CN1913415A - Multi-use method of multi-user signal - Google Patents

Abstract

This invention discloses a multiplexing method for multiple user signals, in which, the transmission end takes out KH bits from bit streams of high S/N ratio users and KL bits from those in low S/N users in terms of preset bit values KH and KL, then combines them into bit words and gets the shortest Euclidean distance among clusters and the smallest Euclidean distance among dots in a constellation map at the same time to get the data symbols corresponding to the current bit words to be output.

Description

A kind of multiplexing method of multiple user signals

Technical field

The present invention relates to the descending multiplex technique of radio honeycomb communication, refer to the multiplexing method of multiple user signals in a kind of radio honeycomb communication especially.

Background technology

The down link of radio honeycomb communication system comes down to the broadcast channel model of point-to-multipoint.A plurality of users' signal is at transmitting terminal, and as base station multiplexing back emission, multiplex mode generally adopts the orthogonal multiplex method, and is multiplexing as the time-frequency in OFDM (OFDM) modulating system.In time-frequency was multiplexing, the signal of different user took different running time-frequency resources, and therefore the signal at the receiving terminal different user can not produce the phase mutual interference.Yet, the channel that signal arrives different user often has different features, and the user away from the base station is also referred to as remote subscriber, the signal that generally receives than near-end user receive more weak, thereby cause lower received signal to noise ratio (SNR) and lower spectrum efficiency.

When occurring a large amount of remote subscribers in the system or remote subscriber need use a large amount of running time-frequency resource on a small quantity, will cause the spectrum efficiency of whole base station system to descend greatly.In addition, when the running time-frequency resource of base station system has in limited time, orthogonal multiplex can cause maximum supported number of users, and promptly user capacity is limited, also promptly has hard capacity problem.

In order to address the above problem, number of patent application is 03137566.9, denomination of invention is in the patent application of " cut apart the multiplexing method that combines nonopiate cutting apart with the time-division ", a kind of nonopiate multiplexing method of introducing the multi-user of going up in constellation mapping (Mapping) has been proposed, Fig. 1 be in the described patent application multi-user based on the constellation mapping principle schematic of stacking method, independently carry out constellation mapping respectively from user 1 bit stream with from user 2 bit stream, and, on same running time-frequency resource, superpose according to separately power respectively through after amplifying.

But, in the method that above-mentioned relevant patent proposes, do not retrain because user power and constellation mapping method imposed a condition, in real application systems, some problems can occur, such as power there not being strict restriction can cause each user inconsistent in the output signal performance of receiving terminal; Each user independently carries out constellation mapping, can not carry out global optimization to the data symbol of output, like this, even under identical transmitting power condition, can not utilize the signal performance of power reforming receiving terminal fully.

Summary of the invention

In view of this, main purpose of the present invention is to provide a kind of multiplexing method of multiple user signals, and each user's receptivity is consistent.

For achieving the above object, technical scheme of the present invention specifically is achieved in that

A kind of multiplexing method of multiple user signals, transmitting terminal is carried out following processing to the bit stream of two different signal to noise ratio users in the bit stream that receives:

A. according to default number of bits value K _LAnd K _H, from low signal-to-noise ratio user's bit stream and high s/n ratio user's bit stream, take out K respectively _LIndividual bit and K _HIndividual bit, and with the K that takes out _LIndividual bit and K _HIndividual bit combination generates a constellation point among the bit words constellation figure; Obtain simultaneously in the planisphere bunch between minimum Eustachian distance and the point between minimum Eustachian distance;

B. according in the described planisphere bunch between minimum Eustachian distance between minimum Eustachian distance and point, export after obtaining the data symbol of current bit words correspondence.

Described in the steps A with the K that takes out _LIndividual bit and K _HThe method that individual bit combination generates a bit words is: first bit to the K that takes out from described low signal-to-noise ratio user's bit stream _LIndividual bit-order is arranged in the preceding K of described bit words _LThe position, first bit to the K that from high s/n ratio user's bit stream, takes out _HIndividual bit-order is arranged in the back K of described bit words _HThe position.

Described planisphere is the planisphere that meets the following conditions:

2 ^KLBunch and every bunch comprise 2 ^KHIndividual constellation point, certain constellation point in the bit that from low signal-to-noise ratio user's bit stream, takes out corresponding bunch number, corresponding bunch of the bit that from high s/n ratio user's bit stream, takes out;

Between bunch minimum Eustachian distance square and the point between minimum Eustachian distance square ratio, equal low signal-to-noise ratio user's the signal to noise ratio and the inverse of the ratio of high s/n ratio user's signal to noise ratio; Perhaps, between bunch minimum Eustachian distance square and the point between minimum Eustachian distance square ratio, equal low signal-to-noise ratio user's the signal to noise ratio and the value of the inverse and the default degree of approximation adjusted value added/subtracted gained of the ratio of high s/n ratio user's signal to noise ratio;

The value of each bit words is corresponding one by one with each constellation point.

Between obtaining described in the steps A in the planisphere bunch between minimum Eustachian distance and some the method for minimum Eustachian distance be: according to described number of bits value K _HAnd K _L, known two users the noise given transmitting power P of transmitting terminal when, obtain in the planisphere bunch between minimum Eustachian distance between minimum Eustachian distance and point.

Described obtain in the planisphere bunch between between minimum Eustachian distance and some the method for minimum Eustachian distance be:

A1. according to described number of bits value K _HAnd K _L, select corresponding constellation mapping table;

A2., the average power that all constellation point in the described planisphere are set equals described given transmitting power P, and according to minimum Eustachian distance between described bunch square and point between minimum Eustachian distance square ratio, equal low signal-to-noise ratio user's the signal to noise ratio and the inverse of the ratio of high s/n ratio user's signal to noise ratio, calculate in the planisphere bunch between Euclidean distance between minimum Eustachian distance and point; Perhaps, according to bunch between minimum Eustachian distance square and the point between minimum Eustachian distance square ratio, equal low signal-to-noise ratio user's the signal to noise ratio and the value of the inverse and the default degree of approximation adjusted value added/subtracted gained of the ratio of high s/n ratio user's signal to noise ratio, calculate in the planisphere bunch between Euclidean distance between minimum Eustachian distance and point.

