CN102118241A - Sample phase selection system based on channel capacities - Google Patents

Abstract

A sample phase selection system based on channel capacities comprises a synchronizer, a digital mixer, an interpolation device, a digital matched filter, a buffer device, a channel estimation device, a sample phase selection device, a downward-sampling device. The sample phase selection system can provide sample phase selection information, so that better symbol synchronization can be obtained under assistance and the integral performance of the system can be promoted.

Description

Sampling phase selective system based on the channel capacity size

Technical field

The present invention relates to the signal transmission technology field, relate in particular to a kind of sampling phase selective system based on the channel capacity size.

Background technology

In existing digital communicating field, often use analog-to-digital converter (ADC) in order to fundamental frequency I and Q signal are carried out over-sampling (Over-sampling).According to sampling thheorem, desirable sample rate must be not less than the twice of system's frequency range.Usually system's frequency range is the symbol rate of the signal that transmits.Make oversample factor (over-sampling factor) κ be defined as the ratio of desirable sample rate and system's frequency range.Desirable κ value commonly used is a positive integer 2 or 4.Yet the sample rate in the actual reception machine, can be because of crystal oscillation frequency factor such as inaccurate or frequency synthesizer different qualities, make actual κ value and this ideal integer that a little deviation be arranged.In the prior art, mostly by symbol sequential (symbol timing) synchronously, in order to the estimation symbol rate and revise the sampling time skew, and then revise indirectly and obtain desirable κ value, and reach the synchronous of sampling time sequence (sample timing).

Generally speaking, symbol sequential (symbol timing) synchronously, be by interpolation device according to symbol rate and sampling time skew that synchronizer estimated, do that interpolation or extrapolation computing finish.Afterwards, the sample rate of interpolation device output, just reducible is correct symbol rate.

Fig. 1 is the exemplary block diagram of the synchronization mechanism of an existing digital receiver.This analog-to-digital converter 310 converts a simulation fundamental frequency signal to the digital baseband signal according to an oversample factor κ.One digital mixer 320 carries out frequency shift according to a carrier frequency shift to this digital baseband signal.One interpolation device 230 carries out interpolative operation according to sampling time skew and sample conversion rate to this skew fundamental frequency signal, and wherein, the sample conversion rate is defined as the ratio that the signals sampling rate is gone in interpolation device 230 outputs.At this, the output signal sample rate of interpolation device 230 is adjusted to symbol rate (Symbol Rate), and the data of input interpolation device 330 are sample rate (Sampling Rate) originally.The output signal of interpolation device 230 is after digital matched filter 340 filtering, just the needs of other algorithm after the foundation as channel equalizer among the figure 370, according to the synchronizing information that synchronizer provided, are pre-stored in a segment signal in one buffer unit 350.

According to the characteristic of system, employed synchronization program will have different with step in the actual synchronizer.The overwhelming majority is with the synchronous of carrier frequency and is compensated for as elder generation, then just makes temporal sign synchronization and frame synchronization.In addition, for the phase place of remaining frequency and time and carrier wave, then followed the trail of elimination with timing recovery loop (timing recovery loop) and phase-locked loop modes such as (phase trackingloop).

In existing digital communicating field, synchronously with its trace mode generally according to maximum similarity (ML) algorithm, and can be divided into data auxiliary (Data Aided), non-data auxiliary (Non-DataAided), and the decision modes such as (Decision-Directed) that leads.

