CN1115006C - Compensation of Doppler shift in mobile communication system - Google Patents

Abstract

In a mobile communication system, signals transmitted from a mobile station moving relatively to a base station are influenced by the Doppler effect. The present invention describes a technique for compensating the Doppler effect, which uses estimating bit judgment for receiving bits which are very close to a zero phase and are not damaged by the Doppler effect. Afterwards, Doppler frequency shifts measured by using the estimating bits are supplied to signal bits far away from a zero phase difference point.

Description

Compensation of Doppler Frequency Shift in Mobile Communication System

Technical field

The invention relates to compensation of Doppler shift in mobile communication stations.

Background technique

In mobile communication systems, signals transmitted from a mobile station moving relative to a base station are subject to what is known as Doppler shift, causing the frequency received by the base station to be shifted relative to the frequency transmitted by the mobile station. This frequency shift is referred to herein as a Doppler shift. Doppler shift is related to the speed and direction of the mobile station moving relative to the base station. Thus, Doppler shift provides an increase or decrease in frequency, depending on the direction in which the mobile station is moving relative to the base station. The magnitude of the Doppler shift depends on the speed at which the mobile station is moving relative to the base station.

Existing mobile communication devices provide a form of Doppler compensation whereby frequency detection circuitry within a base station that selects a particular signal on a particular channel takes into account a certain amount of Doppler shift in the signal.

AU 66426 relates to a method and circuit arrangement for compensating for Doppler shift in radio signal propagation between a base station and a mobile station when the mobile station approaches and moves past the base station. As the mobile station approaches the base station, the propagation delay decreases, and the integral of the propagation delay over the time interval determines how the propagation delay varies with time. This information is used to effect a change in radio signal frequency at a specific time in order to compensate for sudden Doppler shifts in the mobile station as it passes the base station. In TDMA systems, interrogations occur at intervals equal to an integer number of time frames. In the GSM standard, a time frame consists of eight consecutive time slots, in any one time frame a single transmission burst will pass between a particular mobile station and a base station.

Thus, in this method of Doppler compensation, decisions are made by responding to past and incoming signals to improve reception at future times, ie, it is a reactive system. It is desirable to implement a system that can actively compensate for the Doppler shift of the incoming signal in real time.

Contents of the invention

According to one aspect of the invention, there is provided a method of compensating for Doppler shift in a signal transmitted between a mobile station and a base station of a mobile communication system, said signal comprising a sequence of transmission bursts, the method comprising:

For each transmission burst, determine the channel impulse response of the channel receiving the signal;

estimating selected data bits of a transmission burst using a channel impulse response, the selected portion being located near a point of zero phase difference in the transmission burst;

generating a reference vector using the channel impulse response and the estimated data bits;

determining the Doppler characteristic in phase shift per bit period using the selected portion of the transmission burst and the reference vector; and

Doppler shift compensation of the transmission burst is provided using the determined Doppler characteristic for the selected portion.

According to another aspect of the invention, there is provided a system for compensating for Doppler shift of a signal transmitted between a mobile station and a base station in a mobile communication system, said signal comprising a sequence of transmission bursts, the system comprising:

a channel impulse response determining circuit for determining a channel impulse response of a channel receiving a signal for each transmission burst;

an estimation circuit connected to receive a received signal on a channel impulse response and to use the channel impulse response to estimate data bits for a selected portion of a transmission burst, the selected portion being at a point in the transmission burst near zero phase difference s position;

a reference generator for generating a reference vector using the channel impulse response and the estimated data bits;

circuitry for determining the Doppler characteristic in terms of a phase shift per bit period using the selected portion of the transmission burst and the reference vector; and

A Doppler shift compensation circuit for providing Doppler shift compensation for burst transmissions using the Doppler characteristic.

The invention is particularly applicable to TDMA mobile communication systems in which the signals comprise transmission bursts. The selected portion is in the transmission burst close to the point of zero phase difference, so that the effect of the Doppler characteristic is very small and does not corrupt the transmitted bits.