Select the method for corresponding constellation mapping table to be described in the steps A 1: according to the number of bits value K that is pre-stored in transmitting terminal _HAnd K _LWith the one-to-one relationship of constellation mapping table, select corresponding constellation mapping table.

The method of calculating the average power of constellation point in the planisphere described in the steps A 2 is:

A21. from described constellation mapping table, take out the data symbol of each constellation point;

A22. respectively the data symbol of each constellation point is taken absolute value back square, will square after each income value addition, calculate the merchant of income value and constellation point sum after the addition at last.

Between obtaining described in the steps A in the planisphere bunch between minimum Eustachian distance and some the method for minimum Eustachian distance be: minimum Eustachian distance is a normalized value between default described point; According to minimum Eustachian distance between the point in two users' signal to noise ratio and the default planisphere, obtain in the planisphere bunch between minimum Eustachian distance.

Described obtain in the planisphere bunch between the method for minimum Eustachian distance be: the merchant who calculates described high s/n ratio user's signal to noise ratio and low signal-to-noise ratio user's signal to noise ratio, gained merchant minimum Eustachian distance square long-pending again and between the point in the described default planisphere calculated the square root of income value at last.

This method also comprises: default degree of approximation adjusted value;

Described obtain in the planisphere bunch between the method for minimum Eustachian distance be: the merchant who calculates described high s/n ratio user's signal to noise ratio and low signal-to-noise ratio user's signal to noise ratio, minimum Eustachian distance square long-pending again and between the point in the described default planisphere after gained merchant and the default degree of approximation adjusted value added/subtracted calculated the square root of income value at last.

The method of obtaining the data symbol of current bit words correspondence described in the step B is:

B1. the bit words that generates according to steps A, inquiry is used for storing the bit words of planisphere and the constellation mapping table that prestores of the corresponding relation of data symbol computing formula, obtains the data symbol computing formula of correspondence;

B2. according in the planisphere of described acquisition bunch between the data symbol computing formula of Euclidean distance and acquisition between point in the Euclidean distance, default planisphere, calculate the data symbol of described current bit words correspondence.

Described data symbol computing formula comprises real part computing formula and imaginary part computing formula;

Described data symbol computing formula according in the described planisphere of difference bunch between between minimum Eustachian distance and some minimum Eustachian distance draw, and be preset in transmitting terminal.

The method of obtaining the data symbol of current bit words correspondence described in the step B is:

According to described bit words, inquiry is used for storing the bit words of planisphere and the constellation correspondence table of presetting of the corresponding relation of data symbol, obtains the data symbol of current bit words correspondence.

This method further comprises: in the default planisphere bunch between minimum Eustachian distance;

Data symbol in the described constellation correspondence table according in the described planisphere of difference bunch between between minimum Eustachian distance and some minimum Eustachian distance draw, and be preset in transmitting terminal.

Described planisphere further meets the following conditions:

Between different bunches, between all adjacent constellation point bunch between Euclidean distance equate and be bunch between minimum Eustachian distance; In cluster, the Euclidean distance between all adjacent constellation point equates and is minimum Eustachian distance between point;

Described constellation point adopts Gray code Gray to encode.

As seen from the above technical solution, the present invention is to two users that signal to noise ratio is different, and transmitting terminal is according to default number of bits value K _HAnd K _L, from high s/n ratio user's bit stream, take out K respectively _HIndividual bit takes out K from low signal-to-noise ratio user's bit stream _LIndividual bit; Afterwards, transmitting terminal is with the K that takes out _HIndividual bit and K _LIndividual bit combination generates a bit words, and compound mode can be first bit to the K that takes out from low signal-to-noise ratio user's bit stream _LIndividual bit-order is arranged in the preceding K of this bit words _LThe position, first bit to the K that from high s/n ratio user's bit stream, takes out _HIndividual bit-order is arranged in the back K of this bit words _HPosition, i.e. c _{L, 0}C _{L, KL-1}c _{H, 0}C _{H, KN-1}, wherein, c _{L, n}And c _{H, m}Represent n bit and m bit of from low signal-to-noise ratio user and high s/n ratio user's bit stream, taking out respectively; Transmitting terminal is according to number of bits value K simultaneously _HAnd K _L, known two users the noise given transmitting power P of transmitting terminal when, obtain in the planisphere bunch between minimum Eustachian distance and the point between minimum Eustachian distance, perhaps transmitting terminal is according to minimum Eustachian distance between the point in two known users signal to noise ratio and the default planisphere, obtain in the planisphere bunch between minimum Eustachian distance; At last, transmitting terminal according in the planisphere that obtains bunch between minimum Eustachian distance between minimum Eustachian distance and point, inquiry transmitting terminal being used for of prestoring stored the constellation mapping table of the corresponding relation of planisphere bit words and data symbol computing formula, perhaps be used for storing the constellation correspondence table of the corresponding relation of planisphere bit words and data symbol, obtain to export behind the data symbol of current bit words correspondence.The method of multi-user association constellation mapping of the present invention has guaranteed the consistency of each user in the output signal performance of receiving terminal, under the condition of given transmitting power, makes receiving terminal obtain more performance even higher spectrum efficiency simultaneously.

Described planisphere is divided into 2 ^KLBunch, every bunch comprises 2 ^KHIndividual constellation point, certain constellation point in low signal-to-noise ratio user's the bit corresponding bunch number, corresponding bunch of high s/n ratio user's bit.Divide cluster and two levels of point with all constellation point, help receiving terminal and control demodulation performance by the constellation point of control different levels.

In the described planisphere bunch between between least square Euclidean distance and some the ratio of least square Euclidean distance equal the inverse of the ratio of respective user signal to noise ratio, promptly equal low signal-to-noise ratio user's the signal to noise ratio and the inverse of the ratio of high s/n ratio user's signal to noise ratio, guaranteed that each user is behind the data symbol of receiving terminal demodulation from transmitting terminal output, each user's error sign ratio approximately equal, thus make each user's receptivity keep unanimity.In addition, if satisfy between different bunches, between all adjacent constellation point bunch between Euclidean distance equate, promptly be d _{Min, L}In cluster, the Euclidean distance between all adjacent constellation point equates, is d _{Min, H}, the average power minimum of the data symbol of bit words correspondence has further guaranteed under the condition of given transmitting power so, receiving terminal is obtained more performance even higher spectrum efficiency.

