Method for adaptive blind equalization of PSK signal, equalizer and receiver
The present application discloses a method for adaptive blind equalization of a PSK signal, an equalizer and a receiver. According to embodiments as provided, a conjugate product of a current output and a precedent output of an FIR filter is calculated, an equalization coefficient is updated using the conjugate product, and then an input signal is filtered using the FIR filter with the updated equalization coefficient. The embodiments as provided is applicable to adaptive blind equalization of any phase shift keying signal including a BPSK signal.
1. An equalizer for a PSK signal, comprising:
a conjugate product calculator to calculate a conjugate product r of a current output y t and a precedent output y t-n of the equalizer, wherein r=y t y t-n *, t denotes a current time, t−n denotes a precedent time, and n≧1;
an equalization coefficient updator to update the equalization coefficient; and
a finite impact response filter to filter an input signal using the updated equalization coefficient,
wherein the coefficient updator is configured to update the equalization coefficient using a cost function c constructed based on the conjugate product r,
wherein the cost function c is:
c=E{|r m −γ| 2 }
where E{•} denotes expectation, m denotes the number of constellation points of the PSK signal, r denotes the conjugate product, and γ denotes a constant.
2. An equalizer for a PSK signal, comprising:
a conjugate product calculator to calculate a conjugate product r of a current output yt and a precedent output yt−n of the equalizer, wherein r=yt yt−n*, t denotes a current time, t−n denotes a precedent time, and n≧1;
an equalization coefficient updator to update the equalization coefficient; and
a finite impact response filter to filter an input signal using the updated equalization coefficient,
wherein the coefficient updator is configured to update the equalization coefficient using a cost function c constructed based on the conjugate product r,
wherein the equalizer is adapted for a BPSK signal, and the cost function c is:
c=E{∥Re ( r )|−γ|+| Im ( r )|},
where E{•} denotes expectation, r denotes the conjugate product, γ denotes a real constant greater than 0, and Re(•) and Im(•) denote operations for calculating real part and calculating imaginary part, respectively.
3. An equalizer for a PSK signal, comprising:
a conjugate product calculator to calculate a conjugate product r of a current output yt and a precedent output yt−n of the equalizer, wherein r=yt yt−n*, t denotes a current time, t−n denotes a precedent time, and n≧1;
an equalization coefficient updator to update the equalization coefficient; and
a finite impact response filter to filter an input signal using the updated equalization coefficient,
wherein the coefficient updator is configured to update the equalization coefficient using a cost function c constructed based on the conjugate product r and a phase difference between the current output and the precedent output introduced by the frequency difference between an emitting laser and a local oscillation laser,
wherein the cost function c is:
c=E{|r m −e jmΔθ | 2 },
where E{•} denotes expectation, m denotes the number of constellation points of the PSK signal, r denotes the conjugate product, and Δθ denotes the phase difference.
4. An equalizer for a PSK signal, comprising:
a conjugate product calculator to calculate a conjugate product r of a current output yt and a precedent output yt−n of the equalizer, wherein r=yt yt−n*, t denotes a current time, t−n denotes a precedent time, and n≧1;
an equalization coefficient updator to update the equalization coefficient; and
a finite impact response filter to filter an input signal using the updated equalization coefficient,
wherein the coefficient updator is configured to update the equalization coefficient using a cost function c constructed based on the conjugate product r and a phase difference between the current output and the precedent output introduced by the frequency difference between an emitting laser and a local oscillation laser,
wherein the equalizer is adapted for a BPSK signal, and the cost function c is:
c=E{∥Re ( re −jΔθ )|−γ|+| Im ( re −jΔθ )|},
where E{•} denotes expectation, r denotes the conjugate product, Δθ denotes the phase difference, γ denotes a real constant greater than 0, and Re(•) and Im(•) denote operations for calculating real part and calculating imaginary part, respectively.