According to the GSM standard, a TDMA transmission burst includes a training sequence, which is usually used to determine the channel impulse response of the received signal channel. This is done by convolving the received sequence with a stored version of the training sequence.

In existing systems, the channel impulse response is used to remove the influence of the transmission channel on the signal from the received signal, especially the influence of multipath and fading.

According to the GSM standard, the "cleaned" and filtered signal is demodulated to remove the IQ modulation by which the data in the signal was transmitted. The demodulated signal can then be decoded to produce hard bits. This can be done with Viterbi technology.

The Doppler shift is estimated from received signal samples. Therefore, Doppler shift estimation is dependent on channel quality, and generally gets worse when channel quality is poor, and gets better when channel quality improves. Application of Doppler compensation algorithms degrades receiver performance when there is no Doppler shift and channel conditions are close to receiver sensitivity.

On the other hand, if there is a Doppler shift in poor channel conditions close to the receiver sensitivity, the receiver cannot meet the reference sensitivity limit if the Doppler compensation algorithm is implemented.

According to one embodiment of the present invention, a method for compensating for Doppler shift in signals transmitted between a mobile station and a base station in a mobile communication system, comprising:

detecting received signal quality; and

Doppler frequency shift compensation is realized according to the detected signal quality.

Therefore in this embodiment, Doppler shift compensation is only or mainly used for sufficiently good channel conditions. This provides an improvement in receiver performance. This technique is provided to ensure that Doppler compensation does not desensitize the receiver.

The step of detecting signal quality may include estimating a noise energy content of the signal. This itself can be used to generate Doppler correction modifiers for controlling Doppler shift compensation based on detected signal quality, or to generate a signal-to-noise ratio of the received signal, which is then used to generate Doppler corrections modification factor.

Doppler compensation can be performed according to any suitable user-defined signal quality function. For example, it can be a linear function or a step function.

In the described embodiment, first the adaptive Doppler correction method for the GSM system can estimate the channel quality and use the resulting modification factor to scale the calculated phase difference between the reference signal and the actual received signal.

Description of drawings

For a better understanding of the invention and to show how it may be practiced, reference is now made to the accompanying drawings, by way of example, in which:

Fig. 1 is a signal burst diagram in a mobile communication system;

Figure 2 is a block diagram of a circuit implementing modified Doppler shift compensation; and

Figure 2a is a block diagram of a modified Doppler correction modifier generator circuit.

Detailed ways

Figure 1 illustrates a common burst in a mobile communication system according to the GSM standard. This graph represents the bursts received by the base station. For TDMA systems according to the GSM standard, mobile stations transmit bursts as modulated signals on frequency channels assigned by the base station controller. A channel can support up to eight bursts, each burst is associated with an individual call, and each call is assigned a time slot in which to send the burst. The details of the TDMA system according to the GSM standard are not described here since they are known to those skilled in the art.

A common burst includes two 58-bit (DATA, data) packets surrounding a 26-bit training sequence (TRS). Three tail bits (TS) are added to each end of the normal burst. A training sequence (TRS) is a predetermined sequence of bits transmitted by a mobile station (MS) and known to a base station controller (BSC). The base station controller uses it to estimate the impulse response of the channel over which the burst is transmitted. The actual information transmitted is in the data bits (DATA) in the burst.

As explained earlier, the environment through which a signal passes from a mobile station to a base station can vary widely depending on, among other things, the distance between the mobile station and the base station and the buildings and other structures in the area the interference caused. Therefore, the signal strength and signal quality of the signal received by the base station varies greatly. Furthermore, for a mobile station that is moving, the signal received by the base station is affected by the Doppler shift that should be corrected.