Simultaneously, in the planisphere of the present invention, the unique corresponding constellation point of each possibility value of bit words, a kind of value of the also unique corresponding bit word of each constellation point has guaranteed that receiving terminal can correctly carry out demodulation to the data symbol from transmitting terminal.

In addition, the constellation point in the described planisphere is encoded according to Gray code (Gray), that is: with in the cluster, Euclidean distance is that two adjacent pairing bit words of constellation point of minimum Eustachian distance between point only have a bit difference; Between different bunches, Euclidean distance for bunch between two pairing bit words of constellation point of minimum Eustachian distance also have only a bit difference.Like this, taking place under the situation that the symbol erroneous judgement determines, constellation point of the present invention adopts the processing of Gray code coding, has guaranteed that errored bit counts minimum.

Description of drawings

Fig. 1 is the constellation mapping principle schematic of prior art multi-user based on stacking method;

Fig. 2 is the multiplexing principle schematic of multiple user signals of the present invention;

Fig. 3 is the flow chart of the multiplexing method of multiple user signals of the present invention;

Fig. 4 a is the preferable planisphere one of the present invention;

Fig. 4 b is the preferable planisphere two of the present invention;

Fig. 4 c is the preferable planisphere three of the present invention;

Fig. 4 d is the preferable planisphere four of the present invention;

Fig. 4 e is the preferable planisphere five of the present invention;

Fig. 4 f is the preferable planisphere six of the present invention.

Embodiment

Core concept of the present invention is: to two users that signal to noise ratio is different, transmitting terminal is according to default number of bits value K _HAnd K _L, from high s/n ratio user's bit stream, take out K respectively _HIndividual bit takes out K from low signal-to-noise ratio user's bit stream _LIndividual bit; Afterwards, transmitting terminal is with the K that takes out _LIndividual bit and K _HIndividual bit combination generates a bit words; Obtain simultaneously in the planisphere bunch between minimum Eustachian distance and the point between minimum Eustachian distance; At last, transmitting terminal according in the planisphere that obtains bunch between minimum Eustachian distance between point in minimum Eustachian distance and the default planisphere, obtain to export behind the data symbol of current bit words correspondence.

For making purpose of the present invention, technical scheme and advantage clearer, below with reference to the accompanying drawing preferred embodiment that develops simultaneously, the present invention is described in more detail.

Fig. 2 is the multiplexing principle schematic of multiple user signals of the present invention, as can be seen from Figure 2, transmitting terminal generates bit words according to the bit stream from different signal to noise ratio users, and the bit words that generates united after constellation mapping handles the data symbol of output bit words correspondence.In conjunction with Fig. 3, Fig. 3 is the flow chart of the multiplexing method of multiple user signals of the present invention, hypothesis here, and different and signal to noise ratio user 1 of user 1 and user's 2 signal to noise ratio is higher than user 2 signal to noise ratio, claims that user 1 is the high s/n ratio user, and signal to noise ratio is SNR _HClaim that user 2 is the low signal-to-noise ratio user, signal to noise ratio is SNR _L, and transmitting terminal has been known SNR _HAnd SNR _LThe while transmitting terminal has been provided with the number of bits of taking out from high s/n ratio user's bit stream be K _HIndividual, the number of bits of taking out from low signal-to-noise ratio user's bit stream is K _LIndividual, K _HAnd K _LValue and user's the required satisfied error rate of noise receiving terminal demodulation when etc. relevant; After transmitting terminal received two different signal to noise ratio users' bit stream, the method for multi-user association constellation mapping of the present invention specifically described as follows:

Step 300: transmitting terminal is according to the number of bits preset value, always takes out K in user 1 bit stream and the bit stream from user 2 respectively _HIndividual bit and K _LIndividual bit, and two groups of bits, i.e. K that will take out _HIndividual bit and K _LIndividual bit combination generates a bit words.

When two groups of bit combinations generate a bit words, compound mode is a lot, as long as transmitting terminal and receiving terminal are unified and low signal-to-noise ratio user's bit corresponding bunch number, certain constellation point in corresponding bunch of high s/n ratio user's the bit get final product, and wherein a kind of compound mode of act describes here: first bit to the K that takes out from low signal-to-noise ratio user's bit stream _LIndividual bit-order is arranged in the preceding K of this bit words _LThe position, first bit to the K that from high s/n ratio user's bit stream, takes out _HIndividual bit-order is arranged in the back K of this bit words _HPosition, i.e. c _{L, 0}C _{L, KL-1}c _{H, 0}C _{H, KN-1}, wherein, c _{L, n}And c _{H, m}Represent n bit and m bit of from low signal-to-noise ratio user and high s/n ratio user's bit stream, taking out respectively.The distribution situation of the bit words that transmitting terminal generates can adopt planisphere to represent visually, and promptly the label of constellation point correspondence is a bit words, and the corresponding data symbol in the position of each constellation point in planisphere.

Planisphere of the present invention is divided into 2 ^KLBunch, every bunch comprises 2 ^KHIndividual constellation point, certain constellation point in low signal-to-noise ratio user's the bit corresponding bunch number, corresponding bunch of high s/n ratio user's bit.Divide cluster and two levels of point with all constellation point, help receiving terminal and control demodulation performance by the bit of control different levels.

In order to guarantee that receiving terminal can correctly carry out demodulation to the data symbol from transmitting terminal, in the planisphere of the present invention, the necessary unique corresponding constellation point of each possibility value of bit words, a kind of value of the also unique corresponding bit word of each constellation point, promptly the value of bit words is corresponding one by one with constellation point.

Constellation point can adopt natural code, Gray code (Gray) etc. to encode, but, if adopt Gray code to encode, that is: with in the cluster, Euclidean distance only has a bit difference for two adjacent pairing bit words of constellation point of minimum Eustachian distance between point; Between different bunches, Euclidean distance for bunch between two pairing bit words of constellation point of minimum Eustachian distance also have only a bit difference, like this, taking place under the symbol erroneous judgement situation certainly, constellation point of the present invention adopts the processing of Gray code coding, guaranteed that errored bit counts minimum, guaranteed further that under the condition of given transmitting power receiving terminal is obtained more performance even higher spectrum efficiency.

In addition, consider practicality, get K usually _L=1 or 2, and K _L+ K _H≤ 4.

