WO1997024849A1 - Method and apparatus for symbol decoding using a variable number of survivor paths - Google Patents

Method and apparatus for symbol decoding using a variable number of survivor paths Download PDF

Info

Publication number
WO1997024849A1
WO1997024849A1 PCT/US1996/020091 US9620091W WO9724849A1 WO 1997024849 A1 WO1997024849 A1 WO 1997024849A1 US 9620091 W US9620091 W US 9620091W WO 9724849 A1 WO9724849 A1 WO 9724849A1
Authority
WO
WIPO (PCT)
Prior art keywords
value
trelhs
quahty
current
signal
Prior art date
Application number
PCT/US1996/020091
Other languages
French (fr)
Other versions
WO1997024849A9 (en
Inventor
Sandeep Chennakeshu
Ravinder David Koilpillai
John B. Anderson
Original Assignee
Ericsson Inc.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ericsson Inc. filed Critical Ericsson Inc.
Priority to EP96944428A priority Critical patent/EP0872095B1/en
Priority to JP9524372A priority patent/JP2000502859A/en
Priority to DE69635085T priority patent/DE69635085D1/en
Priority to AU14238/97A priority patent/AU716138B2/en
Priority to CA002241691A priority patent/CA2241691C/en
Publication of WO1997024849A1 publication Critical patent/WO1997024849A1/en
Publication of WO1997024849A9 publication Critical patent/WO1997024849A9/en

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L1/00Arrangements for detecting or preventing errors in the information received
    • H04L1/004Arrangements for detecting or preventing errors in the information received by using forward error control
    • H04L1/0056Systems characterized by the type of code used
    • H04L1/0071Use of interleaving
    • HELECTRICITY
    • H03ELECTRONIC CIRCUITRY
    • H03MCODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M13/00Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
    • H03M13/37Decoding methods or techniques, not specific to the particular type of coding provided for in groups H03M13/03 - H03M13/35
    • H03M13/39Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes
    • H03M13/41Sequence estimation, i.e. using statistical methods for the reconstruction of the original codes using the Viterbi algorithm or Viterbi processors
    • HELECTRICITY
    • H03ELECTRONIC CIRCUITRY
    • H03MCODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M13/00Coding, decoding or code conversion, for error detection or error correction; Coding theory basic assumptions; Coding bounds; Error probability evaluation methods; Channel models; Simulation or testing of codes
    • H03M13/65Purpose and implementation aspects
    • H03M13/6502Reduction of hardware complexity or efficient processing
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L1/00Arrangements for detecting or preventing errors in the information received
    • H04L1/0001Systems modifying transmission characteristics according to link quality, e.g. power backoff
    • H04L1/0009Systems modifying transmission characteristics according to link quality, e.g. power backoff by adapting the channel coding
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00Baseband systems
    • H04L25/02Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/03Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
    • H04L25/03006Arrangements for removing intersymbol interference
    • H04L25/03178Arrangements involving sequence estimation techniques
    • H04L25/03203Trellis search techniques
    • H04L25/03216Trellis search techniques using the M-algorithm