The circuit described here provides Doppler shift correction. The particular circuitry described in connection with Figure 2 provides correction only if the channel conditions are good enough to give sufficient received signal quality at the base station. Therefore, the modification factor Sc generated according to the signal quality can be used to control the Doppler shift correction so that the Doppler shift correction is only provided in good enough channel conditions. It should be understood that although the use of the modification factor Sc is beneficial, it is not in itself essential to achieve Doppler correction.

Figure 2 illustrates a circuit 1 suitable for implementing Doppler compensation in a GSM system. It should be understood that although the various modules in FIG. 2 are illustrated as separate interconnected entities, they do not represent separate physical entities, but are intended to schematically represent implemented steps. A module can be implemented as a circuit, or a suitable programmable microprocessor can carry out each function individually assigned to the module.

An antenna 20 receives a signal 11 from a mobile station. Antenna 20 is connected to RF circuit 22 through interconnect means 21 . This circuit 22 operates on received bursts, downshifts the frequency to baseband frequency and samples the bursts to provide digital samples from the analog signal. The output of the RF circuit 22 is a burst r of samples (in digital form), sampled at the expected bit rate of the transmitted signal. Figure 1 illustrates this burst. The output of circuit 22 is provided along line 24 to channel impulse response (C.I.R.) module 10, to variance calculator 16 for estimating the quality of the communication channel (as described later), to filtering and equalization circuit 12, to phase difference calculator 36 And to the transformation circuit 40 for estimation and application of Doppler shift correction for burst r.

The upper part of Figure 2 illustrates the circuitry required to implement the adaptive part of the system in order to generate the Doppler correction modifier Sc. The memory 32 stores a training sequence TRSref, which is a predetermined bit sequence sent by the mobile station MS as a training sequence and at the base station as TRS_received. The reference training sequence TRSref is provided to the reference generator 14 and the channel impulse response (C.I.R) module 10 . The reference generator 14 also receives the estimated channel impulse response h from the channel impulse response module 10 .

The C.I.R module 10 receives the burst r, including the received training sequence TRS_received, and calculates the estimated channel impulse response h by computing the cross-correlation between the received training sequence TRS_received and the known training sequence TRSref. therefore,

h＝xcorr(TRS_received，TRSref) (Formula 1)

It should be understood that the known training sequence TRSref stored in digital form is i,q modulated before cross-correlation is achieved, similar to the way training sequences are modulated for transmission in MSs according to the GSM standard. Cross-correlation is done in a known manner, yielding a channel impulse response in the form of five tap values (h(i) _{i=0 to 4} ).

As is known, the estimated impulse response h is used to compute an estimate of the expected data in the received burst as if the data were subject to the same average noise.

The C.I.R module also generates timing advance information τ, which is used to determine where in the allocated time slot the received burst r is.

For each burst, the channel impulse response h estimated for that burst is calculated by the CIR module 10 and provided to the filtering/equalization circuit in order to recover the data in that burst, DATA(r). As is known, filter/equalization circuit 12 receives the channel impulse response h of the received burst and timing information τ to demodulate, filter and decode the signal to recover the data in a known manner.

A reference generator 14 produces a reference vector, reffi, which is computed using the convolution of the impulse response and the known training sequence. Therefore, the reference generator 14 performs the following calculations:

reffi=h*TRSref (Formula 2)

More specifically, (here reffi _k represents the kth sample of the signal reffi) ${\mathrm{reffi}}_k=\underset{i=0}{\overset{N-1}{\mathrm{\Σ}}}{h}_i\&Center\; Dot;(1-2{\&Center\; Dot;\mathrm{TRS}}_{k-i})$ (Formula 3)

where N represents the number of tap values for the estimated impulse response (N=5 in the described embodiment), and k varies from N-1 to 25.

