CN106878218A - A kind of high reliability demodulation method for IEEE802.15.4 - Google Patents

Abstract

A kind of high reliability BPSK demodulation methods for IEEE802.15.4, useful frequency offset information is extracted first with the corresponding channel received data of the lead code of 32 bits, then receiving sampled signal to the corresponding complex base band of PSDU carries out bit-level difference processing, obtains adjudicating observation, afterwards using extract frequency offset information to judgement observation compensate after carry out detection judgement, the PSDU data that will finally receive send MAC layer to and carry out CRC check.The present invention is based on influence negligible theory of the noise to useful signal when the Taylor series expansion and big signal to noise ratio of arcsin function, there is provided a kind of low-power consumption suitable for IEEE 802.15.4, highly reliable and inexpensive incoherent BPSK receivers.

Description

A kind of high reliability demodulation method for IEEE802.15.4

Technical field

The present invention relates to communication signal waveforms detection technique field, specifically a kind of height for IEEE802.15.4 Reliability BPSK demodulation methods.

Background technology

IEEE 802.15.4 are ZigBee, the basis of the specification such as WirelessHART, describe low rate radio individual office The physical layer and media access control protocol of domain net.It is initially operating in the ISM band of 868/915MHz, 2.4GHz, data Transmission rate reaches as high as 250kbps.Low-power consumption, the advantage of low cost make it in data sampling and processing and analysis, remote control Essence is made the various fields such as agricultural automation, environmental protection and monitoring and is applied widely.The newest standards provided in 2011 In, 314-316MHz, 430-434MHz, 779-787MHz and 950-956MHz working frequency range are added again.

As shown in figure 1,802.15.4 agreements are different with message transmission rate using modulation system on different carrier wave frequency ranges. 48 channels are provided altogether in four typical frequency ranges：1 channel of 868MHz frequency ranges, 10 channels of 915MHz frequency ranges, 2450MHz 16 channels of frequency range, 21 channels of 950MHz frequency ranges.As shown in Fig. 2 in 868/915/950-MHz frequency ranges, signal transacting mistake Cheng Xiangtong, simply data rate is different.Sender is first by the binary data difference of physical layer data protocol element (PPDU) Coding, then again by differential coding after it is each the piece sequence that length is 15 is converted to position, finally using BPSK modulate to letter On road.Differential coding is that the bit that each original bit of data and previous differential coding are generated is carried out into XOR：Wherein E_nIt is the result of differential coding, R_nIt is the original bit to be encoded, E_n-1It is last differential coding As a result.To the packet that each sends, R₁It is first original bit, calculates E₁When assume E₀=0.Differential ference spiral process and volume Code process is similar to：To the packet that each is received, E₁It is first bit of needs demodulation, calculates E₁When vacation Determine E₀=0.As shown in figure 3, each bit after differential coding is converted into the piece sequence that length is 15.Sequence after spread spectrum makes Modulated onto carrier wave with BPSK modulation systems.

As shown in figure 4, the first character section of IEEE 802.15.4 protocol physical layers data frame structures is that four bytes amount to Complete zero lead code of 32, transceiver can complete piece synchronously and symbol during lead code is received according to the feature of preamble sequence Number synchronization.Frame starting separator (SFD) field length is a byte, and its value is fixed as 0xA7, is expressed as physical frame Start, transceiver finishes receiving the bit synchronization that can only accomplish data after lead code, could be same by the value 0xA7 for searching for SFD fields Walk in byte.By low 7 expression of byte, its value is exactly the length of physical frame load to frame length, therefore physical frame is negative The length of load is not over 127 bytes.The payload length of physical frame is variable, referred to as physical layer service data (PSDU), it is generally used to carry mac frame.

