CN101083482A - Signal receiving method for pulse ultra-broadband communication system - Google Patents

Abstract

The invention relates to a signal receive method for the pulse ultra-wide band communications system, the method is the training symbol sequence which supposes specially through the transmission, with its cyclical autocorrelation characteristic, directly pick-ups the initial signal demodulation template from the receive signal; then carries on the energy maximum value search of the initial signal demodulation template by the integrator, obtains the signal synchronized parameter; then, carries on the revision of initial signal demodulation template by the synchronized parameter, removes the pure noise, obtains the final signal demodulation template; finally, using the final signal demodulation template and the synchronized parameter to carry on the demodulation and the decision of the received data mark on the receiving end, completes the signal receive. The method enable to solve the each signal receive question of the existing technology, the sequence of operation is simple, practical, and has feasible performance, has low complexity and the less condition establishment request, has good application value and promotion prospect.

Description

A kind of signal acceptance method that is used for pulse ultra-broadband communication system

Technical field

The present invention relates to a kind of pulse ultra-broad band Ultra-Wideband that is used for (is called for short: the UWB) signal acceptance method of communication system, comprising three main operational steps of demodulation of extraction, signal Synchronization and the data symbol of signal demodulation template, belong to the technical field of the pulse ultra-broad band communication of radio communication.

Background technology

In recent years, the pulse ultra-broad band wireless communication technology becomes one of short distance, physical-layer techniques that fast wireless network is the most popular.Pulse ultra-broadband communication system comes data symbol by the extremely low ultrashort pulse signal of transmitting power, has extremely wide bandwidth and high transmission rate.So-called ultrashort pulse be meant the pulse duration be nanosecond or nanosecond the pulse signal with subordinate.

But because the own characteristic of pulse ultra-broadband communication system, the reception of signal is a bottleneck in its development and the application, and how making up a kind of simple and effective signal acceptance method becomes a hot issue of scientific and technical personnel's concern in the industry.

Impulse ultra-wideband signal method of reseptance commonly used at present is divided into three key steps usually: the extraction of signal Synchronization, signal demodulation template and the demodulation of data symbol.But wherein all there is certain problem in these two operating procedures of the extraction of signal Synchronization and signal demodulation template, have therefore restricted the reception of ultra-broadband signal.Do simple the introduction below:

At first, for the signal Synchronization problem (generally speaking, signal Synchronization is one of prerequisite of data symbol demodulation), existing method for synchronous all need be realized based on following one or multinomial assumed condition: do not exist in (1) transmission because the multipath signal that transmission channel is introduced, perhaps the multipath signal of being introduced is known at receiver end; (2) in the transmission course, there are not the inter-frame-interference or the intersymbol interference of signal; (3) received signal is sampled by high speed, have up to 10GHz, be used for realizing signal synchronously; (4) receiver can be stood high complexity, reaches signal Synchronization by the chip search technique.Obviously, above assumed condition is very harsh, and does not meet actual conditions, and is therefore, nonsensical for practical application.

Secondly, finish after the signal Synchronization, also need to make up signal demodulation template (this is another prerequisite of data symbol demodulation).Now many is the multipath signal information that is produced when adopting the Rake technology to catch transmission, and then structure comprises the signal demodulation template of multipath signal information.But realize the Rake technology, need to signal transmission the channel of process estimate that accurately this just increases the weight of the complexity and the amount of calculation of receiver greatly.In addition, the signal pulse waveform in the transmission course also may change, thereby demodulation causes certain interference to signal.In sum, existing method of reseptance all has problems on the structure of signal Synchronization and signal demodulation template, accomplishes effective unification aspect can not and tallying with the actual situation in complexity, validity.

Summary of the invention

In view of this, the purpose of this invention is to provide a kind of simple and effective signal acceptance method that is used for pulse ultra-broadband communication system.This method can solve the variety of issue that above-mentioned prior art exists preferably in signal receives, thereby can simply, effectively demodulate the pulse ultra-broad band data symbol, correct received signal.

In order to achieve the above object, the invention provides a kind of signal acceptance method that is used for pulse ultra-broadband communication system, it is characterized in that: send ad hoc training symbol sequence, so that utilize its unique cycle autocorrelation performance, from the signal that receives, directly extract initial signal demodulation template; Utilize integrator that this initial signal demodulation template is carried out the energy maximum value search again, with the synchronization parameter of picked up signal; Then, utilize synchronization parameter that this initial signal demodulation template is revised, remove redundant pure noise, obtain final signal demodulation template; At last, utilize final signal demodulation template and synchronization parameter that the data symbol that is received is separated to be in harmonious proportion at receiving terminal and adjudicate, finish the reception of signal.

The major advantage that method of reseptance of the present invention has is: the operating procedure of this method is simple, practical, enforceability can be good, has very low implementation complexity and condition setting requirement still less.For example aspect two prerequisites when demodulating data symbols, the present invention has significant advantage and characteristics: (1) does not need to do the various hypothesis of being mentioned in the above-mentioned prior art, the condition of realistic environment for use in the signal Synchronization part; (2) make up part in signal demodulation template, do not need to use the Rake technology to construct signal demodulation template, but directly from received signal, extract needed signal demodulation template, and can catch the multipath signal information of the overwhelming majority, realize the high unity of operating efficiency performance and received signal reliability.Therefore, the inventive method has good using value and popularizing application prospect.

Description of drawings

Fig. 1 is the operating procedure block diagram of the present invention's signal acceptance method of being used for pulse ultra-broadband communication system.

Fig. 2 is the impulse waveform schematic diagram a kind of commonly used in the pulse ultra-broadband communication system.

Fig. 3 is the waveform schematic diagram of the inventive method at a training symbol of transmitting terminal emission.

Fig. 4 (a) and (b), (c) are respectively the schematic symbol diagram that the inventive method is chosen when extracting three operating procedures of signal demodulation template.

The computational process schematic diagram that Fig. 5 is the inventive method when extracting an alternative signal demodulation template.

