CN1694443A - Blocking bit loading method in selective frequency blocking transmission system - Google Patents

Abstract

This invention discloses a block bit loading method in a frequency selective single carrier block transmission system including the following steps: 1, determining a code rate of transmission after setting up the communication, the receive end blocks usable sub-channels to form block information, 2, assigning the general bit number transmitted by each frame to the usable sub-channels to block the source and decides the modulation way applied in each block to form modulation information 3, the transmission end blocks the source information based on the above information to make modulation image to the points in the block then to make DFT for each block and convert these frequency domain signals to the time domain with IDFT after they get new signal spectrums to get due out signals 4, the receive end receives signals to transform them into the frequency domain for balance and resumes signals of each point to their corresponding blocks based on the block information and makes IDFT in blocks to convert them back to the time domain for judgment.

Description

A kind of blocking bit loading method in the selecting frequency single carrier wave blocking transmission system

(1) technical field

The present invention relates to wideband digital communications method, belong to the broadband connections technical field.

(2) background technology

The communication technology has obtained tremendous development since nearest decades, particularly nineteen nineties, people's daily life and development and national economy have been produced far-reaching influence.And the future communications technology is just developing towards the direction of broadband high-speed, therefore many wideband digital transmission technologys are subjected to paying close attention to widely, OFDM (hereinafter to be referred as OFDM:OrthogonalFrequency Division Multiplexing) and single carrier piecemeal transmission technology (are also referred to as the single carrier of frequency domain equalization, hereinafter to be referred as SC-FDE:Single Carrier with Frequency Domain Equalization) be exactly two kinds of wideband digital transmission technologys of being paid attention to by people, they all belong to the piecemeal transmission technology, and the degree that present OFDM is paid close attention to will be considerably beyond SC-FDE, and in multiple standards, become support technology, for example: the IEEE802.11a in the WLAN (wireless local area network) (WLAN:Wireless Local Area Network), the HiperLAN/2 of ETSI (ETSI:European Telecommunication Standard Institute), the IEEE802.16 in the wireless MAN (WMAN:Wireless Metropolitan Area Network); Various high-speed digital subscriber lines (xDSL:Digital Subscriber Line) in the cable data transmission all are based on the standard of OFDM technology.SC-FDE is not adopted by these standards, just is the physical layer transmission technology with the common suggestion of OFDM in IEEE802.16.

Ofdm system is a kind of multi-carrier transmission technology, and it is divided into N parallel mutually orthogonal narrowband subchannels to whole broad-band channel with N subcarrier.Ofdm system has many noticeable advantages: 1. very high spectrum efficiency; 2. realize fairly simple; 3. anti-multipath interference performance and anti-fading ability are strong; 4. can utilize channel condition information (being self adaptation OFDM technology) further to improve spectrum efficiency or the like.

Self adaptation OFDM technology can be according to given transmission signal power and channel conditions, determine that passes a code check, then according to the modulation system of taking on the different frequency domain points of the adaptive adjusting of channel conditions (just different subchannels), the information that makes each frequency domain point carry different bit numbers according to the channel conditions difference, thereby pass the best systematic function of acquisition under the code check at this, the key technology that realizes Adaptive OFDM system is that bit loads (bit-loading) algorithm.It is certain that the error rate of General System requires, so under the situation that reaches the error rate of system requirement, can improve the biography code check as much as possible, has also just improved spectrum efficiency.

These advantages make OFDM become the research focus over past ten years just, so that are considered to the support technology of future communications, particularly broadband wireless communications.But many shortcomings of ofdm system self, particularly its peak-to-average power is more excessive than (being called for short PAPR:Peak to Average Power Ratio), limiting its practical paces, and existing SC-FDE has above-mentioned all advantages except that the 4th of OFDM, and the PAPR problem that does not have OFDM, performance and efficient are suitable substantially with OFDM.It is people's development on the basis of research OFDM, this SC-FDE system equally takes the piecemeal transmission with OFDM, and adopt CP (if adopt zero padding (being called for short ZP:Zero Padding) mode, and with the be added to front of this frame of every frame hangover, then identical) with the CP effect, so just can be converted into circular convolution to the linear convolution of signal and channel impulse response, and eliminate the inter-frame-interference that multipath causes.Adopt simple frequency-domain equalization technology just can eliminate intersymbol interference at receiving terminal like this, for example: balanced and least mean-square error (the being called for short MMSE:Minimum Mean Square Error) equilibrium of ZF (being called for short ZF:Zero Forcing).

The SC-FDE system compares with OFDM, does not have the PAPR problem.And being ofdm system itself, the PAPR problem is difficult to the problem that solves with low-cost (spectrum efficiency and power efficiency) mode.Therefore the SC-FDE technology is subjected to increasing attention at present.Simply introduce the Mathematical Modeling of traditional SC-FDE system below.