The vector reffi is supplied to the variance calculator 16 from the reference generator. As mentioned above, the variance calculator also receives the burst r, including the received training sequence. The variance calculator calculates the variance (σ ² ) according to the following formula: $\mathrm{var}=\frac{\left(\underset{k=4}{\overset{25}{\mathrm{\Σ}}}\right|r_k-{\mathrm{reffi}}_k|^2)}{\mathrm{reffi}\_\mathrm{length}}$ (Formula 4)

The item reffi_length is a constant representing the length of the reference signal reffi. Calculated by multiplying the number of samples (22) by the bit separation.

In Equation 4, the r _k value is the received training sequence sample value of burst r.

It should be understood that each actual received sample r _k will have a different noise level than the average estimated noise level obtained from the channel impulse response and reflected in the reference sample reffi _k . Thus, the variance gives an indication of the actual received noise energy level, as well as an indication of the signal quality.

The output ^σ2 of variance calculator 16 is supplied to Doppler correction modifier circuit 18. The calculated variance σ ² is used by the Doppler correction modifier circuit 18 so that the modifier Sc can be generated by a user-determined function. In the embodiment shown in Figure 2, the Doppler correction modifier circuit produces the modification Sc as a function of variance ^σ2 , eg a linear function or a non-linear function such as a step function.

In another embodiment of the invention, the Sc value is calculated from the channel signal-to-noise ratio (SNR). Figure 2a illustrates the circuitry implementing the Doppler correction modification circuit 18' of such an embodiment. SNR calculator 42 receives the calculated variance σ ² and energy value E from variance calculator 16 . The signal energy E can be determined from the calculated signal reffi according to: $\mathrm{E.}=\underset{k=4}{\overset{25}{\mathrm{\Σ}}}\frac{{\left|{\mathrm{reffi}}_k\right|}^2}{\mathrm{reffi}\_\mathrm{length}}$ (Equation 4a) or tap values can be used to obtain the signal energy E of the channel according to: $\mathrm{E.}=\underset{i=0}{\overset{4}{\mathrm{\Σ}}}{\left(h\left(i\right)\right)}^2$ The SNR value is calculated by the SNR calculator 42 according to the following formula: $\mathrm{SNR}=\frac{\mathrm{E.}}{{\mathrm{\σ}}^2}$ (Formula 4b)

The technology referenced above is described in our earlier Patent No. PCT/FI96/00461, the contents of which are incorporated herein by reference.

The SNR value is then provided to an improved modification factor generation circuit 44 . This circuit 44 calculates the value of the modification factor Sc as a function of the SNR value, for example a linear function or a non-linear function such as a step function.

The lower portion of the circuit in Figure 2 illustrates a system that implements Doppler shift correction in block diagram form. However, it is clear that the above-described Doppler correction adaptation portion can be used with other implementations of Doppler correction.

The main function of the adaptive part of the above circuit is to provide the Doppler correction modification factor Sc based on the channel condition, so that the Doppler correction is only or mainly used in a sufficiently good channel condition. This is useful in any type of Doppler correction. A specific Doppler correction circuit is described below.

An equalization circuit 30 (eg, a Viterbi equalizer) receives the filtered, demodulated and equalized signal DATA(r) from the filtering and equalization circuit 12 . The equalization circuit 30 operates on the data sequence DATA part of the burst (which part has been obtained from the estimation block in Fig. 1), estimating and outputting the bits sent from the mobile station MS. This output is called estim_bits here, and they vary from k=j to k=j+n. The equalization circuit 30 is used for bit decision, which is the same as in the known mobile communication system, so it will not be described here.

The estimated bit decision estim_bits is provided to the reference circuit 34 . The reference circuit 34 generates the reference vector ref by convolution using the estimated bit decision and the estimated impulse response h according to the following equation:

ref＝estim_bits*h (Formula 5)

Thus, the reference vector ref comprises a set of samples ref _k (k=j→j+n), each having a real and an imaginary part value. The reference vector ref is provided to the phase difference calculator 36 . As previously mentioned, the phase difference calculator 36 also receives the received burst r. As is known in the art, a received burst includes samples _rk each having a real and imaginary value.

