CN105187337A - OFDM decision-aided channel estimation algorithm based on repetition coding - Google Patents

Abstract

The invention relates to an OFDM decision-aided channel estimation algorithm based on repetition coding, comprising the following concrete steps of: (1) setting channel estimation H<k><n>=H<k><0>, wherein H<k><0> is initial estimation of channel; (2) according to the current channel estimation H<k><n>, calculating a symbol value X-hat<k><n> after frequency-domain equalization, and a bit soft decision value {S<k, m><n>} corresponding to the symbol; (3) performing weighted accumulation on the bit soft decision value according to a repetition coding rule; (4) calculating an erasure correction decision value {d<l><n>} according to the current bit soft decision value {LLR<l><n>}; (5) partly coding the {d<l><n>} into a to-be-mapped erasure correction bit stream {d<k, m><n>} corresponding to the nth OFDM symbol through repetition coding; (6) mapping the erasure correction bit stream {d<k, m><n>} into a transmit frequency domain symbol estimation value X-tilde<k><n>; (7) calculating a new channel estimation value; (8) updating the channel estimation value. According to the invention, the number of sample points of estimation is increased, and loss of data bandwidth caused by pilot insertion is avoided.

Description

A kind of OFDM decision-aided channel estimation method based on repeated encoding

Technical field

The present invention relates to a kind of OFDM decision-aided channel estimation method based on repeated encoding, belong to low-voltage powerline carrier communication field.

Background technology

Power-line carrier communication utilizes power line to carry out a kind of communication mode of transfer of data as medium, has natural advantage.Along with the construction of national intelligent grid, power line as a kind of economy easily communication medium favored, but China's power grid environment is complicated, some interference signals that many electrical equipment bands be linked in electrical network produce due to self make power line communication channel to exist multiple interference, and the electrical equipment of access causes the change of load, impedance etc., power line communication channel is made to produce frequency selective fading, power line communication is brought greater impact, especially totally unfavorable to the high-speed transfer of broadband power line communication system.

OFDM (OrthogonalFrequencyDivisionMultiplexing, be abbreviated as OFDM) be use more modulator approach in current high-speed applications, its advantage changes frequency-selective channel into flat fading channel, Cyclic Prefix is utilized to overcome intersymbol interference (Inter-SymbolInterference, ISI) impact, but in power line channel environment, multidiameter delay is unstable, utilize Cyclic Prefix can not eliminate the impact of multipath completely, still there is frequency selective fading in system, and the interference that in electrical network, some electrical equipment is introduced also can cause the deep fade in ofdm system in some frequency range, this just needs effective means to carry out the estimation of channel quality, for the decoding of system, demodulation provides effective information.Simultaneously, the feature of ofdm system can be utilized, different transmitted powers and carrying bit number can be distributed for the channel gain of different subcarriers and channel quality, optimization system transmission performance, this just needs accurate, real-time channel estimating measure, especially in power line communication environment.

Current multiple broadband power line carrier communication agreement, comprise ITUG.hn, HomePlugAV etc., each physical-layer data frame is usually by leading OFDM symbol, multiple OFDM symbol such as frame head OFDM symbol and user data OFDM symbol is formed, and in current multiple broadband power line carrier communication agreement, comprises ITUG.hn, HomePlugAV etc., have employed the channel coding schemes of this repeated encoding cascade forward error correction coding.Repeated encoding is a kind of straightforward procedure improving communication reliability, compare with conventional forward error correction coding, coding gain is lower, but can flexible configuration number of repetition, to meet the requirement of reliable communication under different channels condition, repeated encoding also has the advantage that decoding complex degree is low simultaneously.In systems in practice, repeated encoding often forms serial concatenation of codes, to obtain the balance between coding gain and implementation complexity with forward error correction coding.Repeated encoding provides extra time diversity, except for chnnel coding, also can be used for estimation and the tracking of channel.In order to reduce the impact of frequency selective fading on repeated encoding, the different copies of same transmission bit, generally all can be assigned to different subcarriers as far as possible.Then the soft-decision of the bit copy first received with the bit soft decisions weighted average be currently received, can reconstruct the symbol judgement of corresponding subcarrier, and the decision-aided for channel is estimated.Compared with general symbol judgement assisted channel estimation, due to the existence of multiple bit copy, the correctness of symbol judgement can progressively improve, thus improves the precision of channel estimating.

Summary of the invention

The present invention proposes the diversity performance utilizing the repeated encoding of frame control head and follow-up effective carrier data, the soft-decision result utilizing the bit copy first received and the bit soft decisions result be currently received are weighted on average, then reconstruct the symbol judgement of corresponding subcarrier, the decision-aided for channel is estimated.Both added the sample points of estimation, the loss of inserting the data bandwidth that pilot tone causes can not have been brought again.