Step 301: transmitting terminal obtain in the planisphere bunch between minimum Eustachian distance between point in minimum Eustachian distance and the planisphere.

Usually, minimum Eustachian distance is generally used d between the point _{Min, H}Expression, minimum Eustachian distance is generally used d between bunch _{Min, L}Expression.

According to prior art, based on the down link of OFDM, transmitted power is P, and two different signal to noise ratio users' received signal to noise ratio is respectively so:

${\mathrm{SNR}}_H=\frac{P{\left|h_H\right|}^2}{N_H}---\left(1\right)$

${\mathrm{SNR}}_L=\frac{P{\left|h_L\right|}^2}{N_L}---\left(2\right)$

In formula (1) and (2), h _HBe high s/n ratio user's channel impulse response, h _LChannel impulse response for the low signal-to-noise ratio user; N _HBe the additive noise power in the channel of high s/n ratio user place, N _LBe the additive noise power in the channel of low signal-to-noise ratio user place.Here, the signal to noise ratio just mentioned in the literary composition of received signal to noise ratio.

If receiving terminal adopts the data symbol information of maximum-likelihood criterion demodulation from receiving terminal output, the error sign ratio P after the high s/n ratio user demodulation so _{S, H}With the error sign ratio P after the low signal-to-noise ratio user demodulation _{S, L}Approximate representation is shown in formula (3) and the formula (4) respectively:

$P_{s,H}\&ap;M_{H}Q\left(\sqrt{\frac{d_{\min ,\mathrm{<figure-callout\; id="2"\; label="H"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">H</figure-callout>}}^2{\left|h_H\right|}^2}{N_H}}\right)---\left(3\right)$

$P_{s,L}\&ap;M_{L}Q\left(\sqrt{\frac{d_{\min ,\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">L</figure-callout>}}^2{\left|h_L\right|}^2}{N_L}}\right)---\left(4\right)$

In formula (3) and (4), M _H, M _LBe two constants, d _{Min, L} ²And d _{Min, H} ²Represent respectively in the planisphere bunch between least square Euclidean distance between point in least square Euclidean distance and the planisphere.

Among the present invention, suppose $\frac{d_{\min ,L}^2}{d_{\min ,\mathrm{<figure-callout\; id="2"\; label="H"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">H</figure-callout>}}^2}=\frac{{\mathrm{SNR}}_H}{{\mathrm{SNR}}_L}---\left(5\right)$

So,, be not difficult to draw two users' error sign ratio approximately equal, in the light of actual conditions, can ignore M here according to above formula (1)～(4) _HAnd M _LInfluence.

Based on above analysis, according to formula (5), at transmitting terminal, between adopt satisfying bunch between least square Euclidean distance and some the ratio of least square Euclidean distance equal the inverse of the ratio of respective user signal to noise ratio, promptly equal low signal-to-noise ratio user's the signal to noise ratio and the planisphere of the inverse of the ratio of high s/n ratio user's signal to noise ratio, user 1 and user 2 have been guaranteed behind the data symbol of receiving terminal demodulation, P from transmitting terminal output _{S, H}With P _{S, L}Approximately equal, thus the consistency of each user's receptivity guaranteed.

From last surface analysis as can be seen, since user's error sign ratio mainly by bunch between minimum Eustachian distance decision between minimum Eustachian distance and point, therefore under the condition that guarantees certain error sign ratio, if between different bunches, between all adjacent constellation point bunch between Euclidean distance equate, promptly be d _{Min, L}In cluster, the Euclidean distance between all adjacent constellation point equates, is d _{Min, H}, the average power minimum of the data symbol of bit words correspondence so.If with in the cluster/different bunches between adjacent constellation point distance greater than minimum Eustachian distance between point/bunch between minimum Eustachian distance, so can be the point of correspondence/bunch move to the initial point direction, make Euclidean distance be reduced to minimum range, so both reduced power, guaranteed that also minimum Eustachian distance is constant simultaneously, that is to say, the performance that has guaranteed demodulation is constant, like this, has further guaranteed, under the condition of given transmitting power, can make receiving terminal obtain more performance even higher spectrum efficiency.

Need to prove,, can further improve P by adjustment to formula (5) according to actual conditions _{S, H}With P _{S, L}The degree of approximation such as setting degree of approximation adjusted value Δ, and is adjusted into formula (5):

$\frac{d_{\min ,L}^2}{d_{\min ,\mathrm{<figure-callout\; id="2"\; label="H"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">H</figure-callout>}}^2}=\frac{{\mathrm{SNR}}_H}{{\mathrm{SNR}}_L}\&PlusMinus;\mathrm{\&Delta;}---\left(5^{\&prime;}\right)$

The span of Δ is generally in the formula (5 '): 0＜Δ＜1, and such as Δ=0.5 etc.In this step, d _{Min, H}, SNR _HAnd SNR _LBe known,, draw d easily according to formula (5) or formula (5 ') _{Min, L}, promptly $d_{\min ,L}=\sqrt{d_{\min ,\mathrm{<figure-callout\; id="2"\; label="H"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">H</figure-callout>}}^2\&times;\frac{{\mathrm{SNR}}_H}{{\mathrm{SNR}}_L}}---\left(6\right)$

Or $d_{\min ,L}=\sqrt{d_{\min ,\mathrm{<figure-callout\; id="2"\; label="H"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">H</figure-callout>}}^2\&times;(\frac{{\mathrm{SNR}}_H}{{\mathrm{SNR}}_L}\&PlusMinus;\mathrm{\&Delta;})}---\left(6^{\&prime;}\right)$

In this step, suppose default d _{Min, L}Be normalized value, promptly equal 1, then, can calculate d according to formula (5) or (5 ') and known user's signal to noise ratio _{Min, H}

In addition, receiving terminal can also be according to the given transmitting power P of transmitting terminal, known user's signal to noise ratio and default number of bits value K _HAnd K _L, obtain in the planisphere bunch between minimum Eustachian distance and the point between minimum Eustachian distance, be implemented as follows:

As mentioned above, according to K _HAnd K _LCan determine the planisphere type that adopts, also just determine the constellation mapping table that adopts, at transmitting terminal K _HAnd K _LWith the constellation mapping table be corresponding stored one by one; In addition, the average power of all constellation point is in the planisphere: the data symbol of each constellation point is taken absolute value back square, will square after each income value addition, calculate the merchant of income value and constellation point sum after the addition at last, shown in formula (7) the equation left side, and the average power of all constellation point equals given transmitting power P in the planisphere, that is:

$\frac{1}{M}\underset{i=0}{\overset{M-1}{\mathrm{\&Sigma;}}}{\left|s_i\right|}^2=P---\left(7\right)$

Wherein, $M=2^{K_H+K_L},$ Be used for representing the sum of planisphere constellation point; s _iThe real part of constellation point and imaginary part, i.e. s _iThe value of the value of=real part+i imaginary part, the computing formula of real part and imaginary part can be found out from the described constellation mapping table that prestores.

