 124. Puncturing elements 122 and 124 puncture the power control bits into the packet in predetermined positions. The packets are then provided to covering elements 128 and 130.

The packets consist symbol of ±1 value. The symbols are provided to covering elements 128 and 130 which multiply the symbols by an orthogonal sequence consisting of ±1 values. This orthogonal sequence is dedicated for traffic transmission to the particular remote station user. The orthogonal sequence in the exemplary embodiment is a Walsh sequence the generation of which is well known in the art and is described in detail in U.S. Patent No. 5,103,459, entitled "SYSTEM AND METHOD FOR GENERATING SIGNAL WAVEFORMS IN A CDMA CELLULAR TELEPHONE SYSTEM", assigned to the assignee of the present invention and incorporated by reference herein.

The Walsh covered sequence is then provided to forward link channel summer which sums the data from covering elements 128 and 130 with other similarly modulated data packets for transmission to other users as well as common channel data.

Turning to FIG. 2, the resultant summed data is then provided to PN spreading elements 134, 136, 138 and 140. Each base station is identified by a PN offset that is unique to it in the set of base stations located in its vicinity. In the present invention, two PN sequences are used to spread the data (PN_j and PN_Q). Each PN sequence consists of a sequence of ±1 values. The PN spreading operation illustrated in FIG. 2 is performed to provided the result:

I = I' PN, - Q' PN_Q, (6) Q = F PN_Q + Q' PN,. (7)

The data sequence I' is provided to multipliers 134 and 138. Multiplier 134 multiplies the data (F) by the pseudonoise sequence PN, and provides the result to the summing input of subtractor 142. Multiplier 138 multiplies the data (T) by the pseudonoise sequence PN_Q and provides the result to a first summing input of summer 144.

The data sequence Q' is provided to a first input of multipliers 136 and 140. Multiplier 140 multiplies the data (Q') by the pseudonoise sequence PN, and provides the result to the subtracting input of subtractor 142. Multiplier 136 multiplies the data (Q') by the pseudonoise sequence PN_Q and provides the result to a second summing input of summer 144. Subtractor 142 subtracts the output of multiplier 140 from the output of multiplier 134 and provides the result to baseband filter (BBF) 146. Summer 144 adds the output of multiplier 138 to the output of multiplier 136 and provides the result to baseband filter (BBF) 148. Baseband filters 146 and 148 filter the PN spread sequences and provide the filtered sequences to upconverters 150 and 152 respectively.

Upconverters 150 and 152 upconvert the input data in accordance with quadrature phase shift keyed (QPSK) modulation as is well known in the art. Upconverter 150 upconverts the signal (I) for transmission in accordance with the carrier modulation cos(2_f_c). Upconverter 152 upconverts the signal (Q) in accordance with the carrier modulation sin(2_f_c). The two orthogonal signals are then summed in summing element 154 amplified and transmitted.

In the present invention, a method of spreading data with unique PN sequence spreading for both I and Q channels using a single PN generator is proposed. In order to provide for a unique spreading of the I and Q channels using a single PN generator, the PN sequence used to spread the Q channel data is offset from the PN sequence used to spread the I channel data. A linear feedback shift register (LFSR) is typically used to generate PN sequences for CDMA spreading. The output of the LFSR cycles through a predetermined number of unique state prior to returning to its initial state and then the cycle of outputs begins again. In a maximal length LFSR sequence, the shift register pass through all possible states of its memory prior to returning to the initial state. This is true when the polynomial is selected such that the feedback is maximal length. The output of the LFSR is a function of the state of the LFSR.

The PN sequences (PN, and PN_Q) can either be generated by different polynomials (i.e. PN generators) or by different phases of the same PN generator. In the first exemplary embodiment of the PN generation of the I and Q channels, the I and Q channels are generated using the same PN generator with the sequences offset from one another.

As is known in the art, the phase of the PN generator output can be altered using a masking operation of the performed on the output states of the LFSR. FIG. 5 illustrates the Long code PN generator used in IS-95 systems. LFSR 175 provides the maximal length sequence to masking element 177. Masking element 177 is a series of AND gates that alter the phase of the output. The mask in this case is a 42 bit input sequence that turns on and off taps outputs of the LFSR. The masking operation changes the phase of the output sequence. The masked sequence is then provided to modulo-2 summer 179 which outputs the long code sequence. FIG. 6 illustrates a generalized version of an LFSR with the feed back path determined in accordance with the polynomial:

G=g₀ + &x + g₂ x²+... + g_rx ,. (8)

where g₀ is the rightmost feedback and is always equal to "1", g_n is the a binary value that either enables or disables a corresponding feedback path in the LFSR and x denotes the position of the feedback. It will be understood by one skilled in the art that the feedback polynomial (G) can be harwired where no AND gates would be required. LFSR 206 consists of a set of summers 202, registers 200 and AND gates 204. The state of the LFSR is determined by the values into summers 202. The LFSR starts at an initial state to which it is set by loading the registers 200.

