CN101345529B - Pulse group digital coding and decoding method for multi-channel multi-ary - Google Patents

Abstract

The invention discloses a encoding and decoding method for multi-channel multi-system digital signal. The invention provides a multi-channel transmission channel division and multi-system digital encoding technology based on patent technology of single-channel binary digital signal pulse group duration encoding method, also provides corresponding signal filter and decoding technology. The inventive encoding technology represents improvement compared to digital transmission distance and speed. The invention can also be used for telephone cable and other medium for transmitting digital signal.

Description

Multichannel multi-system pulse group digital coding and coding/decoding method

Technical field

The present invention relates to digital signal transmission field, specifically, the present invention relates to a kind of coding method of multichannel multi-system digital signal, and the filtering and the coding/decoding method that adopt the digital signal of this coding method transmission.

Background technology

(Asymmetric Digital SubscriberLine ADSL) as one of contemporary important broadband technology, has at home and abroad carried out widely and has used ADSL (Asymmetric Digital Subscriber Line).It is transmission medium that the ADSL technology adopts telephone cable, future technology such as will communicate by letter with optical fiber communication, radio communication, camera cable exist simultaneously for a long time.

Current ADSL adopts Discrete Multitone, and (Discrete Multitone DMT) is its core technology.DMT obtains the (ANSI of American National Standards Institute, AmericanNational Standards Institute), (the ETSI of European Telecommunication Standard research institute, European Telecommunications Standards Institute), and (the ITU of Union of International Telecommunication, InternationalTelecommunications Union) acceptance, international and domestic important communications equipment manufacturer are adopted the DMT standard development one after another and are produced the ADSL product.

ANSI T1.413 standard according to DMT, whole frequency band is divided into 256 sub-channels, the bandwidth of each sub-channel is 4kHz, the Digital Transmission speed that each sub-channel is the highest is 60kpbs, preferably on the circuit, the common transmission rate of each sub-channel is 40kpbs, when the signal noise ratio of line difference in signal to noise ratio, the transmission rate of each sub-channel is approximately 16kpbs, even lower.Can reach higher rate when transmission line length is in 2km usually, transmission rate descended much when line length surpassed 2km, can not normal transmission information when line length surpasses 6km.

Patent for invention (the patent No.: ZL03149904X) that in July, 2006, China State Intellectual Property Office authorized " coding method of binary digital signal and method for transmitting signals thereof and circuit ", the foregoing invention patent disclosure a kind ofly be called the digital coding of pulse group duration (PGD) coding and separate decoding transmission method, show on transmission range and speed improvement for current correlation technique.This patent of invention is disclosed to be the binary digit transmission method of a kind of single channel.

Summary of the invention

The present invention is in patent of invention " coding method of binary digital signal and method for transmitting signals thereof and the circuit " (patent No.: on basis ZL03149904X), a kind of multichannel multi-system digital coding method is provided, corresponding signal filtering method and coding/decoding method are provided simultaneously.

A kind of binary digital coding method is disclosed in patent of invention ZL03149904X, this coding method corresponds to a pulse group sequence with binary digital signal, wherein " 0 " in the binary digital signal and " 1 " correspond respectively to two pulse groups that have the different characteristic frequency and have the fixed pulse number, and described two pulse groups include and are at least 2 identical fixed pulse N.This coding method is called pulse group duration (PGD) coding.

In order to improve Digital Transmission speed, this patent provides and has adopted a plurality of subchannels to transmit the PGD coding method of multi-system number simultaneously in the operating frequency range of signal transmission medium, and the binary system PGD coding method that multi-system PGD coding method and patent of invention ZL03149904X provide is similar.In a subchannel, carry out the M system PGD coding of M＞2, at first will be at this subchannel centre frequency f _NcNear choose M characteristic frequency

Correspond respectively to a M system digital " 0 ", " 1 ", " 2 " ... " M-1 ", then the M system number of coding is corresponded to one and have with the corresponding characteristic frequency of this numeral and have the pulse group of fixed pulse number, the fixed pulse of described pulse group is counted N and is at least two.Like this, a M system number just corresponds to a pulse group that has corresponding characteristic frequency and have the fixed pulse number, and the sequence of a M system number just corresponds to an above-mentioned pulse group sequence.Above-mentioned coding method is called multi-system pulse group duration coding (MPGD).