Definitions

  • the present invention relates to decoding received symbols in a communications system, and in particular, describes a decoding technique using a trellis structure where the number of trellis states or paths retained during the decoding process is variable.
  • encoded symbols are transmitted over a communications channel, (e.g., a radio frequency communications channel) which is subject to various types of distortion such as co-channel and adjacent channel interference, noise, dispersion, fading, weak signal strength, etc.
  • Trellis coding is one technique used to encode sequences of symbols to be transmitted, and in the process, introduces memory to the transmitted signal. Because of the introduced memory, the transmitted signal at any instant of time depends on previously transmitted signals. Therefore, in order to demodulate the received signal and recover the transmitted symbols, the decoding process must be performed taking into account the whole sequence of symbols.
  • the preferred method of symbol sequence decoding is maximum likelihood sequence estimation (MLSE) which offers a mechanism for decoding the transmitted signal optimally in the sense of mniimizing the probability of error.
  • MLSE maximum likelihood sequence estimation
  • the received sequence of signals is compared with every possible sequence of symbols that could have been transmitted.
  • the particular sequence out of all of the possible sequences that best matches the received sequence of signals is chosen as the decoded sequence. If the length of the transmitted symbol sequence is N symbols long and there are Q possible values that each symbol can take, then the MLSE pattern matching scheme tries to determine the best match from among Q possible sequences.
  • the number of potential MLSE search sequences is too complex for implementation.
  • Fig. 1 illustrates a simple trellis where each transmitted symbol can take one of four values.
  • Each trellis node corresponds to one value of the transmitted symbol and is often referred to as a state.
  • At each time instant/stage in the trellis there are four nodes defining a four-state trellis.
  • the set of nodes repeats for each transmitted symbol period. Transitions between nodes are called branches, and each branch is associated with a possible symbol that could have been transmitted.
  • This structure of nodes (states) and branches (symbols) defines the trellis.
  • a metric or weight is formed for each branch which indicates the likelihood of the symbol corresponding to that branch being part of the true received symbol sequence.
  • An example of a metric is the squared, absolute difference between a received signal sample and a corresponding estimate of the sample formed using the symbol associated with that branch. This metric is referred to as the squared Eudidean distance metric.
  • a linkage of branches through the nodes forms a path. The path with the best accumulated metric (i.e., the smallest sum of the branch metrics) is chosen, and the symbols corresponding to the branches of the chosen path are output as the decoded symbols.
  • the Viterbi algorithm retains only the branch into each node/state having the bes tvOptimal (i.e., lowest) metric and discards the remaining branches into that node/state. This best branch selection procedure significantly reduces the number of possible paths through the trellis. Then, from those possible paths, the Viterbi algorithm selects the path through the trellis that produces the minimum overall distance. Since it is only necessary to retain as many paths as there are nodes, i.e., there is only one path per state, the paths that remain at each state/node are called "survivors".
  • the Viterbi algorithm is a recursive technique that traces an optimal path of survivors through the trellis.
  • the main problem with the Viterbi algorithm is that its complexity grows exponentially with the number of states (memory) of the transmitted, trellis-coded signal.
  • N represents the memory or dependence of a symbol on other symbols.
  • a "memory of N" implies that each symbol depends on N other symbols, e.g., previous symbols.
  • This complexity can be reduced using what is known as the M-algorithm or T-algorithm, both of which further restrict the number of states retained for decoding.
  • the M-algorithm for each time instant in the trellis, the number of survivors is restricted to a fixed integer number M.
  • the M-algorithm chooses the M most likely states based on the accumulated path metrics.
  • the survivor states having larger accumulated path metrics are discarded.
  • Another variant of the M-algorithm is to retain the best (as defined by the accumulated path metrics) M-paths through the trellis rather than M states.
  • the M-algorithm restricts the search through the trellis to a fixed number of states M at any time instant in the trellis.
  • the selected value for M is too small, there is a high probability of retaining the wrong M states for decoding which results in a high error probability.
  • the selected value of M is too large, the decoding procedure becomes unduly complex.
  • the T-algorithm retains only those states which have an associated metric that is below a predetermined threshold (T).
  • T a predetermined threshold
  • the choice of the fixed threshold T is critical in order to achieve satisfactory performance without undue complexity.
  • the selection of a threshold T for a particular communications application that is suitable for all or most conditions encountered in that application is quite difficult.
  • the problem with a fixed M or a fixed T is that the communications channel, over which encoded symbols are transmitted, received, and then decoded, changes.
  • the radio channels exhibit constantly changing parameters because of fading, multipath dispersion, adjacent channel interference, co-channel interference, noise, etc.
  • the value of M for the M-algorithm or the value of T for the T- algorithm must be set to a relatively large number in order to have a sufficient number of states available to satisfactorily decode the symbols under such worst case channel conditions.
  • this is an inefficient decoding approach because for a large percentage of the time, the channel condition may be quite good, and a much smaller number of states is all that is needed to satisfactorily decode the received signal.
  • the present invention provides low complexity, high performance trellis decoding that adapts trellis decoding complexity based on a current condition of the communications system.
  • a trellis structure of nodes and branches is developed to decode the sequence of symbols received over a communications channel.
  • Each set of nodes in the trellis represents all possible values of a received symbol at one time instant.
  • Each branch defines a specific transition between nodes at different time instants, and each branch has an associated branch weight.
  • a linkage of branches between the nodes and the trellis defines a path through the trellis that represents one possible symbol sequence, and an accumulated weight is generated for each possible path through the trellis.
  • a variable number M of survivor paths (or states) through the trellis is determined based on these respective, accumulated path weights. The value of the variable M depends on the current condition of the communications system. Once the value of M is adapted to the current condition, the sequence of received symbols is trellis decoded using the M survivor paths/states.
  • a first value of M is selected if the current state of communications system condition is less than or equal to a threshold. Otherwise, a second value of M less than the first value of M is selected if the current status of the communications system condition is greater than the threshold.
  • the first value of M corresponds to a maximum value
  • the second value corresponds to a minimum value. If after selecting the minimu ⁇ i value of M, the current communications system deteriorates to less than the threshold, the value of M is increased toward the maximum value in a controlled fashion. In particular, the received signal once again is trellis decoded using an increased value of M.
  • the value of M is "pruned" or reduced in each successive stage of the trellis until it reaches the mi ⁇ mn ⁇ value (assuming the current communications system condition continues to exceed the threshold).
  • One of the communications system conditions may be the quality of the communications channel.
  • a minimum value of M is selected if the current quality of the channel is greater than the threshold, otherwise, a maximum value of M is selected.
  • the communications system condition may relate to the complexity of the encoding scheme used to encode the symbols to be transmitted.
  • a more powerful coding scheme that has more encoding memory can be used to generate a coded signal.
  • this typically involves decoding with a large value of M, such decoding may be performed in accordance with the present invention with a reduced number of states or paths, thereby minimizing decoding complexity.
  • Another communications system condition relates to the current data processing tasks to be performed by the data processing circuitry.
  • the value of M is decreased if the current number of additional tasks to be performed is relatively large. In this way, the data processing circuitry is relieved of the burden of a complex decoding process so that it may devote more processing resources to perform the other pending tasks.
  • the value of M can be increased, thereby improving the performance of the decoder, if the current number of additional pending tasks is relatively small.
  • the value of M can be varied to maintain a predetermined level of decoder performance in response to changes in the quality of the communications channel.
  • M is relatively low, full data processing capacity is not consumed by the operation of the trellis decoder. As a result, the excess data processing capacity may be devoted to other tasks having lower priority.
  • the value of M may be decreased whenever consistent with decoder performance objectives in order to reduce the number of data processing tasks associated with the trellis decoding operation, and thereby, reduce battery drain.
  • the present invention may also be applied to a spread spectrum type receiver in which plural encoded sequences of symbols transmitted by plural transmitters are received on a single communications channel.
  • the value of M may be changed in response to changes in the number of interfering signals. When satisfactory performance of the decoder is degraded because of additional interfering signals, the value of M may be increased to maintain or improve decoder performance. On the other hand, when the number of interfering signals decreases, the value of M may be decreased to a smaller value of M which is still consistent with satisfactory decoder performance in order to conserve processor or battery resources.
  • the present invention adapts the trellis decoding complexity based on a current channel condition in a communications system using a best received signal from plural diversity antennas.
  • Each of the receiver"s plural diversity antennas receives a transmitted signal made up of a sequence of coded symbols.
  • a received signal from one of the plural antennas having the highest signal quality is selected by the receiver and then decoded using a trellis decoding procedure which employs a variable number M of trellis survivor paths or states, where M depends on the quality of the received signal.
  • M depends on the quality of the received signal.
  • another embodiment of the present invention selectively weights and then combines the weighted signals from the plural antennas to generate a combined signal.
  • the received signal from one of the antennas having the highest signal quality is most heavily weighted in the combined signal.
  • variable number M is varied as a function of the current quality of the communications channel.
  • signals received from each of the plural diversity antennas are detected and stored, and a channel quality indicator is determined for each stored signal.
  • the stored signal having the highest channel quality indicator is selected, and the value of the variable M is determined based on that highest channel quality indicator.
  • the stored, received signal is then decoded using that recently determined number of trellis survivor paths or states. More specifically, if the highest channel quaHty indicator is greater than a threshold, the value of M is decreased. On the other hand, if the highest channel quality indicator is less than that threshold, the value of M is increased.
  • Signal quality of each received signal is determined based on received signal strength, and in one example preferred embodiment, is based on an average faded signal amplitude of the received signal.
  • the value of M is adapted with the channel condition as reflected in the received channel quality indicator.
  • the value of M is increased when the channel condition is poor, and the value of M is decreased when the channel condition is good.
  • the complexity of the variable-M decoding procedure is proportional to the ratio of time during which the channel condition is good relative to when the channel condition is poor.
  • the present invention improves this ratio using antenna diversity, and as a result, lowers the transmit power required to achieve a given bit error rate by the receiver decoder relative to use of the same variable M decoding procedure without antenna diversity.
  • Lowering transmit power reduces battery drain for transmitting portable radio transceivers and reduces adjacent channel interference caused by transmitters in fixed, base station type transceivers.
  • the present invention reduces decoding complexity, and therefore reduces battery drain by the receiver, or improves decoding performance in terms of reducing the number of bit errors without increasing battery drain by the receiver.
  • the present invention may also be advantageously applied to a symbol interleaving and or frequency hopping communications system. Both symbol interleaving and frequency hopping improve the quaHty of received signals which permits the use of a smaHer value of M in accordance with the invention.
  • FIGURE 1 is a graphical illustration of a trelHs
  • FIGURE 2 is a function block diagram of one example communications system in which the present invention may be employed;
  • FIGURE 3 is a flowchart diagram outlining basic procedures for implementing the variable stage trelHs decoding algorithm in accordance with the present invention;
  • FIGURE 4 is a flowchart diagram outlining specific procedures for implementing a specific example embodiment of the variable stage trelHs decoding algorithm where the communications system parameter is current communications channel quaHty;
  • FIGURE 5 is a graph illustrating the performance of the reduced complexity trelHs decoding procedure in accordance with the present invention as compared with the traditional, higher complexity Viterbi trelHs decoding procedure.
  • FIGURE 6 is a more detailed function block diagram of the receiver branch shown in FIGURE 2 in accordance with an example embodiment of the invention.
  • FIGURE 7 is a flowchart diagram outlining the trelHs decoding procedure using a variable number of M trelHs states and antenna diversity in accordance with an example embodiment of the invention
  • FIGURE 8 is a graph illustrating the performance of the reduced complexity, variable-M trelHs decoding procedure which incorporates antenna diversity in accordance with an example embodiment of the invention as compared to the variable-M trelHs decoding procedure without antenna diversity.
  • an overaH function block diagram of an example communications system 10 which may employ the present invention is shown. While the present invention is particularly weU suited to radio communications including both fixed and portable radios, it is appHcable to other communications environments as weU.
  • the information signal to be transmitted i.e., a sequence of symbols
  • TCM trelHs code modulation
  • differential encoder 12 If the encoded symbols are being transmitted on a slotted communications channel, for example, using weU-known time division multiplex (TDM) techniques, a symbol interleaver 14 is desirable (although not necessary) to break up or "shuffle" symbol sequence segments into different time slots.
  • Interleaving is particularly effective in communication channels subject to fading because symbol segments received under poor channel conditions are interspersed with symbol segments received under good channel conditions.
  • the symbols may be complex and are represented by two components, namely a real value I (in-phase) and an imaginary value (quadrature).
  • the in-phase and quadrature components are passed through paraUel processing branches. Each component is passed through a transmit filter 18a, 18b, converted in digital- to- analog (D/A) converters 20a, 20b, and frequency shifted by quadrature modulators 22a, 22b.
  • the quadrature modulated signals are then mixed in summer 24 and sent to an RF amplifier 26 which increases the gain of the signal.
  • the RF amplifier 26 passes the amplified signal to one or more antennas 28a...28n for transmission.
  • Conventional frequency hopping techniques may be used to "hop" between multiple carrier frequencies with each frequency carrying only a portion of the symbols.
  • One or more antennas 30A...30N at the receiver side of the communications system 10 receive the signal.
  • diversity antennas 30A...30N may be incorporated in fixed radio transceivers such as base station transceivers/repeaters and in portable/mobile radios.
  • the received signals from each of the plural diversity antennas 30A...30N are processed using respective radio amplification and downconversion paths. However, for simpHcity, only a single downconversion path is illustrated and described in Fig. 2.
  • One antenna signal passes through a filtering and RF preamplifier stage 32 to a first downconverter 34 which reduces the frequency (or frequencies in a frequency hopping system) of the received signal to an intermediate frequency.