As the clock is administered to the circuit, the value stored in the registers 200 is output to a first input of summers 202. The value output by register 200a is output of LFSR 206 and is also fedback on feedback line 208 as one input to AND gates 204. The other inputs to AND gates 204 are the binary values (g₀, g ..g_r).

In the first embodiment, the sequences generated for spreading of the I and Q channels are phase shifted versions of a larger PN sequence. In order to permit a single PN generator to generate non overlapping PN sequences the PN generator must generate a sequence that is at least twice as long as the sequence used to spread the I and Q channels and the sequence must be truncated.

FIG. 7 illustrates the states or phases of the LFSR on the circumference of a circle. A circle makes an effective visualization of the output of an LFSR because like the circle the state of the LFSR starts at an initial state cycle through all output states, return to the initial state and repeats.

In the first embodiment of the present invention, the sequence used to spread the I and Q channels are generated using a truncated PN sequence. In FIG. 7 the LFSR is illustrated as having N unique states. The sequence used to spread the I channel (PN,) is truncated to k states (k<N/2). Thus the PN generator that is designed to cycle through all N possible states is truncated once it reaches the k+1-state after which time it is forced to the initial state.

In the preferred version of the first embodiment, the sequence used to spread the Q channel is offset from the sequence used to spread the I channel by an amount such that the two sequences do not overlap. As shown in FIG. 7, the sequence used to spread the Q channel (PN_Q) consists of the states between m and m+k (where m>k and m+k<N). In the exemplary embodiment, PN_Q is generated by masking the PN, sequence.

In an alternative embodiment, PN, and PN_Q can be generated as truncated PN sequences by different polynomials. The description of the use of a single polynomial fully enables one to realize an embodiment where two distinct polynomials are used. The gernation of distinct PN sequences using multiple polynomials is well known in the art.

FIG. 8 illustrates the PN generator 306 of the present invention for generating a truncated PN sequence. It will be understood by one skilled in the art that the feedback polynomial (G) can be harwired where no AND gates would be required. The initial state (S,) of PN generator 306 is provided on bus line 312 to registers 300. In the exemplary embodiment, registers 300 are disclosed as D-flip flops although other memory structure are equally applicable and are within the scope of the present invention. The output of PN generator 308 is output from register 300a. In addition, the output from register 300a is fedback on line 308 to a first input of AND gates 304. The polynomial G is provided as the second input of AND gates 304 on input line 314. Polynomial G enables or disables the feedback provided on line 308. If the input to an AND gate 304 from input line 314 is high, then the feedback to the corresponding modulo-2 summer 302 is enabled and the output bit is fedback to that summer. If the input to an AND gate 304 from input line 314 is low, then the feedback to the corresponding modulo-2 summer 302 is disabled and the output bit is blocked from reaching that summer. The output of each of registers 300b to 300r is provided to the first input of a corresponding modulo-2 summer 302a to 302(r-l). The second input to modulo-2 summers 302a to 302(r-l) is the fedback output bit from register 300a at the location where the feedback path is enable by the polynomial G. In order to truncate the PN sequence, state identification element 310 is employed to compare the state of the LFSR to the final state of the truncated PN sequence. A variety of methods can be used to implement state detection element 310, the easiest to envision would consist of a plurality or XOR gates each of have one input receiving a bit of the current state and a second input receiving a bit from the final state. The outputs of each of the XOR gates would be provided to an r-input AND gate and the result of this operation would be output on line 318.

Returning briefly the generation of the I-sequence in FIG. 7, PN generator 306 is initially set to the initial state (denoted with a 0). Upon reaching the final state of the truncated sequence (state k+1), state identification element 310 sends a signal to registers 300 which forces the registers to load the initial state (S,). Upon detection by state detection element 310 of the final state S_f, state detection element 310 outputs a binary signal indicative of identification of the final state. The signal on line 318 forces each of registers 300 to switch from output the resultant sum from the preceding summing element to outputting the initial state S,. For example, in most cases register 300c will take the output from summer 302c and provide that output to modulo-2 summer 302b. However, when the signal on line 318 goes high, indicative of the final state having been detected, then register 300c outputs Si₂ to modulo-2 summer 302b. At the next clock interval, state detection element 310 will detect a difference between the state of PN generator and the final state, and set the signal on line 318 to low, which will cause the operation to resume wherein register 300c will provide the output of summer 302c to a first input of summer 302b.