For lacking frame head (frame header) or prime notation (preamble), and on time domain, adopt system in continuous transmission signal (continuous transmission) mode, auxiliary or the decision guidance mode and be not easy to reach synchronously with data is especially in the multi-path transmission channel.If can be by other non-data the assisting of the auxiliary or non-people having the same aspiration and interest (noncoherent) algorithm with the estimated symbols rate, and the symbol rate error is reduced in certain scope, the time detector that Gardner proposed (timing-error detector) is the sign synchronization tracing algorithm that the most frequently used non-data are assisted (Non-Data Aided), it only utilizes the symmetry of signal peak on time domain to obtain the sampling time skew in order to estimation, reach sign synchronization, thereby do not need Given information to assist, and can work in various QAM modulation mode and the transmission channel and (please refer to Floyd M.Gardner, ' A BPSK/QPSK Timing-Error Detector for Sampled Receiver, ' IEEETrans.Commun, vol.COM-34., No.5, May 1986).Moreover, in with communication system receiver based on frame (Frame based), then can be by known transmission signal such as frame head (frame header) or prime notation (preamble) etc. being done relevant (correlation) computing, and the coherent signal of gained is made peak value detect (peak detection), and then obtain symbol and frame synchronization.With upper type, the periodicity that all utilize to detect the time-domain signal peak value is in order to the estimation symbol rate, and by with the symmetry of continuous tracking peak with the skew of decision sampling time.At last, the sign synchronization of being assert to reach via the computing of interpolation device according to this again.

Yet in multipath decline channel (multipath fading channel), transmission performance is responsive for different sampling time skews.Even with identical symbol rate but if adopt different sampling time skews in order to signal is sampled, its channel capacity of experiencing is also inequality, and then will make and produce difference on the performance.Thereby merely with the symmetry of above-mentioned peak value, institute estimate follow the trail of and the sampling time skew sign synchronization result of unreliable gained also, can not guarantee to reach the performance of the best.Therefore, existing simultaneous techniques, especially by with the decision of sampling phase in order to improve sign synchronization, use aspects such as promoting systematic function, still lack and have the space of improvement.

Summary of the invention

Purpose of the present invention mainly provides a kind of sampling phase selective system, selects information so that sampling phase to be provided, and uses and assists and obtain preferable sign synchronization, and then promote overall system performance.Simultaneously, the sampling phase that provided select be size with channel capacity as foundation, it differs from prior art.

According to a characteristic of the present invention, the present invention proposes a kind of sampling phase selective system based on the channel capacity size, comprise one synchronously device, a digital mixer, an interpolation device, a digital matched filter, a buffer unit, to a downsampling device and a sampling phase choice device.This synchronizer is in order to skew (the carrier frequency offset of the carrier frequency in the estimating system, CFO), sampling time skew (sample timing offset), and sample conversion rate (Samplingconversion rate) and frame sequential (frame timing), and counting sampling phase to be chosen (sample phases).This numeral mixer receives a digital baseband signal, according to the frequency shift (FS) of this carrier wave this digital baseband signal is carried out frequency compensation, and then produces a skew fundamental frequency signal.This interpolation device is connected to this numeral mixer, according to this sampling time skew and this sample conversion rate, this skew fundamental frequency signal is carried out interpolative operation, and then compensates and produce interpolation skew fundamental frequency signal.This digital matched filter is connected to this interpolation device, this interpolation skew fundamental frequency signal is carried out filtering, and then produce an over-sampling filtering fundamental frequency signal.This buffer unit is connected to this digital matched filter, according to this frame sequential, with temporary this over-sampling filtering fundamental frequency signal.This is connected to this buffer unit to downsampling device, according to a phase place index for selection, in order to this over-sampling filtering fundamental frequency signal is carried out to the down-sampling computing, and then produces one to the down-sampling fundamental frequency signal.This sampling phase choice device is connected to this to downsampling device, according to a channel frequency response and aforementionedly wait to choose the phase place index, in order to calculating the aforementioned channel capacity of waiting to choose the phase place correspondence, and choose and have maximum phase place to be chosen in this channel capacity, in order to as this phase place index for selection; Wherein, this over-sampling filtering fundamental frequency signal comprises many group fundamental frequency signals, this phase place index for selection according to the fundamental frequency signals of aforementioned many groups one of them, in order to produce this to the down-sampling fundamental frequency signal.

Description of drawings

Fig. 1 is the exemplary block diagram of the synchronization mechanism of an existing digital receiver.

Fig. 2 is the calcspar of a kind of sampling phase selective system based on the channel capacity size of the present invention.