  • the intermediate frequency signal is passed through an intermediate frequency receive filter 36, then through a second downconverter 38 to produce a (low pass) filtered baseband signal.
  • the IF downconversion is desirable, it is of course not necessary to implement the present invention.
  • the filtered baseband signal is then converted into a complex signal having in-phase (I) and quadrature (Q) components by signal processor 39.
  • the in-phase and quadrature signals are digitized by analog-to-digital (A D) converters 40a, 40b and passed through complex signal generator 42 which converts differential phase information into a sequence of symbols.
  • Deinterleaver 44 reconstitutes interleaved slots of information (i.e., it performs the reverse shuffling operation to undo the shuffling operation performed by in terleaver 14).
  • TCM decoder 46 processes this symbol stream using a trellis decoding technique described in further detail below to recover the transmitted signal information.
  • Signal tracker 48 estimates an average faded signal strength of an analog signal output from the IF stage 38 or from samples of the received signal from the deinterleaver 44.
  • deinterleaver 44, TCM decoder 46, and signal tracker 48 are implemented using a suitably programmed microprocessor and/or digital signal processing circuitry.
  • the TCM decoder 46 is a maximum likelihood sequence estimation (MLSE) trelHs decoder that develops a trelHs of nodes and branches (similar to that shown in Fig. 1 and described in the background of the invention) for decoding sequences of symbols received on the communications channel.
  • Each set of nodes in the trelHs represents aH possible values of a received symbol at one time instant. Transition between a node or state at one time instant to a node or state at another time instant is referred to as one stage of the trellis.
  • Each stage typicaUy includes branches with each branch defining a specific transition between nodes at sequential time instances, and each branch has an associated branch weight or metric.
  • a linkage of branches between nodes over multiple stages in the trelHs defines a path through the trelHs that represents a possible received symbol sequence. For each possible trelHs path, the branch metrics associated with that path are accumulated or summed to provide a path metric. The path having the lowest accumulated metric is selected as the best path.
  • the present invention determines which of the survivor paths through the trelHs wiH be maintained for purposes of decoding based on a current condition of the communications system.
  • the other remaining survivor paths are discarded to simplify the decoding operation.
  • a variable number of M survivor paths upon which trelHs decoding of the received sequence of symbols is based depends on a current condition of the communications system.
  • Fig. 3 outlines in flowchart form the general procedures for varying the number of M survivor paths to be used in the trelHs decoding.
  • the current communications system condition is determined in step 50.
  • Various examples of different communications system conditions wiH be described in further detail below.
  • trelHs decoding using, for example, the weU-known Viterbi algorithm is performed to determine the best estimate of the received symbol sequence.
  • the current communications system condition is determined to be sufficiently optimal, which is determined in this example by comparison with a threshold, the number of states or paths needed to satisfactorily (in terms of low bit error rates) decode the received symbol sequence is relatively small.
  • the current communications system condition is less than the threshold, a larger number of M is used to ensure satisfactory decoding performance.
  • a useful communications system parameter is the quaHty of the communications channel.
  • the quaHty of any communications channel varies over time to some degree.
  • the quaHty of the communications channel changes rapidly (especiaHy when the radio user is moving in a car) with the communications channel being subjected to fading, multipath dispersion, adjacent channel interference and co- chan el interference from other users, noise, and other channel impairments.
  • the current quaHty of the communications channel is detected or determined signal tracker 48. If the current quaHty of the channel is greater than the threshold, meaning that the channel quaHty is good, a first relatively smaH value of M is selected. Alternatively, if the current quaHty of the channel is less than a threshold value, indicating poor channel quaHty, another relatively large value of M is selected greater than the value of M 20091
  • Channel quaHty can be measured using a number of 5 conventional channel quaHty indicators such as received signal strength, signal-to-noise ratio (SNR), signal to interference ratio ( SIR ) , bit error rate (BER ) , etc.
  • SNR signal-to-noise ratio
  • SIR signal to interference ratio
  • BER bit error rate
  • One preferred example indicator is averaged faded signal strength (AFSS) which is a smoothed estimate of the channel a pHtude of the received signal which may be obtained using 10 a channel tracking algorithm.
  • AFSS averaged faded signal strength
  • the channel tracking algorithm low-pass filters received signal samples to provide an estimate of the envelope of the ampHtude of the received signal.
  • s(n ) , c n , and ⁇ ( n ) represent a (complex baseband) transmitted 15 symbol, a complex channel gain, and additive Gaussian noise, respectively, at time n. If y(n) is the received signal, and r(n) is the signal at the output of the complex symbol generator 42, then y(n ) and r(n ) may be expressed as foUows:
  • the channel gain c varies with time.
  • the AFSS algorithm obtains the average faded signal strength E n (an averaged value of a n ), and tracks its variation over time by executing the foUowing equations:
  • ⁇ ⁇ min
  • Fig. 4 Olustrates in flowchart diagram format one preferred example embodiment for implementing the present invention where the communications system parameter is channel quaHty. H_iti.aHzati.on procedures are performed in block 100 to select M . .
  • the initiaHzation parameters M . , M ma ⁇ , T, ⁇ , and ⁇ are chosen depending upon the particular communications appHcation and can often be effectively determined using simulation tests.
  • the current channel quaHty is determined in
  • Block 102 using for example an AFSS procedure as described above.
  • One stage of the M state trelHs decoding procedure is performed foUowing the weU-known Viterbi decoding procedures with M set equal to M . (block 104).
  • a decision is made in block 106 whether the current channel quaHty is greater than or equal to the channel
  • the present invention takes into account the fact that these just recently decoded stages of the trelHs were likely to have been adversely impacted by the newly detected poor channel quaHty. Therefore, the decoding operation is repeated for a number ⁇ of last stages by retracing that number ⁇ of stages in the trelHs and restarting the trelHs decoding algorithm (block 110). This retrace procedure ensures a high degree of reHabiHty in decoding accuracy.
  • Control proceeds to block 112, where ⁇ stages of the M- ⁇ £age" ⁇ ⁇ trelHs decoding algorithm are performed with M set equal to ma ⁇ .
  • Another decision is made at block 114 to determine whether the current channel quaHty is less than the threshold T within the ⁇ stages. If it is, the M state decoding procedure is continued at block 112 with M set at M max . However, if the current channel quaHty is greater than or equal to the threshold T, a new, smaUer value of M is generated in block 116 reflecting the current trend that reduced decoding complexity may be appropriate in view of the recently improved channel quaHty. But rather than immediately setting the new value of M to M .
  • the present invention employs a more conservative "pruning" procedure. Namely, the value of M is halved in each successive stag °e of the trelHs until it reaches M ⁇ n ._n . Thus, ' when the channel improves, the present invention employs the poor channel holding window ⁇ and the controUed M-reduction or pruning procedure to ensure that the channel condition remains at the improved level before M is set to M . .
  • the M state decoding procedure is performed with the new smaller value of M in block 118 being reduced again, if possible, in block 116 until it reaches M . (block 120). Control returns to block 102 to repeat the above-described operation recursively decoding received symbols at the lowest, optimal level of decoding complexity to ensure satisfactory decoding performance.
  • Fig. 5 illustrates the performance of the present invention with reduced complexity as compared to the traditional full complexity Viterbi algorithm.
  • Bit error rate (BER) is appHed on the vertical axis with the signal-to-noise ratio (Eb /No [energy per bit to noise power spectral density]) plotted in dB on the horizontal axis.
  • Eb /No [energy per bit to noise power spectral density] the signal-to-noise ratio
  • the radio receiver is moving in a vehicle at a speed of 150 KMPH.
  • M 10 and 11 states
  • the complexity of the variable M algorithm employed by the present invention was six times less complex when measured in terms of number of operations executed to decode the data.
  • the communications system condition may relate to the complexity of the encoding scheme used by the transmitter TCM encoder 12. Codes with more memory potential (sometimes referred to as "constraint length" in the Hterature) exhibit better performance. However, since decoding complexity increases exponentiaHy with memory of the code, these codes cannot be easily used due to data processing constraints. A smaHer value of M may be varied in this instance to adapt the decoding procedure to a level best suited given a particular set of codes.
  • Decoding procedures are performed using data processing circuitry, and oftentimes the data processing circuitry must perform a number of other tasks in a time share relationship.
  • Another communications system condition relates therefore to the current data processing tasks to be performed by the data processing circuitry.
  • the value of M is decreased if the current number of additional tasks to be performed is relatively large. In this way, the data processing circuitry is reHeved of the burden of a complex decoding process so that it may devote more processing resources to perform the other large number of pending tasks.
  • the value of M can be increased, thereby improving the performance of the decoder, if the current number of additional pending tasks is relatively smaH.
  • the value of M can be varied to maintain a predetermined level of decoder performance in response to changes in the quaHty of the communications channel.
  • M is relatively low, the fuH capacity of the data processing circuitry is not consumed by the operation of the trelHs decoder. As a result, the excess capacity of the data processing circuitry can be devoted to other tasks having lower priority.
  • the value of M may be decreased whenever consistent with decoder performance objectives in order to reduce the number of data processing tasks associated with the trelHs decoding operation, and thereby, reduce battery drain.
  • the present invention may also be appHed to a spread spectrum type receiver in which plural encoded sequences of symbols transmitted by plural transmitters are received on a single communications channel.
  • the value of M may be changed in response to changes in the number of interfering signals. When satisfactory performance of the decoder is degraded because of additional interfering signal, the value of M can be increased to maintain or improve decoder performance. On the other hand, when the number of interfering signals decreases, the value of M can be decreased to the smaHest value of M consistent with satisfactory decoder performance in order to conserve processor or battery resources.
  • the receiver may have multiple antennas and obtain relatively uncorrelated copies of the received signal from each antenna.
  • the antennas are rendered uncorrelated by separating them spatiaUy or using orthogonal polarizations.
  • the receiver combines these uncorrelated copies of the received signal in a manner to improve signal quaHty, i.e., selecting the strongest signal based on average faded signal strength, co-phasing the signals to perform maximal- ratio-combining or equal gain combining, adding the branch metrics calculated for each of the multiple antennas, and adding the received signals to suppress interference, etc. If there are N antennas and K interferers and K ⁇ N, then interference can be reduced.
  • Fig. 6 is a function block diagram which shows in more detail the diversity antenna structure and one way of processing the signals received by each of the antennas in accordance with one example embodiment of the present invention.
  • Signals received on each diversity antenna 30A...30N are treated as a separate channel, i.e., channel 1... channel N.
  • Each channel is processed using its own dedicated front end processing circuitry 48A - 48N which performs such conventional tasks as RF preamplification and frequency downconversion.
  • Signal processor 39 in Fig. 2 includes selection logic 52 which determines a signal quaHty indicator for each buffered signal. The selection logic 52 then selects the buffered received signal having the highest signal quaHty indicator for further processing and decoding in the TCM decoder 46. Alternatively, selection logic 52 may selectively weight the buffered signals based on their signal quaHty indicator and then combine the weighted signals into a combined signal. The buffered signal corresponding to the diversity antenna having the highest signal quaHty is most heavily weighted which ensures that the combined signal substantiaHy reflects the best signal quaHty for that particular time instant.
  • Fig. 7 outlines in flowchart form the general procedures for varying the number of M survivor paths to be used in the trelHs decoding based on plural received signals by plural diversity antennas.
  • the baseband signals from each diversity antenna are sampled and then buffered in step 60.
  • the channel quaHties for each demodulated sample corresponding to each of the pluraHty of diversity antennas are then determined in step 62.
  • the buffered sample with the best channel quaHty measure is selected, or alternatively, more heavily weighted in step 64.
  • a variable-M trellis decoding procedure is then performed in step 66 in accordance with the procedures outlined in the flowcharts iHustrated in Figs. 3 and 4 and described above.
  • bit error rate is plotted on the vertical axis of the graph with signal-to-noise ratio (Eb/No) plotted in dB on the horizontal axis.
  • the radio receiver is moving in a vehicle at a speed of 150 KMPH.
  • the variable-M decoding scheme without diversity reception is plotted using plus signs.
  • variable-M decoding procedure with diversity reduces the decoding complexity by a factor of 11.
  • the present invention adapts the complexity of the trelHs decoding procedure in accordance with the current channel conditions. This approach is not only more efficient in the sense of reduced decoding complexity, it also lowers the required transmitted power to achieve a fixed bit error rate performance.
  • the value of M is adapted with the channel condition as reflected in the received channel quaHty indicator. The value of M is increased when the channel condition is poor, and the value of M is decreased when the channel condition is good.
  • the complexity of the variable-M decoding procedure is proportional to the ratio of time during which the channel condition is good relative to when the channel condition is poor.
  • the present invention improves this ratio using antenna diversity, and as a result, lowers the transmit power required to achieve a given bit error rate by the receiver decoder relative to use of the same variable M decoding procedure without antenna diversity.
  • Lowering transmit power reduces battery drain for transmitting portable radio transceivers and reduces adjacent channel interference caused by transmitters in fixed, base station type transceivers.
  • the present invention reduces decoding complexity and therefore reduces battery drain by the receiver or improves decoding performance in terms of reducing the number of bit errors without increasing battery drain by the receiver. Two further example appHcations of the present invention are briefly described.
  • multiple diversity transmitting antennas may be used to transmit the same information to the receiver to improve signal quaHty.
  • the antenna having the better path to the receiver can be selected by the receiver based on the received signal on each antenna. Transmission along a better path offers better signal quaHty, and hence, a lower value of M can be employed.
  • a second further example appHcation of the present invention relates to symbol interleaving and/or frequency hopping. Interleaving is used to "break-up" channel memory. If a sequence of symbols is sent and the symbols are highly correlated in time due to channel memory, decoding performance wiH be poor whenever the channel condition is poor because aU symbols are affected. However, if the symbols are "shuffled” such that they are separated by other unrelated symbols, they are less likely to be equaUy affected by channel memory. Symbol interleaving when used with frequency hopping is particularly effective in improving signal quaHty because, in addition to separating symbols in time, those symbols are also separated in frequency.
  • Groups of time-separated symbols are transmitted at different ca ⁇ ier frequencies.
  • the symbols are suitably deinterleaved and sent to the variable M-algorithm for decoding.
  • the improved signal quaHty due to interleaving (in both time and frequency) aHows the use of a low value of M.
  • trelHs coded modulation it can be used in demodulating block coded modulation, convolutional codes, block codes, partial response modulation such as continuous phase modulation and any set of signals that can be represented by a trelHs.