In order to carry out the multi-channel data transmission, need carry out correct subchannel to the operating frequency range of transmission medium divides, should avoid the mutual signal between the sub-channel signal to disturb when each subchannel carries out transfer of data at the same time, need to know the frequency bandwidth of transmission signals for this reason.According to the disclosed pulse group of patent of invention (patent No. ZL03149904X) duration coding method, binary system PGD code signal can be represented with following mathematical formulae:

$S\left(t\right)=\left[\underset{n=-\∞}{\overset{\∞}{\mathrm{\Σ}}}{a}_n{g}_0(t-T_n)\right]{\cos \mathrm{\ω}}_{0}t+\left[\underset{n=-\∞}{\overset{\∞}{\mathrm{\Σ}}}{\stackrel{\‾}{a}}_n{g}_1(t-T_n)\right]{\cos \mathrm{\ω}}_{1}t---\left(1\right)$

Here,

ω ₀=2 π f ₀, f ₀It is the operating frequency of the binary number that is transmitted PGD coding during for " 0 ";

ω ₁=2 π f ₁, f ₁It is the operating frequency of the binary number that is transmitted PGD coding during for " 1 ";

a _nBe a _nRadix-minus-one complement;

a _n=1, a _n=0 (probability is P);

a _n＝0， ${\stackrel{\‾}{a}}_n=1$ (probability is 1-P);

$T_n=\underset{i=0}{\overset{n}{\mathrm{\Σ}}}(a_i{T}_0+{\stackrel{\‾}{a}}_i{T}_1)$ It is the zero-time of the n time PGD coding;

g ₀(t) be the function of time of single rectangular pulse, its pulse duration is $T_0=\frac{N}{f_0},$ T ₀It is the scramble time when being transmitted number for " 0 ";

g ₁(t) be the function of time of single rectangular pulse, its pulse duration is $T_1=\frac{N}{f_1},$ T ₁It is the scramble time when being transmitted number for " 1 ";

N is the fixed pulse number that PGD coded pulse group is comprised.

Above-mentioned mathematical notation explanation binary system PGD signal is similar with common fsk signal, and the scramble time was when the binary number that just is transmitted was " 0 " $T_0=\frac{N}{f_0},$ The scramble time was when the binary number that is transmitted was " 1 " $T_1=\frac{N}{f_1},$ This is the characteristics of PGD coding method.Usually FSK used scramble time when transmitting binary number " 0 " with " 1 " is identical, and does not stipulate that the used scramble time is the integral multiple of single carrier cycle.

The general approaching characteristic frequency mutually that adopts of PGD, at this moment | f ₁-f ₀The f of |＜＜, so T ₀≈ T ₁, the narrow band transmission of similar FSK can be derived the bandwidth B of PGD code signal thus _PGDBe approximately twice baseband signal bandwidth B:

B _PGD≈2B (2)

Usually the binary number that is transmitted is uncertain, and digital " 0 " and " 1 " alternately are the wideest situations of baseband signal bandwidth, suppose T simultaneously ₀≈ T ₁, the bandwidth of baseband signal at this moment is $B=\frac{1}{T_0}=\frac{f_0}{N},$ Therefore, the bandwidth of PGD code signal can be used following formulate:

$B_{\mathrm{PGD}}\≈\frac{2f_0}{N}---\left(3\right)$

Binary system PGD coded data transfer rate $R=\frac{f_0}{N},$ Therefore,

B _PGD≈2R (4)

During transmission MPGD code signal, the number that is transmitted is uncertain equally, and in the lowest numeric " 0 " of M system number and maximum number " M-1 " alternate transmission in particular cases, baseband signal bandwidth is the wideest.Usually MPGD adopts narrow band transmission, and very approaching mutually corresponding to M characteristic frequency of M system number, in the case, the bandwidth of MPGD code signal is similar to the bandwidth of binary system PGD code signal in a subchannel, can use following formulate:

$B_{\mathrm{MPGD}}\≈\frac{2f_{\mathrm{nc}}}{N}---\left(5\right)$

Here, N is the fixed pulse number of pulse group, f _NcIt is the subchannel centre frequency.B _MPGDIt is the bandwidth of MPGD code signal.Formula (5) illustrates that the bandwidth of MPGD code signal and subchannel centre frequency are in direct ratio, subchannel centre frequency f _NcHigh more, the bandwidth of code signal is wide more.Simultaneously to count N inversely proportional for the bandwidth of code signal and pulse group fixed pulse, and it is narrow more that selected pulse group fixed pulse is counted the bandwidth of the big more then code signal of N, select little pulse group fixed pulse count N then the bandwidth of code signal can widen.