Fig. 3, Fig. 4, Fig. 5 and Fig. 6 are the schematic diagram according to the technology of the present invention analog result.

[primary clustering symbol description]

Analog-to-digital converter 310 digital mixers 320

Interpolation device 330 digital matched filters 340

Buffer unit 350 synchronizers 360

Equalizer 370

Analog-to-digital converter 410 digital mixers 420

Interpolation device 430 digital matched filters 440

Buffer unit 450 is to downsampling device 460

Channel estimating apparatus 470 sampling phase choice devices 480

Synchronizer 490 equalizers 495

Embodiment

Fig. 2 is the calcspar of a kind of sampling phase selective system based on the channel capacity size of the present invention, and this sampling phase selective system applies to a receiving terminal of a wireless communication system, and this wireless communication system uses a carrier wave to transmit signal.The sampling phase selective system comprises: the digital mixer of an analog-to-digital converter 410, one 420, an interpolation device 430, a digital matched filter 440, a buffer unit 450, one are to downsampling device 460, a channel estimating apparatus 470, a sampling phase choice device 480, a device 490 and one equalizer 495 synchronously.

This analog-to-digital converter 410 is connected to this numeral mixer 420, converts a simulation fundamental frequency signal to the digital baseband signal according to an oversample factor κ.The κ value is generally a positive integer 2 or 4.

This synchronizer 490 estimates carrier frequency shift (the carrierfrequency offset in this communication system, CFO), sampling time skew (sample timing offset), and sample conversion rate (sampling conversion rate), frame sequential (frame timing), and count sampling phase all possible to be chosen (candidate sample phases).In the execution mode of reality, synchronizer 490 required input signals are by the algorithm decision of being adopted.

This numeral mixer 420 receives this digital baseband signal, according to carrier frequency shift this digital baseband signal is carried out frequency shift, in order to produce a skew fundamental frequency signal.

This interpolation device 430 is connected to this numeral mixer 420, according to this skew and sample conversion rate in sampling time, in order to this skew fundamental frequency signal is carried out interpolative operation, and then produces an interpolation and is offset fundamental frequency signal.The sample rate of 430 dateouts of this interpolation device is desirable sample rate still, as the κ that is defined as symbol rate doubly.When the κ value is 4, represent that promptly these interpolation device 430 output signals have four sampled points in each symbol period (symbol period).Promptly constitute four with the formed sampled point set of different sampling phases if get same position sampled point in each symbol, for example κ is that 4 o'clock sampling phase is respectively phase place 0, phase place 1, phase place 2, and phase place 3.Therefore, this digital baseband signal comprises 4 groups fundamental frequency signal, corresponds respectively to phase place 0, phase place 1, phase place 2 thereafter, reaches phase place 3.

This digital matched filter 440 is connected to this interpolation device 430, this interpolation skew fundamental frequency signal is carried out filtering, and then produce an over-sampling filtering fundamental frequency signal.

This buffer unit 450 is connected to this digital matched filter 440, according to this frame sequential, in order to temporary this over-sampling filtering fundamental frequency signal.

This is connected to this buffer unit 450 to downsampling device 460, according to a phase place index for selection, in order to this over-sampling filtering fundamental frequency signal is carried out to the down-sampling computing, and then produces one to the down-sampling fundamental frequency signal.

When the κ value is 4, this phase place index for selection corresponding to phase place 0, phase place 1, phase place 2, and phase place 3 one of them.In 4 groups of signals that comprised by this over-sampling filtering fundamental frequency signal, choose wherein one group, as this to the down-sampling fundamental frequency signal.

Because this analog-to-digital converter 410 carries out the over-sampling computing according to an oversample factor κ, thus this after the downward sampling cartridge computing of downsampling device 460 execution, data transfer rate (Data Rate) is reduced to symbol rate (Symbol Rate).So gained to the down-sampling fundamental frequency signal, can be considered the sign synchronization result who obtains based on selected sampling phase.