Being traditionally used for the demodulation method of IEEE802.15.4 has two kinds, and a kind of is 868/915/950- as shown in Figure 5 Tradition typical case's complex base band non-coherent demodulation method of MHz frequency ranges.WithAfter representing channel transmission The complex base band sampled signal for receiving, wherein s (k) is transmission data to be detected, s (k) ∈ {+1, -1 }, ω₀=2 π f₀,f₀With θ is respectively frequency shift (FS) and phase offset, keeps constant in whole data frame, T_cRepresent spreading code chip cycle, η₀K () is Complex base band additive white Gaussian noise, demodulating process can be summarized as：

Step one, receive sampled signal and calculate using the corresponding complex base band of lead code of 32 bits and contain frequency offset information Observation Y：

Wherein, J represents the total number of bits amount of lead code, J=32, and N represents spreading length, N=15,1≤m≤J-1,

0≤n≤N-1, p [n+Nm] represent the channel reception value of corresponding n-th chip of m-th bit of lead code, ()^*Table Show and take conjugate operation, η₁Represent all of noise item.

Step 2, sampled signal is received to the corresponding complex base band of PSDU carry out bit-level difference processing, obtain judgement observation Value A [m]：

Wherein, r [n+Nm] represents the channel reception value of corresponding n-th chip of m-th bit of PSDU, η₂[m] represents all Noise item, E [m] represent send m-th bit data.

Step 3, using the Y in step one₀Frequency offset information is extracted, the A [m] in step 2 is compensated laggard Row detection judgement：

Wherein,Represent that, to m-th court verdict of bit data, q () is quantization function, | | modulo operation is represented, Re () is represented and is taken real part computing, and Im () is represented and taken imaginary-part operation.Bloch, M.R., Hayashi, M. and Thangaraj, A. exist in September, 2010《IEEE Transcactions on Signal Processing》On the article " IEEE that delivers 802.15.4BPSK receiver architecture based on a new efficient detection scheme” In provide a kind of computational methods of the quantization function of YBody,It is the phase of Y, is also N ω₀T_cEstimation Value,Computational methods be specifically described as：

The detection judging process of formula (3) needs to extract frequency offset information N ω from Y in advance₀T_cEstimateThen to A [m] is compensated.As described above, the weak point of tradition typical case's non-coherent demodulation method is：From formula (4), step 3 It is middle to need to obtain frequency offset information N ω by the arctangent cp cp operation of division arithmetic and complexity₀T_cEstimateThis for Energy is supplied for the strictly limited 802.15.4 network terminals, and computation complexity is larger, and energy consumption is larger, and cost of implementation is higher.

Another demodulation method is to reduce the extraction frequency deviation compensated information N ω from Y₀T_cImplementation complexity, by Lee, S., Kwon, H., Jung, Y. and Kim, J.S. exist in August, 2007《Electronics Letter》On the text delivered Chapter " Efficient non-coherent demodulation scheme for IEEE 802.15.4LR-WPAN A kind of demodulation method of the reduced form proposed in systems ", the quantization function to Y is improved, can be described as：

From formula (5), in this reduction procedure, frequency offset information N ω₀T_cEstimateCan be specifically described as：

It can be seen that, the demodulation method of the reduced form is also required to extract frequency offset information N ω from Y in advance₀T_cEstimate Then A [m] is compensated.It is right with formula (5)Approximate processing process larger error, i.e. formula (6) certainly will be produced inclined to frequency Shifting information N ω₀T_cEstimation procedure there is more serious " cross and estimate " or " owing to estimate " phenomenon, reliability can be caused significantly Decline, preferably balance matching is not reached between complicated and performance is realized.As shown in fig. 6, compared to traditional typical case's complex radical Band non-coherent demodulation method, the performance loss of the demodulation method of reduced form is serious, the carrier frequency used in emulation for 924MHz, frequency shift (FS) be maximum 80ppm specified in IEEE 802.15.4 agreements, phase offset θ (0,2 π] in obey It is uniformly distributed, the data length of PSDU is 20 bytes (160 bit), and 3000 frame mistakes are at least gathered under each signal to noise ratio.And 802.15.4 network MAC layer judges the correctness of transmission frame, automatic repeat request (ARQ) using CRC (CRC) Agreement determines therefrom that whether transmission frame needs to retransmit, and without using forward error correction (FEC) mechanism, therefore physical layer modulation is deemed-to-satisfy4 The quality of energy will produce tremendous influence to energy consumption.It is poor in channel condition, communication distance farther out when, received signal power loss It is larger.Now, if using the demodulation method of the reduced form, same PSDU data frames may could succeed by repeatedly re-transmission Received by MAC layer.The communication process of multiple re-transmission will also consume huge energy if data volume is huge, and this can reduce energy Supply the service life of deficient 802.15.4 networks.