FIG. 9 illustrates an exemplary implementation of registers 300 using a loadable D-type flip flop 320. The A input receives the output of a corresponding summer. The B input receives the bit of the initial state S,. The value on line 318 is provided to LD , which determines whether the output (Q) provides the input A or the input B. The clock runs at the PN chip rate. FIG. 10 illustrates an apparatus for generating two pseudonoise sequences PN, and PN_Q which are phase offset from one another. The pseudonoise sequence PN, is generated by PN generator 426 in the same fashion as was described with respect to PN generator 306 of FIG. 9. It will be understood by one skilled in the art that the mask (M) and feedback (G) polynomials can be harwired where no AND gates would be required. The states of PN generator 426,are provided to masking element 428 which phase shifts the output PNQ with respect to the sequence PN,. PN generator 426 of the present invention generates a truncated PN sequence. The initial state (S,) of PN generator 426 is provided on bus line 412 to registers 400. In the exemplary embodiment, registers 400 are disclosed as D-flip flops although other memory structure are equally applicable and are within the scope of the present invention. The output of PN generator 426 is output from register 400a. In addition, the output from register 400a is fedback on line 408 to a first input of AND gates 404. The polynomial G is provided as the second input of AND gates 404 on input line 414. Polynomial G enables or disables the feedback provided on line 408. If the input to an AND gate 404 from bus line 414 is high, then the feedback to the corresponding summer 402 is enabled and the output bit is fedback to that summer. If the input to an AND gate 404 from input line 414 is low, then the feedback to the corresponding summer 402 is disabled and the output bit is blocked from reaching that summer. The output of each of registers 400b to 400r is provided to the first input of a corresponding summer 402a to 402(r-l) (not shown). The second input to summers 402a to 402(r-l) (not shown) is the fedback output bit from register 400a at the location where the feedback path is enable by the polynomial G. In order to truncate the PN sequence, state identification element 410 is employed to compare the state of the LFSR to the final state of the truncated PN sequence. A variety of methods can be used to implement state detection element 410, the easiest to envision would consist of a plurality or XOR gates each of have one input receiving a bit of the current state and a second input receiving a bit from the final state. The outputs of each of the XOR gates would be provided to an r-input AND gate and the result of this operation would be output on line 418.

Returning briefly the generation of the I-sequence in FIG. 7, PN generator 406 is initially set to the zero state. Upon reaching the final state of the truncated sequence (state k), state identification element 410 sends a signal to registers 400 which forces the registers to load the initial state (S,). Upon detection by state detection element 410 of the final state S_f, state detection element 410 outputs a binary signal indicative of identification of the final state. The signal on line 418 forces each of registers 400 to switch from output the resultant sum from the preceding summing element to outputting the initial state S,. For example, in most cases register 400c will take the output from summer 402c and provide that output to summer 402b. However, when the signal on line 418 goes high, indicative of the final state having been detected, then register 400c outputs Si₂ to summer 402b. At the next clock interval, state detection element 310 will detect a difference between the state of PN generator and the final state. And set the signal on line 418 to low, which will cause the operation to resume wherein register 400c will provide the output of summer 402c to a first input of summer 402b.

The states of PN generator 426 (S=[S₀...S_r_ ) are provided on lines 430 to masking element 428. Masking element 428 consists of a set of AND gates 422 and a modulo-2 adder 424. Each of the bits from registers 400 are provided to the first input of a corresponding AND gate 422. The second input of the AND gate is provided by the masking polynomial M which is provided on bus line 420. The results of the AND operation between masking polynomial M and the states S, is provided to modulo-2 summer 424. Modulo-2 summer 424 performs a modulo-2 addition on the inputs from AND gates 422 and outputs the resulting sequence for modulating the Q channel (PN_Q).

Using the present invention, a single generator polynomial can be used for generating the PN spreading sequences regardless of the desired resultant chip rate so long as the polynomial is capable of generating a sufficient number of states for the highest chip rate. This is realized simply by adjusting the truncation of the sequences in accordance with the desired number of states for a given chip rate. For example, when using a 3.6864 Mcps spreading rate, the sequence would be truncated after a predetermined number of states.