This channel estimating apparatus 470 is connected to this sampling phase choice device 480, and the indication of waiting to choose the phase place index of being exported according to synchronizer 490 is with estimated channel, and produces and choose the channel frequency response (CFR) of phase place corresponding to this.

This synchronizer 490 is exported this in regular turn and is waited to choose phase place index i, uses for this channel estimating apparatus 470 estimated channel.When the κ value is 4, this synchronizer 490 export in regular turn respectively to should phase place 0, phase place 1, phase place 2, and this of phase place 3 wait to choose phase place index i, this waits to choose phase place index i 470 foundations of this channel estimating apparatus, calculate respectively to should phase place 0, phase place 1, phase place 2, and the channel frequency response (CFR) of phase place 3.In the execution mode of reality, channel estimating apparatus 470 required input signals are by the algorithm decision of being adopted.

This sampling phase choice device 480 is connected to this to downsampling device 460, this channel estimating apparatus 470 and this synchronizer 490, wait to choose phase place index i according to a channel frequency response (CFR) and, wait to choose the corresponding channel capacity of phase place in order to calculate with this, and to choose that the channel capacity maximum is corresponding waits to choose phase place in order to as this phase place index for selection.

For example: when the κ value is 4, this channel estimating apparatus 470 export in regular turn respectively to should phase place 0, phase place 1, phase place 2, and the channel frequency response (CFR) of phase place 3.This waits to choose phase place index i and this channel frequency response (CFR) 480 foundations of this sampling phase choice device, calculate respectively to should phase place 0, phase place 1, phase place 2, and the channel capacity of phase place 3.When the pairing channel capacities of phase place 2 are maximum, and the phase place 2 of choosing the channel capacity maximum is as this phase place index for selection j.This to downsampling device 460 according to this phase place index for selection, in 4 groups of signals that comprised by this over-sampling filtering fundamental frequency signal, choose with j be 2 corresponding one group of signal, as this to downsampling device 460 exported to the down-sampling fundamental frequency signal.

According to Shannon gauge (Shannon bound) as can be known, but the channel capacity C of addition white Gauss noise (AWGN) channel be:

$R\≤C=B\·{\log }_2(1+\frac{S}{N}),$

In the middle of, R is a maximum attainable data transmission rate (bps), B is the frequency range of channel,

Be signal power to noise power ratio.

When channel was frequency selective channel (frequency selective channel), this was to down-sampling fundamental frequency signal y _nCan be expressed as:

$y_n=x_n\⊗h_n+n_n,$

In the middle of, y _nFor this to down-sampling fundamental frequency signal, x _nFor transmitting signal, n _nBe noise,

Be annular folding long-pending (Circular Convolution), h _nFor the channel impulse response of integral body (channelimpulse response, CIR).The fast Fourier transform of following formula process (Fast FourierTransform FFT) after the computing, can be expressed as:

Y _k＝X _k·H _k+Z _k，

In the middle of, k represents k subcarrier (Subcarrier), H _kIt is the multichannel gain of k subcarrier.Z _kIt is the noise on k the subcarrier.This channel frequency response (CFR) is multichannel gain H _kSet, that is $\mathrm{CFR}=\left\{H_k\right|\∀k\},$ It equals channel impulse response through the fast Fourier conversion, and is provided by channel estimating apparatus.The channel of being considered in this hypothesis has single gain, that is, ∑ _k| H _k| ²=1.The average signal energy S on k subcarrier then _kMay be defined as:

Wherein, E[] represent the computing of probability desired value, σ _X ²Be the average transmission signal power on each subcarrier.Noise energy N on K subcarrier _kMay be defined as:

$N_k\≡E\left[{\left|Z_k\right|}^2\right]={\mathrm{\σ}}_N^2.$

Therefore, when channel was frequency selectivity, the signal power to noise power ratio on k subcarrier was

${\mathrm{\γ}}_k\≡\frac{{\mathrm{\σ}}_X^2\·{\left|H_k\right|}^2}{{\mathrm{\σ}}_N^2}$