The content of the invention

In order to solve deficiency of the prior art, Taylor series expansion and big signal to noise ratio of the present invention based on arcsin function When influence negligible theory of the noise to useful signal, there is provided it is a kind of low computation complexity suitable for IEEE 802.15.4, low The incoherent BPSK demodulation methods of power consumption, high reliability and low cost.

To achieve these goals, the concrete scheme that uses of the present invention for：

A kind of high reliability BPSK demodulation methods for IEEE802.15.4, the data frame of transmitting terminal physical layer by spread spectrum and Channel is transferred to receiving terminal after BPSK modulation, and data frame includes 32 lead codes and physical layer service data of bit PSDU；The complex base band sampled signal that receiving terminal is received is expressed asWherein s (k) is to be detected Transmission data, s (k) ∈ {+1, -1 }, ω₀=2 π f₀,f₀Frequency shift (FS) and phase offset are respectively with θ and in whole data frame Constant, the T of middle holding_cRepresent spreading code chip cycle, η₀K () is complex base band additive white Gaussian noise；Specifically demodulation step is：

Step one, lead code frequency deviation sight of the corresponding channel received data extraction comprising frequency offset information using 32 bits Measured value Y：

Wherein, J represents the total number of bits amount of lead code, and J=32, N represents spreading length, N=15,1≤m≤J-1,0≤n≤N- 1, p [n+Nm] represents the channel reception value of corresponding n-th chip of m-th bit of lead code, ()^*Expression takes conjugation fortune Calculate, η₁Represent all of noise item；

Step 2, bit-level difference processing is carried out to the corresponding complex base band sampled signals of PSDU, obtain judgement observation A [m]：

Wherein, r [n+Nm] represents the channel reception value of corresponding n-th chip of m-th bit of PSDU, η₂[m] represents all Noise item, E [m] represent send m-th bit data；

Step 3, using in step one frequency deviation observation Y extract frequency offset information, the A [m] in step 2 is compensated After carry out detection judgement：

Wherein,M-th bit data that detection judgement is obtained is represented, q () is quantization function, and q (Y) is embodied as：

Wherein, | | modulo operation is represented,Represent A [m] frequency offset N ω₀T_cEstimate；

The PSDU data that step 4, detection will be received after terminating send MAC layer to carries out CRC check；

In the step 3,Computing formula be：

Wherein, Re () is represented and is taken real part computing, and Im () is represented and taken imaginary-part operation.

As a kind of preferred scheme, in the step 3Computational methods simplified, obtain：

Beneficial effect：

1st, the demodulation method and further reduced form that the present invention is provided all are fully able to meet IEEE 802.15.4 agreements to property The requirement of energy, specifies in IEEE 802.15.4 agreements, when it is 20 byte (160 bits) that signal to noise ratio is 5~6dB, PSDU, Packet Error Ratio is less than 1 × 10 less than 1%, i.e. PER^-2, particularly it has been fully able to meet requirement when signal to noise ratio is about 1dB；

2 compare with traditional typical case's complex base band non-coherent demodulation method, and the present invention has lower computation complexity, lower energy Consumption and lower cost of implementation, tradition typical case complex base band non-coherent demodulation method by a division and an arctangent cp cp operation come To frequency offset information N ω₀T_cEstimator, from formula (10), frequency offset information N ω of the invention₀T_cEstimation side Method at most needs three times to compare, multiplication, a division and add operation twice twice；The frequency shift (FS) letter of further reduced form Breath N ω₀T_cMethod of estimation at most need three times to compare, multiplication, a division and an add operation twice, therefore with more Low computation complexity, lower energy consumption and lower cost of implementation；

3 compare with the demodulation method of traditional reduced form, and the present invention has reliability higher, the demodulation of traditional reduced form Method formula (6) is to N ω₀T_cCarry out approximate evaluation, i.e., with 0,- π andFour kinds of phases are to N ω₀T_cEstimated, caused Error is larger, and the present invention is using formula (10) to frequency offset information N ω₀T_cEstimated, " cross and estimate present in estimation procedure Meter " or " owing to estimate " phenomenon are weaker, and more preferably, reliability is higher for performance.

