FI105505B - Method and arrangement for expression of modulation - Google Patents

Description

, 105505

Method and apparatus for detecting modulation

Background of the Invention

The invention relates to a method for detecting modulation, wherein the method comprises taking samples of a received signal in symbol periods over 5 time slots and applying the samples to two or more parallel modulation detectors for detecting bits.

When transmitting information over a radio channel, the signal to be transmitted must be modulated. The purpose of modulation is to obtain the signal in such a form that it can be transmitted at radio frequency. For example, a good modulation-10 method may require that as much information as possible be transmitted in the narrowest possible frequency band. Depending on the intended use, other properties may be emphasized. In addition, the modulation must be such that it causes minimal interference to the neighboring channel.

One modulation method is π / 4-DQPSK (π / 4-shifted, Differential

Quaternary Phase Shift Keying) modulation. There are eight phase states in this modulation method, but only four phase transitions. The allowed phase shifts (symbols) are + π / 4 and + 3π / 4. Figure 3A shows a phase-transition pattern (constellation) of modulation. Each phase offset corresponds to two outgoing 20 bits. In other words, the digital signal modulates the carrier in two-bit sequences such that each combination of two bits is matched by a specific change in step # t * i * during each symbol period. A symbol sequence refers to a signal; · a sequence used to transmit two bits. Bit combinations 00, 01,10 and; ***; 11 corresponding phase changes are π / 4, 3π / 4, -π / 4 and -3π / 4. For example, •% m 25 The symbol frequency used in TETRA (Terrestrial Trunked Radio) is 18 kHz with a bit rate of 36 kHz.

• · · -. ^

When a signal is received, it must be demodulated, i.e., detected by the detector to detect the modulated bits of the signal in order to find out the information contained therein. However, the signal transmitted over the radio may be distorted in various ways, making it difficult to detect modulation. Some of these phenomena that negate the signal are noise and inter-symbol crosstalk (ISI). A signal distortion phenomenon also occurs when a radio signal is reflected from various obstacles such as buildings and In this case, the receiver detects a signal that is the sum of a plurality of 35 propagation paths, each propagation path is of different length and the signals arrive at the receiver at different times, i.e., the delay varies. called Doppler frequencies.

The problem with the above arrangement is that the performance of the different types of detectors is different depending on the type of interference to the signal. For example, a coherent detector works well on a channel that has noise but no crosstalk between symbols. Similarly, a channel having a crosstalk between symbols typically achieves better performance with a channel equalizer type detector. Since it is often difficult to know in advance which types of interference on the channel are mainly occurring, it is preferable to use at least two different types of detectors in the receiver. Indeed, patent application EP 0 605 102 A1 [1] discloses a method for the simultaneous use of two different types of detector arms at reception. The detector branches comprise a multipath compensated branch and a nonpathway compensated branch. According to the method, the detector branch is changed between the received time intervals if the alternative detector branch would achieve better performance. The detector leg may be changed according to the method described only once per time interval. The method described in the application [1] thus does not solve the problem that occurs when the channel interference conditions change several times during the received time slot.

Brief Description of the Invention

It is therefore an object of the invention to provide a method by which the above problems can be solved. The objects of the invention are achieved by a method characterized by dividing a time interval into several parts, by determining, by section, an estimate of the error produced by each detector of the signal produced by each detector, ·: · *: bits corresponding to that part of the time slot produced by an estimator having an estimate for further processing.

The invention is based on dividing the received signal by a, ···. If several parts are selected and the most cost-effective detector is selected, the detection of the received signal can be optimized according to the channel interference conditions.

An advantage of the method according to the invention is that the advantages of two or more detectors, <', in the receiver can be combined efficiently and simply.

The invention also relates to an apparatus for detecting modulation, which comprises means for sampling samples of a received signal.

= I

3 105505 a list of symbol periods over a time slot and two or more parallel modulation detectors indicating bits of a received signal, wherein the apparatus is characterized in that the apparatus is adapted to divide the time interval into a plurality of parts to determine the error estimate of sig-5 signal produced by each detector illustrates the error of the signal produced, to select the detector having the most advantageous error estimate per part and to output to its output the bits produced by the selected detector as the detection plug of the corresponding time slot. With such an apparatus, the advantages of the method of the invention can be achieved in a simple manner.

10 Brief Description of the Figures

The invention will now be further described in connection with preferred embodiments, with reference to the accompanying drawings, in which:

Figure 1 is a block diagram of a reception structure according to a TETRA system; Figure 2 is a simplified schematic diagram of the frame structure of a TETRA system;

Figure 3A shows a phase transition diagram of π / 4-DQPSK modulation; Figure 3B shows constellation points for π / 4-DQPSK modulation;

FIG. 4 is a block diagram of an adaptive MLSE detector and 20 associated channel estimates in accordance with one embodiment.

