Classifications

• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L25/00—Baseband systems
  • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
  • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
  • H04L25/03006—Arrangements for removing intersymbol interference
  • H04L25/03178—Arrangements involving sequence estimation techniques
  • H04L25/03184—Details concerning the metric
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L25/00—Baseband systems
  • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
  • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
  • H04L25/03006—Arrangements for removing intersymbol interference
  • H04L2025/0335—Arrangements for removing intersymbol interference characterised by the type of transmission
  • H04L2025/03375—Passband transmission
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L25/00—Baseband systems
  • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
  • H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
  • H04L25/03006—Arrangements for removing intersymbol interference
  • H04L25/03178—Arrangements involving sequence estimation techniques
  • H04L25/03312—Arrangements specific to the provision of output signals
  • H04L25/03318—Provision of soft decisions

Definitions

• the present invention relates generally to radiotelephones and more particularly to the coherent demodulation of differentially encoded quadrature phase shift keying (DQPSK) signals.
• DQPSK differentially encoded quadrature phase shift keying
• FEC forward error correction
• IS-13S digital cellular system
• ⁇ /4 shifted-DQPSK ⁇ /4 shifted-DQPSK
• Differential encoding of the transmitted signals allows both coherent and differential demodulation of the received signal.
• differential demodulators may not be complex to implement, it is widely accepted that the performance of differential detectors degrades rapidly in the presence of Inter-Symbol Interference (ISI) which may be caused by multi -path propagation. Therefore, coherent demodulators with equalizers are commonly used in many receivers .
• Such a receiver is described in U.S. Patent No. 5,285,480 to Chennakeshu et al . , entitled ADAPTIVE MSLE-VA RECEIVER FOR DIGITAL CELLULAR RADIO.
• Figure 1 depicts a block diagram of a ⁇ /4 shifted-DQPSK system with a conventional differential demodulator receiver.
• the transmitter 105 includes encoder 101 and differential modulator 102.
• Information bits are encoded in encoder 101 to produce coded bits.
• the coded bits are differentially modulated in differential modulator 102 to produce a differentially modulated signal, which is provided to antenna 104 for transmission.
• the transmitted signal reaches the radio receiver after passing through a propagation medium (e.g., a mobile radio channel) .
• the transmitted signal plus any noise are received at the receiver antenna 106 and the received signal provided to receiver 114.
• the received signal is processed by the radio processor 108 which amplifies, mixes, filters, samples and quantizes the received signal to produce a baseband signal .
• a differential demodulator 110 demodulates the received signal and provides symbol or bit values to the decoder 112 which decodes the encoded bits and which may detect and correct possible errors in the received signal.
• the output of the demodulator 110 is preferably soft values which may provide higher performance in decoding.
• FIG. 2 shows a block diagram of a known apparatus for differential demodulation of the DQPSK modulated signals .
• the differential detector uses received samples to get hard and/or soft decision values.
• the present received sample is coupled to the multiplier 203.
• the present received sample is also fed into a delay 201.
• the delay 201 is coupled to a conjugate operator 202, and the output of the conjugate operator 202 is coupled to the multiplier 203.
• the present received sample and the delayed and conjugated received sample are multiplied to undo the effect of the differential encoder at the transmitter.
• the real 204 and imaginary 205 part of the output of the multiplier provide the soft bit values corresponding to the two bits sent in one, di-bit symbol.
• the hard values can be obtained by taking the sign 206 and 207 of the soft values as desired.
• FIG. 3 shows a block diagram of a known apparatus for coherent demodulation of the DQPSK modulated signals.
• the coherent receiver utilizes channel estimation unit 302 which estimates the amplitude and phase of the mobile radio channel. These channel estimates are passed to the coherent QPSK demodulator 301 where the estimates of the QPSK symbols are generated.
• the channel parameters can be estimated using the known data sequences which are periodically inserted into the transmitted information sequences. In systems where the channel parameters change over the transmission of two consecutive known data sequences, like the U.S. digital cellular system (IS-136), it is desirable to adapt the channel parameters during the transmission of unknown data sequences.