Brief description of the drawings

Fig. 1 is four frequency range fundamental characteristics figures of IEEE 802.15.4 protocol physical layers；

Fig. 2 is IEEE 802.15.4 agreement 868/915/950-MHz frequency range physical layer data transmission process figures；

Fig. 3 is IEEE 802.15.4 agreement 868/915/950-MHz frequency ranges spread spectrum mapping mode figure；

Fig. 4 is IEEE 802.15.4 protocol physical layers frame assumption diagrams；

Fig. 5 applies to typical case's tradition non-coherent demodulation method structure chart of 868/915/950-MHz frequency ranges；

Fig. 6 applies to the traditional demodulation method and reduced form demodulation method Performance comparision figure of 868/915/950-MHz frequency ranges；

Fig. 7 is that two kinds of demodulation methods that the present invention is provided and existing two kinds of typical cases demodulation deemed-to-satisfy4 can compare figure；

Fig. 8 is frequency deviation side-play amount f₀Probability distribution graph.

Specific embodiment

Embodiments of the present invention are illustrated below according to accompanying drawing.

A kind of high reliability BPSK demodulation methods for IEEE802.15.4, the data frame of transmitting terminal physical layer is by expanding Channel is transferred to receiving terminal after frequency and BPSK modulation, and data frame includes the lead code of 32 bits and Physical Layer Service Data Unit PSDU；The complex base band sampled signal that receiving terminal is received is expressed asWherein s (k) is to treat The transmission data of detection, s (k) ∈ {+1, -1 }, ω₀=2 π f₀,f₀Frequency shift (FS) and phase offset are respectively with θ and entirely counting According to keeping constant, T in frame_cRepresent spreading code chip cycle, η₀K () is complex base band additive white Gaussian noise；Specific demodulation step For：

Step one, lead code frequency deviation sight of the corresponding channel received data extraction comprising frequency offset information using 32 bits Measured value Y：

Wherein, J represents the total number of bits amount of lead code, J=32, and N represents spreading length, N=15,1≤m≤J-1,

0≤n≤N-1, p [n+Nm] represent the channel reception value of corresponding n-th chip of m-th bit of lead code, ()^*Table Show and take conjugate operation, η₁Represent all of noise item；

Step 2, bit-level difference processing is carried out to the corresponding complex base band sampled signals of PSDU, obtain judgement observation A [m]：

Wherein, r [n+Nm] represents the channel reception value of corresponding n-th chip of m-th bit of PSDU, η₂[m] represents all Noise item, E [m] represent send m-th bit data；

Step 3, using in step one frequency deviation observation Y extract frequency offset information, the A [m] in step 2 is compensated After carry out detection judgement：

Wherein,M-th bit data that detection judgement is obtained is represented, q () is quantization function, and q (Y) is embodied as：

Wherein, | | modulo operation is represented,Represent A [m] frequency offset N ω₀T_cEstimate；

The PSDU data that step 4, detection will be received after terminating send MAC layer to carries out CRC check；

In the step 3,Computing formula be：

Used as the further form for simplifying, the quantization function described in step 3 isWherein：

Theoretical foundation of the invention is described below.

First, to tan at x=0^-1X carries out Taylor series expansion and can obtain：

When | x | is smaller, there is approximation relation tan^-1x≈x。

As frequency offset N ω₀T_cOn the premise of smaller and signal to noise ratio is larger,Value it is smaller.Can direct tan^{- 1}X ≈ x carry out being obtained after simplifying treatment to formula (4):

But as frequency offset N ω₀T_cIn the case that larger or signal to noise ratio snr is not high,Value it is larger.tan^-1x≈x Approximate calculation to formula (4) will bring larger error, cause the extreme loss of final performance.I.e. formula (16) is only applicable to frequency Side-play amount is smaller and two larger constraintss of signal to noise ratio under detection process.