DETAILED DESCRIPTION OF THE INVENTION

The invention is explained below in connection with the TETRA system, but the invention is not intended to be limited to any particular system or ·: ··· modulation method ._____________ 25 In a Tetra system, the information bits obtained from the MAC (Medium Access Layer) layer are encoded and convoluted · '- "··' · ----- - * · thio coding to detect and possibly correct errors in the signal on the radio path ... at reception. The encoded bits are interleaved so that the • • __ ~ __ - * ;; · / incremental bits are far apart. This facilitates error correction if the signal transmitted by '··· * 30 is temporarily interfered with on the radio path. The interleaved bits' Λ are shuffled using a specific color code to transmit to different base stations. ""; The multiplexing combines bits of different logical channels. The multiplexed bits are then formed into a burst. The burst '··] is a structure which is transmitted together in a TDMA (Time Division Multiple Access) * 35 time slot or sub slot, The burst consists of data bit fields 20 and 22 and 105505 4 between them in the middle of the burst training sequence 21 as illustrated in FIG. Differential coding generates modulating symbols from bit pairs in a burst. The symbol modulated carrier is amplified in the transmitter and transmitted to the radio path.

5 Modulation is described above with π / 4-DQPSK (π / 4-shifted, Differential

Quaternary Phase Shift Keying) modulation. There are eight phase states in this modulation method, but only four phase transitions. The allowed phase shifts (symbols) are + π / 4 and + 3π / 4. In practice, therefore, the π / 4-DQPSK constellation varies at symbol intervals between the two 4-point constellations illustrated in Fig. 3B by four black dots (Constellation 1) and four white dots (Constellation 2). As the symbol sequence changes, it is only possible to move from white to black and from black to white. Each of these eight constellation points can be represented by numbers 0-3, as in Figure 3B. Due to the radio channel's non-idealities, the constellation points may shift.

Figure 1 is a block diagram of a receiver structure according to the invention, for example for a TETRA system. Only the parts essential to explaining the invention are described in the receiver. At reception, a signal is received from the antenna (not shown) and the signal is first processed by the RF components. Subsequently, samples of the intermediate frequency signal are taken with A / D converters (not shown). The samples are supplied to the synchronization block 11 as illustrated in FIG. 1 by the RF signal. The synchronization block 11 searches for a training sequence 21 in the frame structure from the samples received. Sen · ;;; allows the synchronization block to accurately determine the sampling time, i.e. I «25, of all symbol positions in the sample stream. The synchronization block also controls the radio frequency portions of the receiver so that the signal to the A / D converter stops at an optimum level. The synchronization block provides a frame for a channel equalizer: T: and a detector block consisting of a control unit 12 and detectors 14 and 15.

The channel equalizer corrects the radio channel channel ideals. · *; 30 and the associated indicator indicate information bits. Finally, a · ···, ··· is formed. in frame 13, a logical channel which is forwarded for further / · * processing.

To illustrate the invention, an example of the general structure of the receiver is described above. The structure of the receiver can. However, this may vary without departing from the present invention, which relates to the channel equalizer / detector 12,14 and 15 of the receiver.

<· 5 105505

In a preferred embodiment of the invention, the detectors 14 and 15 are MLSE detectors equipped with a Viterbi algorithm. Thus, the adaptive MLSE detector comprises a Viterbi detector 41 and at least one adaptive channel estimator 42a (? = 1 ... Mv) as shown in FIG. The Viterbi detector 41 5 estimates the transmitted sequence rn with the aid of a description of the radio channel impulse response generated by the channel estimator 42a. The channel estimator 42a adaptively estimates the impulse response of the radio channel using the decisions Jn or preliminary decisions provided by the Viterbi detector 41. Each Viterbi detector sequence is represented by one channel estimate. It is possible to implement these estimates with a single common channel estimator, but this results in a decrease in the channel estimator's tracking capability. The embodiment shown in Fig. 4 has a plurality of parallel channel estimators 42a, preferably as many as sequences.

It is typical for a radio path that the transmitted signal arrives at the receiver 15 along a plurality of propagation paths, each with its own characteristic time delay, in addition to which the channel characteristics change with time. For example, reflected and delayed beams on the radio path cause inter-symbol crosstalk (ISNIntersymbol interference). The channel frequency response or the impulse response can be estimated as a discrete time filter, a channel pre-20 estimator whose pin coefficients model the radio channel. Channel Estimation is used to describe the status of a radio channel.

. In this specification, channel estimator is commonly understood as mec · ·; nism that estimates and maintains a description of the complex im-

III

· ;;; the pulse. Essentially associated with this mechanism is a method for updating the channel estimate. In the TETRA system, an LMS (Least Mean Square) algorithm can be used to update channel estimates. In order to verify the convergence of the LMS • · v .: algorithm before the start of the actual information bits, the detector 14 and 15 must obtain the best possible estimate of the state of the channel. This estimate is obtained from Synchronization 11, which searches for ···. In the essay, the optimal sampling times of the essay compute a complex cross-correlation between the training signal 21 of the received signal and the stored version • · '· * of the training session. The cross correlation result gives the channel estimate an initial value which; V represents the average state of the channel during the training period. Channel correction and ·., / Symbol detection will only start when the training session is over. ··. 35 taken. This is because symbol synchronization is able to adjust the symbol timing .... as accurately as possible and to form an initial estimate of the channel.