Abstract

A low complexity, high performance trellis decoder adapts the complexity of the trellis decoding process based on a current condition of the communications system. A trellis structure of nodes and branches is developed to decode the sequence of symbols received over a communications channel. Each set of nodes in the trellis represents the possible values of a received symbol at one time instant. Each branch defines a specific transition between nodes at different time instants, and each branch has an associated branch metric or weight. A linkage of branches between the nodes and the trellis defines a path through the trellis that represents one possible symbol sequence, and an accumulated metric/weight is generated for each possible path through the trellis. A variable number M survivor paths through the trellis is determined based on their accumulated path metric/weights. The value of the variable M depends on the current condition of the communications system. Once the value of M is adapted to the current condition, the sequence of received symbols is decoded using M survivor paths through the trellis.

Description

METHOD AND APPARATUS FOR SYMBOL DECODING USING A VARIABLE NUMBER OF SURVIVOR PATHS
The present invention relates to decoding received symbols in a communications system, and in particular, describes a decoding technique using a trellis structure where the number of trellis states or paths retained during the decoding process is variable.
BACKGROUND AND SUMMARY OF THE INVENTION
In coded data communications, encoded symbols are transmitted over a communications channel, (e.g., a radio frequency communications channel) which is subject to various types of distortion such as co-channel and adjacent channel interference, noise, dispersion, fading, weak signal strength, etc. Trellis coding is one technique used to encode sequences of symbols to be transmitted, and in the process, introduces memory to the transmitted signal. Because of the introduced memory, the transmitted signal at any instant of time depends on previously transmitted signals. Therefore, in order to demodulate the received signal and recover the transmitted symbols, the decoding process must be performed taking into account the whole sequence of symbols.
The preferred method of symbol sequence decoding is maximum likelihood sequence estimation (MLSE) which offers a mechanism for decoding the transmitted signal optimally in the sense of mniimizing the probability of error. In the MLSE method, the received sequence of signals is compared with every possible sequence of symbols that could have been transmitted. The particular sequence out of all of the possible sequences that best matches the received sequence of signals is chosen as the decoded sequence. If the length of the transmitted symbol sequence is N symbols long and there are Q possible values that each symbol can take, then the MLSE pattern matching scheme tries to determine the best match from among Q possible sequences. However, even for modest values of Q and N, the number of potential MLSE search sequences is too complex for implementation.
One simplified solution for performing the MLSE search is provided by the well known Viterbi algorithm. To understand the Viterbi algorithm, the search problem is represented as a method of searching a trellis. Fig. 1 illustrates a simple trellis where each transmitted symbol can take one of four values. Each trellis node corresponds to one value of the transmitted symbol and is often referred to as a state. At each time instant/stage in the trellis there are four nodes defining a four-state trellis. The set of nodes repeats for each transmitted symbol period. Transitions between nodes are called branches, and each branch is associated with a possible symbol that could have been transmitted. This structure of nodes (states) and branches (symbols) defines the trellis.
A metric or weight is formed for each branch which indicates the likelihood of the symbol corresponding to that branch being part of the true received symbol sequence. An example of a metric is the squared, absolute difference between a received signal sample and a corresponding estimate of the sample formed using the symbol associated with that branch. This metric is referred to as the squared Eudidean distance metric. A linkage of branches through the nodes forms a path. The path with the best accumulated metric (i.e., the smallest sum of the branch metrics) is chosen, and the symbols corresponding to the branches of the chosen path are output as the decoded symbols.
To streamline the search procedure, the Viterbi algorithm retains only the branch into each node/state having the bes tvOptimal (i.e., lowest) metric and discards the remaining branches into that node/state. This best branch selection procedure significantly reduces the number of possible paths through the trellis. Then, from those possible paths, the Viterbi algorithm selects the path through the trellis that produces the minimum overall distance. Since it is only necessary to retain as many paths as there are nodes, i.e., there is only one path per state, the paths that remain at each state/node are called "survivors". The Viterbi algorithm is a recursive technique that traces an optimal path of survivors through the trellis.
The main problem with the Viterbi algorithm is that its complexity grows exponentially with the number of states (memory) of the transmitted, trellis-coded signal. With the number of states being expressed as Q and with Q as the number of possible values a symbol can take, N represents the memory or dependence of a symbol on other symbols. A "memory of N" implies that each symbol depends on N other symbols, e.g., previous symbols. This complexity can be reduced using what is known as the M-algorithm or T-algorithm, both of which further restrict the number of states retained for decoding. In the basic M-algorithm, for each time instant in the trellis, the number of survivors is restricted to a fixed integer number M. The M-algorithm chooses the M most likely states based on the accumulated path metrics. The survivor states having larger accumulated path metrics are discarded. Another variant of the M-algorithm is to retain the best (as defined by the accumulated path metrics) M-paths through the trellis rather than M states. Hence, the M-algorithm restricts the search through the trellis to a fixed number of states M at any time instant in the trellis. Unfortunately, if the selected value for M is too small, there is a high probability of retaining the wrong M states for decoding which results in a high error probability. On the other hand, if the selected value of M is too large, the decoding procedure becomes unduly complex.
The T-algorithm retains only those states which have an associated metric that is below a predetermined threshold (T). Like the M-algorithm, the choice of the fixed threshold T is critical in order to achieve satisfactory performance without undue complexity. In practice, the selection of a threshold T for a particular communications application that is suitable for all or most conditions encountered in that application is quite difficult. The problem with a fixed M or a fixed T is that the communications channel, over which encoded symbols are transmitted, received, and then decoded, changes. In fact, in radio frequency communications, the radio channels exhibit constantly changing parameters because of fading, multipath dispersion, adjacent channel interference, co-channel interference, noise, etc. In order to compensate for worst case scenarios for a communications channel, the value of M for the M-algorithm or the value of T for the T- algorithm must be set to a relatively large number in order to have a sufficient number of states available to satisfactorily decode the symbols under such worst case channel conditions. However, this is an inefficient decoding approach because for a large percentage of the time, the channel condition may be quite good, and a much smaller number of states is all that is needed to satisfactorily decode the received signal.
What is needed therefore is a trellis decoding procedure with reduced complexity that can adapt the number of survivor paths or states in the trellis based upon one or more conditions of the communications system. In radio communication applications, and especially battery operated, portable radios, transmit power is at a premium. However, lower transmit power may result in poor decoding performance at the receiver. There is also a need to reduce the transmit power of fixed transceivers like base station/repeaters to minimize adjacent channel interference. But this reduction is only feasible if receivers can accurately and efficiently decode the signals transmitted at lower power.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a trellis decoding method and apparatus with reduced complexity and high performance. It is an object of the present invention to provide a trellis decoding method and apparatus with reduced complexity, low bit error rate, and reduced power requirements.
It is a further object to provide a reduced complexity trellis decoder and decoding procedure which adapts to a current condition of the communications system.
It is yet another object to improve the efficiency and effectiveness of trellis decoding procedures using antenna diversity, frequency hopping, and/or symbol interleaving. The present invention provides low complexity, high performance trellis decoding that adapts trellis decoding complexity based on a current condition of the communications system. A trellis structure of nodes and branches is developed to decode the sequence of symbols received over a communications channel. Each set of nodes in the trellis represents all possible values of a received symbol at one time instant. Each branch defines a specific transition between nodes at different time instants, and each branch has an associated branch weight. A linkage of branches between the nodes and the trellis defines a path through the trellis that represents one possible symbol sequence, and an accumulated weight is generated for each possible path through the trellis. A variable number M of survivor paths (or states) through the trellis is determined based on these respective, accumulated path weights. The value of the variable M depends on the current condition of the communications system. Once the value of M is adapted to the current condition, the sequence of received symbols is trellis decoded using the M survivor paths/states.
A first value of M is selected if the current state of communications system condition is less than or equal to a threshold. Otherwise, a second value of M less than the first value of M is selected if the current status of the communications system condition is greater than the threshold. In one embodiment, the first value of M corresponds to a maximum value, and the second value corresponds to a minimum value. If after selecting the minimuπi value of M, the current communications system deteriorates to less than the threshold, the value of M is increased toward the maximum value in a controlled fashion. In particular, the received signal once again is trellis decoded using an increased value of M. On the other hand, when the value of M is at a maximum and the current communications system condition improves and now exceeds the threshold, the value of M is "pruned" or reduced in each successive stage of the trellis until it reaches the miτιιτmnτι value (assuming the current communications system condition continues to exceed the threshold).
One of the communications system conditions may be the quality of the communications channel. A minimum value of M is selected if the current quality of the channel is greater than the threshold, otherwise, a maximum value of M is selected.
Alternatively, the communications system condition may relate to the complexity of the encoding scheme used to encode the symbols to be transmitted. For example, a more powerful coding scheme that has more encoding memory can be used to generate a coded signal. Although this typically involves decoding with a large value of M, such decoding may be performed in accordance with the present invention with a reduced number of states or paths, thereby minimizing decoding complexity.
Sometimes the trellis decoding procedures are performed using data processing circuitry and that data processing circuitry usually must perform a number of other tasks in a time share relationship, another communications system condition relates to the current data processing tasks to be performed by the data processing circuitry. The value of M is decreased if the current number of additional tasks to be performed is relatively large. In this way, the data processing circuitry is relieved of the burden of a complex decoding process so that it may devote more processing resources to perform the other pending tasks. On the other hand, the value of M can be increased, thereby improving the performance of the decoder, if the current number of additional pending tasks is relatively small.
When decoder performance is of primary concern, the value of M can be varied to maintain a predetermined level of decoder performance in response to changes in the quality of the communications channel. When M is relatively low, full data processing capacity is not consumed by the operation of the trellis decoder. As a result, the excess data processing capacity may be devoted to other tasks having lower priority. In the context of portable, battery-operated radios, the value of M may be decreased whenever consistent with decoder performance objectives in order to reduce the number of data processing tasks associated with the trellis decoding operation, and thereby, reduce battery drain.
The present invention may also be applied to a spread spectrum type receiver in which plural encoded sequences of symbols transmitted by plural transmitters are received on a single communications channel. The value of M may be changed in response to changes in the number of interfering signals. When satisfactory performance of the decoder is degraded because of additional interfering signals, the value of M may be increased to maintain or improve decoder performance. On the other hand, when the number of interfering signals decreases, the value of M may be decreased to a smaller value of M which is still consistent with satisfactory decoder performance in order to conserve processor or battery resources. In one specific example embodiment, the present invention adapts the trellis decoding complexity based on a current channel condition in a communications system using a best received signal from plural diversity antennas. Each of the receiver"s plural diversity antennas receives a transmitted signal made up of a sequence of coded symbols. A received signal from one of the plural antennas having the highest signal quality is selected by the receiver and then decoded using a trellis decoding procedure which employs a variable number M of trellis survivor paths or states, where M depends on the quality of the received signal. By selecting the antenna producing a received signal with the highest signal quality, the variable number M is reduced to a greater extent than it otherwise would be if a weaker, lower quality signal received by another of the antennas were used. As an alternative to selecting the received signal from only one of the plural diversity antennas, another embodiment of the present invention selectively weights and then combines the weighted signals from the plural antennas to generate a combined signal. The received signal from one of the antennas having the highest signal quality is most heavily weighted in the combined signal.
In general, the variable number M is varied as a function of the current quality of the communications channel. In a receiver using diversity antennas, signals received from each of the plural diversity antennas are detected and stored, and a channel quality indicator is determined for each stored signal. The stored signal having the highest channel quality indicator is selected, and the value of the variable M is determined based on that highest channel quality indicator. The stored, received signal is then decoded using that recently determined number of trellis survivor paths or states. More specifically, if the highest channel quaHty indicator is greater than a threshold, the value of M is decreased. On the other hand, if the highest channel quality indicator is less than that threshold, the value of M is increased. Signal quality of each received signal is determined based on received signal strength, and in one example preferred embodiment, is based on an average faded signal amplitude of the received signal.
In essence, the value of M is adapted with the channel condition as reflected in the received channel quality indicator. The value of M is increased when the channel condition is poor, and the value of M is decreased when the channel condition is good. The complexity of the variable-M decoding procedure is proportional to the ratio of time during which the channel condition is good relative to when the channel condition is poor. The present invention improves this ratio using antenna diversity, and as a result, lowers the transmit power required to achieve a given bit error rate by the receiver decoder relative to use of the same variable M decoding procedure without antenna diversity. Lowering transmit power reduces battery drain for transmitting portable radio transceivers and reduces adjacent channel interference caused by transmitters in fixed, base station type transceivers. For a given transmit power, the present invention reduces decoding complexity, and therefore reduces battery drain by the receiver, or improves decoding performance in terms of reducing the number of bit errors without increasing battery drain by the receiver.
Similar to the diversity antenna selection example described above, the present invention may also be advantageously applied to a symbol interleaving and or frequency hopping communications system. Both symbol interleaving and frequency hopping improve the quaHty of received signals which permits the use of a smaHer value of M in accordance with the invention. These and other features and advantages of the present invention will become apparent from the following description of the drawings and from the claims.
BRIEF DESCRIPTION OF THE DRAWINGS
FIGURE 1 is a graphical illustration of a trelHs;
FIGURE 2 is a function block diagram of one example communications system in which the present invention may be employed; FIGURE 3 is a flowchart diagram outlining basic procedures for implementing the variable stage trelHs decoding algorithm in accordance with the present invention;
FIGURE 4 is a flowchart diagram outlining specific procedures for implementing a specific example embodiment of the variable stage trelHs decoding algorithm where the communications system parameter is current communications channel quaHty; and
FIGURE 5 is a graph illustrating the performance of the reduced complexity trelHs decoding procedure in accordance with the present invention as compared with the traditional, higher complexity Viterbi trelHs decoding procedure.
FIGURE 6 is a more detailed function block diagram of the receiver branch shown in FIGURE 2 in accordance with an example embodiment of the invention;
FIGURE 7 is a flowchart diagram outlining the trelHs decoding procedure using a variable number of M trelHs states and antenna diversity in accordance with an example embodiment of the invention; and FIGURE 8 is a graph illustrating the performance of the reduced complexity, variable-M trelHs decoding procedure which incorporates antenna diversity in accordance with an example embodiment of the invention as compared to the variable-M trelHs decoding procedure without antenna diversity.
DETAILED DESCRIPTION OF THE DRAWINGS
In the foUowing description, for purposes of explanation and not limitation, specific details are set forth, such as particular circuit arrangements, techniques, etc. and in order to provide a thorough understanding of the present invention. However, it wiH be apparent to one skilled in the art that the present invention may be practiced in other embodiments that depart from these specific details. In other instances, detailed description of weU-known methods, devices, and circuits are omitted so as not to obscure the description of the present invention with unnecessary detail.
Referring to Fig. 2, an overaH function block diagram of an example communications system 10 which may employ the present invention is shown. While the present invention is particularly weU suited to radio communications including both fixed and portable radios, it is appHcable to other communications environments as weU. On the transmit side, the information signal to be transmitted, i.e., a sequence of symbols, is encoded by a trelHs code modulation (TCM) and differential encoder 12. If the encoded symbols are being transmitted on a slotted communications channel, for example, using weU-known time division multiplex (TDM) techniques, a symbol interleaver 14 is desirable (although not necessary) to break up or "shuffle" symbol sequence segments into different time slots. Interleaving is particularly effective in communication channels subject to fading because symbol segments received under poor channel conditions are interspersed with symbol segments received under good channel conditions. The symbols may be complex and are represented by two components, namely a real value I (in-phase) and an imaginary value (quadrature). The in-phase and quadrature components are passed through paraUel processing branches. Each component is passed through a transmit filter 18a, 18b, converted in digital- to- analog (D/A) converters 20a, 20b, and frequency shifted by quadrature modulators 22a, 22b. The quadrature modulated signals are then mixed in summer 24 and sent to an RF amplifier 26 which increases the gain of the signal. The RF amplifier 26 passes the amplified signal to one or more antennas 28a...28n for transmission. Conventional frequency hopping techniques may be used to "hop" between multiple carrier frequencies with each frequency carrying only a portion of the symbols.
One or more antennas 30A...30N at the receiver side of the communications system 10 receive the signal. In a receive antenna diversity appHcation, diversity antennas 30A...30N may be incorporated in fixed radio transceivers such as base station transceivers/repeaters and in portable/mobile radios. The received signals from each of the plural diversity antennas 30A...30N are processed using respective radio amplification and downconversion paths. However, for simpHcity, only a single downconversion path is illustrated and described in Fig. 2.