Channel capacity on k subcarrier is

$C_k\≡\frac{W}{M}\·{\log }_2(1+\frac{{\mathrm{\σ}}_X^2\·{\left|H_k\right|}^2}{{\mathrm{\σ}}_N^2})$

In the middle of, W is a channel width, M is (subcarrier) number between the spectrum region.Simultaneously, average signal power to noise power ratio is

$\mathrm{\γ}\≡\frac{{\mathrm{\Σ}}_k^M{S}_k}{{\mathrm{\Σ}}_k^M{N}_k}=\frac{{\mathrm{\σ}}_X^2}{{\mathrm{\σ}}_N^2}\·\frac{1}{M}\·{\mathrm{\Σ}}_k^M{\left|H_k\right|}^2.$

When channel was frequency selectivity, this overall channel capacity was the summation of this subcarrier upper signal channel capacity, that is

$C={\mathrm{\Σ}}_k^M{C}_k=\frac{W}{M}\·{\mathrm{\Σ}}_k^M{\log }_2(1+\frac{{\mathrm{\σ}}_X^2\·{\left|H_k\right|}^2}{{\mathrm{\σ}}_N^2}).$

When changing channel frequency response in order to comparison channel capacity size, the proportionality constant of following formula can omit, so can be rewritten into:

$C\∝{\mathrm{\Σ}}_k^M{\log }_2(1+\frac{{\mathrm{\σ}}_X^2\·{\left|H_k\right|}^2}{{\mathrm{\σ}}_N^2})\·\·\·\·\·\·\·\·\·\·\·\·\·\·\·\·\·\·\left(1\right)$

$={\mathrm{\Σ}}_k^M\left({\log }_{2}e\right)\·1n(1+\frac{{\mathrm{\σ}}_X^2\·{\left|H_k\right|}^2}{{\mathrm{\σ}}_N^2})$

$=\left({\log }_{2}e\right)\·{\mathrm{\Σ}}_k^{M}1n(1+\frac{{\mathrm{\σ}}_X^2\·{\left|H_k\right|}^2}{{\mathrm{\σ}}_N^2})\·\·\·\·\·\·\·\·\·\·\·\left(2\right),$

And when (1＜x≤1), $1n(1+x)={\mathrm{\Σ}}_{n=1}^{\∞}{(-1)}^{n+1}\·\frac{x^n}{n},$ So formula (2) can be rewritten into:

$C=\left({\log }_{2}e\right)\·{\mathrm{\Σ}}_k^M{\mathrm{\Σ}}_{n=1}^{\∞}\frac{{(-1)}^{n+1}}{n}\·{\left(\frac{{\mathrm{\σ}}_X^2\·{\left|H_k\right|}^2}{{\mathrm{\σ}}_N^2}\right)}^n\·\·\·\·\·\·\·\·\·\·\·\left(3\right).$

When high signal noise ratio, that is

The time, formula (1) can be rewritten into:

$C\stackrel{\left(1\right)}{\≈}{\mathrm{\Σ}}_k^M{\log }_2\left(\frac{{\mathrm{\σ}}_X^2\·{\left|H_k\right|}^2}{{\mathrm{\σ}}_N^2}\right)\∝{\mathrm{\Σ}}_k^M{\log }_2\left|H_k\right|,$

Therefore, when high signal noise ratio, this channel capacity C is proportional to ∑ _k ^MLog ₂(| H _k|).

When the low signal noise ratio, the high-order term of formula (3) can be ignored, so can be rewritten into:

$C\stackrel{\left(3\right)}{\≈}\left({\log }_{2}e\right)\·{\mathrm{\Σ}}_k^M\left(\frac{{\mathrm{\σ}}_X^2\·{\left|H_k\right|}^2}{{\mathrm{\σ}}_N^2}\right)\∝{\mathrm{\Σ}}_k^M{\left|H_k\right|}^2,$

Therefore, when the low signal noise ratio, this channel capacity is proportional to ∑ _k ^M| H _k| ²