The phase difference calculator uses the t_diff value, which represents the time between the zero phase difference point h_time and the middle of the estimated block, as shown in Figure 1. The zero phase difference point h_time is the zero phase difference point within the training sequence where the calculated impulse response is true. In fact, it is generally in the middle of the training sequence. The t_diff and h_time values can be determined during the system design phase and are constant thereafter. Of course, they can also be reprogrammed if necessary during system use.

Furthermore, the starting position (j) of the estimation block estim_block and its length (n) are determined and programmed into the equalization circuit 30 . The estimation block is chosen such that the Doppler bias has not corrupted the received bits.

The phase change (ph_diff) of each bit period produced by the Doppler effect (Doppler characteristic) is calculated from the reference signal ref and the actual received signal r by the phase difference calculator 36 according to one of the following formulas: $\mathrm{pH}\_\mathrm{diff}=\frac{1}{t\_\mathrm{diff}}{\mathrm{the\; tan}}^{-1}\left\{\frac{{\mathrm{\Σ}}_k\mathrm{imag}(r_k\&Center\; Dot;\mathrm{ref}{*}_k)}{{\mathrm{\Σ}}_k\mathrm{real}(r_k\&Center\; Dot;\mathrm{ref}{*}_k)}\right\}$ (Formula 6) $\mathrm{pH}\_\mathrm{diff}=\frac{1}{t\_\mathrm{diff}}\underset{k}{\mathrm{\Σ}}\·{\mathrm{the\; tan}}^{-1}\left\{\frac{\mathrm{imag}(r_k\&Center\; Dot;\mathrm{ref}{*}_k)}{\mathrm{real}(r_k\·\mathrm{ref}{*}_k)}\right\}/\mathrm{length}\left(k\right)$ (Formula 7)

Here k varies from j to j+n, and length(k) represents the number of different values of k in the summation, ie n.

The Doppler shift estimated from the received samples is then corrected using Doppler correction circuitry 38 . The Doppler correction circuit 38 receives the zero phase difference point h_time and the Doppler correction modification factor Sc. Furthermore, it receives the phase difference calculated by the phase difference calculator 36 . Knowing that the h_time point has zero phase offset, the actual Doppler phase offset φ can be calculated for each bit as follows:

φ _k =S _c ·ph_diff·(k-h_time) (Formula 8)

Here k is the bit index of the received sample r. When the index k<h_time, the phase offset has the opposite sign than when k>h_time. Obviously, Sc will default to 1 in Equation 8 if the Doppler correction circuit operates independently of the adaptive circuit generating Sc.

Doppler shift correction on the received burst r is then effected using the transformation circuit 40 to produce a corrected signal. The conversion circuit receives the estimated Doppler shift vector φ (including the value of φ _k ) and the sampled values of the received burst r. It performs a CORDIC operation to correct for Doppler shift per sample according to: $\left[\begin{array}{c}N\_\mathrm{real}\_\mathrm{sample}\left(k\right)\\ N\_\mathrm{imag}\_\mathrm{sample}\left(k\right)\end{array}\right]=\left[\begin{array}{cc}\cos {\mathrm{\&phi;}}_k& \sin {\mathrm{\&phi;}}_k\\ \sin {\mathrm{\&phi;}}_k& \cos {\mathrm{\&phi;}}_k\end{array}\right]\left[\begin{array}{c}\mathrm{real}\left(r_k\right)\\ \mathrm{imag}\left(r_k\right)\end{array}\right]$ (Formula 9)

The Doppler shift correction factor DCV output from the conversion circuit 40 is supplied to the filter/equalization circuit 12 for recovering data from the signal using the Doppler corrected signal.