Fig. 6 is the curve comparison diagram of the present invention's embodiment pairing bit error rate and signal to noise ratio when adopting different accumulative frequency of being used for the pulse ultra-broadband communication system received signal.

Fig. 7 be embodiment that the present invention is used for the pulse ultra-broadband communication system received signal when whether having intersymbol interference different bit error rates and the curve comparison diagram of signal to noise ratio.

Embodiment

For making the purpose, technical solutions and advantages of the present invention clearer, the present invention is described in further detail below in conjunction with accompanying drawing.

The present invention is a kind of signal acceptance method that is used for pulse ultra-broadband communication system, this method is: (this training symbol sequence is meant in order to reach the purpose of prior setting to send ad hoc training symbol sequence, the special symbol sequence of setting quantity that has in the signal sending end emission, signal receiving end is known in advance to the special symbol sequence of this setting quantity), so that utilize its unique cycle autocorrelation performance, from the signal that receives, directly extract initial signal demodulation template; Utilize integrator that this initial signal demodulation template is carried out the energy maximum value search again, with the synchronization parameter of picked up signal; Then, utilize synchronization parameter that this initial signal demodulation template is revised, remove redundant pure noise, obtain final signal demodulation template; At last, utilize final signal demodulation template and synchronization parameter that the data symbol that is received is separated to be in harmonious proportion at receiving terminal and adjudicate, finish the reception of signal.

Referring to Fig. 1, introduce the concrete operations step of the inventive method:

(1) extract initial signal demodulation template:

Send ad hoc training symbol sequence,, from the signal that receives, directly make up 4 alternative signal demodulation template so that utilize its unique cycle autocorrelation performance; Described training symbol sequence be the symbol sebolic addressing that sends of circulation 1,1,1-1}, or by 31 and 1-1 other symbol sebolic addressing signals of forming; Adopt integrator from these 4 alternative signal demodulation template, to select energy the maximum then, be the initial signal demodulation template V that will extract _K1(t).

This step (one) can refinement be divided into following 4 little steps again:

(11) after the parameters of initialized setting system, begin to send training symbol sequence and received signal: the system parameters of setting comprises: the moment t of transmitter proceed-to-send signal ₀Be 0, system or receiver begin the moment t of received signal ₁, each symbol is by N ₁Individual pulse is formed, and its duration is T, and T blanking time of each pulse in each symbol ₁For: T ₁=T/N ₁The number of times that signal adds up in finding the solution the time average process is an IV, and the time span of every section of choosing training symbol sequence signal all is 4 symbol durations; In making up 4 alternative signal demodulation template processes employed 4 group code sequences be respectively 1,1,1 ,-1), 1,1 ,-1,1}, 1 ,-1,1,1), 1,1,1,1); Time scouting interval in the signal Synchronization process is T ₂, then Sou Suo number of times is T/T ₂, promptly T is T ₂Integral multiple; Decision device thresholding in the data symbol demodulating process is 0:

(12) compacting noise: initial signal demodulation template is directly to extract from the received signal that includes noise contribution, in order to reduce noise to the signal Synchronization of carrying out subsequently with make up the influence of final signal demodulation template, the noise in must be is to received signal suppressed.Concrete steps are: x (t) carries out time average to the received signal, promptly from beginning the moment t of received signal ₁Initial, receiving IV section time span continuously is the signal of training symbol sequence cycle (i.e. 4 symbol durations), and after adding up, asks its mean value, obtains the signal behind the time average

Its length also is 4 symbol durations, should

In noise eliminate fully as yet, but with respect to x (t), noise wherein is pressed and weakens greatly.

If the signal x (t) that receiver receives is: $x\left(t\right)=\sqrt{\mathrm{\ϵ}}\underset{i=1}{\overset{+\∞}{\mathrm{\Σ}}}{s}_i\underset{l=1}{\overset{L}{\mathrm{\Σ}}}{\mathrm{\α}}_l{w}_s(t-(i-1)T-{\mathrm{\τ}}_l+t_1)+n\left(t\right);$

In the formula, the span of time variable t be [0 ,+∞), t ₁For receiver begins moment of received signal, ε is the energy of each burst transmissions, s _iBe the training symbol sequence that adopts pulse amplitude modulation (PAM) mode to send, its value is ± 1, and natural number i is the sequence number of this training symbol; α _lAnd τ _lBe respectively the attenuation coefficient and the time of delay in each footpath in the multi-path channel, natural number l is the sequence number of multipath channel, and L is the number of multipath channel.It should be noted that: because time t from beginning to transmit ₀Start the time t of received signal to receiver ₁Interval (t ₁-t ₀) the common the shortest time delay τ that all arrives receiver much larger than multipath signal from transmitter ₁So, establish τ ₁=0.Clear for the ease of analyzing, being provided with in the transmission course of this symbol sebolic addressing does not have intersymbol interference, and promptly multipath signal arrives the longest time delay τ of receiver from transmitter _LDuration T with each pulse ₃Sum is less than T blanking time between each pulse ₁: T _L+ T ₃＜T ₁So,, the duration of the multipath signal of previous symbol can not extend to the time of reception of next symbol.w _s(t) for transmitting terminal send by N ₁(referring to the second order derived function of a kind of impulse waveform-Gaussian function commonly used shown in Figure 2, the duration of waveform is T to the waveform of the symbol that individual pulse is formed ₃Illustrated waveform is energy normalized, and its energy is 1.Fig. 3 is the schematic diagram of a symbol waveform of transmitter emission, and the impulse waveform in this symbol is exactly pulse shown in Figure 1, and the duration of this symbol is T, and be T the blanking time between each pulse ₁, the number of pulse is N ₁), its duration is T.N (t) is the white Gaussian noise in the channel.