One antenna signal (or a combination of signals from the diversity antennas) passes through a filtering and RF preamplifier stage 32 to a first downconverter 34 which reduces the frequency (or frequencies in a frequency hopping system) of the received signal to an intermediate frequency. The intermediate frequency signal is passed through an intermediate frequency receive filter 36, then through a second downconverter 38 to produce a (low pass) filtered baseband signal. Although the IF downconversion is desirable, it is of course not necessary to implement the present invention. The filtered baseband signal is then converted into a complex signal having in-phase (I) and quadrature (Q) components by signal processor 39. The in-phase and quadrature signals are digitized by analog-to-digital (A D) converters 40a, 40b and passed through complex signal generator 42 which converts differential phase information into a sequence of symbols. Deinterleaver 44 reconstitutes interleaved slots of information (i.e., it performs the reverse shuffling operation to undo the shuffling operation performed by in terleaver 14). TCM decoder 46 processes this symbol stream using a trellis decoding technique described in further detail below to recover the transmitted signal information. Signal tracker 48 estimates an average faded signal strength of an analog signal output from the IF stage 38 or from samples of the received signal from the deinterleaver 44. In a preferred embodiment, deinterleaver 44, TCM decoder 46, and signal tracker 48 are implemented using a suitably programmed microprocessor and/or digital signal processing circuitry.
The TCM decoder 46 is a maximum likelihood sequence estimation (MLSE) trelHs decoder that develops a trelHs of nodes and branches (similar to that shown in Fig. 1 and described in the background of the invention) for decoding sequences of symbols received on the communications channel. Each set of nodes in the trelHs represents aH possible values of a received symbol at one time instant. Transition between a node or state at one time instant to a node or state at another time instant is referred to as one stage of the trellis. Each stage typicaUy includes branches with each branch defining a specific transition between nodes at sequential time instances, and each branch has an associated branch weight or metric. Only the "survivor" branch into each node having the best (smaUest) metric is retained. A linkage of branches between nodes over multiple stages in the trelHs defines a path through the trelHs that represents a possible received symbol sequence. For each possible trelHs path, the branch metrics associated with that path are accumulated or summed to provide a path metric. The path having the lowest accumulated metric is selected as the best path.
The present invention determines which of the survivor paths through the trelHs wiH be maintained for purposes of decoding based on a current condition of the communications system. The other remaining survivor paths are discarded to simplify the decoding operation. In other words, a variable number of M survivor paths upon which trelHs decoding of the received sequence of symbols is based depends on a current condition of the communications system. Fig. 3 outlines in flowchart form the general procedures for varying the number of M survivor paths to be used in the trelHs decoding. The current communications system condition is determined in step 50. Various examples of different communications system conditions wiH be described in further detail below. In decision block 52, a decision is made whether the current communications system condition is greater than or equal to a predetermined threshold. If it is, a lower or smaH value of M is selected, or alternatively, the current M value is reduced (block 54). Otherwise, a larger (or large) value of M is selected, or alternatively, the current value of M is increased (block 56). Once the current value of M is determined, trelHs decoding using, for example, the weU-known Viterbi algorithm is performed to determine the best estimate of the received symbol sequence. In essence, if the current communications system condition is determined to be sufficiently optimal, which is determined in this example by comparison with a threshold, the number of states or paths needed to satisfactorily (in terms of low bit error rates) decode the received symbol sequence is relatively small. On the other hand, if the current communications system condition is less than the threshold, a larger number of M is used to ensure satisfactory decoding performance.
One example of a useful communications system parameter is the quaHty of the communications channel. As already described above, the quaHty of any communications channel varies over time to some degree. In radio communications in particular, the quaHty of the communications channel changes rapidly (especiaHy when the radio user is moving in a car) with the communications channel being subjected to fading, multipath dispersion, adjacent channel interference and co- chan el interference from other users, noise, and other channel impairments. Hi this situation, the current quaHty of the communications channel is detected or determined signal tracker 48. If the current quaHty of the channel is greater than the threshold, meaning that the channel quaHty is good, a first relatively smaH value of M is selected. Alternatively, if the current quaHty of the channel is less than a threshold value, indicating poor channel quaHty, another relatively large value of M is selected greater than the value of M 20091
17
selected if the channel quaHty is good. H practice, there are minimum Main) and maximum (Mmaτ) limits on the value of M selected depending upon the particular appHcation.
Channel quaHty can be measured using a number of 5 conventional channel quaHty indicators such as received signal strength, signal-to-noise ratio (SNR), signal to interference ratio (SIR), bit error rate (BER), etc. One preferred example indicator is averaged faded signal strength (AFSS) which is a smoothed estimate of the channel a pHtude of the received signal which may be obtained using 10 a channel tracking algorithm. Preferably, the channel tracking algorithm low-pass filters received signal samples to provide an estimate of the envelope of the ampHtude of the received signal. More specificaHy, in a differentiaUy encoded Phase Shift Keying (PSK) system, s(n), cn, and η(n) represent a (complex baseband) transmitted 15 symbol, a complex channel gain, and additive Gaussian noise, respectively, at time n. If y(n) is the received signal, and r(n) is the signal at the output of the complex symbol generator 42, then y(n) and r(n) may be expressed as foUows:
y(π) = Cnsin) + ηn (1) r(π) = y{n)y'{n - 1) = [cns( ) + ηn] [cn_ιs(π - 1) + Tjn-i]" , (2)
= Cnc .! s{n) s'{n - I) + cn'_lS * {n - l)η(n) 4- cns{n)η'(n - 1) + ηnηn"-.^
r(π) = αne*^ + η'n, (3)
25 where Δφn is the differential phase angle at time n (satisfying the relation s(n) i = s =(_"n - 1 -)) e c*Λ*n., a annda a^ - - r c^ r c* n l - 1 I cn ι I2 is_, a measure o ef the energy of the faded signal. In a fading channel, the channel gain c varies with time. The AFSS algorithm obtains the average faded signal strength En (an averaged value of an), and tracks its variation over time by executing the foUowing equations:
Δ Λ = min |r(π) - an- έ''t"' |2 , k = 0, 1, • • • (all possible transmitted phase angles) (4) a'n = Real [r(π)e-J'Δ**] (5) α„ = 2n-ι + (1 - 7) a'n (6) where γ is a real number in the range (0,1). The value of γ controls the extent of smoothing (to minimize the effect of noise). A typical
10 value is γ = 0.8.
Fig. 4 Olustrates in flowchart diagram format one preferred example embodiment for implementing the present invention where the communications system parameter is channel quaHty. H_iti.aHzati.on procedures are performed in block 100 to select M . .
15 M , the channel quaHty threshold value T, a poor channel holding window Δ, and a retrace stage depth δ . The initiaHzation parameters M . , Mmaγ, T, Δ, and δ are chosen depending upon the particular communications appHcation and can often be effectively determined using simulation tests. The current channel quaHty is determined in
20 block 102 using for example an AFSS procedure as described above. One stage of the M state trelHs decoding procedure is performed foUowing the weU-known Viterbi decoding procedures with M set equal to M . (block 104). A decision is made in block 106 whether the current channel quaHty is greater than or equal to the channel
25 quaHty threshold T. If it is, the decoding procedure is continued with M set to its minimum value M . thereby reducing the complexity of the trelHs decoding operation. However, if the current channel quaHty is less than the channel quaHty threshold T, the value of M is set to its maximum value M max (block 108).
Rather than continuing on in the trelHs decoding operation assuming that the just recently decoded stages of the trelHs were accurately decoded, the present invention takes into account the fact that these just recently decoded stages of the trelHs were likely to have been adversely impacted by the newly detected poor channel quaHty. Therefore, the decoding operation is repeated for a number δ of last stages by retracing that number δ of stages in the trelHs and restarting the trelHs decoding algorithm (block 110). This retrace procedure ensures a high degree of reHabiHty in decoding accuracy.
Control proceeds to block 112, where Δ stages of the M-§£age"~ σ trelHs decoding algorithm are performed with M set equal to maχ. Another decision is made at block 114 to determine whether the current channel quaHty is less than the threshold T within the Δ stages. If it is, the M state decoding procedure is continued at block 112 with M set at M max . However, if the current channel quaHty is greater than or equal to the threshold T, a new, smaUer value of M is generated in block 116 reflecting the current trend that reduced decoding complexity may be appropriate in view of the recently improved channel quaHty. But rather than immediately setting the new value of M to M . , the present invention employs a more conservative "pruning" procedure. Namely, the value of M is halved in each successive stag °e of the trelHs until it reaches M τn ._n . Thus, ' when the channel improves, the present invention employs the poor channel holding window Δ and the controUed M-reduction or pruning procedure to ensure that the channel condition remains at the improved level before M is set to M . . The M state decoding procedure is performed with the new smaller value of M in block 118 being reduced again, if possible, in block 116 until it reaches M . (block 120). Control returns to block 102 to repeat the above-described operation recursively decoding received symbols at the lowest, optimal level of decoding complexity to ensure satisfactory decoding performance.
Fig. 5 illustrates the performance of the present invention with reduced complexity as compared to the traditional full complexity Viterbi algorithm. Bit error rate (BER) is appHed on the vertical axis with the signal-to-noise ratio (Eb /No [energy per bit to noise power spectral density]) plotted in dB on the horizontal axis. In this practical iHustration, the radio receiver is moving in a vehicle at a speed of 150 KMPH. The fuH Viterbi search corresponds to a 64-state (M=64) demodulation scheme (indicated by asterisks), while the graph of the decoding scheme in accordance with the present invention uses a variable M (indicated by plus signs). Virtually the same decoding performance was achieved by both procedures. However, the present invention accompHshed this decoding using an average of 10 and 11 states (M=10-ll) as compared to the fuH 64-state Viterbi algorithm. Thus, the complexity of the variable M algorithm employed by the present invention was six times less complex when measured in terms of number of operations executed to decode the data.
Other example embodiments using different communications system conditions are now described. For example, the communications system condition may relate to the complexity of the encoding scheme used by the transmitter TCM encoder 12. Codes with more memory potential (sometimes referred to as "constraint length" in the Hterature) exhibit better performance. However, since decoding complexity increases exponentiaHy with memory of the code, these codes cannot be easily used due to data processing constraints. A smaHer value of M may be varied in this instance to adapt the decoding procedure to a level best suited given a particular set of codes.
Decoding procedures are performed using data processing circuitry, and oftentimes the data processing circuitry must perform a number of other tasks in a time share relationship. Another communications system condition relates therefore to the current data processing tasks to be performed by the data processing circuitry. The value of M is decreased if the current number of additional tasks to be performed is relatively large. In this way, the data processing circuitry is reHeved of the burden of a complex decoding process so that it may devote more processing resources to perform the other large number of pending tasks. On the other hand, the value of M can be increased, thereby improving the performance of the decoder, if the current number of additional pending tasks is relatively smaH.
When decoder performance is of primary concern, the value of M can be varied to maintain a predetermined level of decoder performance in response to changes in the quaHty of the communications channel. When M is relatively low, the fuH capacity of the data processing circuitry is not consumed by the operation of the trelHs decoder. As a result, the excess capacity of the data processing circuitry can be devoted to other tasks having lower priority. In the context of portable, battery-operated receivers, the value of M may be decreased whenever consistent with decoder performance objectives in order to reduce the number of data processing tasks associated with the trelHs decoding operation, and thereby, reduce battery drain. The present invention may also be appHed to a spread spectrum type receiver in which plural encoded sequences of symbols transmitted by plural transmitters are received on a single communications channel. The value of M may be changed in response to changes in the number of interfering signals. When satisfactory performance of the decoder is degraded because of additional interfering signal, the value of M can be increased to maintain or improve decoder performance. On the other hand, when the number of interfering signals decreases, the value of M can be decreased to the smaHest value of M consistent with satisfactory decoder performance in order to conserve processor or battery resources.
Another example embodiment of the present invention using diversity reception is now described in detail. As shown in Fig. 2, the receiver may have multiple antennas and obtain relatively uncorrelated copies of the received signal from each antenna. The antennas are rendered uncorrelated by separating them spatiaUy or using orthogonal polarizations. The receiver combines these uncorrelated copies of the received signal in a manner to improve signal quaHty, i.e., selecting the strongest signal based on average faded signal strength, co-phasing the signals to perform maximal- ratio-combining or equal gain combining, adding the branch metrics calculated for each of the multiple antennas, and adding the received signals to suppress interference, etc. If there are N antennas and K interferers and K < N, then interference can be reduced. All these methods improve signal quaHty, aUow the use of a smaH value of M, and reduce the variance of signal quaHty aUowing using a smaHer rang °e of M max' , M τnι .n . Fig. 6 is a function block diagram which shows in more detail the diversity antenna structure and one way of processing the signals received by each of the antennas in accordance with one example embodiment of the present invention. Signals received on each diversity antenna 30A...30N are treated as a separate channel, i.e., channel 1... channel N. Each channel is processed using its own dedicated front end processing circuitry 48A - 48N which performs such conventional tasks as RF preamplification and frequency downconversion. Once the signals are converted to baseband, they are stored in a memory 50 which individuaHy stores each baseband signal in a respective buffer 50A...50N. Signal processor 39 in Fig. 2 includes selection logic 52 which determines a signal quaHty indicator for each buffered signal. The selection logic 52 then selects the buffered received signal having the highest signal quaHty indicator for further processing and decoding in the TCM decoder 46. Alternatively, selection logic 52 may selectively weight the buffered signals based on their signal quaHty indicator and then combine the weighted signals into a combined signal. The buffered signal corresponding to the diversity antenna having the highest signal quaHty is most heavily weighted which ensures that the combined signal substantiaHy reflects the best signal quaHty for that particular time instant.
Fig. 7 outlines in flowchart form the general procedures for varying the number of M survivor paths to be used in the trelHs decoding based on plural received signals by plural diversity antennas. The baseband signals from each diversity antenna are sampled and then buffered in step 60. The channel quaHties for each demodulated sample corresponding to each of the pluraHty of diversity antennas are then determined in step 62. As described above, the buffered sample with the best channel quaHty measure is selected, or alternatively, more heavily weighted in step 64. A variable-M trellis decoding procedure is then performed in step 66 in accordance with the procedures outlined in the flowcharts iHustrated in Figs. 3 and 4 and described above.
Accordingly, even further reductions in complexity corresponding to an even further reduced value of M are achieved using diversity antenna reception in accordance with the present invention as described above. Again, bit error rate is plotted on the vertical axis of the graph with signal-to-noise ratio (Eb/No) plotted in dB on the horizontal axis. And again, the radio receiver is moving in a vehicle at a speed of 150 KMPH. The variable-M decoding scheme without diversity reception is plotted using plus signs. The plot of the variable-M decoding procedure using diversity reception (indicated by asterisks) reveals that using an average of 5 to 7 states (M=5-7), approximately an additional 50% reduction in decoding complexity is achieved. Moreover, relative to the full 64-state Viterbi algorithm, the variable-M decoding procedure with diversity reduces the decoding complexity by a factor of 11. Thus, using diversity reception in conjunction with a variable-M decoding procedure, the present invention adapts the complexity of the trelHs decoding procedure in accordance with the current channel conditions. This approach is not only more efficient in the sense of reduced decoding complexity, it also lowers the required transmitted power to achieve a fixed bit error rate performance. In essence, the value of M is adapted with the channel condition as reflected in the received channel quaHty indicator. The value of M is increased when the channel condition is poor, and the value of M is decreased when the channel condition is good. The complexity of the variable-M decoding procedure is proportional to the ratio of time during which the channel condition is good relative to when the channel condition is poor. The present invention improves this ratio using antenna diversity, and as a result, lowers the transmit power required to achieve a given bit error rate by the receiver decoder relative to use of the same variable M decoding procedure without antenna diversity. Lowering transmit power reduces battery drain for transmitting portable radio transceivers and reduces adjacent channel interference caused by transmitters in fixed, base station type transceivers. For a given transmit power, the present invention reduces decoding complexity and therefore reduces battery drain by the receiver or improves decoding performance in terms of reducing the number of bit errors without increasing battery drain by the receiver. Two further example appHcations of the present invention are briefly described. First, multiple diversity transmitting antennas (see antennas 28a...28n) may be used to transmit the same information to the receiver to improve signal quaHty. For example, the antenna having the better path to the receiver can be selected by the receiver based on the received signal on each antenna. Transmission along a better path offers better signal quaHty, and hence, a lower value of M can be employed.
A second further example appHcation of the present invention relates to symbol interleaving and/or frequency hopping. Interleaving is used to "break-up" channel memory. If a sequence of symbols is sent and the symbols are highly correlated in time due to channel memory, decoding performance wiH be poor whenever the channel condition is poor because aU symbols are affected. However, if the symbols are "shuffled" such that they are separated by other unrelated symbols, they are less likely to be equaUy affected by channel memory. Symbol interleaving when used with frequency hopping is particularly effective in improving signal quaHty because, in addition to separating symbols in time, those symbols are also separated in frequency.
Groups of time-separated symbols are transmitted at different caπier frequencies. At the receiver, the symbols are suitably deinterleaved and sent to the variable M-algorithm for decoding. The improved signal quaHty due to interleaving (in both time and frequency) aHows the use of a low value of M.
While the invention has been described in connection with what is presently considered to be the most practical and preferred embodiment, it is to be understood that the invention is not to be limited to the disclosed embodiment, but on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims. For example, although the present invention is described in terms of post-detection diversity where the buffered samples were selected after the signals from the diversity antennas were demodulated to baseband, the signal selection could be performed using predetectioii/ demodulation diversity, i.e., at RF or IF frequencies. Although the invention is described for trelHs coded modulation, it can be used in demodulating block coded modulation, convolutional codes, block codes, partial response modulation such as continuous phase modulation and any set of signals that can be represented by a trelHs.