The time-domain signal that the frame that the SC-FDE system sends at transmitting terminal does not contain CP is s (n), (n=0,1,, N-1), pass through multipath channel, wherein the impulse response of channel is h (n), (n=0,1, L-1), be subjected to the interference of additive white Gaussian noise (AWGN:Additive White Gaussian Noise) in the signals transmission, establishing noise is w (n), (n=0,1 ... N-1), remove after the CP, the time-domain signal r (n) that receives is:

r(n)＝s(n)h(n)+w(n)，(n＝0，1，…，N-1)????????????????????(1)

Wherein, "  " expression circular convolution computing.

At receiving terminal signal is done discrete Fourier transform (DFT) (hereinafter to be referred as DFT:Discrete Fourier Transform) and transform to frequency domain, according to the time domain convolution theorem of DFT, resulting frequency-region signal is:

R(k)＝S(k)·H(k)+W(k)，(k＝0，1，…，N-1)????????????????????(2)

Wherein, R (k), S (k), H (k), W (k) they are respectively r (n), s (n), h (n), w (n) does the frequency domain symbol of N point DFT, and, H (k), (k=0,1 ..., N-1) be the frequency domain response of channel.Through frequency-region signal after the zero forcing equalization be:

$\tilde{S}\left(k\right)=S\left(k\right)+\frac{W\left(k\right)}{H\left(k\right)}=S\left(k\right)+\tilde{W}\left(k\right),(k=\mathrm{0,1},\·\·\·,N-1)---\left(3\right)$

At last, signal is done inverse discrete Fourier transform (hereinafter to be referred as IDFT:Inverse Discrete Fourier Transform) become time domain again and adjudicate, obtain the data of transmitting terminal transmission.

From (3) formula as can be seen, there is error in the signal that finally obtains with the actual signal of transmission, and this error is caused by noise, especially exists under the situation about declining deeply a little at channel and can too amplify noise, can make signal produce distortion during in addition with the MMSE equilibrium.If utilized channel condition information in the SC-FDE system, these problems just can obtain fine solution.Therefore, the applicant has proposed a kind of single carrier block transmission method of frequency selection method and (has applied for national inventing patent, number of patent application: 200410036439.6), overcome the shortcoming that traditional SC-FDE system can not utilize channel condition information, this new SC-FDE system has higher systematic function and efficient.

The performing step of the single carrier block transmission method of this frequency selection method is divided into:

The first step is found out available subchannels, and whether channel be can be used as mark, then the subchannel label information is sent to transmitting terminal by backward channel.

The channel condition information H (k) that the receiving terminal basis estimates, (k=0,1 ..., N-1), from N subchannel, (the individual available subchannels of M≤N), the label of establishing this M available subchannels is k to select M from big to small according to amplitude gain _i(i=0,1 ..., M-1), and with remaining subchannel forbidding, use 1 bit information, promptly each subchannel of " 0 " or one token is available subchannels or forbidding subchannel, Here it is the needed subchannel label information of transmitting terminal, if receiving terminal is the DFT that N is ordered, promptly total N subchannel, the subchannel label information that feeds back to transmitting terminal has the N bit, then this N bit information is beamed back transmitting terminal by backward channel.

In second step, change signal spectrum according to the subchannel label information

After transmitting terminal receives that receiving terminal sends the subchannel label information of returning, just can come transmission signals with M available subchannels, like this to a frame M SC-FDE symbol s (n), (n=0,1 ..., M-1), be M point DFT and transform to frequency domain:

$S\left(i\right)=\underset{n=0}{\overset{M-1}{\mathrm{\&Sigma;}}}s\left(n\right)e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="A20051004366500131.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{2\mathrm{\&pi;}}{M}\mathrm{ni}},(i=\mathrm{0,1},\&CenterDot;\&CenterDot;\&CenterDot;,M-1)---\left(4\right)$

Just obtain the frequency-region signal that M is ordered, with the k that elects _i, (i=0,1 ..., M-1) individual available subchannels H (k _i), (i=0,1,, M-1) transmission i frequency-region signal S (i), (i=0,1 ..., M-1), promptly on the signal spectrum point of available subchannels correspondence, place and want frequency domain signal transmitted, and the signal spectrum point zero setting that will forbid the subchannel correspondence also can be filled some non-information datas, so just obtain the new frequency-region signal S ' of a frame (k), (k=0,1,, N-1), counting is N:

$S^{\&prime;}\left(k\right)=\left\{\begin{array}{cc}S\left(i\right),& k=k_i\\ 0,& k\&NotEqual;k_i\end{array}\right.,(k=\mathrm{0,1},\&CenterDot;\&CenterDot;\&CenterDot;,N-1;i=\mathrm{0,1},\&CenterDot;\&CenterDot;\&CenterDot;,M-1)---\left(5\right)$

Then (k) to S ', (k=0,1 ..., N-1) be N point IDFT:

$s^{\&prime;}\left(n\right)=\frac{1}{N}\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{S}^{\&prime;}\left(k\right){\mathrm{<figure-callout\; id="2"\; label="e\; j"\; filenames="A20051004366500131.png"\; state="\{\{state\}\}">e</figure-callout>}}^j,(n=\mathrm{0,1},\&CenterDot;\&CenterDot;\&CenterDot;,N-1)---\left(6\right)$

Become time-domain signal, IDFT counts and is greater than N during oversampling, HFS zero setting, this time-domain signal made D/A (digital to analog conversion) after, modulate again and send.