For the formula of the received signal x (t) that makes above-mentioned receiver simpler and clearer, the waveform w of the symbol that the present invention receives at this definition receiving terminal _R(t) be: $w_R\left(t\right)=\underset{l=1}{\overset{L}{\mathrm{\Σ}}}{\mathrm{\α}}_l{w}_s(t-{\mathrm{\τ}}_l);$ This w _R(t) signal packet contains the attenuation coefficient α in each footpath in the multipath channel _lAnd delay time T _lIf, with waveform w _R(t) formula of formula substitution received signal x (t), then received signal x (t) formula can be reduced to: $x\left(t\right)=\sqrt{\mathrm{\ϵ}}\underset{i=1}{\overset{+\∞}{\mathrm{\Σ}}}{s}_i{w}_R(t-(i-1)T+t_1)+n\left(t\right).$

In order to suppress noise, need carry out temporal average to the received signal x (t) of this formulate: from the moment T of beginning received signal ₁Beginning, the N segment length of choosing continuously among the received signal x (t) is the signal of 4 symbol durations, and after adding up, asks its mean value, obtains the signal behind the time average

Its computing formula is: $\stackrel{\&OverBar;}{x}\left(t\right)=\frac{1}{N}\underset{{\mathrm{<figure-callout\; id="1"\; label="n"\; filenames="A20071011873000191.png,A20071011873000211.png"\; state="\{\{state\}\}">n</figure-callout>}}_1=1}{\overset{N}{\mathrm{\&Sigma;}}}x(t+4(n_1-1)T);$ In the formula, the span of time variable t be [0,4T), represent the signal behind this time average Be 4 symbol durations, N is an accumulative frequency, n ₁Expression the sequence number of the number of winning the confidence hop count.

(13) make up 4 alternative signal demodulation template: in order to extract initial signal demodulation template, ignoring under the prerequisite of noise, according to the cycle autocorrelation performance of training symbol sequence, making up 4 alternative signal demodulation template V earlier _k(t), wherein natural number k is the sequence number of signal demodulation template, and span is [1,4].Concrete steps are: with the signal behind the time average that obtains in the step (12)

Be equally divided into 4 sections, every segment signal has the duration of a symbol, can be expressed as: x _m(t), wherein natural number m (referring to the schematic diagram of the received signal x that does not add noise (t) shown in Fig. 4 (a), the moment that begins to receive is t for the sequence number of this 4 segment signal ₁).Again this 4 segment signal is combined in twos by end to end order, constituted the signal that 4 segment length are two symbol durations: X _m(t), the order of this combination is followed successively by: x ₁(t) and x ₂(t) combination, x ₂(t) and x ₃(t) combination, x ₃(t) and x ₄(t) combination, x ₄(t) and x ₁(t) combination.Then, with 4 segment signal X after this combination _m(t) { 1,1,1, a symbol the among-1}, and after it is added up is averaged and is obtained an alternative signal demodulation template V to multiply by symbol sebolic addressing respectively in order ₁(t), this computational process can be referring to shown in Figure 5.Obtain other 3 alternative signal demodulation template according to above-mentioned steps again, but this moment 4 segment signal X _m(t) symbol sebolic addressing that is multiplied each other is different from the former, be respectively 1,1 ,-1,1}, 1 ,-1,1, and 1}, 1,1,1,1}.

Easy for what illustrate, temporarily ignore earlier noise section n (t).From Fig. 4 (a) as can be seen, generally speaking, with signal

After being equally divided into 4 sections, every segment signal x _m(t) be made up of two parts: the front end of a tail end of previous training symbol (among the figure be the part of background with the grid) and a back training symbol (among the figure be the part of background with the vertical element) is total up to the duration of a symbol.After again this 4 segment signal being made up in twos by above-mentioned setting order, obtain the signal of 4 sections two symbol durations.From Fig. 4 (a), can find out every segment signal X of this moment _m(t) comprise three parts: the fore-end of the end section of first training symbol, second complete training symbol and the 3rd training symbol.(note: X ₄(t) by x ₄(t) and x ₁(t) combine, do not meet this combination rule originally, but because training symbol sequence has periodicity, so can use x ₁(t) replace coming x ₄(t) so signal of the back symbol duration is X ₄(t) also meet above-mentioned rule).So,, be somebody's turn to do the signal X that forms by three parts through the derivation of equation of the received signal x (t) after simplifying _m(t) can be expressed as: $X_m\left(t\right)=\sqrt{\mathrm{\&epsiv;}}\{{\mathrm{<figure-callout\; id="2"\; label="s\; N"\; filenames="A20071011873000211.png,A20071011873000221.png"\; state="\{\{state\}\}">s</figure-callout>}}_{N_{}}{w}_R(t+e)+{\mathrm{<figure-callout\; id="2"\; label="s\; N"\; filenames="A20071011873000211.png,A20071011873000221.png"\; state="\{\{state\}\}">s</figure-callout>}}_{N_{}}{w}_R(t-T+e)+s_{{\mathrm{<figure-callout\; id="2"\; label="N"\; filenames="A20071011873000211.png,A20071011873000221.png"\; state="\{\{state\}\}">N</figure-callout>}}_2+m+2}{w}_R(t-2T+e)\};$ In the formula, the span of time variable t be [0,2T), expression signal X _m(t) duration is two symbol durations, and three parts in this formula correspond respectively to the fore-end of the end section of first training symbol, second complete training symbol and the 3rd training symbol; N ₂Expression is t constantly ₁The number of Fa She symbol before, e=t ₁-N ₂T represents t constantly ₁Before Fa She last symbol from original position to moment t ₁Duration; Again with described 4 segment signal X _m(t) { 1,1,1, a symbol the among-1} is averaged after adding up, and just can obtain an alternative signal demodulation template V to multiply by symbol sebolic addressing respectively ₁(t); This resolving is: $V_1\left(t\right)=\frac{1}{4\sqrt{\mathrm{\&epsiv;}}}\underset{m=1}{\overset{4}{\mathrm{\&Sigma;}}}\left\{X_m\left(t\right)s_m\right\};$ In the formula, the span of time variable t be [0,2T), expression signal V ₁(t) duration is two symbol durations, s _m, preceding 4 of training symbol sequence are got in m ∈ [1,4] expression: and 1,1,1 ,-1}; This moment V ₁(t) can be expressed as:

$V_1\left(t\right)=\frac{1}{4}\underset{m=1}{\overset{4}{\mathrm{\&Sigma;}}}\left\{\{{\mathrm{<figure-callout\; id="2"\; label="s\; N"\; filenames="A20071011873000211.png,A20071011873000221.png"\; state="\{\{state\}\}">s</figure-callout>}}_{N_{}}{w}_R(t+e)+{\mathrm{<figure-callout\; id="2"\; label="s\; N"\; filenames="A20071011873000211.png,A20071011873000221.png"\; state="\{\{state\}\}">s</figure-callout>}}_{N_{}}{w}_R(t-T+e)+{\mathrm{<figure-callout\; id="2"\; label="s\; N"\; filenames="A20071011873000211.png,A20071011873000221.png"\; state="\{\{state\}\}">s</figure-callout>}}_{N_{}}{w}_R(t-2T+e)\}{s}_m\right\}$

$=\frac{1}{4}\left\{\underset{m=1}{\overset{4}{\mathrm{\&Sigma;}}}{\mathrm{<figure-callout\; id="2"\; label="s\; N"\; filenames="A20071011873000211.png,A20071011873000221.png"\; state="\{\{state\}\}">s</figure-callout>}}_{N_{}}{s}_m\right\}{w}_R(t+e)+\frac{1}{4}\left\{\underset{m=1}{\overset{4}{\mathrm{\&Sigma;}}}{\mathrm{<figure-callout\; id="2"\; label="s\; N"\; filenames="A20071011873000211.png,A20071011873000221.png"\; state="\{\{state\}\}">s</figure-callout>}}_{N_{}}{s}_m\right\}{w}_R(t-T+e)+\frac{1}{4}\left\{\underset{m=1}{\overset{4}{\mathrm{\&Sigma;}}}{\mathrm{<figure-callout\; id="2"\; label="s\; N"\; filenames="A20071011873000211.png,A20071011873000221.png"\; state="\{\{state\}\}">s</figure-callout>}}_{N_{}}{s}_m\right\}{w}_{R(t-2T+e)}\}.$

In order to make this formula simpler and clearer, the present invention defines the periodic auto-correlation function R of training symbol sequence _s(j) be: $R_s\left(j\right)=\frac{1}{4}\underset{m=v+1}{\overset{v+4}{\mathrm{\&Sigma;}}}{s}_m{s}_{m+j},$ In the formula, integer j is the parameter of periodic auto-correlation function, the original position of selected symbol sebolic addressing when nonnegative integer v is used to determine the computing cycle auto-correlation function; In the above-mentioned formula of this periodic auto-correlation function formula substitution, signal V then ₁(t) can be reduced to: V ₁(t)=R _s(N ₂) w _R(t+e)+R _s(N ₂+ 1) w _R(t-T+e)+R _s(N ₂+ 2) w _R(t-2T+e)

Other three alternative demodulation templates V ₂(t), V ₃(t) and V ₄(t) calculate according to above-mentioned same step and obtain, wherein the span of time variable t also be [0,2T), represent that the duration of these 3 signals is two symbol durations, but this moment will acquisition V ₁(t) s in the resolving _mUse s respectively _M+1, s _M+2And s _M+3Replace it, to represent this 4 segment signal X _m(t) sequence that is multiplied each other be respectively 1,1 ,-1,1}, 1 ,-1,1,1}, 1,1,1,1}; The alternative demodulation templates V that obtains like this ₂(t), V ₃(t) and V ₄(t) be respectively:

V ₂(t)＝R _s(N ₂-1)w _R(t+e)+R _s(N ₂)w _R(t-T+e)+R _s(N ₂+1)w _R(t-2T+e)

V ₃(t)＝R _s(N ₂-2)w _R(t+e)+R _s(N ₂-1)w _R(t-T+e)+R _s(N ₂)w _R(t-2T+e)

V ₄(t)＝R _s(N ₂-3)w _R(t+e)+R _s(N ₂-2)w _R(t-T+e)+R _s(N ₂-1)w _R(t-2T+e)。

(14) obtain initial signal demodulation template: in 4 alternative signal demodulation template that in step (13), obtain, have only one to be satisfactory initial signal demodulation template, it has comprised the complete information of a symbol that receives, thereby energy is maximum, and other three alternative signal demodulation template have included only the partial information of a symbol that receives or do not comprised any signal message.Therefore, the concrete step that obtains initial demodulation templates is: utilize 4 alternative signal demodulation templates V of integrator to obtaining in the step (13) _k(t) carry out energy measuring and comparison respectively, the signal demodulation template of choosing energy maximum wherein is as required initial signal demodulation template V _K1(t).Fig. 4 (b) is this initialize signal demodulation templates of choosing, can see having only on the symbol duration position having symbolic information, and other parts are pure noise.τ _eBe synchronization point.

According to the training symbol sequence that sends, the periodic auto-correlation function of the training symbol sequence of the present invention's definition $R_s\left(j\right)=\frac{1}{4}\underset{m=v+1}{\overset{v+4}{\mathrm{\&Sigma;}}}{s}_m{s}_{m+j}$ Should satisfy following characteristic: $R_s\left(j\right)=\left\{\begin{array}{cc}1& j=4q\\ 0& j\&NotEqual;4q\end{array}\right.q=0,\&PlusMinus;1,\&PlusMinus;2\&CenterDot;\&CenterDot;\&CenterDot;;$ Promptly have only when parameter value is 4 integral multiple, the periodic auto-correlation function value of training symbol sequence is 1, and other situation values are 0.According to this characteristic, 4 alternative signal demodulation template V that from step (13), obtain _k(t) in the expression formula as can be seen: work as N ₂Numerical value determine after, have only an alternative signal demodulation template to comprise the information of a complete receiving symbol, promptly make this V _k(t) w in the expression formula _R(t-T+e) periodic auto-correlation function that is multiplied each other is 1, and the periodic auto-correlation function that all the other two parts multiplied each other is 0.Simultaneously, in other 3 alternative signal demodulation template, w _R(t-T+e) periodic auto-correlation function that is multiplied each other is 0 certainly, so they do not comprise a complete received signal information; Therefore, the template that comprises a complete receiving symbol information is exactly needed initialize signal demodulation templates, and its energy also is maximum with respect to other signal templates.Therefore, the present invention adopts comparison energy size to obtain it, and its computing formula is: ${\mathrm{<figure-callout\; id="1"\; label="k"\; filenames="A20071011873000191.png,A20071011873000211.png"\; state="\{\{state\}\}">k</figure-callout>}}_1=\underset{k\&Element;\left[\mathrm{1,4}\right]}{\arg \max }J\left(k\right),$ In the formula, integral formula $J\left(k\right)={\&Integral;}_0^{2T}{V}_k^2\left(t\right)\mathrm{dt}$ Expression is to the pairing template signal V of different value of K _k(t) ask energy, $\underset{k\&Element;\left[\mathrm{1,4}\right]}{\arg \max }J\left(k\right)$ Pairing k value when J (k) maximum is got in expression.