Claims

WHAT IS CLAIMED TS:
1. In a communications system, a method for decoding encoded symbols transmitted over a communications channel, comprising the steps of:
(a) developing a trelHs structure of nodes and branches for decoding a sequence of symbols received on the communications channel, each set of nodes in the trelHs representing possible values of a received symbol at one time instant, each branch defining a specific transition between nodes at different time instants and having an associated branch weight, wherein a linkage of branches between nodes in the trelHs defines a path through the trelHs that represents a possible symbol sequence;
(b) accumulating for each trelHs path an accumulated weight;
(c) determining a current condition of the communications system; (d) determining M survivor paths through the trelHs based on respective accumulated path weights for each trelHs path, the value of M depending on the condition of the communications system determined in step (c); and
(e) decoding the sequence of received symbols using the M survivor paths.
2. The method in claim 1, step (d) including: selecting a first value of M if the current communications system condition is greater than or equal to a threshold, and selecting a second value of M greater than the first value of M if the current status of the communications system condition is less than the threshold.
3. The method in claim 1, step (d) including: increasing the value of M if the current condition is less than a threshold, and decreasing the value of M if the current condition is greater than or equal to the threshold.
4. The method in claim 1, wherein the condition is a quaHty of the communications channel, said determining step (c) including: detecting the current quaHty of the communications channel, selecting a first value of M if the current quaHty of the communications channel is greater than or equal to a threshold, and selecting a second value of M greater than the first value of M if the current quaHty of the communications channel is less than the threshold.
5. The method in claim 4, wherein the current quaHty of the communications channel is determined based on a signal strength of the received signal.
6. The method in claim 5, wherein the communications channel is a fading channel and the signal strength corresponds to an average faded signal ampHtude of the received signal.
7. The method in claim 4, wherein the first value of M corresponds to a τm'niττnrm value of M and the second value of M corresponds to a maximum value of M, and wherein after selecting the minimum value of M and the current quaHty of the channel drops to less than the threshold, increasing the value of M in a controUed fashion.
8. The method in claim 3, wherein after decreasing the value of M and the current quaHty of the channel drops to less than the threshold, repeating steps (d) and (e) for the received sequence of symbols using an increased value of M.
9. The method in claim 7, wherein after increasing the value of M and the current quaHty of the channel improves to greater than the threshold, maintaining the value of M for one or more symbol periods before determining whether to decrease the value of M.
10. The method in claim 9, wherein after the one or more symbol periods, the value of M is halved in successive stages of the trelHs to a τm'τπτmrm value while the current condition of the communications channel continues to equal or exceed the threshold.
11. The method in claim 1, wherein the condition is a size of a coding memory of an encoding scheme used to encode symbols to be transmitted, step (d) including adjusting the value of M to decode symbols encoded using an encoding scheme based on the size of the coding memory.
12. The method in claim 1, wherein steps (a)-(e) are performed using data processing circuitry, the data processing circuitry performing additional other tasks in time-shared fashion, and wherein the condition is current data processing tasks to be performed by the data processing circuitry, step (d) including: decreasing the value of M if a current number of the additional other tasks to be performed is relatively large, and increasing the value of M if the current number of additional other tasks to be performed is relatively smaU.
13. The method in claim 1, wherein the communications system includes a portable, battery-operated receiver and steps (a)-(e) are performed using data processing circuitry, and wherein the condition is a quaHty of the communications channel, step (d) including: decreasing the value of M if the quaHty of the communications channel equals or exceeds a threshold to reduce battery drain, and increasing the value of M if the quaHty of the communications channel is less than said threshold to improve decoding performance.
14. The method in claim 1, further comprising: receiving a number of signals including encoded sequences of symbols transmitted from plural users on the communications channel at a single receiver, wherein the condition is the number of signals; increasing the value of M for a relatively large number of received signals; and decreasing the value of M for a relatively smaH number of received signals.
15. A communications system, comprising: a transmitter having an encoder for trelHs encoding a sequence of symbols and transmitting the trelHs encoded symbols over a communications channel; a receiver for receiving the trelHs encoded symbols transmitted over the communications channel and having a trellis decoder including electronic circuitry for performing the steps of:
(a) developing a trelHs structure of nodes and branches for decoding a sequence of symbols received on the communications channel, each set of nodes in the trelHs representing possible values of a received symbol at one time instant, each branch defining a specific transition between nodes at different time instants and having an associated branch weight, wherein a linkage of branches between nodes in the trelHs defines a path through the trelHs that represents a possible symbol sequence;
(b) accumulating for each trelHs path an accumulated weight;
(c) determining a current condition of the communications system; (d) deteπnining M survivor paths through the trelHs based on respective accumulated path weights for each trelHs path, the value of M depending on the condition of the communications system determined in step (c); and
(e) decoding the sequence of received symbols using the M survivor paths.
16. The system in claim 15, wherein the condition is a quaHty of the communications channel and the electronic circuitry detects the current quaHty of the communications channel, selects a first value of M if the current quaHty of the communications channel is greater than or equal to a threshold, and selects a second value of M greater than the first value of M if the current quaHty of the communications channel is less than the threshold.
17. The system in claim 15, wherein the electronic circuitry increases the value of M if the current condition is less than a threshold, and decreases the value of M if the current condition is greater than or equal to the threshold.
18. The system in claim 17, wherein after decreasing the value of M and the current quaHty of the channel drops to less than the threshold, the electronic circuitry repeats steps (d) and (e) for the received sequence of symbols using an increased value of M.
19. The system in claim 18, wherein after increasing the value of M and the current quaHty of the channel improves to greater than the threshold, the electronic circuitry maintains the value of M at its current value for one or more symbol periods before determining whether to decrease the value of M.
20. The system in claim 19, wherein after the one or more symbol periods, the electronic circuitry halves the value of M in successive stages of the trelHs to a minimum value while the current condition of the communications channel continues to equal or exceed the threshold.
21. The system in claim 15, wherein the condition is a size of a coding memory of an encoding scheme used to encode symbols to be transmitted, step (d) including adjusting the value of M to decode symbols encoded using an encoding scheme based on the size of the coding memory.
22. The system in claim 15, wherein said electronic circuitry performs additional tasks other than tasks related to performing steps (a)-(e) in time share fashion, and wherein the condition is current data processing tasks to be performed by the electronic circuitry, the electronic circuitry decreasing the value of M if a current number of the additional other tasks to be performed is relatively large, and increasing the value of M if the current number of additional other tasks to be performed is relatively smaH.
23. The system in claim 15, wherein the receiver includes a battery for portable battery operation and the electronic circuitry performs additional tasks other than tasks related to performing steps (a)-(e) and the condition is current data processing tasks to be performed by the electronic circuitry, the electronic circuitry decreasing the value of M if a current number of total tasks to be performed is relatively large to decrease power required by the electronic circuitry, and increasing the value of M if the current number of total tasks to be performed is relatively smaH.
24. The system in claim 15, further comprising plural transmitters, wherein the receiver receives a number of signals including encoded sequences of symbols transmitted from the plural transmitters on the communications channel, wherein the condition is the number of signals; increasing the value of M for a first number of received signals; and decreasing the value of M for second number of received signals smaHer than the first number.
25. In a communications system, a method for decoding encoded symbols transmitted over a communications channel, comprising the steps of:
(a) developing a trelHs structure of nodes and branches for decoding a sequence of symbols received on the communications channel, each set of nodes in the trelHs representing possible values of a received symbol at one time instant, each branch defining a specific transition between nodes of nodes sets at different time instants and having an associated branch weight, wherein a linkage of branches between nodes in the trelHs defines a path through the trelHs that represents a possible symbol sequence;
(b) accumulating for each trelHs path an accumulated weight;
(c) determining M survivor paths through the trelHs based on respective accumulated path weights for each path; (d) determining a current quaHty of the communications channel and comparing the channel quaHty to a threshold;
(e) if the compared channel quaHty equals or exceeds the threshold, setting the value of M to a relatively smaH value, performing one trelHs decoding stage of the sequence of received symbols, and returning to step (d); (f) if the compared channel quaHty is less than the threshold, setting the value of M to a relatively large value and performing one or more stages of trelHs decoding before returning to step (d).
26. The method in claim 25, wherein step (f) further comprises: retracing a preset number of stages in the trelHs before performing trelHs decoding.
27. The method in claim 26, further comprising: after retracing, performing a predetermined number of stages of trelHs decoding; checking whether the current channel quaHty is less than the threshold; and if the current channel quaHty is less than the threshold after performing the predetermined number of stages of trelHs decoding, performing a next predetermined number of stages of trelHs decoding.
28. The method in claim 25, wherein if after setting the value of M to the relatively large value, a current channel quaHty is determined to be greater than or equal to the threshold for a predetermined number of trelHs decoding stages, setting a new value of M less than the relatively large value.
29. The method in claim 28, wherein the new value of M is set to the larger of the relatively large value of M divided by 2 and a minimum value of M.
30. The method in claim 25, wherein the current quaHty of the communications channel is determined based on a signal strength of the received signal.
31. The method in claim 25, wherein the communications channel is a fading channel and the signal strength corresponds to an average faded signal strength of the received signal.
32. The method in claim 25, wherein the communications channel is slotted and the sequence of encoded symbols is divided into portions with different portions being interleaved into different slots on the communications channel before transmission, and wherein before step (a), the method includes deinterleaving the received sequence of symbols.
33. In a communications system, a method for decoding in a radio receiver encoded symbols transmitted over a communications channel, the radio receiver having plural antennas for receiving a transmitted signal including a sequence of coded symbols, comprising trelHs decoding the received signal using a variable number M of trelHs survivor paths or states where the variable number M is reduced by selecting a received signal from one of the plural antennas having a highest signal quaHty thereby reducing the complexity of the trelHs decoding procedure.
34. The method in claim 33, further comprising: detecting and storing signals received from each of the plural antennas; dete_πnining a channel quaHty indicator for each stored signal; selecting the stored signal with the highest channel quaHty indicator.
35. The method in claim 34, further comprising: decreasing the value of M if the highest signal quaHty is greater than or equal to a threshold, and increasing the value of M if the highest signal quaHty is less than the threshold.
36. The method in claim 34, further comprising: selecting a first value of M if a current highest signal quaHty is greater than or equal to a threshold, and selecting a second value of M greater than the first value of M if the current highest signal quaHty is less than the threshold.
37. The method in claim 36, wherein signal quaHty is determined based on a signal strength of the received signal.
38. The method in claim 33, further comprising:
(a) developing a trelHs structure of nodes and branches for decoding the sequence of symbols received on the co muni cations channel, each set of nodes in the trelHs representing possible values of a received symbol at one time instant, each branch defining a specific transition between nodes of nodes sets at different time instants and having an associated branch weight, wherein a linkage of branches between nodes in the trelHs defines a path through the trelHs that represents a possible symbol sequence; (b) accumulating for each trelHs path an accumulated weight; and
(c) determining the M survivor paths through the trelHs based on respective accumulated path weights for each trelHs path, wherein M is varied in accordance with a condition of the communications channel.
39. A communications system, comprising: a transmitter having an encoder for trelHs encoding a sequence of symbols and transmitting the trelHs encoded symbols over a communications channel; and a receiver having plural antennas for receiving transmitted signals including a sequence of trelHs encoded symbols transmitted over the communications channel and a trelHs decoder including electronic circuitry for trelHs decoding a received signal using a variable number M of trelHs survivor paths or states and optimally varying the variable number M by selecting the signal from one of the plural antennas having a highest signal quaHty to reduce the complexity of the trellis decoding procedure.
40. The communications system in claim 39, wherein the signal quaHty is determined based on a signal strength of the received signal.
41. The communications system in claim 39, wherein the receiver determines a current signal quaHty condition of the communications channel for signals received by each of the antennas and the trelHs decoder varies M based on the determined current condition for the selected signal.
42. The communications system in claim 41, wherein the trelHs decoder decreases M if the highest signal quaHty is greater than or equal to a threshold, and increases M if the highest signal quaHty is less than the threshold.
43. La a communications system, a method for decoding in a receiver encoded symbols transmitted over a communications channel, the receiver having plural antennas for receiving a transmitted signal including a sequence of coded symbols, comprising trelHs decoding the received signal using a variable number M of trelHs survivor paths or states where the variable number M is optimaUy varied by selectively weighting and then combining the received signal from the plural antennas to generate a combined signal such that an antenna signal having the highest signal quaHty is the most heavily weighted in the combination to thereby reduce the complexity of the trelHs decoding procedure.
44. The method in claim 43, further comprising: decreasing the value of M if an overaU signal quaHty of the combined signal is greater than or equal to a threshold, and increasing the value of M if the highest signal quaHty is less than the threshold.
45. The method in claim 43, further comprising: selecting a first value of M if an overaU signal quaHty of the combined signal is greater than or equal to a threshold, and selecting a second value of M greater than the first value of M if the overaH signal quaHty of the combined signal is less than the threshold.
46. In a communications system, a method comprising: encoding a sequence of symbols; separating portions of the sequence of encoded symbols in time or frequency; transmitting the separated portions; receiving the separated portions and combining the separated portions into the sequence of symbols; and trelHs decoding the received signal using a variable number M of trelHs survivor paths or states where the variable number M is reduced as a resulting of the separating step.
47. The method in claim 46, wherein the separating step includes separating the portions in both time and frequency.
PCT/US1996/020091 1995-12-27 1996-12-23 Method and apparatus for symbol decoding using a variable number of survivor paths WO1997024849A1 (en)