The 3rd step, select the signal that transmits on the available subchannels, then the signal of electing is carried out equilibrium, and conversion returns time domain and adjudicates, finally obtain the data of transmitting.

Receiving terminal receives signal, and the time domain discrete signal that removes behind the CP is:

r′(n)＝s′(n)h(n)+w(n)，(n＝0，1，…，N-1)??????????????????????????(7)

It is the DFT that N is ordered:

$R^{\&prime;}\left(k\right)=\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{r}^{\&prime;}\left(n\right)e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="A20051004366500131.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{2\mathrm{\&pi;}}{N}\mathrm{nk}},(k=\mathrm{0,1},\&CenterDot;\&CenterDot;\&CenterDot;,n-1)---\left(8\right)$

And:

R′(k)＝S′(k)H(k)+W(k)，(k＝0，1，…，N-1)????????????????????????????(9)

So just can select M the signal R ' (k on the available subchannels according to the subchannel label information _i), (i=0,1 ..., M-1), then with available subchannels Parameter H (k in the channel condition information that estimates _i), (i=0,1 ..., M-1), the signal of electing is carried out equilibrium; Can select one of following three kinds of balanced ways:

1, zero forcing equalization,

2, least mean-square error equilibrium,

3, mixed equilibrium, promptly a part of subchannel zero forcing equalization, and the least mean-square error equilibrium of another part subchannel; With the zero forcing equalization is that example is described:

$\tilde{S}\left(k_i\right)=\frac{R^{\&prime;}\left(k_i\right)}{H\left(k_i\right)},(i=\mathrm{0,1},\&CenterDot;\&CenterDot;\&CenterDot;,M-1)---\left(10\right)$

Order

$\tilde{S}\left(i\right)=\tilde{S}\left(k_i\right),(i=\mathrm{0,1},\&CenterDot;\&CenterDot;\&CenterDot;,M-1)---\left(11\right)$

It is the IDFT that M is ordered:

$\tilde{s}\left(n\right)=\frac{1}{M}\underset{i=0}{\overset{M-1}{\mathrm{\&Sigma;}}}\tilde{S}\left(i\right){\mathrm{<figure-callout\; id="2"\; label="e\; j"\; filenames="A20051004366500131.png"\; state="\{\{state\}\}">e</figure-callout>}}^j,(n=\mathrm{0,1},\&CenterDot;\&CenterDot;\&CenterDot;,M-1)---\left(12\right)$

These group data are just adjudicated and can be recovered initial data.

The single carrier block transmission method of frequency selection method utilizes channel condition information, can avoid declining a little deeply to frequency selective fading channels, thereby significantly improve the error performance of system.Communication system generally has certain error performance requirement, and systematic function is by the snr loss's decision that causes of channel estimation errors and synchronous error when of noise after the equilibrium, and balanced back signal to noise ratio is meant the ratio of balanced back signal power and noise power.And balanced back noise is the snr loss that causes of channel estimation errors and the synchronous error channel capacity of going back decision systems when, so according to the snr loss and the systematic function requirement that cause of channel estimation errors and synchronous error when of the balanced back noise of system, the modulation system of adaptive regulating system can further improve spectrum efficiency.

Bit loading technique (" A new loading algorithm for discrete multitonetransmission " Fischer, R.F.H. by Fischer and Huber proposition; Huber, J.B.; Global Telecommunications Conference, 18-22 Nov.1996Page (s): 724-728 vol.1) be the extraordinary OFDM bit of a kind of performance loading technique, but because this complicated technology realization Du Taigao, be not applied in the practical communication system, just there are some suboptimums (promptly to realize simplifying, performance descends to some extent) technology in the part wired communication system, be applied, but owing to reasons such as its complexity height still are difficult to be applied in the wireless environment so far.

(3) summary of the invention

At the single carrier block transmission method of above-mentioned frequency selection method, the present invention proposes can make full use of transmitted power, further improve a kind of blocking bit loading method of system spectrum utilance based on this transmission method.