(2) obtain the signal Synchronization parameter:

Because of the middle initial signal demodulation template that makes up of step () is the duration of two symbols, but wherein have only on the continuous position of a symbol duration and include a received symbolic information, all the other positions are all pure noise, this can have influence on the accuracy of data symbol demodulation, must before the data symbol demodulation, delete this pure noise (referring to Fig. 4 (b)) earlier, could carry out demodulation to the data symbol exactly.This just requires to obtain the synchronization parameter of signal, promptly finds the original position of received symbolic information from initial signal demodulation template.Concrete steps are: from the original position of initial signal demodulation template, by integrator the signal of a symbol duration is carried out the energy maximum value search, be T the blanking time of this search ₂, the number of times of search is the blanking time T of the duration T of each symbol divided by this search ₂The merchant: T/T ₂, wherein the pairing position of energy maximum is exactly the original position of the symbolic information that receives, also is the synchronization parameter τ of signal _e, can be referring to Fig. 4 (b).

The computing formula of the peaked search procedure of energy that this step is carried out is: $n_e=\underset{n_{}}{\arg \mathrm{<figure-callout\; id="2"\; label="max\; n"\; filenames="A20071011873000211.png,A20071011873000221.png"\; state="\{\{state\}\}">max</figure-callout>}}{\&Integral;}_0^{\mathrm{<figure-callout\; id="1"\; label="T\; V\; k"\; filenames="A20071011873000191.png,A20071011873000211.png"\; state="\{\{state\}\}">T</figure-callout>}}{V}_{k_{}}^{}{{}_1}_2^{}(t+(n_2-1)T_2)\mathrm{dt};$ In the formula, natural number n ₂Be the sequence number of searching times, its span be [1, T/T ₂], n _eThe sequence number of pairing searching times for can measure maximum the time, the synchronization parameter τ that is found the solution _eComputing formula be: τ _e=(n _e-1) T ₂Need to prove: the precision of synchronization parameter depends on the scouting interval time T ₂Setting, T ₂Big more, precision is low more, but complexity is low more; T ₂More little, precision is high more, and complexity is also high more thereupon, will take into account the unification of the two in the practical application.

(3) make up final signal demodulation template:

Behind the picked up signal synchronization parameter, need revise, remove pure noise wherein, obtain final signal demodulation template initial signal demodulation template.Concrete steps are: with the signal Synchronization parameter that obtained in the step (two) is starting point, one section signal that duration is a symbol of intercepting on initial signal demodulation template, and this segment signal is final signal demodulation template V _e(t): $V_e\left(t\right)=V_{k_1}(t+{\mathrm{\&tau;}}_e);$ In the formula, the span of time variable t be [0, T), the expression V _e(t) be that a symbol is long, can be referring to the final signal demodulation template ideally shown in Fig. 4 (c), during practical application, because T ₂The selection of size, it is residual to produce certain noise section, but does not influence overall situation.

(4) utilize synchronization parameter and final signal demodulation template that the data symbol that receives is carried out demodulation:

With the synchronization parameter is starting point, chooses the signal of a symbol duration from received signal x (t), allows this selected signal and final signal demodulation template do integral operation by integrator, obtains judgement amount; Rule out the numerical value of the data symbol that receives again by decision device: judgement amount is greater than decision threshold 0, then the data of Fa Songing be+1; Judgement amount is less than decision threshold 0, and then the data of Fa Songing are-1; Choose the received signal of next symbol duration then, according to the data symbol that receives with the quadrat method judgement next one, so cycling finishes up to Data Receiving.

The computing formula of the judgement amount of a symbol is: $D_i={\&Integral;}_0^T{V}_e\left(t\right)x(t+(i-1)T)\mathrm{dt};$ In the formula, natural number i is the sequence number of the symbol that will adjudicate, the symbol that rules out

Then be expressed as: ${\hat{s}}_i=\mathrm{sign}\left[D_i\right];$ In the formula, sign[] the expression decision device, when parameter during greater than decision threshold, judgement is 1, when parameter during less than decision threshold, adjudicates and is-1.

In order to verify the validity of this method of reseptance, the inventor has carried out implementing test and performance simulation to method of the present invention with computer, and the test parameters during emulation is set as follows: T=800 nanosecond; N ₁=16; Use impulse waveform shown in Figure 1, T ₃=1 nanosecond; The CM1 model that multipath channel adopts IEEE 802.15.3a working group to formulate; T ₂=25 nanoseconds, N ₂=32.Whether adopt different accumulative frequencies then in an embodiment and exist the different situations of intersymbol interference to carry out emulation testing, its performance simulation result as shown in Figure 6 and Figure 7.

Fig. 6 is the curve comparison diagram of the present invention's embodiment pairing bit error rate and signal to noise ratio when adopting different accumulative frequency of being used for the pulse ultra-broadband communication system received signal.Wherein trunnion axis is symbol signal to noise ratio E _b/ N ₀, vertical axis is bit error rate BER, and N is the accumulative frequency in the signal averaging process, and along with the increase of N, bit error rate descends as can be seen.When N=200, performance of BER is very near theoretical limit.