To achieve these goals, technical scheme of the present invention is as follows.

Based on an OFDM decision-aided channel estimation method for repeated encoding, its system frequency-domain model simplified is,

$Y_k^m=H_{k,ideal}^n*X_k^n+W_k^n;$

Wherein, for frequency domain sends symbol, for channel transfer functions, for the interference of frequency domain equivalent noise, n is OFDM symbol sequence number, and k is subcarrier sequence number; These algorithm concrete steps are as follows:

(1) signaling channel is estimated wherein for channel initial estimation; Initialization codes bit soft decisions value $\left\{{\mathrm{LLR}}_l^0\right\}=0;$

(2) estimate according to present channel calculate the value of symbol after frequency domain equalization the bit soft decisions value corresponding with this symbol wherein, n is OFDM symbol sequence number, and k is subcarrier sequence number, m ∈ 0,1,2 ..., M _kbe bit sequence corresponding to this sub-carrier, M _kit is the bit number of carrying on subcarrier k;

(3) according to repeated encoding rule, weighted accumulation coded-bit soft-decision-value:

${\mathrm{LLR}}_{f(k,m,n)}^n=(1-w)*{\mathrm{LLR}}_{f(k,m,n)}^{n-1}+w*s_{k,m}^n;$

Wherein w is weight coefficient, and w ∈ (0,1], f (k, m, n) is ratio in corresponding forward error correction code word

Special sequence number; (4) according to current bit soft-decision-value decision value is deleted in entangling of calculation code bit

$d_l^n=\left\{\begin{array}{cc}-1,& {\mathrm{LLR}}_l^n\&le;-t\\ 0,& \left|{\mathrm{LLR}}_l^n\right|\&le;t\\ 1,& {\mathrm{LLR}}_l^n\&GreaterEqual;t\end{array}\right.;$

Wherein t is positive number, determines whether current bit is entangle the threshold value of deleting position;

(5) part repeated encoding is that to be mapped entangling corresponding to the n-th OFDM symbol deletes bit stream

(6) bit stream is deleted entangling the transmission frequency domain symbol estimated value be mapped as if corresponding the entangling of transmission frequency domain symbol is deleted bit and comprised 0 value, then this transmission frequency domain symbol valuation Euclidean decision value of balanced symbol replaces;

(7) new channel estimation value is calculated:

${\tilde{H}}_k^n=Y_k^n/{\tilde{X}}_k^n;$

(8) channel estimation value is upgraded:

$H_k^{n+1}=(1-\mathrm{\&mu;})*H_k^n+\mathrm{\&mu;}*{\tilde{H}}_k^n;$

Wherein μ is smoothing factor, μ ∈ (0,1];

(9) if n<N, then n=n+1, goes to 2) start to process next OFDM symbol; Otherwise terminate equalization algorithm and enter the decoding FEC stage.

The beneficial effect of the invention is: algorithm of the present invention utilizes the diversity performance of the repeated encoding of frame control head and follow-up effective carrier data, the soft-decision result utilizing the bit copy first received and the bit soft decisions result be currently received are weighted on average, then the symbol judgement of corresponding subcarrier is reconstructed, decision-aided for channel is estimated, both added the sample points of estimation, the loss of inserting the data bandwidth that pilot tone causes can not have been brought again.The application in a power line communication system of this algorithm.Repeated encoding and map and can design separately at receiving terminal, also can the coding of multiplexing transmitting terminal and mapping block.

Accompanying drawing explanation

Fig. 1 be in the embodiment of the present invention use system principle diagram.

Embodiment

Below in conjunction with accompanying drawing, the specific embodiment of the present invention is described, better to understand the present invention.

Embodiment

Be illustrated in figure 1 the schematic diagram of this system, wherein, based on the OFDM decision-aided channel estimation method of repeated encoding, its system frequency-domain model simplified is,

$Y_k^m=H_{k,ideal}^n*X_k^n+W_k^n;$

Wherein, for frequency domain sends symbol, for channel transfer functions, for the interference of frequency domain equivalent noise, n is OFDM symbol sequence number, and k is subcarrier sequence number; These algorithm concrete steps are as follows:

(1) signaling channel is estimated wherein for channel initial estimation; Initialization codes bit soft decisions value $\left\{{\mathrm{LLR}}_l^0\right\}=0;$

(2) estimate according to present channel calculate the value of symbol after frequency domain equalization the bit soft decisions value corresponding with this symbol wherein, n is OFDM symbol sequence number, and k is subcarrier sequence number, m ∈ 0,1,2 ..., M _kbe bit sequence corresponding to this sub-carrier, M _kit is the bit number of carrying on subcarrier k;

(3) according to repeated encoding rule, weighted accumulation coded-bit soft-decision-value:

${\mathrm{LLR}}_{f(k,m,n)}^n=(1-w)*{\mathrm{LLR}}_{f(k,m,n)}^{n-1}+w*s_{k,m}^n;$

Wherein w is weight coefficient, and w ∈ (0,1], f (k, m, n) is ratio in corresponding forward error correction code word

Special sequence number; (4) according to current bit soft-decision-value decision value is deleted in entangling of calculation code bit

$d_l^n=\left\{\begin{array}{cc}-1,& {\mathrm{LLR}}_l^n\&le;-t\\ 0,& \left|{\mathrm{LLR}}_l^n\right|\&le;t\\ 1,& {\mathrm{LLR}}_l^n\&GreaterEqual;t\end{array}\right.;$

Wherein t is positive number, determines whether current bit is entangle the threshold value of deleting position;

(5) part repeated encoding is that to be mapped entangling corresponding to the n-th OFDM symbol deletes bit stream

(6) bit stream is deleted entangling the transmission frequency domain symbol estimated value be mapped as if corresponding the entangling of transmission frequency domain symbol is deleted bit and comprised 0 value, then this transmission frequency domain symbol valuation Euclidean decision value of balanced symbol replaces;

(7) new channel estimation value is calculated:

${\tilde{H}}_k^n=Y_k^n/{\tilde{X}}_k^n;$

(8) channel estimation value is upgraded:

$H_k^{n+1}=(1-\mathrm{\&mu;})*H_k^n+\mathrm{\&mu;}*{\tilde{H}}_k^n;$

Wherein μ is smoothing factor, μ ∈ (0,1];

(9) if n<N, then n=n+1, goes to 2) start to process next OFDM symbol; Otherwise terminate equalization algorithm and enter the decoding FEC stage.

The above is the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, under the premise without departing from the principles of the invention; can also make some improvements and modifications, these improvements and modifications are also considered as protection scope of the present invention.

Claims (1)

1. based on an OFDM decision-aided channel estimation method for repeated encoding, it is characterized in that: the system frequency-domain model of the simplification adopted is,

$Y_k^m=H_{k,ideal}^n*X_k^n+W_k^n;$

Wherein, for frequency domain sends symbol, for channel transfer functions, for the interference of frequency domain equivalent noise, n is OFDM symbol sequence number, and k is subcarrier sequence number; These algorithm concrete steps are as follows:

(1) signaling channel is estimated wherein for channel initial estimation; Initialization codes bit soft decisions value $\left\{{\mathrm{LLR}}_l^0\right\}$ $=0;$

(2) estimate according to present channel calculate the value of symbol after frequency domain equalization the bit soft decisions value corresponding with this symbol wherein, n is OFDM symbol sequence number, and k is subcarrier sequence number, m ∈ 0,1,2 ..., M _kbe bit sequence corresponding to this sub-carrier, M _kit is the bit number of carrying on subcarrier k;

(3) according to repeated encoding rule, weighted accumulation coded-bit soft-decision-value:

${\mathrm{LLR}}_{f(k,m,n)}^n=(1-w)*{\mathrm{LLR}}_{f(k,m,n)}^{n-1}+w*s_{k,m}^n;$

Wherein w is weight coefficient, and w ∈ (0,1], f (k, m, n) is bit sequence in corresponding forward error correction code word;

(4) according to current bit soft-decision-value decision value is deleted in entangling of calculation code bit

$d_l^n=\left\{\begin{array}{cc}-1,& {\mathrm{LLR}}_l^n\&le;-t\\ 0,& \left|{\mathrm{LLR}}_l^n\right|\&le;t\\ 1,& {\mathrm{LLR}}_l^n\&GreaterEqual;t\end{array}\right.;$

Wherein t is positive number, determines whether current bit is entangle the threshold value of deleting position;

(5) part repeated encoding is that to be mapped entangling corresponding to the n-th OFDM symbol deletes bit stream

(6) bit stream is deleted entangling the transmission frequency domain symbol estimated value be mapped as if corresponding the entangling of transmission frequency domain symbol is deleted bit and comprised 0 value, then this transmission frequency domain symbol valuation Euclidean decision value of balanced symbol replaces;

(7) new channel estimation value is calculated:

${\tilde{H}}_k^n=Y_k^n/{\tilde{X}}_k^n;$

(8) channel estimation value is upgraded:

$H_k^{n+1}=(1-\mathrm{\&mu;})*H_k^n+\mathrm{\&mu;}*{\tilde{H}}_k^n;$

Wherein μ is smoothing factor, μ ∈ (0,1];

(9) if n<N, then n=n+1, goes to 2) start to process next OFDM symbol; Otherwise terminate equalization algorithm and enter the decoding FEC stage.