WO2003046600A2

WO2003046600A2 - Positioning system and method

Info

Publication number: WO2003046600A2
Application number: PCT/EP2002/013474
Authority: WO
Inventors: Burkhart Dietrich; Rolf Kraemer
Original assignee: Ihp Gmbh - Innovations For High Performance Microelectronics / Institut Für Innovative Mikroelektronik
Priority date: 2001-11-27
Filing date: 2002-11-26
Publication date: 2003-06-05
Also published as: DE10159226A1; WO2003046600A3; EP1451609A2

Abstract

The invention relates to a system for determining the position of a mobile terminal, which comprises a base station (1) and at least two time measuring stations (3). Said time measuring stations (3) comprise one time acquisition unit each with which the arrival time of a signal sent by the mobile terminal (5) can be detected, said time acquisition unit (3) being provided with a slide register and a correlator. The system is further provided with a positioning unit for calculating the position on the basis of the acquired arrival times. In order to position the mobile terminal (5), the base station (1), triggered by a position-related inquiry of the mobile terminal (5), transmits a source signal to the mobile terminal (5). The mobile terminal (5) responds with a response signal that is transmitted to at least two time measuring stations (3). The time measuring stations detect the arrival time of the respective response signal received. The arrival times of the response signal in the various time measuring stations (3) is used to calculate the position of the mobile terminal (5).

Description

System and procedure for location determination

The present invention relates to a method for determining the location of a mobile terminal, in particular in a Bluetooth network, and a system for determining the location of a mobile terminal, in particular based on a Bluetooth network.

In wireless networks with mobile terminals, it is often desirable or even necessary to determine the location of the mobile terminal, e.g. B. to determine the location of a mobile phone in a building. Special positioning systems are used for this. Many of these location determination systems are based on the determination of transit times or transit time differences of signals which originate from the mobile terminal and are received by stationary stations. The location of the mobile terminal or the mobile phone is then determined from the transit times or the differences in the transit times. Such methods are e.g. B. described in DE 199 36 846. Location data is also often required for wireless networks operating in the high-frequency range. With the help of such location data z. B. Send a service provider location-related information to mobile terminals.

A mobile terminal in the sense of the invention can in addition to a mobile phone also any other conceivable mobile communication unit, such as. B. a notebook or a Personal Digital Assistant (PDA).

A wireless network in the high frequency range is e.g. B. a network based on the Bluetooth standard. Bluetooth is a technology for wireless communication between different devices. The communication between the devices takes place in the license-free ISM band, a frequency band that is reserved for general industrial, scientific and medical use (ISM, Industrial, Scientific, Medical).

Bluetooth networks often have a range of less than 100 meters. If location-resolved information is to be made available in such networks, high demands are placed on the accuracy of the location determination. In order to be able to carry out location determination methods that meet these requirements, mobile terminals for Bluetooth networks are provided with elements specifically provided for location determination, such as, for example, B. a GPS receiver.

The object of the present invention is to provide an improved method and an improved system for location determination, with which the location determination can be carried out without the need for elements provided specifically for the location determination in the mobile terminals.

This object is achieved by a method for determining the location of a mobile terminal according to claim 1 and a system, that is to say an arrangement for determining the location of a mobile terminal according to claim 17. Further refinements of the method according to claim 1 are specified in claims 2 to 16. Claims 18 to 24 contain further refinements of the arrangement according to Claim 17.

In the method according to claim 1, a base station, triggered by a location-related request from a mobile terminal, sends an originating signal to this mobile terminal. The mobile terminal responds with a response signal that is received by at least three time measuring stations. The arrival time of the response signal received in each case is determined in the time measuring stations. The location of the mobile terminal is then calculated from the arrival times of the response signal at the various time measuring stations.

This is preferably done by determining the respective arrival times of the original signal at the time measuring stations. The difference between the arrival times of the response signal and the original signal at the respective time measuring station is determined.

Alternatively, the transmission of the response signal in the time measuring stations can activate a counter for counting the time measuring clock cycles and the arrival of a stop signal sent from a known location can stop counting in the time measuring station (3). In principle, any transmitter whose position is determined in a predetermined coordinate system can be considered as the transmitter of the stop signal.