This blocking bit loading method may further comprise the steps:

(1) after receiving-transmitting sides was set up communication, system determined that passes the bit number R that code check is promptly determined every frame transmission _T, receiving terminal is counted the amplitude gain of M and these available subchannels according to the available subchannels of choosing | H (k _i) |, (i=0,1 ..., M-1), it is divided into the Q piece, form and divide block message;

(2) total bit number R that every frame is transmitted _TBe assigned on the Q piece available subchannels, the bit number of every transmission is R _i, (i=0,1 ..., Q-1), and $\underset{i=0}{\overset{Q-1}{\mathrm{\&Sigma;}}}{R}_i=R_T,$ Count Q and every information bit R that goes up transmission according to piece _i, (i=0,1 ..., Q-1) determine information source is carried out the rule of piecemeal, and determine the every modulation system that interior each point is taked to form modulation system information;

(3) according to a minute block message, modulation system information and the modulation system taked by each point in the piece of their decisions, transmitting terminal is with information source information bit piecemeal, each point in the piece is done the modulation mapping, be DFT separately for every then, again these frequency-region signals are placed on the point of available subchannels correspondence and obtain new signal spectrum, the frequency-region signal that will newly obtain with IDFT transforms to time domain again, obtains the signal that will send;

(4) after receiving terminal receives signal, it is changed to frequency domain carry out equilibrium, according to minute block message the each point signal is returned in the piece of corresponding place then, and piecemeal is IDFT and is become time domain again and detect judgement.

Below above step is elaborated:

The first step, after receiving-transmitting sides was set up communication, system determined to pass code check, and with the available subchannels piecemeal.

After receiving-transmitting sides is set up communication, system determines that according to snr loss's situation and corresponding host-host protocol that channel status, received signal to noise ratio, various error cause passes a code check, it is available choosing M sub-channel tag according to channel condition information from all N subchannel, and this M available subchannels is divided into the Q piece, the available subchannels number on every is respectively M _i, (i=0,1 ... Q-1), and $\underset{i=0}{\overset{Q-1}{\mathrm{\&Sigma;}}}{M}_i=M;$

The present invention only relates to blocking bit loading method wherein, the piecemeal here has several different methods, for example according to the height piecemeal of channel gain, if available subchannels is divided into two, a thresholding can be set, the available subchannels that channel gain is higher than this thresholding is divided into one, and the available subchannels that channel gain is lower than this thresholding is divided into one; For realizing conveniently can be according to channel gain just with preceding 2 ^kIndividual available subchannels is divided into one, and remaining available subchannels is divided into one, and wherein 2 ^kCount M less than available subchannels; If available subchannels is divided into more than two, the branch that channel gain in the available subchannels can be more or less the same is gone in same, and in order to realize that conveniently every size can be elected 2 integral number power as;

The piece that available subchannels is divided is counted Q, every big or small M _i, (i=0,1 ... Q-1) and the label information of each available subchannels place piece form to divide a block message.

Second step is with the total bit number R of every frame information transmitted _TBe assigned on each piece, and determine every modulation system of taking.

Every frame is needed the total bit number R of information transmitted _TBe assigned on the Q piece available subchannels, and the every bit number of going up distribution is respectively R _i, (i=0,1 ..., Q-1), and $\underset{i=0}{\overset{Q-1}{\mathrm{\&Sigma;}}}{R}_i=R_T;$

The rule of allocation bit is a lot, for example accounts for the pro rate of overall channel capacity according to the channel capacity of each piece;

Available subchannels according to every is counted M _i, (i=0,1 ... Q-1) and every information transmitted bit number R _i, (i=0,1 ..., Q-1) determine the piecemeal situation of time domain point and the modulation system that the interior each point of piece is taked; At first according to the piecemeal situation of available subchannels M time domain point of every frame is divided into the Q piece, wherein the piecemeal situation of time domain point is consistent with the piecemeal situation of available subchannels, then with every available subchannels information transmitted bit number R _i, (i=0,1 ..., Q-1) correspondence is distributed to Q piece time domain point, determines the modulation system that each time domain point is taked in the piece;

If the bit number that each the time domain point on every carries is respectively R _Ij, subscript i represents that this point is in the i piece, j represents that this point is in j point of i piece, and $\underset{j=0}{\overset{M_i-1}{\mathrm{\&Sigma;}}}{R}_{\mathrm{ij}}=R_i;$ Under the common situation, the SC-FDE communication system is taked the linear modulation mode, so the modulation system that this time domain point is taked is 2 ^RijThe QAM of system or MPSK modulation;

What each time domain point carried bit number determines that a variety of rules are also arranged, and may carry a kind of with different time domain point in every below or two kinds of bit numbers are that example is introduced the bit loading method in the time domain piece; If the bit number that each point carries in the i piece is respectively K _I1Or K _I2, wherein,

K _i2＝K _i1+1???????????????????????????????????????????????(14)

Round under the expression;

And M _iCarry K in the individual point _I2Individual bit information count for:

M _i2＝R _i-K _i1·M _i???????????????????????????????????(15)

M then _iCarry K in the individual point _I1Individual bit information count for:

M _i1＝M _i-M _i2???????????????????????????????????????(16)

Like this, in the i piece M is arranged _I1Individual point carries K _I1Bit information, M _I2Individual point carries K _I2Bit information, and can be from M _iAppoint in the individual point and get M _I1Individual point makes it carry K _I1Bit information, all the other M _I2Individual point carries K _I2Bit information; But in order to realize conveniently can making M _iPreceding M in the individual point _I1Individual point carries K _I1Bit information, back M _I2Individual point carries K _I2Bit information; At last with the bit number R of each piece transmission _i, (i=0,1 ..., Q-1), and the modulation system of being taked by the bit number decision that each time domain point in the piece carries forms modulation system information;

Divide determining and also can finish this a series of work by the agreement that receiving-transmitting sides is reached, and pass to the other side at transmitting terminal of block message, modulation system information at receiving terminal by forward direction or backward channel; If this work is undertaken by receiving terminal, then need minute block message and modulation system information are passed to transmitting terminal by backward channel; If this work is undertaken by transmitting terminal, then need with channel gain H (k), (k=0,1 ..., N-1) pass to transmitting terminal.