Like this, combinatorial formula (5) and formula (7), or combinatorial formula (5 ') and formula (7), can calculate in the planisphere bunch between minimum Eustachian distance and the point between minimum Eustachian distance.

For example: suppose K _H=1, K _L=1, SNR _H=4, SNR _L=1, P=1, so according to number of bits value KH and KL, the constellation mapping table of determining the planisphere correspondence is constellation mapping table 1 and the formula (7) in the step 302, is not difficult to draw average power and equals $\frac{1}{4}(d_{\min ,\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">L</figure-callout>}}^2+d_{\min ,H}^2),$ If this power equals given transmitting power 1, just have

$\frac{1}{4}(d_{\min ,\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">L</figure-callout>}}^2+d_{\min ,H}^2)=1$

Simultaneously, have according to formula (5):

$\frac{d_{\min ,L}^2}{d_{\min ,\mathrm{<figure-callout\; id="2"\; label="H"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">H</figure-callout>}}^2}=\frac{{\mathrm{SNR}}_H}{{\mathrm{SNR}}_L}=4$

According to two unknown number d of above-mentioned two solution of equations _{Min, L}And d _{Min, H}, then calculate and can get:

$d_{\min ,H}=2/\sqrt{5},d_{\min ,L}=4/\sqrt{5}$

Need to prove that step 300 and step 301 can be regardless of execution sequence.

Step 302: transmitting terminal according in the planisphere that obtains bunch between minimum Eustachian distance between minimum Eustachian distance and point, obtain the data symbol of bit words correspondence after, with the data symbol output that obtains.

In order to obtain the data symbol of bit words correspondence, the invention provides two kinds of methods:

Method one: the default constellation mapping table of inquiry.

The constellation mapping table is used for storing the corresponding relation of planisphere bit words and data symbol computing formula, and wherein the data symbol computing formula comprises real part computing formula and two parts of imaginary part computing formula.Planisphere draws and is preset in transmitting terminal to the constellation mapping table according to the present invention, the constellation mapping table that different planispheres is corresponding different.

Fig. 4 a is the preferable planisphere one of the present invention, and among Fig. 4 a, the composition of bit words is: c _{L, 0}c _{H, 0}, i.e. K _HBe 1, K _LBe 1, planisphere one is divided into 2 ^KLBunch, promptly 2 bunches; Every bunch comprises 2 ^KH, i.e. 2 constellation point.As seen, bit words adopts Gray code to encode from Fig. 4 a, and wherein cluster comprises bit words 10 and bit words 11, and another bunch comprises bit words 00 and bit words 01; Euclidean distance shown in 11 of bit words 01 and the bit words is d _{Min, L}, the Euclidean distance shown in 10 of bit words 11 and the bit words is d _{Min, H}, according to the planisphere one shown in Fig. 4 a, draw the constellation mapping table of this planisphere correspondence easily, as shown in table 1:

Bit words	The real part of constellation point	The imaginary part of constellation point
			00	d min，L/2	-d min，H/2
01	d mim，L/2	d min，H/2
			10	-d min，L/2	-d min，H/2
11	-d min，L/2	d min，H/2

Table 1

With planisphere shown in Fig. 4 a is example, supposes that the bit words that generates in the step 300 is 01, and in this step, by inquiring about default table 1, the data symbol computing formula that can obtain bit words 01 correspondence is respectively: real part equals d _{Min, L}/ 2, imaginary part equals d _{Min, H}/ 2; Then according to the d that obtains in the step 301 _{Min, L}And d _{Min, H}, can calculate the data symbol of bit words 01 correspondence.

Fig. 4 b is the preferable planisphere two of the present invention, and among Fig. 4 b, the composition of bit words is: c _{L, 0}c _{H, 0}c _{H, 1}, i.e. K _HBe 2, K _LBe 1, planisphere two is divided into 2 ^KLBunch, promptly 2 bunches; Every bunch comprises 2 ^KH, i.e. 4 constellation point.As seen, bit words adopts Gray code to encode from Fig. 4 b, and wherein cluster comprises bit words 110, bit words 111, and bit words 101 and bit words 100, another bunch comprises bit words 010, bit words 011, bit words 001 and bit words 000; Euclidean distance shown in 101 of bit words 001 and the bit words is d _{Min, L}, the Euclidean distance shown in 111 of bit words 101 and the bit words is d _{Min, H}, according to the planisphere two shown in Fig. 4 b, draw the constellation mapping table of this planisphere correspondence easily, as shown in table 2:

Bit words	The real part of constellation point	The imaginary part of constellation point
			000	d min，L/2	3d min，H/2
001	d min，L/2	d min，H/2
			010	d min，L/2	-3d min，H/2
011	d min，L/2	-d min，H/2
			100	-d min，L/2	3d min，H/2

101	-d min，L/2	d min，H/2
			110	-d min，L/2	-3d min，H/2
111	-d min，L/2	-d min，H/2

Table 2

With planisphere shown in Fig. 4 b is example, supposes that the bit words that generates in the step 300 is 010, and in this step, by inquiring about default table 2, the computing formula that can obtain the data symbol of bit words 010 correspondence is respectively: real part equals d _{Min, L}/ 2, imaginary part equals-3d _{Min, H}/ 2; Then according to the d that obtains in the step 301 _{Min, L}And d _{Min, H}, can calculate the data symbol of bit words 010 correspondence.