• Such an adaptive coherent receiver is described in U.S. Patent No.5, 285, 480 to Chennakeshu et al . entitled ADAPTIVE MLSE-VA RECEIVER FOR DIGITAL CELLULAR RADIO.
• the output values of the coherent QPSK demodulator 301 which are the hard coherent symbols, are passed through a differential detector 303 to undo the effect of the differential encoder in the transmitter.
• the outputs of the differential detector are the hard decision values corresponding to the transmitted information bits.
• DIGITAL COMMUNICATION AND METHOD AND APPARATUS in which only the phase and frequency offset are estimated using the received signal. After compensating the phase and frequency offset, the likelihood of each possible QPSK symbol value for each sample is calculated. Therefore, a likelihood vector for each sample is obtained and this likelihood vector is provided to the decoder. The decoder uses the likelihood vectors to estimate the transmitted symbol values.
• Another conventional method for the soft decoding of differentially encoded QPSK signal is described in U.S. Patent No. 5,754,600 to Rahnema entitled METHOD AND APPARATUS FOR OPTIMUM SOFT-DECISION VITERBI DECODING OF CONVOLUTIONAL DIFFERENTIAL ENCODED QPSK DATA IN COHERENT DETECTION.
• the differential and Viterbi decoders are integrated, i.e., differential decoding is part of the convolutional decoding process.
• Soft information for maximum likelihood sequence estimation (MLSE) for frequency selective fading channels has been extensively studied, for example as described in J. Hagenauer and P. Hoeher, "A Viterbi algorithm with soft-decision outputs and its applications", Proceeding of IEEE Globecom Conference , pp.47.1.1-47.1.7, Dallas, TX, USA, November 1989. These techniques have been extended to ⁇ /4 shifted-DQPSK systems for example as described in Jong Park, Stephan B. Wicker and Henry L. Owen, “Soft Output Equalization Techniques for ⁇ /4 DQPSK Mobile Radio", IEEE Interna tional Conference on Communications, pp.1503- 1507, Dallas, TX, USA, 1997.
• a further object of the present invention is to provide soft values in a manner which does not require complex implementation.
• Still another object of the present invention is to provide soft values from coherent detection of differentially encoded bits or symbols.
• determining a soft value is done by summing the metrics for each possible current and previous coherent symbol value that corresponds to a first differential bit value to provide a first value metric sum and summing the metrics for each possible current and previous symbol value corresponding to a second differential bit value to provide a second value metric sum.
• the first value metric sum may be divided by the second value metric sum to provide a ratio of probabilities of the differential bit value being the first value and differential bit value being the second value.
• the metrics may be exponentiated and the exponentiated metrics summed for each possible symbol pair corresponding to a first bit value to provide a first value metric sum.
• the exponentiated metrics are also summed for each possible symbol pair corresponding to a second bit value to provide a second value metric sum.
• the logarithm of the ratio of probabilities may be taken to provide the soft value for the bit .
• metrics may be determined for each possible coherent symbol value which indicate the probability of a possible symbol value being encoded in the signal using the estimated channel coefficients and the signal for each antenna.
• each metric is a sum of metrics for each antenna.
• an impairment correlation may be determined for the plurality of antennas. Metrics may then be determined for each possible symbol value which indicate the probability of a possible coherent symbol value being encoded in the signal using the estimated channel coefficients, the impairment correlation and the signal for each antenna.
• a noise power associated with the received signal may be determined and the metrics determined based in part on the noise power.
• a detected value is determined from the signal samples and the channel coefficients.
• the detected values may be bit values for symbols encoded in the differentially encoded signal.
• the determined soft values may be associated the detected bit values .
• the signal samples are generated by a radiotelephone receiving and processing a differentially encoded signal .
• soft values are generated from signal samples by estimating channel coefficients associated with the received signal and determining metrics for each possible coherent symbol value which indicate the probability of a possible symbol value being encoded in the signal samples of the signal using the estimated channel coefficients and the signal samples of the received signal .