The 3rd step, transmitting terminal is according to the available subchannels of choosing, divide block message and modulation system information, to each time domain point piecemeal and do corresponding modulation mapping, each piece is DFT separately then, again the frequency-region signal that obtains is placed on and obtains new frequency-region signal on the available subchannels, transform to time domain at last and obtain the signal that will send.

Transmitting terminal is according to every modulation system information of taking, and each point is done corresponding modulation mapping, and each point is assigned in pairing, and counting in every is respectively M _i, (i=0,1 ... Q-1); If the signal after the mapping is s _i(n), (n=0,1 ... M _i-1), subscript i represents that this point is in the i piece, and n represents that this point is in n point of i piece; Be M respectively for every then _i, (i=0,1 ... Q-1) DFT, the frequency-region signal that obtains is:

$S_i\left(k\right)=\underset{n=0}{\overset{M_i-1}{\mathrm{\&Sigma;}}}{s}_i\left(n\right)\exp (-\frac{\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="A20051004366500131.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;kn}}{M_i}),(k=\mathrm{0,1},\&CenterDot;\&CenterDot;\&CenterDot;,M_i-1)---\left(17\right)$

Discrete frequency domain signal after the above-mentioned piecemeal DFT is placed on the M of corresponding blocks _iOn the individual available subchannels, zero setting or transmit non-information signal on other subchannels in all N subchannel beyond M available subchannels, finish the spectrum transformation of signal, note finish frequency-region signal after the spectrum transformation be S ' (k), (k=0,1 ... N-1), the frequency-region signal after utilizing IDFT with conversion again becomes time domain again and sends.

In the 4th step, receiving terminal changes to frequency domain with received signal, and balanced back returns to original frequency spectrum state according to minute block message with signal, and piecemeal is IDFT and becomes time domain again, detects judgement according to modulation system information.Receiving terminal to frequency domain, is selected the frequency-region signal on the available subchannels with the signal transformation that receives after the equilibrium, according to minute block message these frequency-region signals are distributed in the corresponding piece, and piecemeal is IDFT and is become time domain again and detect judgement according to modulation system information then.

The present invention is based on the single carrier block transmission method of frequency selection method, according to certain regular piecemeal, the bit that carries out at time-domain adaptive behind the piecemeal loads with available subchannels.With respect to traditional single-carrier frequency domain equalization system, the method that the present invention proposes can further improve the spectrum efficiency of system, and implementation complexity increases little.Can reach the good system performance passing certain and transmitted power one timing of code check, and when passing code check and systematic function once regularly saving transmitted power significantly, same transmitted power and the error rate one timing, the biography code check of system can improve a lot than existing system.Self adaptation OFDM complicated technology realization Du Taigao does not ideally obtain practical application.Comparatively speaking, though the systematic function that the method that the present invention proposes obtains descends a little to some extent, realize fairly simplely, be convenient to be applied in the cable and wireless environment, especially can be applied in the wireless communications environment, for example WLAN (wireless local area network), wireless MAN or the like.

(4) description of drawings

Fig. 1 is a system block diagram of realizing method proposed by the invention.

Fig. 2 is that the present invention proposes method and the comparison of the self adaptation OFDM technology error performance of the bit loading algorithm that proposes based on Fischer and Huber when every frame transmits 1000 bits.

Among the figure: 1. information source module, 2. piecemeal module, 3. sign map module, 4. piecemeal FFT (Fast Fourier Transform: fast fourier transform, the fast algorithm of DFT) module, 5. spectrum transformation module, 6.IFFT (Inverse Fast FourierTransform: fast adverse Fourier transform, the fast algorithm of IDFT) module (N point), 7. add Cyclic Prefix (CP) module, 8.D/A module, 9. intermediate frequency and rf modulations module, 10. channel, 11. radio frequency and intermediate frequency demodulation module, 12.A/D module, 13. go the CP module, 14.FFT module (N point), 15. balance module, 16. the frequency spectrum inverse transform block, 17. piecemeal IFFT modules, 18. judging module, 19. piecemeal output module, 20. channel estimating and subchannel are selected module, 21. piecemeals and modulation system determination module, 22. synchronization modules

(5) embodiment

Embodiment:

Fig. 1 has provided the system block diagram of realizing method proposed by the invention, and each module effect is as follows:

Information source module 1: the data that generation will be transmitted.

