US 2026/0205345 A1 · SYSTEM FOR WIRELESS COMMUNICATION IN DATA LINKS

Abstract

The invention proposes a radio communication system for data links. On the transmit side, the system uses single-carrier modulation in order to simplify the modulation design and reduce computational complexity on the transmit side so that it can be installed on airborne platforms. On the receive side, the system addresses crucial issues in high-speed data transmission and reception, including frame synchronization, frequency-offset estimation and compensation, and channel equalization in the frequency domain based on the MMSE method. As a result, the proposed system can adapt well to diverse environments and is suitable for communication systems in various fields, such as military communication systems.

A system for wireless communication in data links, comprising:

a transmit (Tx) side including the following blocks:
a bit-encoding block, whose input is data to be transmitted in binary (bit) form and whose output is error-correcting encoded data, the bit-encoding block performing turbo coding (Turbocodes) with a code rate (coderate) of ⅓, such that each input bit is encoded into three output bits;
in order to achieve desired code rates such as ½, ⅔ or ¾, the system uses a puncturing technique, to reduce the number of bits to be transmitted while still maintaining the error-correction capability of the code;
furthermore, an interleaving technique is performed after puncturing, to improve decoder performance and reduce burst errors caused by the transmission channel;
an input size of an interleaving block equals a length of a bit stream after puncturing;
an M-PSK/M-QAM modulation block, in which binary data after bit encoding are converted into decimal symbols depending on a type of modulation;
to reduce autocorrelation and repeating patterns in a transmitted signal, decimal data are scrambled according to a scrambling table and then mapped onto M-PSK/M-QAM constellations (M-ary phase shift keying (M-PSK) or M-ary quadrature amplitude modulation (M-QAM));
output of this block is IQ data after modulation;
a cyclic-prefix (CP) insertion block, in which input IQ data are divided into NFFT−Nn segments of samples, where NFFT is the size of a fast Fourier transform and Nn is a number of Null samples inserted for a purpose of protecting data at the two ends of each OFDM symbol;
a CP (cyclic prefix) part consists of Ng samples;
an output of this block is signal samples after CP insertion;
each OFDM signal symbol includes Ng CP samples, (NFFT−Nn) data samples after M-PSK/M-QAM modulation, and Null samples that protect two ends of the data of each OFDM symbol;
a data-multiplexing block, which multiplexes multiple signal samples and pilot samples that are formed from data OFDM samples and pilot OFDM samples with a synchronization part and a frame-end part to form a complete waveform frame;
in this frame, a pilot signal and data are divided into time slots;
each time slot comprises one pilot OFDM symbol and data OFDM symbols;
each pilot OFDM symbol includes Ng CP samples and NFFT pilot samples;
an interpolation and transmit-filtering block, which upsamples (interpolates) the signal after the data-multiplexing block by a factor of two, such that a sampling rate of the signal is doubled relative to an input sampling rate;
after the signal is upsampled by a factor of two it is passed through a transmit filter, then modulated onto a carrier and transmitted into the radio environment;
a radio-frequency block, which is a radio-frequency channel between a transmitter and the receiver;
and a receive (Rx) side including the following blocks:
a receive-filter block, which filters a received signal;
a frame-synchronization block, which finds a start point of each transmitted frame based on a synchronization signal by adjusting a threshold according to the power of the received signal, and performs the following steps:
step 301:
calculating the correlation R between the received signal sequence after filtering and a reference signal which is the synchronization signal inserted on the transmit side;
step 302:
calculating the power P of the input signal;
step 303:
determining a threshold value based on a power value (denoted thresholdP), the purpose of step 303 being to reduce computational complexity in hardware by adapting a threshold to the power of the received signal;
specifically, when the signal power is high the threshold can be increased to reduce noise and avoid false detections, whereas when the signal power is low the threshold is lowered to maintain detection capability;
step 304:
comparing a maximum value obtained from the correlation in step 301 (denoted maxValueR) with an adaptive threshold thresholdP obtained from step 302 in order to determine a synchronization flag:
if maxValueR≥ThresholdP, the synchronization flag indicates successful synchronization and is set to 1 (Yes);
otherwise, the synchronization flag indicates unsuccessful synchronization and is set to 0 (No), and the algorithm then returns to step 301 to continue calculating the correlation R;
step 305:
after successful synchronization, once a start point of the frame has been determined, downsampling the received signal by a factor of two so that the sampling rate is reduced by one half compared with an original sampling rate;