This sampling phase choice device 480 can calculate and the pairing channel capacity C of this sampling phase to be selected according to a channel frequency response (CFR) and a sampling phase index to be selected (i) ⁽ⁱ⁾This sampling phase choice device 480 is also chosen the maximum corresponding sampling phase to be selected of channel capacity as sampling phase.That is this sampling phase is:

$\widehat{\mathrm{\ξ}}=\arg \underset{i}{\mathrm{Max}}{C}^{\left(i\right)}=\arg \underset{i}{\mathrm{Ma}}x{\mathrm{\Σ}}_k^M{\mathrm{\ψ}}_k^{\left(i\right)},$

In the middle of,

Be these sampling phase choice device 480 last selected sampling phases, C ⁽ⁱ⁾Be the channel capacity of sampling phase correspondence to be selected, i is a sampling phase index to be selected, ψ _k ⁽ⁱ⁾Channel capacity comparison expression for the pairing equivalent-simplification of sampling phase i to be selected.These sampling phase choice device 480 output and sampling phases

Corresponding index j is as this phase place index for selection.Make { H _k ⁽ⁱ⁾It is channel frequency response (CFR) corresponding to i sampling phase.When high signal noise ratio, get this ψ _k ⁽ⁱ⁾Be log ₂(| H _k ⁽ⁱ⁾|).When the low signal noise ratio, then get this ψ _kFor | H _k ⁽ⁱ⁾| ²

In other embodiments, 470 of this channel estimating apparatus produce the part pairing channel frequency response of sampling phase index to be selected i (CFR), all the other pairing channel frequency responses of sampling phase index to be selected (CFR) then use interpolation (interpolation or extrapolation) method to produce, and reduce system thus because of the essential calculating processing delay (processingdelay) that all CFR produced.

This equalizer 495 is connected to this to downsampling device 460 and this channel estimating apparatus 470, according to the result of this channel estimating, this is carried out balancing operational to the down-sampling fundamental frequency signal, in order to eliminate intersymbol interference (Inter-Symbol Interference).In the execution mode of reality, the result of this channel estimating apparatus 470 channel estimating of this equalizer 495 that exports to is by the algorithm decision of being adopted.

This numeral mixer 420, this interpolation device 430, this digital matched filter 440, this buffer unit 450, this is to downsampling device 460, this channel estimating apparatus 470, this sampling phase choice device 480, the interior receiver (inner receiver) that this synchronizer 490 and described equalizer 495 constitute in the receiving system, the output of equalizer 495 then is connected to outer receiver (the outer receiver in the receiving system, figure does not show), carry out subsequent treatment, for example inverse correspondence (de-mapping), reciprocal cross mistake (de-interleaving), channel decoding computings such as (channel decoding).

Fig. 3 to Fig. 6 is the schematic diagram according to the technology of the present invention analog result.System parameters is according to the DTMB default.Suppose perfect channel estimating and synchronous simultaneously.System's frequency range is 7.56MHz, and oversample factor κ is 4, so sample rate is 30.24MHz.During simulation, in four candidate's phase places (phase place 0, phase place 1, phase place 2, phase place 3), only calculate the corresponding channel frequency response (CFR) of phase place 0 and phase place 2, use a desirable interpolating method simultaneously in the hope of the corresponding channel frequency response (CFR) of phase place 1 and phase place 3.When the channel capacity size compares, all adopt the comparison expression when high signal noise ratio, that is, get ψ _k ⁽ⁱ⁾Be log ₂(| H _k ⁽ⁱ⁾|).Fig. 3 and Fig. 4 consider awgn channel, and Fig. 5 and Fig. 6 consider multipath decline channel (multipath fading channel) channel.At this, we consider the SARFT-8 channel, and its characteristic of channel is as shown in table 1, and table 1 lies in the various parameters of SARFT-8 multi-path channel.Fig. 3 and Fig. 5 be multi-carrier mode (Multi-carrier mode, MC), Fig. 4 and Fig. 6 be single carrier mode (Single carrier mode, SC).Among Fig. 3 to Fig. 6, (Uncoded Bit Error Rate, UBER), transverse axis is signal noise ratio (SNR) to the longitudinal axis for bits of coded error rate not.