The bit error rate when solid line among Fig. 7 represents that intersymbol interference does not exist and the curve of symbol signal to noise ratio, the curve of bit error rate and symbol signal to noise ratio when dotted line is represented to have intersymbol interference.Establish L=45 earlier, at this moment τ _L+ T ₃＜T ₁The time, intersymbol interference does not exist.Establish L=65 again, at this moment τ _L+ T ₃＞T ₁The time, intersymbol interference exists.As can be seen from the figure, though when intersymbol interference existed, performance had reduction slightly, performance is not about the same when not existing with intersymbol interference, so this method has good tolerance to intersymbol interference.

Claims (9)

1, a kind of signal acceptance method that is used for pulse ultra-broadband communication system is characterized in that: send ad hoc training symbol sequence, so that utilize its unique cycle autocorrelation performance, directly extract initial signal demodulation template from the signal that receives; Utilize integrator that this initial signal demodulation template is carried out the energy maximum value search again, with the synchronization parameter of picked up signal; Then, utilize synchronization parameter that this initial signal demodulation template is revised, remove redundant pure noise, obtain final signal demodulation template; At last, utilize final signal demodulation template and synchronization parameter that the data symbol that is received is separated to be in harmonious proportion at receiving terminal and adjudicate, finish the reception of signal.

2, signal acceptance method according to claim 1, it is characterized in that: described training symbol sequence is meant in order to reach the purpose of prior setting, in the special symbol sequence of setting quantity that has of signal sending end emission, signal receiving end is known in advance to the special symbol sequence of this setting quantity.

3, signal acceptance method according to claim 1 is characterized in that: described method comprises following operating procedure:

(1) extracts initial signal demodulation template: send ad hoc training symbol sequence,, from the signal that receives, directly make up 4 alternative signal demodulation template so that utilize its unique cycle autocorrelation performance; Described training symbol sequence be the symbol sebolic addressing that sends of circulation 1,1,1-1}, or by 31 and 1-1 other symbol sebolic addressing signals of forming; Adopt integrator from these 4 alternative signal demodulation template, to select energy the maximum then, be the initial signal demodulation template V that will extract _K1(t);

(2) obtain the signal Synchronization parameter: because of the initial signal demodulation template that makes up is the duration of two symbols, wherein have only on the continuous position of a symbol duration and include a received symbolic information, all the other positions are all pure noise, have only this pure noise of deletion earlier, could carry out demodulation to the data symbol exactly; Must obtain the synchronization parameter of signal for this reason, promptly from initial signal demodulation template, find the original position of the described symbolic information that receives; Concrete steps are: from the original position of initial signal demodulation template, by integrator the signal of a symbol duration is carried out the energy maximum value search, the number of times of this search is the blanking time T of the duration T of each symbol divided by this search ₂The merchant: T/T ₂, wherein the pairing position of energy maximum is the original position of the symbolic information that receives, also is the synchronization parameter τ of signal _e

(3) make up final signal demodulation template: behind the picked up signal synchronization parameter, initial signal demodulation template is revised, removed pure noise wherein, obtain final signal demodulation template; Concrete steps are: with the signal Synchronization parameter that is obtained is starting point, and the signal that one section duration of intercepting is a symbol on initial signal demodulation template is final signal demodulation template V _e(t);

(4) utilize synchronization parameter and final signal demodulation template that the data symbol that receives is carried out demodulation: to be starting point with the synchronization parameter, from received signal x (t), choose the signal of a symbol duration, make the signal of this symbol duration and final signal demodulation template do integral operation by integrator, obtain judgement amount; Rule out the numerical value of the data symbol that receives again by decision device: judgement amount is greater than decision threshold, then the data of Fa Songing be+1; Judgement amount is less than decision threshold, and then the data of Fa Songing are-1; Choose the received signal of next symbol duration then, according to the data symbol that receives with the quadrat method judgement next one, so cycling finishes up to Data Receiving.

4, signal acceptance method according to claim 3 is characterized in that: described step (1) further comprises following content of operation:

(11) after the parameters of initialized setting system, begin to send training symbol sequence and received signal: the system parameters of setting comprises: the moment t of transmitter proceed-to-send signal ₀Be 0, system or receiver begin the moment t of received signal ₁, each symbol is by N ₁Individual pulse is formed, and its duration is T, and T blanking time of each pulse in each symbol ₁For: T ₁=T/N ₁The number of times that signal adds up when seeking time mean value is N, and the time span of every segment signal of the training symbol sequence of choosing all is 4 symbol durations; In making up 4 alternative signal demodulation template processes employed 4 group code sequences be respectively 1,1,1 ,-1}, 1,1 ,-1,1}, 1 ,-1,1,1}, 1,1,1,1}; Time scouting interval in the signal Synchronization process is T ₂, then Sou Suo number of times is T/T ₂, promptly T is T ₂Integral multiple; Decision device thresholding in the data symbol demodulating process is 0;

(12) compacting noise: x (t) carries out time average to the received signal, promptly from beginning the moment t of received signal ₁Initial, receiving N section time span continuously is the training symbol sequence cycle, i.e. the signal of 4 symbol durations, and after adding up, ask its mean value obtains the signal behind the time average Should

In noise eliminate fully as yet, but with respect to x (t), noise wherein is pressed and weakens greatly;

(13) make up 4 alternative signal demodulation template: ignoring under the prerequisite of noise, according to the cycle autocorrelation performance of training symbol sequence, making up 4 alternative signal demodulation template V earlier _k(t), wherein natural number k is the sequence number of signal demodulation template, and span is [1,4]; Concrete steps are: with the signal behind the time average that obtains in the step (12)

Be equally divided into 4 sections, every section is the signal of a symbol duration: x _m(t), wherein natural number m is the sequence number of this 4 segment signal; Again this 4 segment signal is combined in twos by end to end order, constituted the signal that 4 segment length are two symbol durations: X _m(t), the order of this combination is followed successively by: x ₁(t) and x ₂(t) combination, x ₂(t) and x ₃(t) combination, x ₃(t) and x ₄(t) combination, x ₄(t) and x ₁(t) combination; Then, with 4 segment signal X after this combination _m(t) { 1,1,1, each symbol the among-1}, and after it is added up is averaged and is obtained an alternative signal demodulation template V to multiply by symbol sebolic addressing respectively in order ₁(t); Obtain other 3 alternative signal demodulation template according to above-mentioned steps again, but this moment 4 segment signal X _m(t) symbol sebolic addressing that is multiplied each other is different from the former, be respectively 1,1 ,-1,1}, 1 ,-1,1, and 1}, 1,1,1,1};

(14) obtain initial signal demodulation template: utilize 4 the alternative signal demodulation templates Vs of integrator to obtaining in the step (13) _k(t) carry out energy measuring and comparison respectively, the signal demodulation template of choosing energy maximum wherein is as required initial signal demodulation template V _K1(t).