Fig. 4 c is the preferable planisphere three of the present invention, and among Fig. 4 c, the composition of bit words is: c _{L, 0}c _{H, 0}c _{H, 1}c _{H, 2}, i.e. K _HBe 3, K _LBe 1, planisphere three is divided into 2 ^KLBunch, promptly 2 bunches; Every bunch comprises 2 ^KH, i.e. 8 constellation point.From Fig. 4 c as seen, bit words adopts Gray code to encode, wherein cluster comprises bit words 1110,1010,1111,1011,1101,1001,1100 and bit words 1000, and another bunch comprises bit words 0010,0110,0011,0111,0001,0101,0000 and bit words 0100; Euclidean distance shown in 1000 of bit words 0000 and the bit words is d _{Min, L}, the Euclidean distance shown in 1001 of bit words 1000 and the bit words is d _{Min, H}, according to the planisphere three shown in Fig. 4 c, draw the constellation mapping table of this planisphere correspondence easily, as shown in table 3:

Bit words	The real part of constellation point	The imaginary part of constellation point
			0000	d min，L/2	3d min，H/2
0001	d min，L/2	d min，H/2
			0010	d min，L/2	-3d min，H/2
0011	d min，L/2	-d min，H/2
			0100	d min，L/2+d min，H	3d min，H/2
0101	d min，L/2+d min，H	d min，H/2
			0110	d min，L/2+d min，H	-3d min，H/2
0111	d min，L/2+d min，H	-d min，H/2
			1000	-d min，L/2	3d min，H/2

1001	-d min，L/2	d min，H/2
			1010	-d min，L/2	-d min，H/2
1011	-d min，L/2	-d min，H/2
			1100	-d min，L/2-d min，H	3d min，H/2
1101	-d min，L/2-d min，H	d min，H/2
			1110	-d min，L/2-d min，H	-d min，H/2
1111	-d min，L/2-d min，H	-d min，H/2

Table 3

With planisphere shown in Fig. 4 c is example, supposes that the bit words that generates in the step 300 is 0101, and in this step, by inquiring about default table 3, the computing formula that can obtain the data symbol of bit words 0101 correspondence is respectively: real part equals d _{Min, L}/ 2+d _{Min, H}, imaginary part equals d _{Min, H}/ 2; Then according to the d that obtains in the step 301 _{Min, L}And d _{Min, H}, can calculate the data symbol of bit words 0101 correspondence.

Fig. 4 d is the preferable planisphere four of the present invention, and among Fig. 4 d, the composition of bit words is: c _{L, 0}c _{L, 1}c _{H, 0}, i.e. K _HBe 1, K _LBe 2, planisphere four is divided into 2 ^KLBunch, promptly 4 bunches; Every bunch comprises 2 ^KH, i.e. 2 constellation point.From Fig. 4 d as seen, bit words adopts Gray code to encode, and wherein cluster comprises that bit words 000 and bit words comprise bit words 100 and bit words 101 for 001, the second bunch, the 3rd bunch comprises bit words 111 and bit words 110, and last cluster comprises bit words 010 and bit words 011; Euclidean distance shown in 000 of bit words 100 and the bit words is d _{Min, L}, the Euclidean distance shown in 101 of bit words 100 and the bit words is d _{Min, H}, and the line of 101 of bit words 100 and bit words and transverse axis are in angle of 45 degrees, according to the planisphere four shown in Fig. 4 d, draw the constellation mapping table of this planisphere correspondence easily, as shown in table 4:

Bit words	The real part of constellation point	The imaginary part of constellation point
			000	d min，L/2	d min，L/2+d min，H/
001	d min，L/2+d min，H/	d min，L/2
			010	d min，L/2	-d min，L/2-d min，H/
011	d min，L/2+d min，H/	-d min，L/2

100	-d min，L/2	d min，L/2+d min，H/
			101	-d min，L/2-d min，H/	d min，L/2
110	-d min，L/2	-d min，L/2-d min，H/
			111	-d min，L/2-d min，H/	-d min，L/2

Table 4

With planisphere shown in Fig. 4 d is example, supposes that the bit words that generates in the step 300 is 101, and in this step, by inquiring about default table 4, the computing formula that can obtain the data symbol of bit words 101 correspondences is respectively: real part equals-d _{Min, L}/ 2-d _{Min, H}/ , imaginary part equals d _{Min, L}/ 2; Then according to the d that obtains in the step 301 _{Min, L}And d _{Min, H}, can calculate the data symbol of bit words 101 correspondences.

Fig. 4 e is the preferable planisphere five of the present invention, and among Fig. 4 e, the composition of bit words is: c _{L, 0}c _{L, 1}c _{H, 0}c _{H, 1}, i.e. K _HBe 2, K _LBe 2, planisphere five is divided into 2 ^KLBunch, promptly 4 bunches; Every bunch comprises 2 ^KH, i.e. 4 constellation point.From Fig. 4 e as seen, bit words adopts Gray code to encode, wherein cluster comprises bit words 0000,0010,0011 and bit words 0001, second bunch comprises bit words 1010,1000,1001 and bit words 1011, the 3rd bunch comprises bit words 1101,1111,1100 and bit words 1110, and last cluster comprises bit words 0111,0101,0110 and bit words 0100; Euclidean distance shown in 0011 of bit words 1011 and the bit words is d _{Min, L}, the Euclidean distance shown in 0000 of bit words 0010 and the bit words is d _{Min, H}, according to the planisphere five shown in Fig. 4 e, draw the constellation mapping table of this planisphere correspondence easily, as shown in table 5:

Bit words	The real part of constellation point	The imaginary part of constellation point
			0000	d min，L/2+d min，H	d min，L/2+d min，H
0001	d min，L/2+d min，H	d min，L/2
			0010	d min，L/2	d min，L/2+d min，H
0011	d min，L/2	d min，L/2
			0100	d min，L/2+d min，H	-d min，L/2-d min，H
0101	d min，L/2+d min，H	-d min，L/2
			0110	d min，L/2	-d min，L/2-d min，H

0111	-d min，L/2	-d min，L/2
			1000	-d min，L/2-d min，H	d min，L/2+d min，H
1001	-d min，L/2-d min，H	d min，L/2
			1010	-d min，L/2	d min，L/2+d min，H
1011	-d min，L/2	d min，L/2
			1100	-d min，L/2-d min，H	-d min，L/2-d min，H
1101	-d min，L/2-d min，H	-d min，L/2
			1110	-d min，L/2	-d min，L/2-d min，H
1111	-d min，L/2	-d min，L/2

Table 5

With planisphere shown in Fig. 4 e is example, supposes that the bit words that generates in the step 300 is 1110, and in this step, by inquiring about default table 5, the computing formula that can obtain the data symbol of bit words 1110 correspondences is respectively: real part equals-d _{Min, L}/ 2, imaginary part equals-d _{Min, L}/ 2-d _{Min, H}Then according to the d that obtains in the step 301 _{Min, L}And d _{Min, H}, can calculate the data symbol of bit words 1110 correspondences.