• the maximum of the combined metrics for each possible current and previous symbol value corresponding to first bit value is determined to provide a first maximum metric value.
• the maximum metric of the combined metrics for each possible current and previous symbol value corresponding to a second bit value is also determined to provide a second minimum metric value.
• the second maximum metric value is then subtracted from the first maximum metric value to provide a difference of log likelihoods of the bit value being the first bit value and the bit value being the second bit value.
• soft values are generated from signal samples by estimating channel coefficients associated with the signal and determining metrics for detected coherent symbol values which indicate the probability of the possible symbol values being encoded in the signal using the estimated channel coefficients and the signal samples of the signal.
• a detected term is determined from the metrics associated with the detected symbols which provides a first value metric corresponding to the detected differential bit.
• a second value metric which corresponds to the non- detected bit, is obtained by obtaining the two different metric values by either flipping the current detected coherent symbol value or the previous detected coherent symbol value and keeping the other coherent symbol value as the detected value and selecting the maximum metric value among these two values.
• This alternative embodiment may further reduce the complexity of the soft value determination by basing the soft value determination on a detected symbol value .
• Figure 2 is a block diagram of a conventional system for demodulation of DQPSK signals
• Figure 3 is a block diagram of a conventional system for coherent demodulation of DQPSK signals
• Figure 4 is a block diagram of a radiotelephone incorporating the present invention
• Figure 5 is a block diagram of one embodiment of the present invention
• FIG. 6 is a block diagram of an alternative embodiment of the present invention.
• Figure 7 is a block diagram of a second alternative embodiment of the present invention
• Figure 8 is a flow chart illustrating operations of one embodiment of the present invention
• Figure 9 is a flow chart illustrating operations of an alternative embodiment of the present invention.
• Figure 10 is a flowchart illustrating operation of the second alternative embodiment of the present invention.
• radiotelephone 10 which includes a soft value generator 15 according to the present invention is depicted in the block diagram of Figure 4.
• radiotelephone 10 typically includes a transmitter 12, a receiver 14, a user interface 16 and an antenna system 18.
• the antenna system 18 may include an antenna feed structure 22 and an antenna 20.
• transmitter 12 converts the information which is to be transmitted by radiotelephone 10 into an electromagnetic signal suitable for radio communications.
• Receiver 14 demodulates electromagnetic signals which are received by radiotelephone 10 so as to provide the information contained in the signals to user interface 16 in a format which is understandable to the user.
• transmitters 12, receivers 14, user interfaces 16 e.g., microphones, keypads, rotary dials
• radiotelephone 10 other than the software generator 15 is well known to those of skill in the art and will not be further described herein.
• the present invention may be utilized in a base station and such an embodiment of the present invention need not be further described herein.
• Figure 5 depicts a coherent demodulator for
• the baseband received signal is coupled to the likelihood information generator 404 and the channel estimator 402.
• the output of the channel estimator 402 is coupled to the input of the likelihood information generator 404.
• the output of the likelihood information generator 404 is coupled to the input of the likelihood information processor 406.
• the channel estimator 402 estimates the complex channel coefficient corresponding to the mobile radio channel.
• channel estimation refers to estimating a single channel coefficient associated with a received value as shown in the equation (a) above.
• the baseband received signal and the channel estimate are used in likelihood information generator 404 to calculate the likelihood functions corresponding to different QPSK symbol hypotheses.
• the likelihood information generator 404 may provide likelihood or log-likelihood functions.
• the output of the likelihood information generator 404 is passed to the likelihood information processor 406 to calculate the soft information values corresponding to each bit.