In the preferred embodiment of the present invention, The same initial phase offset is used for all chip rates. As shown in FIG. 13 the initial phase of the PN sequences is the same regardless of the resultant chip rate. As illustrated in FIG. 13 a PN sequence that is large enough for any of the predetermined chip rates is selected. In this case a maximal length sequence of 2²⁰-l is used. In the example, the states of the PN sequence used to spread the I channel do not overlap with the states of the PN sequence used to spread the Q channel. However, all sequences contain a common initial state. By having common initial phases hardware savings can be achieved by aligning the initial (or final phase) which automatically results in fixing the offset between the I and Q channels which allows for the use of a common mask for all chip rates to distinguish between the I and Q spreading sequences.

Each base station would spread its forward link signal in accordance with an offset version of this same truncated sequence. Then, in order to realize a chip rate of 3.6864 Mcps, the sequence would not be truncated until approximately 3 times as many states later using the same hardware and same generator polynomial.

As described previously, in cdmaOne and proposed cdma2000 systems, each base station transmits a phase offset versions of a common PN sequence. In the case of a truncated PN sequence, this leads to a unique problem with respect to the masking operation used to generate the offset truncated sequences. FIGS. 11a and lib illustrates the problem with generating an offset truncated PN sequence.

FIG. Ha illustrates the generation of an offset PN sequence that is not truncated. The sequence illustrated with circle 500 begins at state 501, returns to state 501 and repeats. The offset sequence illustrated with circle 502 begins at PN offset position 503, returns to initial state 503 and repeats. When the sequence is not truncated this can be accomplished using a simple masking operation as described earlier, which shifts the sequence to a state earlier or later in its generation.

However, turning now to FIG. lib, when the sequence is truncated traditional masking function is inadequate. Arc 504 illustrates the unshifted PN sequence. The generator starts at initial state 505 and moves to final state 507, after which the generator is returned to state 505 as described previously. Arc 506 represents the truncated PN sequence 504 shifted using traditional masking techniques. Using traditional masking methods, the shifted sequence would begin at state 508 and continue to 509 before being returned to state 508. This is not a correctly phase shifted version of the first PN sequence. A correctly phase shifted version the truncated PN sequence would start at initial state 508 proceed to the final state 507, return to initial state 505 and progress through state 508 to final state 507 and repeat.

FIG. 12 illustrates a method for providing a correctly phase shifted version of an offset PN sequence. Truncated sequence generator 600 generates the truncated PN sequence as described with respect to PN generator 306. The truncated PN sequence is provided to mask operator 606 on line 608. Mask operator 608 performs a masking operation in accordance with a mask provided from multiplexer 604 as described with respect to masking element 428. Mask operator 608 shifts the truncated PN sequence as described previously.

Initially mask operator 606 use a first mask denoted Ml to provide a shift to the sequence corresponding to the shift denoted PN offset in FIG. lib. Upon reaching the final state of the truncated sequence (505 in FIG. lib), a second mask is used by mask operator 606 to phase shift the truncated sequence S to provide the initial portion of the truncated sequence.

Phase detector 602 is used to detect the time to switch between masks. Referring back to FIG. lib, phase detector 602 detects the points of mask transition 507 and 508. Mask 1 is used to generate the advanced sequence from state 508 to state 507. Upon detection of state 507, Mask 2 is employed to generate the retarded sequence between states 505 and 508. Upon detection of state 508, Mask 1 is again used.

In the exemplary embodiment, the states of truncated sequence generator 600 is provided on bus line 609 phase detector 602. Phase detector 602 compares the state with the appropriate state at which to make the next mask change as described above. In the exemplary embodiment, the state at which a mask change will be made is stored in memory element 608 and provided by microprocessor 610 on a bus line to multiplexer 612 which provides the phase information to phase detector 602.

Upon detecting the need for a mask change, phase detector 602 sends a signal to multiplexer 604. In response to the signal from phase detector 602, multiplexer 604 switches the mask that it provides to mask operator 606. In the exemplary embodiment, the masking values are stored in memory 608 and provided through microprocessor 610 on a common microprocessor bus line to registers 612. Registers 612 provides mask 1 to a first input of multiplexer 604 and provide mask 2 to a second input of multiplexer 604.

Although described in the context of LFSR PN sequences, the present invention is extendable to use with other classes of PN sequences such as Gold codes and other configurations such as Fibonacci.

The previous description of the preferred embodiments is provided to enable any person skilled in the art to make or use the present invention. The various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without the use of the inventive faculty. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

WE CLAIM:

Claims

1. An apparatus for providing a truncated PN sequence comprising:

PN sequence generator for generating a PN sequence containing a number of states greater than said truncated PN sequence; final state detector mean for detecting a final state in said states of said PN sequence generator and for providing a signal indicative of the detection of said final state; and initial state loading means for loading an initial state into said PN sequence generator in response to said signal indicative of the detection of said final state.