By the simulation schematic diagram of Fig. 3 to Fig. 6 as can be known, even with identical symbol rate but if adopt different sampling time skews in order to signal is sampled, its channel capacity of experiencing is also inequality, and then will make and produce difference on the performance.Therefore existing as the method that Gardner is mentioned or the method for related operation, institute estimate follow the trail of and sampling time be offset also instability.By the simulation schematic diagram of Fig. 3 to Fig. 6 as can be known, the selected candidate's phase place of the technology of the present invention has minimum not bits of coded error rate, that is with UBER as system performance index, the technology of the present invention always can be selected best candidate's phase place.

Number of path	1	2	3	4	5	6
							Extend in the path	-1.8	0.0	0.15	1.8	5.7	30
Path attenuation (dB)	-18	0	-20	-20	-10	0
							Path phase	0	0	0	0	0	0

Table 1

The present invention is desirable sample rate (according to the κ that is defined as symbol rate doubly) at this data transfer rate before downsampling device 460, then is symbol rate (Symbol Rate) at this data transfer rate after downsampling device 460.Yet estimate when inaccurate in sampling time skew (Sampling time Offset), the present invention but can provide via the suitable selection of many groups sampling phase (sample phase) intercropping, and reaches preferable performance.

By above stated specification as can be known, the present invention proposes that a kind of it makes the signal sampling rate of interpolation device 430 outputs in order to determine the method for preferable sampling time skew, still remains this ideal sample rate (by definition, it be κ a times of symbol rate).At this moment, after this signal is done κ times of down-sampling (down-sampling) computing according to different sampling phases (sample phase), just can get κ unlike signal based on symbol rate.So, select different sampling phases promptly to be equal to and select different sampling time skews., just can choose preferable sampling time skew, and then reach sign synchronization, thereby improve performance the selection of sampling phase by suitably.The present invention then proposes the size with channel capacity, as the foundation of sampling phase selection.

In sum, the data of prior art after interpolation device are symbol rate (SymbolRate), when the sampling time that synchronizer is estimated, skew can't make the systematic function optimization, there is no and take any remedial measure, it is synchronous with auxiliary symbol that yet the present invention can provide sampling phase to select than prior art, also provides simultaneously and take to use the sampling phase selection technology of channel capacity as foundation.That is the invention provides a brand-new sampling phase selection technology and a system, to remedy the deficiency of existing sign synchronization.Because prior art there is no and utilizes the suitable choice of sampling phase (sample phase) intercropping, to obtain best performance, also do not propose to select the system of sampling phase according to the size of channel capacity, the present invention is according to the suitable choice of many groups sampling phase (sample phase) intercropping in event, and then obtains preferable performance.

From the above, no matter the present invention all show it totally different in the feature of prior art, has practical value with regard to purpose, means and effect.But it should be noted that above-mentioned many embodiment only give an example for convenience of explanation, the interest field that the present invention advocated should be as the criterion so that claims are described, but not only limits to the foregoing description.

Claims (11)

1. the sampling phase selective system based on the channel capacity size applies in the receiving terminal of a communication system, and this communication system uses a carrier wave to transmit signal, and this sampling phase selective system comprises:

One installs synchronously, in order to frequency shift (FS), sampling time skew, a sample conversion rate of estimating this carrier wave in this communication system, and a frame sequential, and counting sampling phase to be chosen;

One digital mixer is connected to this synchronizer, and receives a digital baseband signal, according to the frequency shift (FS) of this carrier wave this digital baseband signal is carried out frequency compensation, and then produces a skew fundamental frequency signal;

One interpolation device is connected to this numeral mixer and this synchronizer, according to this sampling time skew and this sample conversion rate, this skew fundamental frequency signal is carried out interpolative operation, and then produces interpolation skew fundamental frequency signal;

One digital matched filter is connected to this interpolation device, this interpolation skew fundamental frequency signal is carried out filtering, and then produce an over-sampling filtering fundamental frequency signal;