• the partition of the functions in blocks such as the likelihood information generator 404 and likelihood information processor 406 can differ from implementation to implementation and, therefore, the present invention should not be limited to the configuration illustrated in Figure 5. Also, blocks 404 and 406 can be combined to obtain a single function which generates soft information using the channel parameters and the received baseband signal. In the present description, only one partition is described, however, those of skill in the art will recognize other suitable partitions.
• Real (in-phase) and imaginary (quadrature-phase) bits of the differential symbols are represented as b R ik) and b ⁇ ⁇ k) respectively.
• Table 1 summarizes the relation between the phase of bik) , [aik) , aik-1) ⁇ , and the sign of b R ik) and b x (k) ( ⁇ 1 in value) for DQPSK.
• b(k) a(k) [a(k-l] * and Table 1 below would be modified accordingly.
• Soft information can be obtained by taking the log-likelihood ratio of the probability of the detecting +1 (0) and the probability of the detecting -1 (1), given the received signal.
• the log- likelihood ratio can be written as:
• Equation 1 Equation 1 becomes:
• the indicator function is on half the time, so that of the 16 possible combinations, there are 8 that correspond to the numerator and 8 that correspond to the denominator.
• the terms that correspond to the numerator and the denominator can be obtained using Table 1.
• Table 1 Relation between coherent symbols, differential symbols and differential bits
• Equation 2 Using the Bayes rule, assuming equi-likely bit values, Equation 2 can be re-written as:
• Equation 3 which uses coherent QPSK symbol hypotheses, gives the optimal soft information.
• the probabilities in Equation 3 for multiple receive antennas can be written as :
• Equation 3 Equation 3 can be re-written as seen in Equation 5.
• Equation 6 can be further simplified as
• Equation 7 the following term appears throughout :
• This term is referred to as a coherent (QPSK) symbol metric corresponding to the log likelihood of a potential or hypothetical coherent symbol.
• the term metric generally refers to a value which indicates the probability of a possible symbol value being encoded in the signal samples of the received signal .
• the transmitted signal plus the impairment (thermal noise + interference) are received at the receiver antennas 501 and 502.
• the received signals are processed by the radio processors 503 and 504, each of which amplify, mix, filter, sample and quantize to produce a baseband signal for a different antenna.
• the channel coefficients are estimated in channel estimator 505 for each signal on different antennas using the baseband signals at the output of the radio processors 503 and 504. Such channel estimation is also well known to those of skill in the art and will not be described further herein.
• Impairment correlation among the antennas is estimated in impairment correlation estimator 507 which can operate as described in U.S. Patent No. 5,680,419 to Bottomley entitled METHOD AND APPARATUS FOR INTERFERENCE REJECTION COMBINING IN MULTI-ANTENNA DIGITAL CELLULAR COMMUNICATION SYSTEMS the disclosure of which is incorporated herein by reference as if set forth fully.
• the likelihood information is processed in likelihood information processor 508 to produce soft and/or hard differential bit or differential symbol values (such as Equation 5) .
• FIG. 7 An alternative embodiment of the present invention for uncorrelated noise on each antenna is shown in Figure 7.
• the present value of the baseband signal is multiplied 601 with the conjugate 602 of the channel estimate and the result is divided 604 by the value of the noise power estimate 603.
• the noise power estimate is obtained by using the baseband signal and the channel estimates as is well known to those of skill in the art. Relative noise powers may be used, and noise power estimation may be omitted, if the relative noise powers are known. If the relative noise powers are known to be equal, division by the noise power may be omitted.
• the output of the divider is coupled with the metric computer 606 and a delay block 605.
• the present value and the previous (delayed) values plus the hypothesis of the present and previous coherent symbols are used to compute the metrics as described in Equation 8.
• the metric values from different antennas are combined as shown in Equation 7 in adder 608.
• the blocks of Figure 7 except the adder 608 may be repeated for each antenna. If there is only one antenna, adder 608 may be eliminated.
• These combined metric values are used to obtain soft bit values for the differential bits in likelihood information processor 406 or 508.
• the combination of the present values and previous values results in 16 different metric combinations. Out of these 16 combinations only half results in detection of a +1 value for the differential bit, and the remaining half results in -1.