Piecemeal module 2: the branch block message that transmits according to piecemeal and modulation system determination module 21, the data allocations that information source is produced is in each piece.

Mapping block 3: according to the modulation system information that piecemeal and modulation system determination module 21 transmit, select different modulation system (QAM or MPSK), the data map that information source is produced is to the point of planisphere correspondence.

Piecemeal FFT module 4: according to the branch block message that piecemeal and modulation system determination module 21 transmit, do the FFT conversion respectively for every, every data are transformed to frequency domain respectively.

Spectrum transformation module 5: the available subchannel information of selecting module 20 to transmit according to channel estimating and subchannel, piecemeal is done the frequency-region signal that FFT obtains, be placed on the Q piece available subchannels according to the frequency-region signal of minute block message every, and the corresponding frequency spectrum point zero setting of forbidding subchannel.This module need (number of patent application: 200410036439.6) method of the method for Jie Shaoing and the present invention proposition be programmed, and is realized by the general digital signal processing chip according to the patent of invention of mentioning in the background technology.

N point IFFT module 6: the frequency-region signal that will newly obtain transforms to time domain.

Add CP module 7: every frame data that will obtain add Cyclic Prefix.

D/A module 8: with digital signal conversion is analog signal.

Intermediate frequency and rf modulations module 9: if use this system under wireless environment, need make rf modulations to signal could the antennas emission.Need sometimes earlier signal to be modulated to carry out the intermediate frequency amplification on the intermediate frequency, remake rf modulations, at last the modulated signal antennas is launched.(for example: xDSL) use down this system, then do not need to do rf modulations, do not need antenna to transmit yet, but will move signal spectrum beyond the voice channel frequency band yet, guaranteeing does not influence speech transmissions in the transmission data if at cable environment.

Channel 10: the wire message way of transmission signals or wireless channel.

Radio frequency and intermediate frequency demodulation module 11: in wireless environment, the frequency spectrum that reception antenna is received signal is moved low frequency from radio frequency or intermediate frequency.Before demodulation, need the frequency deviation that causes with in the Frequency Synchronization data correction signal transmission course.

A/D module 12: analog signal after the demodulation is transformed to digital signal.A/D need sample to analog signal, provides the crystal oscillator of clock signal need follow the crystal oscillator frequency of transmitter D/A module identical, otherwise will cause the sampling rate error.Therefore it is synchronous to carry out sampling rate before A/D.

Go CP module 13: Cyclic Prefix is removed.At this moment just have the problem of judging when frame data begin, therefore going needs to do regularly synchronously before the CP.

N point FFT module 14: the signal transformation that will remove CP is to frequency domain.

Balance module 15: the available subchannel information of selecting module 20 to transmit according to channel estimating and subchannel, to the signal equalization on the available subchannels, balanced way can be selected one of following three kinds of balanced ways: zero forcing equalization, least mean-square error equilibrium, mixed equilibrium (that is: a part of subchannel zero forcing equalization, and the least mean-square error equilibrium of another part subchannel).

Frequency spectrum inverse transform block 16: with balanced data according to the piecemeal information distribution in the piece at place separately.

Piecemeal IFFT module 17: the branch block message that transmits according to piecemeal and modulation system determination module 21, equilibrium and frequency spectrum inverse transform block 16 are transmitted data behind the piecemeal, do the IFFT conversion respectively for every, obtain the piecemeal time-domain signal.

Judging module 18: according to planisphere, the modulation system information that piecemeal and modulation system determination module 21 transmit is finished the judgement of every time-domain signal.

Piecemeal output module 19: the piecemeal time-domain signal that judging module 18 transmits is exported according to minute block message, recovered the data that information source produces.

Channel estimating and subchannel are selected module 20: obtain CSI by parameter Estimation, blind Channel Estimation that commonly used generally is and based on the channel estimating of auxiliary data.Select available subchannels after estimating CSI, give balance module 15 these available subchannels parameters; Simultaneously whether available according to channel, with 1 bit information (" 0 " or " 1 ") mark, form the subchannel label information, give signal spectrum conversion module 5 and signal spectrum inverse transform block 16 simultaneously with the subchannel label information.This module need (number of patent application: 200410036439.6) method of introducing be programmed, and is realized by the general digital signal processing chip according to the patent of invention of mentioning in the background technology.

Piecemeal and modulation system determination module 21: the available subchannel information according to channel estimating and subchannel select module 20 to obtain, these available subchannels of selecting are divided into the Q piece according to certain rule, divided block message accordingly.And the data that every frame need be transmitted according to certain regular allocation to this Q piece, obtain modulation system information simultaneously.These information are passed to module 2,3,4,17,18,19 respectively.This module need be programmed according to the method that the present invention proposes, and is realized by the general digital signal processing chip.