5, signal acceptance method according to claim 4 is characterized in that: described step (12) further comprises following content of operation:

If the signal x (t) that receiver receives is: $x\left(t\right)=\sqrt{\mathrm{\&epsiv;}}\underset{i=1}{\overset{+\&infin;}{\mathrm{\&Sigma;}}}{s}_i\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_l{w}_s(t-(i-1)T-{\mathrm{\&tau;}}_1+t_1)+n\left(t\right);$ In the formula, the span of time variable t be [0 ,+∞), t ₁Be the moment of beginning received signal, ε is the energy of each burst transmissions, s _iBe the training symbol sequence that adopts the pulse amplitude modulation mode to send, its value is ± 1, and natural number i is the sequence number of this training symbol; α _lAnd τ _lBe respectively the attenuation coefficient and the time of delay in each footpath in the multipath channel, natural number l is the sequence number of multipath channel, and L is the number of multipath channel; Because time t from beginning to transmit ₀Start the time t of received signal to receiver ₁Interval (t ₁-t ₀) the common the shortest time delay τ that all arrives receiver much larger than multipath signal from transmitter ₁So, establish τ ₁=0; And analyze for convenience, being provided with in the transmission course of this symbol sebolic addressing does not have intersymbol interference, and promptly multipath signal arrives the longest time delay τ of receiver from transmitter _LDuration T with each pulse ₃Sum is less than T blanking time between each pulse ₁: τ _L+ T ₃＜T ₁So,, the duration of the multipath signal of previous symbol can not extend to the time of reception of next symbol; w _s(t) for transmitting terminal send by N ₁The waveform of the symbol that individual pulse is formed, its duration is T; N (t) is the white Gaussian noise in the channel;

If the waveform w of the symbol that receiving terminal receives _R(t) be: $w_R\left(t\right)=\underset{l=1}{\overset{L}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_l{w}_s(t-{\mathrm{\&tau;}}_l),$ This w _R(t) signal packet contains the attenuation coefficient α in each footpath in the multipath channel _lAnd delay time T _lThen above-mentioned received signal x (t) then can be reduced to: $x\left(t\right)=\sqrt{\mathrm{\&epsiv;}}\underset{i=1}{\overset{+\&infin;}{\mathrm{\&Sigma;}}}{s}_i{w}_R(t-(i-1)T+t_1)+n\left(t\right);$

Be the compacting noise, this received signal x (t) carried out time average: from the moment t of beginning received signal ₁Beginning, the N segment length of choosing continuously among the received signal x (t) is the signal of 4 symbol durations, and after adding up, asks its mean value, obtains the signal behind the time average Its computing formula is: $\stackrel{\&OverBar;}{x}\left(t\right)=\frac{1}{N}\underset{n_1=1}{\overset{N}{\mathrm{\&Sigma;}}}x(t+4(n_1-1)T);$ In the formula, the span of time variable t be [0,4T), represent the signal behind this time average

Be 4 symbol durations, N is an accumulative frequency, n ₁Expression the sequence number of the number of winning the confidence hop count.

6, signal acceptance method according to claim 4 is characterized in that: described step (13) further comprises following content of operation: with the signal behind the time average that obtains

Be divided into 4 sections, every segment signal x _m(t) all comprise two parts: the fore-end of the end section of previous training symbol and a back training symbol is total up to the duration of a symbol; After again this 4 segment signal being made up in twos by above-mentioned setting order, obtain the signal of 4 sections two symbol durations; Every segment signal X of this moment _m(t) comprise three parts: the fore-end of the end section of first training symbol, second complete training symbol and the 3rd training symbol, this signal X _m(t) can be expressed as: $X_m\left(t\right)=\sqrt{\mathrm{\&epsiv;}}\{s_{N_2+m}{w}_R(t+e)+s_{N_2+m+1}{w}_R(t-T+e)+s_{N_2+m+2}{w}_R(t-2T+e)\};$

In the formula, the span of time variable t be [0,2T), expression signal X _m(t) duration is two symbol durations, and three parts in this formula correspond respectively to the fore-end of the end section of first training symbol, second complete training symbol and the 3rd training symbol; N ₂Expression is t constantly ₁The number of Fa She symbol before, e=t ₁-N ₂T represents t constantly ₁Before Fa She last symbol from original position to moment t ₁Duration; Again with described 4 segment signal X _m(t) { 1,1,1, a symbol the among-1} is averaged after adding up, and just can obtain an alternative signal demodulation template V to multiply by symbol sebolic addressing respectively ₁(t); This resolving is: $V_1\left(t\right)=\frac{1}{4\sqrt{\mathrm{\&epsiv;}}}\underset{m=1}{\overset{4}{\mathrm{\&Sigma;}}}\left\{X_m\left(t\right)s_m\right\};$ In the formula, the span of time variable t be [0,2T), expression signal V ₁(t) duration is two symbol durations, s _m, preceding 4 of training symbol sequence are got in m ∈ [1,4] expression: and 1,1,1 ,-1}; This moment V ₁(t) can be expressed as:

$V_1\left(t\right)=\frac{1}{4}\underset{m=1}{\overset{4}{\mathrm{\&Sigma;}}}\left\{\right\{s_{N_2+m}{w}_R(t+e)+s_{N_2+m+1}{w}_R(t-T+e)+s_{N_2+m+2}{w}_R(t-2T+e)\left\}{s}_m\right\}$