• likelihood information processor 406 or 508 the 8 combinations which provide a +1 bit value are used to obtain the first log term as shown in Equation 7 by first taking the exponential of these terms, then summing them all, and finally taking the logarithm of the sum. In the same way, the second log term is obtained by considering the other 8 terms. Finally, the soft differential bit value is obtained by taking the difference of the two log terms.
• these soft values can also be used to obtain the hard values by taking the sign of the soft values. Also, for the generation of the hard values, the logarithm operation can be avoided by computing the numerator and the denominator and then finding the larger of these two values, which provides the hard decision.
• the determination of the soft values begins by determining symbol log likelihoods (block 700) which corresponds to calculating the metrics of Equations 5b or 8. These metrics are then combined (block 702) and exponentiated (block 704) . The process is repeated for all symbol pairs (blocks 706 and 708) . Once a combined metric has been determined for each hypothetical symbol pair, then the combined metrics corresponding to a potential +1 are summed and the combined metrics corresponding to a potential -1 are summed (block 710) . The +1 sum is then divided by the -1 sum and the logarithm of that value taken to provide the soft values (block 712) .
• the first alternative approach requires the computation of sixteen combined metrics, C ik, x, y) , at each symbol index k .
• sixteen combined metrics are obtained by using the coherent metrics at time k and k- 1 using Equation 10. Note that, even if a hard decision is determined, the most likely values for the detected coherent symbols are those which maximizes the coherent metrics at each time index.
• the 8 combination terms are not required to find the maximum of them.
• the two QPSK symbols may be detected and the metrics associated with the detected values utilized.
• the 8 combination terms and their maximum is required in the first alternative approach.
• One realization of the first alternative approach can be obtained using the same block diagram as shown in Figure 5.
• the likelihood information generator provides the symbol metrics corresponding to the present values and the symbol metrics corresponding to the previous values.
• the maximum of the 8 terms which is the combination of the metrics corresponding to the present values and previous values which provide a +1 or -1 for the differential bit value
• the soft information is obtained by taking the difference of these two maximum values. Therefore, the alternative approach does not require exponentiation and logarithm operations. Moreover, the knowledge of the noise variance is not needed for the calculation of soft information as long as it is constant. However, it can be used if desired, should the noise be dominated by time-varying interference .
• the determination of the soft values begins by determining the log likelihood of a possible symbol pair value for the current and previous symbols (block 800) which corresponds to calculating the metrics of Equations 5b or 8. These metrics are then combined (block 802) . If there are more possible symbol pairs, then the process is repeated for the next symbol pair (blocks 806 and 808) . Otherwise, when a combined metric has been determined for each hypothetical symbol pair, then the maximum combined metric corresponding to a potential +1 is determined and the maximum combined metric corresponding to a potential -1 is determined (block 810) . The -1 maximum combined metric is then subtracted from the +1 maximum combined metric to provide the soft value (block 812) .
• the detected differential bit values are decided by finding the most likely coherent QPSK symbol values (by finding the coherent symbol which minimizes the symbol metric) for the present and previous values .
• the sum of the minimum symbol metric values at the present and previous time provides the numerator or the denominator (depending on the detected bit value) terms.
• the term corresponding to the non-detected bit value (opposite of the detected bit value) can be obtained by finding the maximum of the 8 terms corresponding to the non-detected bit values.
• the approach can be generalized by taking the best K x terms for the detected values and K 2 terms for the non-detected values.
• the second alternative embodiment is obtained by further simplification of the previous alternative approach, in which it is assumed that the dominant terms include at least one detected symbol value. Therefore, in the second alternative embodiment the minimum of the 8 terms for the non-detected case is not found. Instead, it is assumed that the maximum of the 8 terms includes one of the detected QPSK symbols with high probability.
• the non-detected term by assuming that the dominating terms will include one of the detected symbols, one of the symbols is fixed as the detected symbol and the other term toggled with the closest border symbol and then the corresponding metric sum is found,