a frequency-offset estimation block 109, which estimates a frequency offset based on an accumulation algorithm applied to the cyclic-prefix data and performs the following steps:
step 401:
extracting the cyclic-prefix values of each OFDM symbol;
step 402:
calculating a correlation between cyclic-prefix sequences;
step 403:
accumulating the correlation values calculated in step 402 over multiple OFDM symbols and multiple frames to reduce noise and increase accuracy;
step 404:
calculating a phase angle based on the accumulated value stored in step 403 and, based on the phase angle, calculating a frequency offset of the system;
step 405:
from the frequency offset, calculating a phase array, then calculating a complex exponential of this phase array to obtain cosine and sine values, and performing a pointwise multiplication between the input signal and a complex conjugate of the cosine and sine array so as to obtain an output signal in which the frequency offset has been compensated;
a CP-removal block, which, after frequency-offset compensation, removes the cyclic prefix (CP) before channel estimation and equalization;
a channel-estimation and equalization block, which estimates channel characteristics based on pilot reference values in the frequency domain using the MMSE estimation method, and performs the following steps:
step 501:
extracting received data after frequency-offset compensation, extracted data being pilot signals or data signals;
step 502:
transforming the extracted signal into the frequency domain using a fast Fourier transform R(k)=FFT[r(n)], and then normalizing an amplitude of the FFT output by reducing the amplitude by a factor of √{square root over (NFFT)} to compensate for the FFT gain;
R
⁡
(
k
)
=
R
⁡
(
k
)
*
1
N
F
⁢
F
⁢
T
step 503:
calculating an inverse (reciprocal) of the pilot reference values in the frequency domain;
step 504:
depend on the type of extracted data (pilot signal or data signal), selecting the input data of multiplier block in step 505, if the extracted data is pilot, output of selector is a reciprocal of the pilot reference values P, if the extracted data is data signal, output of selector is channel-equalization coefficients W;
step 505:
performing a multiplication operation, in which, depending on the output of the selector in step 504, input to the multiplier block is either the reciprocal of the pilot reference values in the frequency domain P or channel-equalization coefficients W;
if the input data are the values P, the output is the channel frequency response, denoted Hest(k);
if the input data are the coefficients W, the output is the data after channel equalization in the frequency domain;
step 506:
transforming the output signal of step 503 from the frequency domain to the time domain by an inverse FFT;
if the input signal is the value Hest(k), the IFFT yields a channel impulse response in the time domain according to following equation;
h
e
⁢
s
⁢
t
(
n
)
=
IFFT
⁢
{
H
e
⁢
s
⁢
t
(
k
)
}
if the input signal is the equalized data in the frequency domain, the IFFT yields the data in the time domain, which are then passed to the M-PSK/M-QAM demodulation block;
afterwards, normalizing the amplitude by amplifying the amplitude of the signal after IFFT by a factor of √{square root over (NFFT)};
step 507:
removing data outside the cyclic-prefix interval of the channel impulse response by setting them to zero so as to reduce an influence of noise on accuracy of the channel-estimation result according to following formula;
h
ˆ
e
⁢
s
⁢
t
(
n
)
=
⁢
{
h
est
(
n
)
,
0
≤
n
<
N
g
0
,
N
g
≤
n
<
N
FFT
In which,
hest(n):
in which denotes the channel impulse response in the time domain calculated in step 504;
step 508:
calculating a noise σ2 according to a predetermined formula:
σ
2
=
1
2
⁢
(
N
F
⁢
F
⁢
T
-
N
g
)
⁢
∑
n
=
N
g
N
F
⁢
F
⁢
T
-
1
❘
"\[LeftBracketingBar]"
h
e
⁢
s
⁢
t
(
n
)
|
2
step 509:
transforming the signal into the frequency domain using an FFT and thereafter normalizing the amplitude by reducing an amplitude of the FFT output by a factor of √{square root over (NFFT)};
step 510:
calculating the channel-equalization coefficients, the equalization coefficients being calculated according to a predetermined formula:
W
⁡
(
k
)
=
(
H
est
′
)
*
⁢
(
k
)
|
H
e
⁢
s
⁢
t
′
(
k
)
|
2
+
2
⁢
σ
2
In which,
W(k) denotes the MMSE (minimum mean square error) channel-equalization coefficients;
(⋅)* and σ2 denote respectively complex conjugation and a variance of the estimation noise
step 511:
storing the channel-equalization coefficients W;
an M-PSK/M-QAM demodulation block, in which the signal processed by the channel-estimation and equalization block is first descrambled and then demodulated by the M-PSK/M-QAM demodulator, the demodulation process in M-PSK and M-QAM reconstructing the original data from a signal received after propagation through the channel, output of the M-PSK/M-QAM demodulation block being soft-decision data (soft bits);
and a bit-decoding block, which performs deinterleaving to rearrange the data back into an original order before depuncturing, and finally performs turbo decoding (Turbocodes) on the binary data to produce the decoded output data.

234 more elements

System for wireless communication in data links

Life of this patent

System for wireless communication in data links

Life of this patent

Abstract

Claims1