$=\frac{1}{4}\left\{\underset{m=1}{\overset{4}{\mathrm{\&Sigma;}}}{s}_{N_2+m}{s}_m\right\}{w}_R(t+e)+\frac{1}{4}\left\{\underset{m=1}{\overset{4}{\mathrm{\&Sigma;}}}{s}_{N_2+m+1}{s}_m\right\}{w}_R(t-T+e)+\frac{1}{4}\left\{\underset{m=1}{\overset{4}{\mathrm{\&Sigma;}}}{s}_{N_2+m+2}{s}_m\right\}{w}_R(t-2T+e)\}$

In order to make above-mentioned formula simpler and clearer, the periodic auto-correlation function of definition training symbol sequence is: $R_s\left(j\right)=\frac{1}{4}\underset{m=v+1}{\overset{v+4}{\mathrm{\&Sigma;}}}{s}_m{s}_{m+j},$ In the formula, integer j is the parameter of periodic auto-correlation function, the original position of selected symbol sebolic addressing when nonnegative integer v is used to determine the computing cycle auto-correlation function; In the above-mentioned formula of this periodic auto-correlation function formula substitution, signal V then ₁(t) can be reduced to:

V ₁(t)＝R _s(N ₂)w _R(t+e)+R _s(N ₂+1)w _R(t-T+e)+R _s(N ₂+2)w _R(t-2T+e)

Calculate according to above-mentioned same step again, obtain other three alternative signal demodulation template V ₂(t), V ₃(t) and V ₄(t), the span of their time variable t also be [0,2T), represent that the duration of these three signals is two symbol durations, but will obtain V this moment ₁(t) s in the resolving _mUse s respectively _M+1, s _M+2And s _M+3Replace it, to represent this 4 segment signal X _m(t) sequence that is multiplied each other be respectively 1,1 ,-1,1}, 1 ,-1,1,1}, 1,1,1,1}; The alternative signal demodulation template V that obtains like this ₂(t), V ₃(t) and V ₄(t) be respectively:

V ₂(t)＝R _s(N ₂-1)w _R(t+e)+R _s(N ₂)w _R(t-T+e)+R _s(N ₂+1)w _R(t-2T+e)

V ₃(t)＝R _s(N ₂-2)w _R(t+e)+R _s(N ₂-1)w _R(t-T+e)+R _s(N ₂)w _R(t-2T+e)

V ₄(t)＝R _s(N ₂-3)w _R(t+e)+R _s(N ₂-2)w _R(t-T+e)+R _s(N ₂-1)w _R(t-2T+e)。

7, according to claim 4 or 6 described signal acceptance methods, it is characterized in that: described step (14) further comprises following content of operation: according to the training symbol sequence that sends, the periodic auto-correlation function of the training symbol sequence of described definition $R_s\left(j\right)=\frac{1}{4}\underset{m=v+1}{\overset{v+4}{\mathrm{\&Sigma;}}}{s}_m{s}_{m+j}$ Should satisfy following characteristic: $R_s\left(j\right)=\left\{\begin{array}{cc}1& j=4q\\ 0& j\&NotEqual;4q\end{array}\right.$ Q=0, ± 1, ± 2 Promptly have only when parameter value is 4 integral multiple, the periodic auto-correlation function value of training symbol sequence is 1, and other situation values are 0; According to this characteristic, 4 alternative signal demodulation template V that from step (13), obtain _k(t) in the expression formula as can be seen: work as N ₂Numerical value determine after, have only an alternative signal demodulation template to comprise the information of a complete receiving symbol, promptly make this V _k(t) w in the expression formula _R(t-T+e) periodic auto-correlation function that is multiplied each other is 1, and the periodic auto-correlation function that all the other two parts multiplied each other is 0; Simultaneously, in other 3 alternative signal demodulation template, w _R(t-T+e) periodic auto-correlation function that is multiplied each other is 0 certainly, so they do not comprise the information of a complete receiving symbol; Therefore, the template that comprises a complete receiving symbol information is exactly needed initialize signal demodulation templates, and its energy also is maximum with respect to other signal templates; The energy size obtains it so adopt relatively, and its computing formula is: $k_1=\underset{k\&Element;\left[\mathrm{1,4}\right]}{\arg \max }J\left(k\right),$ In the formula, formula $J\left(k\right)={\&Integral;}_0^{2T}{V}_k^2\left(t\right)\mathrm{dt}$ Expression is to the pairing template signal V of different value of K _k(t) integration is asked energy,

Pairing k value when J (k) maximum is got in expression.

8, signal acceptance method according to claim 3 is characterized in that: in the described step (2), the computing formula of the peaked search procedure of energy of carrying out is: $n_e=\underset{n_2\&Element;[1,T/T_2]}{\arg \max }{\&Integral;}_0^T{V}_{k_1}^2(t+(n_2-1)T_2)\mathrm{dt};$

In the formula, natural number n ₂Be the sequence number of searching times, its span be [1, T/T ₂], n _eThe sequence number of pairing searching times for can measure maximum the time, the synchronization parameter τ that is found the solution _eComputing formula be: τ _e=(n _e-1) T ₂Need to prove: the precision of synchronization parameter depends on scouting interval T ₂Setting, T ₂Big more, precision is low more, but complexity is low more; T ₂More little, precision is high more, and complexity is also high more thereupon, and practical application will take into account the two.

9, signal acceptance method according to claim 3 is characterized in that: in the described step (4), the computing formula of the judgement amount of each symbol is: $D_i={\&Integral;}_0^T{V}_e\left(t\right)x(t+(i-1)T)\mathrm{dt};$ In the formula, natural number i is the sequence number of the symbol that will adjudicate, the symbol  that rules out _iThen be expressed as:  _i=sign[D _i] in the formula, sign[] the expression decision device, when parameter during greater than decision threshold, judgement is 1, when parameter during less than decision threshold, adjudicates and is-1.

Images