To obtain the low complex degree frequency offset estimation methods under a kind of unconfined condition, we using arcsin function and The equivalence relation of arctan function can be obtained：

If in first and fourth quadrant of complex plane coordinate system, i.e., Y fallsThen have：

If in second and third quadrant of complex plane coordinate system, i.e., Y fallsThen have：

According to formula (17) and formula (18), formula (4) can equivalence be changed into：

Secondly, to sin at x=0^-1X carries out Taylor series expansion and can obtain：

Then when | x | is smaller, there is approximation relation sin^-1x≈x。

Due toIt is permanent under the conditions of any frequency offset and signal to noise ratio to set up, therefore can be directly sharp Use sin^-1X ≈ x are in formula (19)Item carries out simplifying available:

3rd, the calculating process of the fraction of formula (20), it is contemplated that further simplify to it.When signal to noise ratio is larger, we There is following approximation relation:

4th, with formula (21) it is only approximate to the molecule Im (Y) of formula (20) fraction after can obtain：

As shown in fig. 7, the demodulation method that the present invention is provided compares with traditional canonical reduction form demodulation method, performance is more excellent More, reliability is higher.The carrier frequency used in emulation is 924MHz, frequency shift (FS) f₀Angular distribution as shown in Figure 8 is obeyed, Phase offset θ (0,2 π] in obey and be uniformly distributed, the data length of PSDU is 20 bytes (160 bit), each signal to noise ratio Under at least gather 3000 frame mistakes.It can be seen that, two kinds of demodulation method performances that the present invention is provided be in two kinds of traditional demodulation methods it Between.Simplify demodulation method with tradition to compare, when Packet Error Ratio is 1 × 10-3, the gain not less than 1.3dB can be obtained.Therefore this hair The two schemes of bright offer have reached more preferable balance matching between implementation complexity and performance.

Further the theoretical foundation of reduced form is：With formula (21) to the molecule Re (Y) and denominator Im in formula (20) fraction (Y) while can be obtained after approximate：

Calculating process can further be simplified, the consumption of energy is reduced.

Claims (2)

1. a kind of high reliability BPSK demodulation methods for IEEE802.15.4, the data frame of transmitting terminal physical layer is by spread spectrum Receiving terminal is transferred to channel after BPSK modulation, data frame includes the lead code of 32 bits and Physical Layer Service Data list First PSDU；The complex base band sampled signal that receiving terminal is received is expressed asWherein s (k) is to be checked The transmission data of survey, s (k) ∈ {+1, -1 }, ω₀=2 π f₀,f₀Frequency shift (FS) and phase offset are respectively with θ and in whole data Constant, T is kept in frame_cRepresent spreading code chip cycle, η₀K () is complex base band additive white Gaussian noise；Specific demodulation step For：Step one, lead code frequency deviation sight of the corresponding channel received data extraction comprising frequency offset information using 32 bits Measured value Y：

Wherein, J represents the total number of bits amount of lead code, and J=32, N represents spreading length, N=15,1≤m≤J-1,0≤n≤N- 1, p [n+Nm] represents the channel reception value of corresponding n-th chip of m-th bit of lead code, and () * is represented and taken conjugation fortune Calculate, η₁Represent all of noise item；

Step 2, bit-level difference processing is carried out to the corresponding complex base band sampled signals of PSDU, obtain judgement observation A [m]：

Wherein, r [n+Nm] represents the channel reception value of corresponding n-th chip of m-th bit of PSDU, η₂[m] represents all of Noise item, E [m] represents m-th bit data for sending；

Step 3, using in step one frequency deviation observation Y extract frequency offset information, the A [m] in step 2 is compensated After carry out detection judgement：

Wherein,M-th bit data that detection judgement is obtained is represented, q () is quantization function, and q (Y) is embodied as：

Wherein, | | modulo operation is represented,Represent A [m] frequency offset N ω₀T_cEstimate；

The PSDU data that step 4, detection will be received after terminating send MAC layer to carries out CRC check；

It is characterized in that：

In the step 3,Computing formula be：

Wherein, Re () is represented and is taken real part computing, and Im () is represented and taken imaginary-part operation.

2. a kind of high reliability BPSK demodulation methods for IEEE802.15.4 as claimed in claim 1, it is characterised in that： To in the step 3Computational methods simplified, obtain：

。