• the detected bit, b i/det is first determined by using the minimum metrics. Then, the above method is applied as if the detected bit is +1.
• the detected differential bit values are decided by finding the most likely coherent symbol values (by finding the coherent symbol which minimizes the metric) for the present and previous values .
• the sum of the maximum metric values at the present and previous time provides the numerator or the denominator (depending on the detected bit value) terms.
• the non-detected term by assuming that the dominating terms will include one of the detected symbols, one of the symbols is fixed as the detected coherent symbol and the other term toggled with the closest border symbol.
• the second alternative embodiment of the present invention is illustrated in Figure 10.
• the detected symbols are determined and their corresponding metrics (block 900) .
• the detected combined metric is determined by summing the maximum metric of the detected symbols (block 902) .
• a first potential non-detected combined metric is determined (block 904) and a second potential non- detected combined metric is also determined (block 906) .
• the maximum of these non-detected combined metrics is selected (block 908) and the soft value determined based on the selected non-detected combined metric and the detected combined metric (block 910) .
• the multiple channels could correspond to antennas in an antenna array, such as a phased array or diversity array, possibly with different polarizations.
• the multiple channels could also correspond to multiple beams or multiple carriers or multiple codes or multiple time slots.
• the present invention can also be used to produce soft information on differential symbols (di- bits) .
• likelihoods for each of the different symbol values would be produced.
• DQPSK differential symbols
• four soft values would be produced, one for each possible differential symbol value.
• the log likelihood of a differential symbol value would be the sum of four terms, i . e .
• the symbol likelihood would include one of the detected symbol terms .
• the present invention is applicable to a mixture of differential and normal encoding, providing soft information or detected values to the differentially encoded bits or symbols. In essence, coherent symbol pairs are hypothesized and symbol pair (combined) metrics are produced. Combinations corresponding to one differential bit value are grouped as are combinations corresponding to the other differential bit value. The likelihood ratio or an approximation to it are then determined.
• FIGS. 8 through 10 are flowchart illustrations of embodiments of the present invention. It will be understood that each block of the flowchart illustrations, and combinations of blocks in the flowchart illustrations, can be implemented by computer program instructions. These program instructions may be provided to a processor to produce a machine, such that the instructions which execute on the processor create means for implementing the functions specified in the flowchart block or blocks. The computer program instructions may be executed by a processor to cause a series of operational steps to be performed by the processor to produce a computer implemented process such that the instructions which execute on the processor provide steps for implementing the functions specified in the flowchart block or blocks.
• blocks of the flowchart illustrations support combinations of means for performing the specified functions and combinations of steps for performing the specified functions. It will also be understood that each block of the flowchart illustrations, and combinations of blocks in the flowchart illustrations, can be implemented by special purpose hardware-based systems which perform the specified functions or steps, or combinations of special purpose hardware and computer instructions.

Landscapes

• Engineering & Computer Science (AREA)
• Power Engineering (AREA)
• Computer Networks & Wireless Communication (AREA)
• Signal Processing (AREA)
• Radio Transmission System (AREA)

Priority Applications (3)

Applications Claiming Priority (2)

Publications (1)

Family

ID=22505447

Family Applications (1)

Country Status (6)

Cited By (3)

Families Citing this family (23)

Citations (6)

Family Cites Families (11)

• 1998
  • 1998-08-31 US US09/143,754 patent/US6487255B1/en not_active Expired - Fee Related
• 1999
  • 1999-08-24 CN CN99810173A patent/CN1120603C/zh not_active Expired - Fee Related
  • 1999-08-24 EP EP99941246A patent/EP1110359A1/en not_active Withdrawn
  • 1999-08-24 WO PCT/US1999/019221 patent/WO2000013386A1/en not_active Application Discontinuation
  • 1999-08-24 AU AU54937/99A patent/AU5493799A/en not_active Abandoned
• 2002
  • 2002-03-22 HK HK02102218.0A patent/HK1040855B/zh not_active IP Right Cessation

Patent Citations (6)

Also Published As

Similar Documents

Legal Events