As part of the Doppler correction technique described above, estimation block bit decisions have been made. These make up the data section. Therefore, it is not necessary to evaluate the same bits again, although it can be done. Instead, the equalization circuit 30 can stop at the end of the estimation block and retain the existing state. Then, Doppler correction is performed on the remaining bits, and then Viterbi equalization is performed on the Doppler corrected bits until the end of the slot. In the second part, Doppler correction can be performed on the first part of the time slot, and then a Viterbi estimation is performed on the data 58 of the first part of the time slot. This approach reduces the computations required in the receiver.

A limit can be imposed on the difference value ph_diff such that no correction is performed if the difference is below a certain threshold.

Typical environments where Doppler correction can be used are fast trains or motorways. In this case, there may be a line-of-sight path from the base station to the mobile station such that the Doppler correction is constant across time slots if the speed of the mobile station is the same. In this case, an average phase difference value can be calculated from several different time slots and this average value used as the correction value.

In the case of a diversity receiver with several different branches, the phase difference can be calculated for all branches using the same estimate as received samples, each branch having its own impulse response.

Claims (11)

1. A method for compensating for a Doppler shift in a signal transmitted between a mobile station and a base station of a mobile communication system, said signal comprising a sequence of transmission bursts, the method comprising: For each transmission burst, determine the channel impulse response of the channel receiving the signal; estimating data bits for a selected portion of a transmission burst using a channel impulse response, the selected portion being located near a point of zero phase difference in the transmission burst; generating a reference vector using the channel impulse response and the estimated data bits; determining the Doppler characteristic in phase shift per bit period using the selected portion of the transmission burst and the reference vector; and Doppler shift compensation of the transmission burst is provided using the determined Doppler characteristic for the selected portion. 2. The method according to claim 1, characterized in that it is used in a TDMA mobile communication system. 3. The method according to claim 1, characterized in that the Doppler shift compensation φ _k of each bit is calculated as follows, wherein k is a bit index, h_time is a point of zero phase difference, and ph_diff is a phase difference per bit period, ph_diff (k-h_time) 4. A method according to claim 2 or 3, characterized in that the channel impulse response is determined from the training sequence of said transmission burst. 5. The method according to claim 1, characterized in that the data bits are estimated by the following steps: Use the channel impulse response to remove the influence of the transmission channel on the signal from the received signal; demodulating the resulting signal; and The demodulated signal corresponding to the selected portion of the transmission burst is decoded to estimate data bits. 6. A method according to claim 5, characterized in that the decoding step is performed by the Viterbi method. 7. The method of claim 1, wherein the step of using the Doppler characteristic comprises: for each of the plurality of indexed received signal samples based on the Doppler characteristic and the position of the introduced sample in the received signal One to determine the Doppler phase shift for that sample. 8. A method according to claim 7, wherein the samples are indexed according to the bit rate of the received signal. 9. A system for compensating for Doppler shift of a signal transmitted between a mobile station and a base station in a mobile communication system, said signal comprising a sequence of transmission bursts, the system comprising: a channel impulse response determining circuit for determining a channel impulse response of a channel receiving a signal for each transmission burst; an estimation circuit connected to receive a received signal on a channel impulse response and to use the channel impulse response to estimate data bits for a selected portion of a transmission burst, the selected portion being at a point in the transmission burst near zero phase difference s position; a reference generator for generating a reference vector using the channel impulse response and the estimated data bits; Circuitry for determining the Doppler characteristic in terms of phase shift per bit period using selected portions of the transmission burst and the reference vector; and A Doppler shift compensation circuit for providing Doppler shift compensation for transmission bursts using the Doppler characteristic. 10. The system of claim 9, wherein the estimation circuit comprises: filtering and equalization circuitry for removing the influence of the transmission channel on the signal from the received signal by using the channel impulse response; a demodulation circuit for demodulating the resulting signal; and Decoding circuitry for decoding the demodulated signal corresponding to the selected portion of the transmission burst to estimate data bits. 11. A system according to claim 10, characterized in that the decoding circuit is a Viterbi decoder.

Images