Synchronization module 22: the method by parameter Estimation (for example: blind estimation and based on the estimation of auxiliary data) obtains the various synchrodatas that system needs.Synchronization module is given radio frequency and intermediate frequency demodulation module 11 with the Frequency Synchronization data; Give analog-to-digital conversion module 12 with the sampling rate synchrodata; Regularly synchrodata is given CP module 13.

This embodiment simulation parameter:

Simulated environment: Matlab7.0

This emulation is to carry out under the situation of ideal synchronisation, and the receiving terminal that is formed on of channel estimating, subchannel piecemeal, modulation system information is finished, and hypothesis transmits accurate when passing to transmitting terminal by backward channel.

Subchannel sum: N=256

The available subchannels number, promptly every frame SC-FDE data symbol number: M=224

The piece number that available subchannels is divided: Q=3

Counting of different masses is respectively: 128,64,32 (for FFT/IFFT realizes simply, every size all is 2 integral number power, and according to channel gain piecemeal from high to low)

The bit number of every frame transmission: R _T=1000

The modulation system of taking:

First (totally 128 points)		Second (totally 64 points)		The 3rd (totally 32 points)
						Carry 677 bits		Carry 236 bits		Carry 87 bits
Preceding 37 points	91 points in back	Preceding 44 points	20 points in back	Preceding 23 points	9 points in back
						??64QAM	??32QAM	??16QAM	??8QAM	??8QAM	??4QAM

CP length: 32

The selected receipts signal to noise ratio of emulation scope: snr=1-22 (dB)

The channel model that present embodiment adopts is a sample of SUI-5 channel (one of test channel of advising in IEEE 802.16 standards).

As can be seen from Figure 2 the method for the present invention's proposition is very close with the self adaptation OFDM technology error performance of Fischer and Huber proposition bit loading algorithm, the self adaptation OFDM technology that Fischer and Huber propose the bit loading algorithm is the extraordinary self adaptation OFDM technology of a kind of performance, but implementation complexity is too high, so far be not applied, have only some suboptimal block adaptive OFDM technology to be applied in the wire communication.The method that the present invention proposes realizes simply more than optimum self adaptation OFDM technology, can be applied to wired fairly simplely and wireless communication system in, and decreased performance is very small, so the method advantage that the present invention proposes is very obvious.

Claims (4)

1. a kind of blocking bit loading method in the selecting frequency single carrier wave blocking transmission system is characterized in that:

This method may further comprise the steps:

(1) after receiving-transmitting sides was set up communication, system determined that passes the bit number R that code check is promptly determined every frame transmission _T, receiving terminal is counted the amplitude gain of M and these available subchannels according to the available subchannels of choosing | H (k _i) |, (i=0,1 ..., M-1), it is divided into the Q piece, form and divide block message;

(2) total bit number R that every frame is transmitted _TBe assigned on the Q piece available subchannels, the bit number of every transmission is R _i, (i=0,1 ..., Q-1), and $\underset{i=0}{\overset{Q-1}{\mathrm{\&Sigma;}}}{R}_i=R_T,$ Count Q and every information bit R that goes up transmission according to piece _i, (i=0,1 ..., Q-1) determine information source is carried out the rule of piecemeal, and determine the every modulation system that interior each point is taked to form modulation system information;

(3) according to a minute block message, modulation system information and the modulation system taked by each point in the piece of their decisions, transmitting terminal is with information source information bit piecemeal, each point in the piece is done the modulation mapping, be DFT separately for every then, again these frequency-region signals are placed on the point of available subchannels correspondence and obtain new signal spectrum, the frequency-region signal that will newly obtain with IDFT transforms to time domain again, obtains the signal that will send;

(4) after receiving terminal receives signal, it is changed to frequency domain carry out equilibrium, according to minute block message the each point signal is returned in the piece of corresponding place then, and piecemeal is IDFT and is become time domain again and detect judgement.

2. the blocking bit loading method in the selecting frequency single carrier wave blocking transmission system according to claim 1 is characterized in that: described (1) step adopts following method to realize:

After receiving-transmitting sides is set up communication, system determines that according to snr loss's situation and corresponding host-host protocol that channel status, received signal to noise ratio, various error cause passes a code check, it is available choosing M sub-channel tag according to channel condition information from all N subchannel, and this M available subchannels is divided into the Q piece, the available subchannels number on every is respectively M _i, (i=0,1 ... Q-1), and $\underset{i=0}{\overset{Q-1}{\mathrm{\&Sigma;}}}{M}_i=M;$

The present invention only relates to blocking bit loading method wherein, the piecemeal here has several different methods, for example according to the height piecemeal of channel gain, if available subchannels is divided into two, a thresholding can be set, the available subchannels that channel gain is higher than this thresholding is divided into one, and the available subchannels that channel gain is lower than this thresholding is divided into one; For realizing conveniently can be according to channel gain just with preceding 2 ^kIndividual available subchannels is divided into one, and remaining available subchannels is divided into one, and wherein 2 ^kCount M less than available subchannels; If available subchannels is divided into more than two, the branch that channel gain in the available subchannels can be more or less the same is gone in same, and in order to realize that conveniently every size can be elected 2 integral number power as;

The piece that available subchannels is divided is counted Q, every big or small M _i, (i=0,1 ... Q-1) and the label information of each available subchannels place piece form to divide a block message.