Be example only below, the specific implementation method of step 302 of the present invention has been described,, enumerate no longer one by one here for other planisphere that meets requirement of the present invention with five preferable planispheres that meet requirement of the present invention.All can adopt the inventive method to obtain the data symbol of transmitting terminal bit words correspondence so long as meet the planisphere of planisphere requirement of the present invention, reaching the receptivity that the present invention guaranteed each user simultaneously is consistent, simultaneously under the condition of given transmitting power, make receiving terminal obtain the purpose of more performance even higher spectrum efficiency.

Method two: the default constellation correspondence table of inquiry.

The constellation correspondence table is used for storing the corresponding relation of planisphere bit words and data symbol, and wherein data symbol comprises the value of real part and two parts of value of imaginary part.Constellation correspondence table constellation mapping table according to the present invention calculates and is preset in transmitting terminal, and the parameter in certain planisphere of the present invention is as d _{Min, H}And d _{Min, L}Deng the corresponding different constellation correspondence table of different values, that is to say that certain opens constellation mapping table described in the method one, according to d _{Min, H}And d _{Min, L}Different value, corresponding different constellation correspondence table.

Like this, in step 302, only need the d that obtains according in the bit words that generates in the step 300, the step 301 _{Min, L}And d _{Min, H},, can directly draw the data symbol of bit words correspondence by the default constellation correspondence table of inquiry transmitting terminal.

Generally can adopt the processing of method two, store the constellation correspondence table of several parameters commonly used of several planisphere correspondences at transmitting terminal in advance, can save loaded down with trivial details computational process like this according to actual conditions.

Method from above multi-user association constellation mapping of the present invention, as can be seen, according to pairing constellation mapping table of the planisphere that satisfies requirement of the present invention or constellation correspondence table, the data symbol that after transmitting terminal is united constellation mapping to multi-user's combined bit word, obtains, when having guaranteed this data symbol of receiving terminal demodulation, each user's error sign ratio approximately equal, thus the consistency of receptivity guaranteed; In addition, under the condition of given transmitting power, can make receiving terminal obtain more performance even higher spectrum efficiency.

Need to prove, more than be that example is described in detail the inventive method only with two users' associating constellation mapping, method of the present invention also is applicable to three or three above users' associating constellation mapping.

The preferable planisphere six of the present invention that Fig. 4 f is is example with three users, the signal to noise ratio of supposing user 1 is SNR ₁, user 2 signal to noise ratio is SNR ₂, user 3 signal to noise ratio is SNR ₃, and SNR ₁＜SNR ₂＜SNR ₃, then:

$\frac{d_{\min ,1}^2}{d_{\min ,2}^2}=\frac{{\mathrm{<figure-callout\; id="2"\; label="SNR"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">SNR</figure-callout>}}_2}{{\mathrm{<figure-callout\; id="1"\; label="SNR"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">SNR</figure-callout>}}_1},\frac{d_{\min ,2}^2}{d_{\min ,3}^2}=\frac{{\mathrm{<figure-callout\; id="3"\; label="SNR"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">SNR</figure-callout>}}_3}{{\mathrm{<figure-callout\; id="2"\; label="SNR"\; filenames="A20051009124300301.png"\; state="\{\{state\}\}">SNR</figure-callout>}}_2}$

Among Fig. 4 f, the composition of bit words is: c ₁c ₂c ₃, i.e. K ₁Be 1, K ₂Be 1, K ₃Be 1, as seen, bit words adopts Gray code to encode among Fig. 4 f, and planisphere six is divided into 2 ^KLBig bunch, promptly 2 big bunches; First bunch of bit words is 000,001,010,011, and second largest bunch comprises bit words 100,101,110,111; Each big bunch comprises two 2 ^K2Tuftlet, i.e. 2 each tuftlet, first bunch first tuftlet is 000,001, first bunch second tuftlet is 010,011; Second largest bunch first tuftlet is 100,101, and second largest bunch second tuftlet is 110,111.Tuftlet number in user 1 the bit identification big bunch number, one big bunch of user 2 bit identification, user 3 then identifies certain point under the tuftlet in one big bunch.。From the Euclidean distance shown in 110 of bit words 010 and the bit words is d _{Min, 1}, the Euclidean distance shown in 010 of bit words 011 and the bit words is d _{Min, 3}, the Euclidean distance shown in 001 of bit words 011 and the bit words is d _{Min, 2}, according to the planisphere six shown in Fig. 4 e, draw the constellation mapping table of this planisphere correspondence easily, as shown in table 6:

Bit words	The real part of constellation point	The imaginary part of constellation point
			000	d min，2/2	d min，3/2
001	d min，2/2	d min，3/2+d min，1
			010	-d min，2/2	d min，3/2
011	-d min，2/2	d min，3/2+d min，1
			100	d min，2/2	-d min，3/2
101	d min，2/2	-d min，3/2-d min，1
			110	-d min，2/2	-d min，3/2
111	-d min，2/2	-d min，3/2-d min，1

Table 6

The above is preferred embodiment of the present invention only, is not to be used to limit protection scope of the present invention.

Claims (15)

1. the multiplexing method of a multiple user signals is characterized in that, transmitting terminal is carried out following processing to the bit stream of two different signal to noise ratio users in the bit stream that receives:

A. according to default number of bits value K _LAnd K _H, from low signal-to-noise ratio user's bit stream and high s/n ratio user's bit stream, take out K respectively _LIndividual bit and K _HIndividual bit, and with the K that takes out _LIndividual bit and K _HIndividual bit combination generates a constellation point among the bit words constellation figure; Obtain simultaneously in the planisphere bunch between minimum Eustachian distance and the point between minimum Eustachian distance;

B. according in the described planisphere bunch between minimum Eustachian distance between minimum Eustachian distance and point, export after obtaining the data symbol of current bit words correspondence.

