QLM demodulation

A method for implementation of orthogonal frequency division multiple access (OFDMA) QLM for communications using frequency offset as a differentiating parameter, said method comprising the steps:

• generating a first communications signal over a frequency bandwidth 1/Ts at a carrier frequency for a first set of N channels by modulating a first stream of N data symbols at a data symbol rate 1/Ts with an N-point inverse fast fourier transform (FFT−1) waveform over a FFT−1 block length NTs to generate N orthogonal frequency harmonics modulated with the respective data symbols and which modulated harmonics are the N channels for communications,
• generating a second communications signal over the same block length over the same frequency bandwidth at the same carrier frequency for a second set of N channels by modulating a second stream of N data symbols at a data symbol rate 1/Ts with the same FFT−1 waveform with a frequency offset Δk equal to Δk=1/np wherein “np” is the number of QLM sets of channels in said frequency bandwidth and FFT−1 harmonics are e^j2π(k+Δk)n/N for normalized frequencies k=0,1, . . . N−1 corresponding to channels 0,1, . . . , N−1 wherein “j”=√(−1), “n=pi”, and “n” is a time index,
• for any additional sets of channels, continuing generation of communication signals over the same block length over the same frequency bandwidth at the same carrier frequency by modulating additional streams of data symbols with the same waveform at the same data symbol rate as the first and second streams of data symbols with frequency offsets increasing in each communication signal in increments of Δk=1/np until the np signals are generated for np QLM sets of channels,
• repeating this generation of the QLM signal for the next FFT−1 block length NTs and continuing for subsequent block lengths,
• transmitting and receiving said communications signals over a QLM communications link consisting of the np QLM sets of channels,
• recovering data symbols of the communications signals in a receiver using a demodulation algorithm, and
• combining said algorithm with error correction decoding to recover the transmitted information whereby frequency offset has been used as a differentiating parameter to enable QLM parallel sets of channels for communications over the same frequency bandwidth at the same carrier frequency with a data symbol rate independent of the Nyquist rate, to be demodulated.

A method for implementation of QLM demodulation for communications over the same frequency bandwidth of a carrier frequency using time and/or frequency and/or other offsets as the differentiating parameters and using trellis symbol demodulation, said method comprising the steps:

• receiving QLM data symbols indexed on k=1,2, . . . , n for n data symbols,
• evaluating correlation coefficients for received QLM data symbols,
• establishing trellis states and trellis paths for each data symbol whereby
• a) there are ns^(2nc−2) trellis states wherein “ns” is the number of states of each data symbol, and “nc” is the number of one-sided correlation coefficients for symmetrical correlation functions,
• b) there are ns^(2nc−1) trellis paths from trellis state Sk−1 for data symbol k−1 to a new trellis state Sk for data symbol k,
• c) index jxi=0,1,2, . . . , ns^(2nc−1)−1 is a trellis path index from Sk−1 to Sk using ns-ary index symbols jxi reading from left to right wherein “j” is an index for a new data symbol, “i” is an index for a last data symbol and “x” is a set of ns-ary index data symbols between “j” and “i”,
• d) index jx is over ns^(2nc−2) trellis states Sk(jx),
• e) index xi is over ns^(2nc−2) trellis states Sk−1(xi),
• f) creating a D row by ns^(2nc−2) column memory “M” for storing trellis path decisions wherein D is a multiple of nc,
• initializing the trellis algorithm for k=0
• implementing the trellis algorithm for each step k starting with k=1 by
• 1) evaluating hypothesized values {circumflex over (X)}k(jxi) of received QLM correlated data symbols for all paths jxi,
• 2) measuring a received QLM data symbol Yk for data symbol k,
• 3) evaluating a logarithm state transition decisioning metric R(jxi) which is a function of {circumflex over (X)}k(jxi) and Yk for symbol k for all possible paths jxi from Sk−1(xi) to Sk(jx),
• 4) finding a best path metric αk(jx) and corresponding path jxi from Sk−1(xi) to Sk(jx) by using a logarithm decisioning equation αk(jx)=max{αk−1(xi)+Rk(jxi)} which finds xi that maximizes “max” the sum “αk−1(xi)+Rk(jxi)” for a given jx,
• 5) using said αk(jx) and corresponding jxi to define a data symbol estimate {circumflex over (Z)}k for Sk(jx),
• 6) for k<D continuing to fill memory M from the top down by replacing column xi with column jx after moving the row elements of column xi down by one symbol and placing {circumflex over (Z)}k in the vacant first row, for all jx,
• 7) for k≧D continuing to replenish memory M from the top down by replacing column xi with column jx after moving the row elements of column xi down by one symbol and placing {circumflex over (Z)}k in the vacant first row, for all jx,
• 8) for k≧D selecting the last symbol {circumflex over (Z)}k−D in column jx in the deleted bottom row corresponding to the maximium value of αk(jx) over all jx, as the best estimate of data symbol k−D,
• 9) for k>n continuing step 8 without replenishing memory M to complete the estimates of the n data symbols,
• 10) performing error correction decoding of these estimated data symbols,
• wherein said QLM demodulated signal in a communications receiver has been generated in a communications transmitter, up-converted to a radio frequency RF, power amplified, and transmitted as a RF QLM signal in a communications transmitter, and
• received in said communications receiver, amplified, down-converted, and QLM demodulated as described in this claim followed by decoding to recover the data symbol information.