B _MPGDBe in the transmission medium operating frequency range, to carry out the important parameter that subchannel is divided foundation.Adopt the transmission rate of MPGD coding transmission digital signal to be:

$R={\log }_{2}M\×\frac{f_0}{N}---\left(6\right)$

Can see that the transmission rate that adopts the MPGD code signal has improved log than the transmission rate that binary system PGD encodes ₂M doubly.

The present invention also provides a kind of coding/decoding method corresponding to above-mentioned multi-system pulse group duration coding method.For a pulse group sequence through the MPGD coding, pulse group sequence is divided into the set of pulse group according to the fixed pulse number, measure the duration of each pulse group in the described set, judge the feature work frequency of pulse group according to the difference of described pulse group duration, according to the feature work frequency of its pulse group with the corresponding multi-system number that returns original coding of this pulse group.The described duration is whole accumulated time in the cycle of pulses in the described pulse group, perhaps is the accumulated time in partial pulse cycle specific in the pulse group.

The MPGD coded data of transmitting simultaneously for a plurality of subchannels of reception that can be correct, need carry out bandpass filtering to each sub-channel signal, the MPGD code signal of this subchannel and the code signal of other passages are made a distinction, can carry out effective MPGD signal decoding like this.This patent provides multistage RLC filter circuit, has realized sub-channel signal is carried out effective bandpass filtering.The RLC filter circuit of described each grade can be the RLC parallel circuits, it also can be the RLC series circuit, when choosing the parameter of described rlc circuit, make the centre frequency of its resonance frequency and MPGD coding have certain frequency shift (FS) Δ f, Δ f is chosen at the straight line portion of received signal frequency characteristics, because the frequency characteristic signal curve is with respect to the centre frequency symmetry, therefore Δ f can be that just to be offset also can be that negative bias moves, filtering through above-mentioned multistage rlc circuit, the frequency change of received signal presents and the corresponding relation of code signal, measure the work period of this each pulse of signal, and carry out arithmetic average for the work period of N pulse of each pulse group, promptly can obtain the average period of this pulse group pulse, can extrapolate the duration of the characteristic frequency and the pulse group of this pulse group thus, the duration of this pulse group is corresponded to the multi-system number that MPGD encodes, promptly reached the purpose of signal decoding.We deserve to be called the subchannel filtering stated and the method for decoding is " resonance frequency shift filtering decoding " method.

The invention provides above-mentioned RLC filter circuit method for digitizing, described RLC filter circuit is carried out mathematical simulation, calculate the output variable that changes with input variable, use above-mentioned counting to learn result of calculation and carry out computer programming, promptly realized the digitlization of RLC filter circuit.Have identical filter action with the RLC filter circuit of simulation for input signal through the later RLC filter circuit of digitlization, difference is the input signal of the RLC filter circuit treatment of simulated simulated, is through digitized input signal and digitlization RLC filter circuit is handled.

Aforementioned M system number sequence row can be represented with a pulse group sequence, if adjoining pulse group is represented different multi-system numbers, their corresponding characteristic frequencies can be different, be difficult to occur the sudden change of characteristic frequency in the pulse of two pulse group intersections, the characteristic frequency of pulse group changes needs a transit time, in order to measure the characteristic frequency of pulse group accurately, can give up each one or several pulse of pulse group and front and back pulse group intersection, only measure the cycle of each pulse of pulse group mid portion, carry out arithmetic average then and can obtain the average period of this each pulse of pulse group, can extrapolate the duration of the characteristic frequency and the pulse group of this pulse group thus, the duration of this pulse group is corresponded to the multi-system number that MPGD encodes, can reach the purpose of signal decoding more accurately.