3. the blocking bit loading method in the selecting frequency single carrier wave blocking transmission system according to claim 1 is characterized in that: described (2) step adopts following method to realize:

Every frame is needed the total bit number R of information transmitted _TBe assigned on the Q piece available subchannels, and the every bit number of going up distribution is respectively R _i, (i=0,1 ..., Q-1), and $\underset{i=0}{\overset{Q-1}{\mathrm{\&Sigma;}}}{R}_i=R_T;$

The rule of allocation bit is a lot, for example accounts for the pro rate of overall channel capacity according to the channel capacity of each piece;

Available subchannels according to every is counted M _i, (i=0,1 ... Q-1) and every information transmitted bit number R _i, (i=0,1 ..., Q-1) determine the piecemeal situation of time domain point and the modulation system that the interior each point of piece is taked; At first according to the piecemeal situation of available subchannels M time domain point of every frame is divided into the Q piece, wherein the piecemeal situation of time domain point is consistent with the piecemeal situation of available subchannels, then with every available subchannels information transmitted bit number R _i, (i=0,1 ..., Q-1) correspondence is distributed to Q piece time domain point, determines the modulation system that each time domain point is taked in the piece;

If the bit number that each the time domain point on every carries is respectively R _Ij, subscript i represents that this point is in the i piece, j represents that this point is in j point of i piece, and $\underset{j=0}{\overset{M_i-1}{\mathrm{\&Sigma;}}}{R}_{\mathrm{ij}}=R_i;$ Under the common situation, the SC-FDE communication system is taked the linear modulation mode, so the modulation system that this time domain point is taked is 2 ^RijThe QAM of system or MPSK modulation;

What each time domain point carried bit number determines that a variety of rules are also arranged, and may carry a kind of with different time domain point in every below or two kinds of bit numbers are that example is introduced the bit loading method in the time domain piece; If the bit number that each point carries in the i piece is respectively K _I1Or K _I2, wherein,

K _i2＝K _i1+1

Round under the expression;

And M _iCarry K in the individual point _I2Individual bit information count for:

M _i2＝R _i-K _i1·M _i

M then _iCarry K in the individual point _I1Individual bit information count for:

M _i1＝M _i-M _i2

Like this, in the i piece M is arranged _I1Individual point carries K _I1Bit information, M _I2Individual point carries K _I2Bit information, and can be from M _iAppoint in the individual point and get M _I1Individual point makes it carry K _I1Bit information, all the other M _I2Individual point carries K _I2Bit information; But in order to realize conveniently can making M _iPreceding M in the individual point _I1Individual point carries K _I1Bit information, back M _I2Individual point carries K _I2Bit information; At last with the bit number R of each piece transmission _i, (i=0,1 ..., Q-1), and the modulation system of being taked by the bit number decision that each time domain point in the piece carries forms modulation system information;

Divide determining and also can finish this a series of work by the agreement that receiving-transmitting sides is reached, and pass to the other side at transmitting terminal of block message, modulation system information at receiving terminal by forward direction or backward channel; If this work is undertaken by receiving terminal, then need minute block message and modulation system information are passed to transmitting terminal by backward channel; If this work is undertaken by transmitting terminal, then need with channel gain H (k), (k=0,1 ..., N-1) pass to transmitting terminal.

4. the blocking bit loading method in the selecting frequency single carrier wave blocking transmission system according to claim 1 is characterized in that: described (3) step adopts following method to realize:

Transmitting terminal is according to every modulation system information of taking, and each point is done corresponding modulation mapping, and each point is assigned in pairing, and counting in every is respectively M _i, (i=0,1 ... Q-1); If the signal after the mapping is s _i(n), (n=0,1 ... M _i-1), subscript i represents that this point is in the i piece, and n represents that this point is in n point of i piece; Be M respectively for every then _i, (i=0,1 ... Q-1) DFT, the frequency-region signal that obtains is:

$S_i\left(k\right)=\underset{n=0}{\overset{M_i-1}{\mathrm{\&Sigma;}}}{s}_i\left(n\right)\exp (-\frac{j2\mathrm{\&pi;kn}}{M_i}),(k=\mathrm{0,1},\&CenterDot;\&CenterDot;\&CenterDot;,M_i-1)$

Discrete frequency domain signal after the above-mentioned piecemeal DFT is placed on the M of corresponding blocks _iOn the individual available subchannels, zero setting or transmit non-information signal on other subchannels in all N subchannel beyond M available subchannels, finish the spectrum transformation of signal, note finish frequency-region signal after the spectrum transformation be S ' (k), (k=0,1 ... N-1), the frequency-region signal after utilizing IDFT with conversion again becomes time domain again and sends.

Images