I «----- 6 105505

Both forward and backward channel correction is performed such that after initialization of the estimates, a motion is always made by teaching the detector 14 and 15 over the training period 21 towards the end or beginning of the burst, as shown in Figure 2.

5 The Viterbi algorithm is a method of finding a trellis for a signal path corresponding to a maximum likelihood among a plurality of signal paths in which one channel estimate corresponds to each signal path. By "signal paths" is meant herein various combinations of successive modulation symbols. At each step of the trellis search, the detector proceeds with an ML sequence, each of which has its own Euclidean distance metric. Based on the information available about the current state of the channel, i.e. the channel estimate, reference constellation points are constructed. Once the reference constellation points have been calculated, the difference between the reference point and the received sample for each channel estimate can be calculated. This error can be used to update the channel estimate. The error metric for a given signal path is calculated by summing the error metrics of all points in the signal path, i.e. the squared errors.

According to the basic idea of the invention, the detector block control unit 12 transmits the detectable signal to the detectors 14 and 15 which operate in parallel.

There may be more than two parallel detectors, without departing from the spirit of the invention. Detectors 14 and 15 perform bit detection independently of the received signal independently of each other. Preferably, the detectors are of different types, wherein the second detector can achieve good performance under reception conditions where the second detector is not operating optimally. For example, a coherent detector works well on a channel that has noise but not slight crosstalk. Correspondingly, a channel that has inter-symbol overlap: typically provides better performance with a channel equalizer type detector. Detectors 14 and 15 determine the addition of the detected bits · **; 30 error metric of the bit information it produces, which reflects it. ···. inaccuracy, as previously described in this specification. The error metric values produced by the different detectors are preferably co-linear, whereby the performance of two different detectors: r1 can be directly compared with each other. The received 11 «time slot 20, 21 and 22 is divided into several parts. The time slot typically comprises 220 35 bits, which are divided into, for example, ten sections (sub-slots), the section ....: length being 22 bits. The control unit 12 compares the parallel detectors 14 and «·

3 I

105505 15 error metric hops one by one (sub slot) at a time and selects one bit (sub slot) at a time by a better performance detector, i.e. a bit having a lower error metric value for that part (sub slot) to be forwarded to the frame 13.

It is obvious to a person skilled in the art that as technology advances, the basic idea of the invention can be implemented in many different ways. The invention and its embodiments are thus not limited to the examples described above, but may vary within the scope of the claims.

Hl «im

I 1 I

I - -

«III« «I« I

I I "----------- 4« «f ...........

• · ...

I · - · ------: --- f · * ....-------- · • · · ........

• · IM -------- • <· - - ------ • · * «·· * ----- - - t · - .......

• · · · · · · · · · · · · · · · · · · · · · · · · · · · · · ·

r i *. _ («11«

I

«« 11 «.....-« 4 «f ·« ·

4 I I

11 ·· «• ·

Claims (11)

A method for detecting a modulation, in which method samples of a received signal are taken symbolically over a time interval and the samples are measured in two or more parallel modulation detectors for detecting the bits, characterized in that the time interval is divided into several parts, for each part a the error estimate of the signals from each detector, which describes the error of the generated signal on the part of the time interval, and for each part, further processing bits corresponding to the relevant part of the time interval are selected and are generated by the detector with the the most likely error estimate. 2. A method according to claim 1, characterized by the error estimate of the signal generated by the detector, an error metric formed by means of the square errors calculated on the basis of test points for each individual symbol interval and corresponding to the reference constellation points. 3. A method according to claim 2, characterized by being a detector for part of a signal's time interval, the detector whose error metrics on the part in question is the least selected. 4. A method according to any one of claims 1-3, characterized in that one of the parallel modulation detectors is a coherent detector. i * t>: 25 5. A method according to any one of claims 1-4, characterized in that one of the parallel modulation detectors is a channel corrector type detector. · Apparatus for detecting a modulation, comprising apparatus for sampling the received signal symbolically interval ... 30 over a time interval and two or more parallel modulation detectors (14, 15) detecting the bits of the received signal , characterized in that the apparatus is arranged to: • "": divide the time interval (20, 21, 22) into several parts, define for each part an error estimate of the signals from each and the detector, which describes the error of the generated signal, for each part select the detector with the most probable error estimate and, as the detection result for each part of the time interval, output the bits generated by the detector selected for each part. Apparatus according to claim 6, characterized in that it is arranged, as error estimate for the signal generated by the detector (14, 15), to use an error metric formed by means of the square errors calculated on the basis of test points for each individual symbol interval. and corresponding reference constellation points. Apparatus according to claim 7, characterized in that it is arranged, as a detector for part of the time interval of the signal (20, 21, 22), to select the detector whose error metrics on the part in question are the least. Apparatus according to any of claims 6-8, characterized in that one of the parallel modulation detectors (14, 15) is a coherent detector. 10. Apparatus according to any one of claims 6-9, characterized in that one of the parallel modulation detectors (14, 15) is a channel corrector type detector. 20 <HM III a <a · · · · · · 1 · · · · ···· · · · · · · · · · · · · · · · · · · - ··· • · 9 · • · ··· (I _. Ml «· · • ·>« · fa · · «a · • · • · •« · ·