The bandwidth B of aforementioned MPGD code signal _MPGDBe that the transmission medium operating frequency range is carried out the important parameter that subchannel is divided.Notice $B_{\mathrm{MPGD}}\≈\frac{2f_{\mathrm{nc}}}{N},$ B _MPGDWith f _NcIn direct ratio, different subchannels has different bandwidth, simultaneously B _MPGDInversely proportional with N, the numerical value of N is big more, and the bandwidth of subchannel is narrow more, and the frequency band division of MPGD code signal can not be divided according to simple five equilibrium.The invention provides a kind of method of multichannel frequency band division.Consider improvement through the later frequency characteristic signal of filtering, utilize filter circuit resonance frequency in aforesaid " resonance frequency shift filtering decoding " method to depart from the characteristics of the centre frequency of encoding simultaneously, it is certain overlapped that two adjoining subchannels can be had on bandwidth, for two adjoining subchannels, adopt circuit resonant frequencies to move respectively for the just skew and the negative bias of coding central feature frequency, make the resonance frequency of filter circuit be in outside the bandwidth of contiguous subchannel, the overlapping phase mutual interference that can not cause signal on such two sub-bandwidth chahnels, when carrying out efficient coding signal filtering and decoding, can in the transmission medium operating frequency range, mark off more subchannel, thereby improve total transmission rate.

Description of drawings

Fig. 1: multichannel multi-system MPGD coding transmission and multi-channel digital filter solution code system block diagram;

Fig. 2: multichannel MPGD Digital Transmission channel division example.

Embodiment

Below in conjunction with the drawings and specific embodiments the present invention is done and to describe in further detail.

The course of work of multichannel multi-system MPGD coding transmission and corresponding digital filtering and decode system is referring to Fig. 1.The binary number that is transmitted at first will carry out packing data and channel allocation according to Data Transport Protocol, carry out the MPGD coding of multi-system respectively according to the selected frequency of each subchannel through the data of packing and channel allocation, each number of subchannels behind the coding is converted into analog signal through D/A after according to addition, and this analog passband signal is overdrived and sent into transmission medium.The data receiver, the signal that is transmitted by transmission medium carries out frequency-division filter and signal amplification according to the upstream or downstream signal, and be converted into digital signal by A/D, this digital signal is carried out the digital filtering of subchannel, each subchannel carries out the MPGD decoding to filtered signal respectively, data to all subchannels unpack and ressemble at last, obtain receiving data.

A centre frequency is f _NcSubchannel in carry out 4 system PGD coding, at first will be at this subchannel centre frequency f _NcNear choose 4 characteristic frequencies

Correspond respectively to one 4 system digital " 0 " or " 1 ", " 2 ", " 3 ", the 4 system numbers that will encode then correspond to one to have with the corresponding characteristic frequency of this numeral and has the pulse group of fixed pulse number.Like this, one 4 system number just corresponds to a pulse group that has corresponding characteristic frequency and have the fixed pulse number, and the sequence of one 4 system number is just corresponding to the sequence of an above-mentioned pulse group.

The method that multichannel MPGD Digital Transmission subchannel is divided is please referring to Fig. 2.Having selected the fixed pulse of pulse group to count N among Fig. 2 is 8, and (1100kHz is to 2000kHz) divided 4 subchannels in given frequency range, and the centre frequency that 4 subchannels divide is respectively f _1c, f _2c, f _3c, f _4c, the division of subchannel at first is the centre frequency that will determine the sub-channel of most significant end, Fig. 2. at first to select f in the example _1c, choose the resonance frequency f of the filter circuit of the 1st subchannel then _1d, the resonance frequency of present disclosure is aforementioned RLC filter circuit is chosen in the straight line portion of frequency characteristics both sides, selects f in this example _1d＞f _1cThe centre frequency f of the 2nd subchannel _2cSelection should make the most significant end frequency of bandwidth of the 2nd subchannel be lower than f _1d, can avoid of the interference of the 2nd sub-channel signal like this for the 1st sub-channel signal, select f simultaneously _2d＜f _2c, last time selecting f _2cThe basis on adjust the centre frequency f of the 2nd subchannel once more _2cAnd the resonance frequency f of corresponding filter circuit _2d, make the low side frequency of bandwidth of the 1st subchannel be higher than f _2d, avoid of the interference of the 1st sub-channel signal for the 2nd sub-channel signal.Like this, the bandwidth of the 1st and the 2nd subchannel just can have certain overlapping, but so overlapping phase mutual interference that can not cause between above-mentioned two sub-channel signal of bandwidth.From Fig. 2. can see that the 2nd is relative with the distance of 3 subchannels wideer, this is because selected f _2d＜f _2c, selected f simultaneously again _3d＞f _3c, the centre frequency f of the 3rd subchannel _3cSelection should make the most significant end frequency of bandwidth of the 3rd subchannel be lower than f _2d, adjust the centre frequency f of the 3rd subchannel _3c, make the low side frequency of bandwidth of the 2nd subchannel be higher than f _3d, avoid of the interference of the 2nd sub-channel signal for the 3rd sub-channel signal.Can determine the centre frequency f of the 4th subchannel according to the method described above equally _4cResonance frequency f with filter circuit _4dSo just determined the centre frequency of 4 sub-channels.The following frequency range of 1100kHz can also continue to divide more sub-channel according to the method described above in the drawings.Choose bigger M system number and carry out Digital Transmission, can obtain higher transmission rate, but can in the actual transmissions process, increase the difficulty of decoding.Choose bigger pulse group fixed pulse and count N, can reduce the bandwidth of each subchannel, in same frequency range, mark off more subchannel, but the speed of each subchannel to reduce in proportion but, therefore in the process that the physical channel is divided, can weigh the selection of above-mentioned parameter.