One buffer unit is connected to this digital matched filter and this synchronizer, according to this frame sequential, with temporary this over-sampling filtering fundamental frequency signal;

One to downsampling device, is connected to this buffer unit, according to a phase place index for selection, in order to this over-sampling filtering fundamental frequency signal is carried out to the down-sampling computing, and then produces one to the down-sampling fundamental frequency signal; And

One sampling phase choice device, be connected to this to downsampling device, according to a channel frequency response and aforementionedly wait to choose the phase place index, in order to calculate the aforementioned channel capacity of waiting to choose the phase place correspondence, and choose and have maximum phase place to be chosen in this channel capacity, in order to as this phase place index for selection;

Wherein, this over-sampling filtering fundamental frequency signal comprises many group fundamental frequency signals, this phase place index for selection according to the fundamental frequency signals of aforementioned many groups one of them, in order to produce this to the down-sampling fundamental frequency signal.

2. sampling phase selective system according to claim 1, it also comprises:

One analog-to-digital converter is connected to this numeral mixer, converts a simulation fundamental frequency signal to this digital baseband signal according to an oversample factor.

3. sampling phase selective system according to claim 2, it more comprises:

One channel estimating apparatus is connected to this sampling phase choice device, waits to choose the phase place index in order to estimated channel according to this, and produces pairing this channel frequency response.

4. sampling phase selective system according to claim 1, wherein, when this channel was a frequency selective channel, this channel capacity was:

$C=\frac{W}{M}\·{\mathrm{\Σ}}_k^M{\log }_2(1+\frac{{\mathrm{\σ}}_X^2\·{\left|H_k\right|}^2}{{\mathrm{\σ}}_N^2}),$

In the middle of, C is this channel capacity, and W is a channel width, and M is a number between the spectrum region, σ _X ²Be average signal power interior between each spectrum region, σ _N ²Be average noise power interior between each spectrum region, H _kBe multichannel gain interior between k spectrum region, this channel frequency response is H _kSet.

5. sampling phase selective system according to claim 4, wherein, when this channel was high signal noise ratio, this channel capacity was proportional to ∑ _k ^MLog ₂(| H _k|).

6. sampling phase selective system according to claim 4, wherein, this channel is when the low signal noise ratio, and this channel capacity is proportional to ∑ _k ^M| H _k| ²

7. sampling phase selective system according to claim 4, wherein, this sampling phase is:

$\widehat{\mathrm{\ζ}}=\arg \underset{i}{\mathrm{Max}}{C}^{\left(i\right)}=\arg \underset{i}{\mathrm{Max}}{\mathrm{\Σ}}_k^M{\mathrm{\ψ}}_k^{\left(i\right)},$

In the middle of, C ⁽ⁱ⁾Be the channel capacity of this sampling phase correspondence to be selected, i waits to choose phase place index, ψ for this _k ⁽ⁱ⁾Comparison expression for this channel capacity of the pairing equivalent-simplification of this sampling phase to be selected.

8. sampling phase selective system according to claim 7, wherein, when this channel is high signal noise ratio, ψ _k ⁽ⁱ⁾For being proportional to log ₂(| H _k ⁽ⁱ⁾|), and { H _k ⁽ⁱ⁾It is channel frequency response corresponding to i sampling phase.

9. sampling phase selective system according to claim 7, wherein, when this channel is the low signal noise ratio, ψ _k ⁽ⁱ⁾For being proportional to | H _k ⁽ⁱ⁾| ²

10. sampling phase selective system according to claim 3, wherein, this channel estimating apparatus produces the pairing channel frequency response of this sampling phase index to be selected of part, and the pairing channel frequency response of all the other these sampling phase indexs to be selected then produces according to interpolating method.

11. sampling phase selective system according to claim 3, it also comprises:

One equalizer is connected to this to downsampling device and this channel estimating apparatus, according to this channel frequency response, this is carried out balancing operational to the down-sampling fundamental frequency signal, and then eliminate intersymbol interference.

Images