2. method according to claim 1 is characterized in that, described in the steps A with the K that takes out _LIndividual bit and K _HThe method that individual bit combination generates a bit words is: first bit to the K that takes out from described low signal-to-noise ratio user's bit stream _LIndividual bit-order is arranged in the preceding K of described bit words _LThe position, first bit to the K that from high s/n ratio user's bit stream, takes out _HIndividual bit-order is arranged in the back K of described bit words _HThe position.

3. method according to claim 1.It is characterized in that described planisphere is the planisphere that meets the following conditions:

2 ^KLBunch and every bunch comprise 2 ^KHIndividual constellation point, certain constellation point in the bit that from low signal-to-noise ratio user's bit stream, takes out corresponding bunch number, corresponding bunch of the bit that from high s/n ratio user's bit stream, takes out;

Between bunch minimum Eustachian distance square and the point between minimum Eustachian distance square ratio, equal low signal-to-noise ratio user's the signal to noise ratio and the inverse of the ratio of high s/n ratio user's signal to noise ratio; Perhaps, between bunch minimum Eustachian distance square and the point between minimum Eustachian distance square ratio, equal low signal-to-noise ratio user's the signal to noise ratio and the value of the inverse and the default degree of approximation adjusted value added/subtracted gained of the ratio of high s/n ratio user's signal to noise ratio;

The value of each bit words is corresponding one by one with each constellation point.

4. method according to claim 3 is characterized in that, between obtaining described in the steps A in the planisphere bunch between minimum Eustachian distance and some the method for minimum Eustachian distance be: according to described number of bits value K _HAnd K _L, known two users the noise given transmitting power P of transmitting terminal when, obtain in the planisphere bunch between minimum Eustachian distance between minimum Eustachian distance and point.

5. method according to claim 4 is characterized in that, described obtain in the planisphere bunch between between minimum Eustachian distance and some the method for minimum Eustachian distance be:

A1. according to described number of bits value K _HAnd K _L, select corresponding constellation mapping table;

A2., the average power that all constellation point in the described planisphere are set equals described given transmitting power P, and according to minimum Eustachian distance between described bunch square and point between minimum Eustachian distance square ratio, equal low signal-to-noise ratio user's the signal to noise ratio and the inverse of the ratio of high s/n ratio user's signal to noise ratio, calculate in the planisphere bunch between Euclidean distance between minimum Eustachian distance and point; Perhaps, according to bunch between minimum Eustachian distance square and the point between minimum Eustachian distance square ratio, equal low signal-to-noise ratio user's the signal to noise ratio and the value of the inverse and the default degree of approximation adjusted value added/subtracted gained of the ratio of high s/n ratio user's signal to noise ratio, calculate in the planisphere bunch between Euclidean distance between minimum Eustachian distance and point.

6. method according to claim 5 is characterized in that, selects the method for corresponding constellation mapping table to be described in the steps A 1: according to the number of bits value K that is pre-stored in transmitting terminal _HAnd K _LWith the one-to-one relationship of constellation mapping table, select corresponding constellation mapping table.

7. method according to claim 5 is characterized in that, the method for calculating the average power of constellation point in the planisphere described in the steps A 2 is:

A21. from described constellation mapping table, take out the data symbol of each constellation point;

A22. respectively the data symbol of each constellation point is taken absolute value back square, will square after each income value addition, calculate the merchant of income value and constellation point sum after the addition at last.

8. method according to claim 3 is characterized in that, between obtaining described in the steps A in the planisphere bunch between minimum Eustachian distance and some the method for minimum Eustachian distance be: minimum Eustachian distance is a normalized value between default described point; According to minimum Eustachian distance between the point in two users' signal to noise ratio and the default planisphere, obtain in the planisphere bunch between minimum Eustachian distance.

9. method according to claim 8, it is characterized in that, described obtain in the planisphere bunch between the method for minimum Eustachian distance be: the merchant who calculates described high s/n ratio user's signal to noise ratio and low signal-to-noise ratio user's signal to noise ratio, gained merchant minimum Eustachian distance square long-pending again and between the point in the described default planisphere calculated the square root of income value at last.

10. method according to claim 8 is characterized in that, this method also comprises: default degree of approximation adjusted value;

Described obtain in the planisphere bunch between the method for minimum Eustachian distance be: the merchant who calculates described high s/n ratio user's signal to noise ratio and low signal-to-noise ratio user's signal to noise ratio, minimum Eustachian distance square long-pending again and between the point in the described default planisphere after gained merchant and the default degree of approximation adjusted value added/subtracted calculated the square root of income value at last.

11. method according to claim 3 is characterized in that, the method for obtaining the data symbol of current bit words correspondence described in the step B is:

B1. the bit words that generates according to steps A, inquiry is used for storing the bit words of planisphere and the constellation mapping table that prestores of the corresponding relation of data symbol computing formula, obtains the data symbol computing formula of correspondence;

B2. according in the planisphere of described acquisition bunch between the data symbol computing formula of Euclidean distance and acquisition between point in the Euclidean distance, default planisphere, calculate the data symbol of described current bit words correspondence.

12. method according to claim 11 is characterized in that: described data symbol computing formula comprises real part computing formula and imaginary part computing formula;

Described data symbol computing formula according in the described planisphere of difference bunch between between minimum Eustachian distance and some minimum Eustachian distance draw, and be preset in transmitting terminal.

13. method according to claim 3 is characterized in that, the method for obtaining the data symbol of current bit words correspondence described in the step B is:

According to described bit words, inquiry is used for storing the bit words of planisphere and the constellation correspondence table of presetting of the corresponding relation of data symbol, obtains the data symbol of current bit words correspondence.

14. method according to claim 13 is characterized in that: this method further comprises: in the default planisphere bunch between minimum Eustachian distance;

Data symbol in the described constellation correspondence table according in the described planisphere of difference bunch between between minimum Eustachian distance and some minimum Eustachian distance draw, and be preset in transmitting terminal.

15. method according to claim 3 is characterized in that, described planisphere further meets the following conditions:

Between different bunches, between all adjacent constellation point bunch between Euclidean distance equate and be bunch between minimum Eustachian distance; In cluster, the Euclidean distance between all adjacent constellation point equates and is minimum Eustachian distance between point;

Described constellation point adopts Gray code Gray to encode.

Images