Claims (6)

1. the coding method of a multi-system digital signal, M system numeral for M＞2, at first near the Digital Transmission channel central frequency, choose M characteristic frequency f0, f1, f2...fM-1, make it to correspond respectively to a M system digital " 0 ", " 1 ", " 2 ", " M-1 ", then the M system number of coding being corresponded to one has with the corresponding characteristic frequency of this numeral and has the pulse group of fixed pulse number, the fixed pulse of described pulse group is counted N and is at least two, such M system number just corresponds to a pulse group that has corresponding characteristic frequency and have the fixed pulse number, and the sequence of a M system number just corresponds to the sequence of an above-mentioned pulse group.

2. method that the digital signal of the described method of claim 1 coding is transmitted simultaneously by a plurality of passages, it is characterized in that the operating frequency range of transmission medium is divided into a plurality of subchannels, each subchannel has different channel central frequencies, near every channel central frequency, choose each numeral that a plurality of characteristic frequencies make it to correspond respectively to multi-system respectively, then the multi-system number of coding being corresponded to one has with the corresponding characteristic frequency of this numeral and has the pulse group of fixed pulse number, multi-system number in each subchannel corresponds to the pulse group that has corresponding characteristic frequency and have fixed pulse number separately respectively like this, the sequence of the multi-system number in each subchannel just corresponds to the sequence of corresponding pulse group, the pulse group signal of above-mentioned all subchannels is carried out superposition send to transmission medium later on, transmission when carrying out a plurality of subchannel multi-system number, described a plurality of subchannel can transmit different multi-system numbers respectively, and described a plurality of subchannels can be selected different pulse group fixed pulse numbers respectively.

3. a plurality of channel signals to claim 2 transmission carry out the method for bandpass filtering respectively, it is characterized in that the bandpass filtering centre frequency that adopts and the centre frequency of this sub-channel signal have certain frequency shift (FS), described frequency shift (FS) is chosen at the straight line portion of received signal frequency characteristics.

4. specify the later signal of subchannel bandpass filtering to carry out coding/decoding method for claim 3 pair for one kind corresponding to the multilevel code signal of claim 1, it is characterized in that carrying out cutting apart of pulse group for specifying the later signal of subchannel bandpass filtering to count paired pulses group sequence according to fixed pulse, obtain the set of pulse group, have identical pulse number according to described pulse group and carry out the measurement of pulse group duration with different pulse frequencies, the measured pulse group duration is divided into a plurality of different pulse group duration according to their difference, above-mentioned a plurality of different pulse group duration are corresponded to the multi-system number respectively.

5. coding/decoding method according to claim 4 is characterized in that the described pulse group duration is whole adding up the time of the cycle of pulses in the pulse group, perhaps adding up the time for cycle of specific part pulse in the described pulse group.

6. the method for a multichannel frequency band division, it is characterized in that utilizing filter circuit resonance frequency in the claim 3 to depart from the characteristics of the centre frequency of encoding, it is certain overlapped that two adjoining subchannels are had on bandwidth, for two adjoining subchannels, the employing circuit resonant frequencies is moved for the just skew and the negative bias of coding central feature frequency, makes the filter circuit resonance frequency be in outside the broadband of contiguous subchannel.

Images