Priority Applications (5)

Application Number Priority Date Filing Date Title
EP96944428A EP0872095B1 (en) 1995-12-27 1996-12-23 Method and apparatus for symbol decoding using a variable number of survivor paths
JP9524372A JP2000502859A (en) 1995-12-27 1996-12-23 Code decoding method and apparatus using variable number of surviving paths
DE69635085T DE69635085D1 (en) 1995-12-27 1996-12-23 ARRANGEMENT AND METHOD FOR SYMBOL DECODING WITH VARIABLE NUMBER OF SURVIVOR PATHS
AU14238/97A AU716138B2 (en) 1995-12-27 1996-12-23 Method and apparatus for symbol decoding using a variable number of survivor paths
CA002241691A CA2241691C (en) 1995-12-27 1996-12-23 Method and apparatus for symbol decoding using a variable number of survivor paths

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US08/579,349 US5905742A (en) 1995-12-27 1995-12-27 Method and apparauts for symbol decoding
US08/579,349 1995-12-27

Publications (2)

Publication Number Publication Date
WO1997024849A1 true WO1997024849A1 (en) 1997-07-10
WO1997024849A9 WO1997024849A9 (en) 1997-11-20

Family

ID=24316537

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US1996/020091 WO1997024849A1 (en) 1995-12-27 1996-12-23 Method and apparatus for symbol decoding using a variable number of survivor paths

Country Status (8)

Country Link
US (1) US5905742A (en)
EP (1) EP0872095B1 (en)
JP (1) JP2000502859A (en)
KR (1) KR100367812B1 (en)
CN (1) CN1211366A (en)
AU (1) AU716138B2 (en)
DE (1) DE69635085D1 (en)
WO (1) WO1997024849A1 (en)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO1999004537A1 (en) 1997-07-21 1999-01-28 Ericsson Inc. Determination of the length of a channel impulse response
GB2328592A (en) * 1997-06-06 1999-02-24 Nec Corp Viterbi decoder having a variable number of states
WO1999023766A2 (en) * 1997-10-31 1999-05-14 At & T Wireless Services, Inc. Maximum likelihood detection of concatenated space-time codes for wireless applications with transmitter diversity
EP1152576A1 (en) * 2000-05-05 2001-11-07 Lucent Technologies Inc. Joint estimation using the M-algorithm or T-algorithm in multiantenna systems
US6501803B1 (en) 1998-10-05 2002-12-31 At&T Wireless Services, Inc. Low complexity maximum likelihood detecting of concatenated space codes for wireless applications
US6741635B2 (en) 1997-12-23 2004-05-25 At&T Wireless Services, Inc. Near-optimal low-complexity decoding of space-time codes for fixed wireless applications
US6775329B2 (en) 1997-09-16 2004-08-10 At&T Wireless Services, Inc. Transmitter diversity technique for wireless communications
EP1510010A1 (en) * 2002-05-31 2005-03-02 Nokia Corporation Method and arrangement for enhancing search through trellis
KR100710743B1 (en) * 1998-05-26 2007-04-24 코닌클리케 필립스 일렉트로닉스 엔.브이. Transmission system having a simplified channel decoder and operating method thereof
US7274752B2 (en) 1998-09-17 2007-09-25 Cingular Wireless Ii, Llc Maximum ratio transmission
US7515659B2 (en) 2001-05-04 2009-04-07 Agere Systems Inc. Decoding techniques for multi-antenna systems

Families Citing this family (33)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0931388B1 (en) 1996-08-29 2003-11-05 Cisco Technology, Inc. Spatio-temporal processing for communication
US6192503B1 (en) * 1997-08-14 2001-02-20 Ericsson Inc. Communications system and methods employing selective recursive decording
US6215827B1 (en) * 1997-08-25 2001-04-10 Lucent Technologies, Inc. System and method for measuring channel quality information in a communication system
US6151370A (en) * 1998-02-12 2000-11-21 Lucent Technologies Inc. Path-oriented decoder for signal-dependent noise
US6654365B1 (en) * 1998-02-24 2003-11-25 L-3 Communications Corporation Reduced complexity trellis-based multiuser detector for CDMA
US6408418B1 (en) * 1998-10-29 2002-06-18 Lucent Technologies Inc. Reduced-state device and method for decoding data
JP3519291B2 (en) * 1998-11-06 2004-04-12 松下電器産業株式会社 OFDM communication apparatus and method
US6618451B1 (en) 1999-02-13 2003-09-09 Altocom Inc Efficient reduced state maximum likelihood sequence estimator
US6597743B1 (en) 1999-12-07 2003-07-22 Ericsson Inc. Reduced search symbol estimation algorithm
DE19959409A1 (en) * 1999-12-09 2001-06-21 Infineon Technologies Ag Turbo code decoder and turbo code decoding method with iterative channel parameter estimation
US6788750B1 (en) * 2000-09-22 2004-09-07 Tioga Technologies Inc. Trellis-based decoder with state and path purging
JP2003141820A (en) * 2001-11-01 2003-05-16 Fujitsu Ltd Data reproducing device
FI111887B (en) * 2001-12-17 2003-09-30 Nokia Corp Procedure and arrangement for enhancing trellis crawling
US6704376B2 (en) * 2002-01-23 2004-03-09 Bae Systems Information And Electronic Systems Integration Inc. Power and confidence ordered low complexity soft turbomud with voting system
US7986672B2 (en) * 2002-02-25 2011-07-26 Qualcomm Incorporated Method and apparatus for channel quality feedback in a wireless communication
KR100859865B1 (en) * 2002-05-28 2008-09-24 삼성전자주식회사 OFDM Equalizer capable of equalizing adaptively according to the channel state
JP4490265B2 (en) * 2002-07-24 2010-06-23 ビーエイイー・システムズ・インフォメーション・アンド・エレクトロニック・システムズ・インテグレイション・インコーポレーテッド Co-channel interference receiver
WO2004086762A1 (en) * 2003-02-11 2004-10-07 Electronics And Telecommunications Research Institute Decision feedback equalizer for digital tv and method thereof
HUE043293T2 (en) 2003-08-20 2019-08-28 Panasonic Corp Radio communication apparatus and subcarrier assignment method
DE10351051A1 (en) * 2003-10-31 2005-06-09 Infineon Technologies Ag Verification method for a bit vector sent by a mobile telephone to a base station with two antennas uses weighting factors for amplitudes and/or phases for signals to be transmitted by the antennas
US7460583B2 (en) * 2003-12-15 2008-12-02 Telefonaktiebolaget Lm Ericsson (Publ) Method for path searching and verification
US7684521B2 (en) * 2004-02-04 2010-03-23 Broadcom Corporation Apparatus and method for hybrid decoding
JP2005311717A (en) * 2004-04-21 2005-11-04 Matsushita Electric Ind Co Ltd Decoding device and receiver of communication system
US7499452B2 (en) * 2004-12-28 2009-03-03 International Business Machines Corporation Self-healing link sequence counts within a circular buffer
GB0504483D0 (en) * 2005-03-03 2005-04-13 Ttp Communications Ltd Trellis calculations
US7676736B2 (en) * 2006-09-13 2010-03-09 Harris Corporation Programmable continuous phase modulation (CPM) decoder and associated methods
US7657825B2 (en) * 2006-09-13 2010-02-02 Harris Corporation Programmable trellis decoder and associated methods
JP4482835B2 (en) 2008-03-14 2010-06-16 ソニー株式会社 Data processing apparatus, data processing method, and program
US10075195B2 (en) * 2014-08-29 2018-09-11 Samsung Electronics Co., Ltd. Electronic system with Viterbi decoder mechanism and method of operation thereof
CN104537202B (en) * 2014-10-31 2017-12-22 哈尔滨工业大学深圳研究生院 Space antenna array synthetic method based on satellites formation cooperation
US9385896B1 (en) * 2015-07-09 2016-07-05 Huawei Technologies Co., Ltd. Method and apparatus for low-complexity quasi-reduced state soft-output equalizer
WO2018014738A1 (en) * 2016-07-22 2018-01-25 深圳超级数据链技术有限公司 Fast decoding method and device suitable for ovxdm system, and ovxdm system
CN107645360B (en) * 2016-07-22 2022-02-18 深圳汇思诺科技有限公司 OvXDM system decoding method and device and OvXDM system

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0441731A1 (en) * 1990-02-06 1991-08-14 ETAT FRANCAIS représenté par le Ministre des P.T.T. (Centre National d'Etudes des Télécommunications-CNET) Method of data broadcasting using time-frequency interleaving and coherent demodulation
EP0443997A1 (en) * 1990-02-16 1991-08-28 Telefonaktiebolaget L M Ericsson A method of reducing the influence of fading of a Viterbi-receiver having at least two antennas
EP0457460A2 (en) * 1990-05-17 1991-11-21 Orbitel Mobile Communications Limited Spacediversity switching receiver
EP0496152A2 (en) * 1991-01-24 1992-07-29 Roke Manor Research Limited Viterbi equaliser with variable number of states
EP0691770A2 (en) * 1994-07-07 1996-01-10 Nec Corporation Maximum-likelihood sequence estimator with variable number of states

Family Cites Families (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5020056A (en) * 1989-05-02 1991-05-28 General Electric Company Reduction of synchronous fading effects by time hopping of user slots in TDMA frames
US5202901A (en) * 1991-05-21 1993-04-13 General Electric Company Digital discriminator for pulse shaped π/4 shifted differentially encoded quadrature phase shift keying
US5151926A (en) * 1991-05-21 1992-09-29 General Electric Company Sample timing and carrier frequency estimation circuit for sine-cosine detectors
US5283815A (en) * 1991-05-21 1994-02-01 General Electric Company Tangental type differential detector for pulse shaped PI/4 shifted differentially encoded quadrature phase shift keying
US5349589A (en) * 1991-07-01 1994-09-20 Ericsson Ge Mobile Communications Inc. Generalized viterbi algorithm with tail-biting
US5249205A (en) * 1991-09-03 1993-09-28 General Electric Company Order recursive lattice decision feedback equalization for digital cellular radio
US5177740A (en) * 1991-09-03 1993-01-05 General Electric Company Frame/slot synchronization for U.S. digital cellular TDMA radio telephone system
US5283811A (en) * 1991-09-03 1994-02-01 General Electric Company Decision feedback equalization for digital cellular radio
US5311552A (en) * 1992-06-15 1994-05-10 General Electric Company Trellis coding technique to improve adjacent channel interference protection ratio in land mobile radio systems
US5311553A (en) * 1992-06-15 1994-05-10 General Electric Company Trellis coding technique to increase adjacent channel interference protection ratio in land mobile radio systems under peak power constraints
US5363407A (en) * 1992-09-02 1994-11-08 General Electric Company Transmitter optimization for spectrally congested radio communication systems
AU5550694A (en) * 1992-11-06 1994-06-08 Pericle Communications Company Adaptive data rate modem
US5343498A (en) * 1993-03-08 1994-08-30 General Electric Company Sample timing selection and frequency offset correction for U.S. digital cellular mobile receivers
US5400362A (en) * 1993-03-29 1995-03-21 General Electric Company Double sided slot traversing decoding for time division multiple access (TDMA) radio systems
US5351274A (en) * 1993-08-20 1994-09-27 General Electric Company Post detection selection combining diversity receivers for mobile and indoor radio channels
US5406593A (en) * 1993-08-20 1995-04-11 General Electric Company Adaptive phase-locked loop employing channel state information estimation from received signal phase angles
US5371471A (en) * 1993-10-29 1994-12-06 General Electric Company Low complexity adaptive equalizer radio receiver employing direct reference state updates
US5586128A (en) * 1994-11-17 1996-12-17 Ericsson Ge Mobile Communications Inc. System for decoding digital data using a variable decision depth

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0441731A1 (en) * 1990-02-06 1991-08-14 ETAT FRANCAIS représenté par le Ministre des P.T.T. (Centre National d'Etudes des Télécommunications-CNET) Method of data broadcasting using time-frequency interleaving and coherent demodulation
EP0443997A1 (en) * 1990-02-16 1991-08-28 Telefonaktiebolaget L M Ericsson A method of reducing the influence of fading of a Viterbi-receiver having at least two antennas
EP0457460A2 (en) * 1990-05-17 1991-11-21 Orbitel Mobile Communications Limited Spacediversity switching receiver
EP0496152A2 (en) * 1991-01-24 1992-07-29 Roke Manor Research Limited Viterbi equaliser with variable number of states
EP0691770A2 (en) * 1994-07-07 1996-01-10 Nec Corporation Maximum-likelihood sequence estimator with variable number of states

Non-Patent Citations (8)

* Cited by examiner, † Cited by third party
Title
AFTELAK S B ET AL: "Adaptive reduced-state Viterbi algorithm detector", JOURNAL OF THE INSTITUTION OF ELECTRONIC AND RADIO ENGINEERS, MAY 1986, UK, vol. 56, no. 5, ISSN 0267-1689, pages 197 - 206, XP000670561 *
ANDERSON J B ET AL: "Sequential coding algorithms: a survey and cost analysis", IEEE TRANSACTIONS ON COMMUNICATIONS, FEB. 1984, USA, vol. COM-32, no. 2, ISSN 0090-6778, pages 169 - 176, XP000670570 *
ANDERSON J B: "Limited search trellis decoding of convolutional codes", IEEE TRANSACTIONS ON INFORMATION THEORY, SEPT. 1989, USA, vol. 35, no. 5, ISSN 0018-9448, pages 944 - 955, XP000100926 *
BENELLI G ET AL: "Some digital receivers for the GSM pan-European cellular communication system", IEE PROCEEDINGS-COMMUNICATIONS, JUNE 1994, UK, vol. 141, no. 3, ISSN 1350-2425, pages 168 - 176, XP000438038 *
CHANG K C ET AL: "An adaptive reduced-state channel equalizer with T-algorithm", VTC 1994. 'CREATING TOMORROW'S MOBILE SYSTEMS'. 1994 IEEE 44TH VEHICULAR TECHNOLOGY CONFERENCE (CAT. NO.94CH3438-9), PROCEEDINGS OF IEEE VEHICULAR TECHNOLOGY CONFERENCE (VTC), STOCKHOLM, SWEDEN, 8-10 JUNE 1994, ISBN 0-7803-1927-3, 1994, NEW YORK, NY, USA, IEEE, USA, pages 1237 - 1240 vol.2, XP000497617 *
COX R V ET AL: "An efficient adaptive circular Viterbi algorithm for decoding generalized tailbiting convolutional codes", IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, FEB. 1994, USA, vol. 43, no. 1, ISSN 0018-9545, pages 57 - 68, XP000450947 *
SIMMONS S J: "Breadth-first trellis decoding with adaptive effort", IEEE TRANSACTIONS ON COMMUNICATIONS, JAN. 1990, USA, vol. 38, no. 1, ISSN 0090-6778, pages 3 - 12, XP000102576 *
WEN-TA LEE ET AL: "A single-chip Viterbi decoder for a binary convolutional code using an adaptive algorithm", IEEE TRANSACTIONS ON CONSUMER ELECTRONICS, FEB. 1995, USA, vol. 41, no. 1, ISSN 0098-3063, pages 150 - 159, XP000529220 *

Cited By (32)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2328592A (en) * 1997-06-06 1999-02-24 Nec Corp Viterbi decoder having a variable number of states
GB2328592B (en) * 1997-06-06 2000-04-12 Nec Corp Apparatus for receiving data
US6240145B1 (en) 1997-06-06 2001-05-29 Nec Corporation Apparatus for receiving data
KR100568647B1 (en) * 1997-07-21 2006-04-07 에릭슨 인크. Determination of the length of a channel impulse response
AU747699B2 (en) * 1997-07-21 2002-05-16 Ericsson Inc. Determination of the length of a channel impulse response
WO1999004537A1 (en) 1997-07-21 1999-01-28 Ericsson Inc. Determination of the length of a channel impulse response
US6775329B2 (en) 1997-09-16 2004-08-10 At&T Wireless Services, Inc. Transmitter diversity technique for wireless communications
US7120200B2 (en) 1997-09-16 2006-10-10 Cingular Wireless Ii, Llc Transmitter diversity technique for wireless communications
US9749032B2 (en) 1997-09-16 2017-08-29 At&T Mobility Ii Llc Transmitter diversity technique for wireless communications
US9203499B2 (en) 1997-09-16 2015-12-01 At&T Mobility Ii Llc Transmitter diversity technique for wireless communications
US6853688B2 (en) 1997-10-31 2005-02-08 Cingular Wireless Ii, Llc Low complexity maximum likelihood detection of concatenated space codes for wireless applications
WO1999023766A2 (en) * 1997-10-31 1999-05-14 At & T Wireless Services, Inc. Maximum likelihood detection of concatenated space-time codes for wireless applications with transmitter diversity
EP1808969A3 (en) * 1997-10-31 2010-04-14 Cingular Wireless II, LLC Maximum likehood detection of concatenated space-time codes for wireless applications with transmitter diversity
EP2285011A1 (en) * 1997-10-31 2011-02-16 AT & T Mobility II, LLC Maximum likelihood detection of concatenated space codes for wireless applications
US9065516B2 (en) 1997-10-31 2015-06-23 At&T Mobility Ii, Llc Low complexity maximum likelihood detection of concatenated space codes for wireless applications
WO1999023766A3 (en) * 1997-10-31 1999-07-08 At & T Wireless Services Inc Maximum likelihood detection of concatenated space-time codes for wireless applications with transmitter diversity
US6807240B2 (en) 1997-10-31 2004-10-19 At&T Wireless Services, Inc. Low complexity maximum likelihood detection of concatenate space codes for wireless applications
US7046737B2 (en) 1997-12-23 2006-05-16 Cingular Wireless Ii, Llc Near-optimal low-complexity decoding of space-time codes for wireless applications
US8179991B2 (en) 1997-12-23 2012-05-15 At&T Mobility Ii Llc Near-optimal low-complexity decoding of space-time codes for fixed wireless applications
US6741635B2 (en) 1997-12-23 2004-05-25 At&T Wireless Services, Inc. Near-optimal low-complexity decoding of space-time codes for fixed wireless applications
US7526040B2 (en) 1997-12-23 2009-04-28 At&T Mobility Ii Llc Near-optimal low-complexity decoding of space-time codes for fixed wireless applications
KR100710743B1 (en) * 1998-05-26 2007-04-24 코닌클리케 필립스 일렉트로닉스 엔.브이. Transmission system having a simplified channel decoder and operating method thereof
US7274752B2 (en) 1998-09-17 2007-09-25 Cingular Wireless Ii, Llc Maximum ratio transmission
US7362823B2 (en) 1998-09-17 2008-04-22 Cingular Wireless Ii, Llc Maximum ratio transmission
US6501803B1 (en) 1998-10-05 2002-12-31 At&T Wireless Services, Inc. Low complexity maximum likelihood detecting of concatenated space codes for wireless applications
EP2139183A1 (en) * 2000-05-05 2009-12-30 Agere System Inc. Joint estimation using the M-algorithm or T-algorithm in multiantenna systems
US7180954B2 (en) 2000-05-05 2007-02-20 Agere Systems Inc. Decoding techniques for multi-antenna systems
US8416890B2 (en) 2000-05-05 2013-04-09 Agere Systems Llc Decoding techniques for multi-antenna systems
US6968013B2 (en) 2000-05-05 2005-11-22 Agere Systems Inc. Decoding techniques for multi-antenna systems
EP1152576A1 (en) * 2000-05-05 2001-11-07 Lucent Technologies Inc. Joint estimation using the M-algorithm or T-algorithm in multiantenna systems
US7515659B2 (en) 2001-05-04 2009-04-07 Agere Systems Inc. Decoding techniques for multi-antenna systems
EP1510010A1 (en) * 2002-05-31 2005-03-02 Nokia Corporation Method and arrangement for enhancing search through trellis

Also Published As

Publication number Publication date
KR19990076857A (en) 1999-10-25
JP2000502859A (en) 2000-03-07
US5905742A (en) 1999-05-18
KR100367812B1 (en) 2003-04-21
AU716138B2 (en) 2000-02-17
EP0872095A1 (en) 1998-10-21
AU1423897A (en) 1997-07-28
CN1211366A (en) 1999-03-17
DE69635085D1 (en) 2005-09-22
EP0872095B1 (en) 2005-08-17

Similar Documents

Publication Publication Date Title
US5905742A (en) Method and apparauts for symbol decoding
WO1997024849A9 (en) Method and apparatus for symbol decoding using a variable number of survivor paths
CA2152315C (en) Soft decision signal outputting receiver
EP1516432B1 (en) Method and receiver for processing a multi-user signal
KR960006476B1 (en) Maximum likelihood sequence estimation for a mobile digital cellular radio receiver
US7277498B2 (en) Mapping method of code word with QAM modulation
US6483821B1 (en) CDMA mobile communications system and method with improved channel estimation and pilot symbol transmission
US7593492B1 (en) Combinational hybrid turbo-MUD
US5235621A (en) Receiver systems
JP2002532006A (en) Characterization of adaptive channels using decoded symbols.
CA2108494C (en) A frequency hopping code division multiple access radio communication unit
Liu et al. A soft-output bidirectional decision feedback equalization technique for TDMA cellular ratio
EP0660978B1 (en) A data decoder and method for use therein using a dynamically indexed channel state metric
Arslan et al. Iterative co-channel interference cancellation in narrowband mobile radio systems
WO2003092170A1 (en) Method and apparatus for the reception of digital communication signals
CA2241691C (en) Method and apparatus for symbol decoding using a variable number of survivor paths
JP4495596B2 (en) Signal processing method and apparatus using bit confidence factor
Zummo et al. Performance analysis of binary coded systems over Rician block fading channels
Femenias et al. Dual MRC diversity reception of TCM-MPSK signals over Nakagami fading channels
AU2003233263B2 (en) Method and receiver for processing a multi-user signal
Femenias A low frequency hopping TCM-8 PSK adaptive RSSE receiver for TDMA wideband cellular mobile radio
Wibisono et al. Trellis coded PSK modulation with diversity on correlated Rayleigh fading channel
Binia et al. Robust soft decision metrics for a turbo decoder in a partial band jammer environment
Checked KI/ERA/SRF/BT Girum Alebachew
Kazi et al. Adaptive LMS filter receivers for a convolutionally coded CDMA system

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 96180036.4

Country of ref document: CN

AK Designated states

Kind code of ref document: A1

Designated state(s): AL AM AT AU AZ BA BB BG BR BY CA CH CN CZ DE DK EE ES FI GB GE HU IL IS JP KE KG KP KR KZ LC LK LR LS LT LU LV MD MG MK MN MW MX NO NZ PL PT RO RU SD SE SG SI SK TJ TM TR TT UA UG UZ VN AM AZ BY KG KZ MD RU TJ TM

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): KE LS MW SD SZ UG AT BE CH DE DK ES FI FR GB GR IE IT LU MC NL PT SE BF BJ CF CG CI CM

DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application
COP Corrected version of pamphlet

Free format text: PAGES 1/6-6/6,DRAWINGS,REPLACED BY NEW PAGES 1/7-7/7;DUE TO LATE TRANSMITTAL BY THE RECEIVING OFFICE

ENP Entry into the national phase

Ref document number: 2241691

Country of ref document: CA

Ref document number: 2241691

Country of ref document: CA

Kind code of ref document: A

WWE Wipo information: entry into national phase

Ref document number: 1019980704984

Country of ref document: KR

WWE Wipo information: entry into national phase

Ref document number: 1996944428

Country of ref document: EP

WWP Wipo information: published in national office

Ref document number: 1996944428

Country of ref document: EP

REG Reference to national code

Ref country code: DE

Ref legal event code: 8642

WWP Wipo information: published in national office

Ref document number: 1019980704984

Country of ref document: KR

WWG Wipo information: grant in national office

Ref document number: 1019980704984

Country of ref document: KR

WWG Wipo information: grant in national office

Ref document number: 1996944428

Country of ref document: EP