The method according to claim 1 offers the advantage that the location determination can be carried out as part of a special Bluetooth profile. A Bluetooth profile is a set of rules that specify how the Bluetooth protocols should run. The steps carried out in the method according to claim 1 can all be implemented by means of commands from the Bluetooth protocols. Therefore, there are no modules specifically required for determining the location required in the mobile terminals. The integration of the method into a Bluetooth profile is the subject of claim 16. In one embodiment of the method according to the invention, the time measuring stations compare the response signal bit by bit with the previously stored original signal in order to determine the arrival time. A correlation method is used for the comparison.

In order to increase the accuracy of the determination of the arrival times, the determination of the arrival times can be carried out with an increased timing clock rate compared to the signal clock rate resulting from the bit rate of the signal (oversampling or oversampling). Local resolutions of a few meters are possible.

An even more precise location determination is possible through the additional application of an interpolation method for interpolation between the correlation values obtained with the correlation method.

The system according to the invention for determining the location of a mobile terminal comprises a base station and at least three time measuring stations. The time measuring stations each comprise a time recording unit with which the arrival time of the original signal and a signal emanating from the mobile terminal can be recorded, a shift register and a correlator each being present in the time recording unit. The system also includes a location determination unit for calculating the location from the recorded arrival times.

The system according to the invention, that is to say the arrangement according to the invention, offers the advantage that an exact location can be determined for each existing mobile Bluetooth terminal without the mobile Bluetooth terminal having to be specially equipped for this.

The time measuring stations of the system can be clocked for detecting the arrival time of the signal coming from the mobile terminal via a common time clock. This makes it easy to ensure a common time scale for all time measuring stations. The timer can provide a clock rate that is higher than the signal clock rate resulting from the bit rate. This improves the location resolution of the location determination.

In the system according to the invention, in a preferred exemplary embodiment, each time measuring station is each connected to a receiver which has a receiving device for receiving a signal and a baseband processor. The received signal can be picked up in the receiver in front of the baseband processor. For this purpose, a corresponding signal connection or a branch is provided in the circuit of the receiver, via which the received signal is routed to the respective time measuring station. Timing station and receiver are preferably integrated in one device.

Due to the possibility of tapping the received signal in front of the baseband processor, that is to say before the received signal has been synchronized to the clock of the baseband processor, the receiver can deliver a signal which still contains the information about the transit time of the signal. The signal can be evaluated by a time measuring station. Bluetooth stations with such a receiver can therefore be used as time measuring stations for the system according to the invention without an additional receiver for determining the location. This provides a receiver, in particular for Bluetooth devices, which enables the location of a mobile terminal to be determined in a simple manner. In principle, the receiver can also be operated independently of the system according to the invention.

In a preferred embodiment, a pulse shaper is present in the receiver. The received signal can be tapped between the pulse shaper and the baseband processor. As a result, the structural deviations of the receiver according to the invention from standard receivers for Bluetooth stations can be reduced to a minimum. In embodiments that use a transmitter with a precisely defined position, this is preferably the base station. For a position determination based on the difference between the arrival times of the response signal and the stop signal at the time measuring stations, the base station is preferably designed as a transmitter of the stop signal. This means that when you install the base station, your exact position is determined using known positioning methods.

The method according to the invention, the system according to the invention and the receiver according to the invention are described in detail below using an exemplary embodiment and with reference to the figures.

1 schematically shows an exemplary embodiment for the construction of the system according to the invention for determining the location of a mobile terminal.

2 shows a possible sequence of the method according to the invention for determining the location of a mobile terminal.

3a shows schematically the reading in of the response signal into a shift register, the signal being oversampled.

3b shows the assignment of a pulse to the shift register cells.

3c shows the assignment of a pulse to the shift register cells.

Fig. 4 shows schematically an embodiment for the construction of a receiver according to the invention.

An exemplary embodiment of the system and the method according to the invention is described in more detail below on the basis of determining the location of a mobile terminal in a Bluetooth network. 1 shows a system for determining the location in a network based on Bluetooth technology. Bluetooth is a standard for wireless communication between different devices. Communication between the devices takes place in the license-free ISM band, a frequency band that is reserved for general industrial, scientific and medical use (ISM, Industrial, Scientific, Medical).

The network shown is a so-called piconet, a network based on Bluetooth technology, which has a master which controls the piconet and which can comprise up to seven further devices, so-called slaves. According to FIG. 1, the piconet comprises a base station 1 as a master and two timing stations 3 as slaves, the positions of which are known. In the exemplary embodiment, the base station 1 takes over the function of a further, third time measuring station. A common clock generator 7, hereinafter referred to as the clock generator, provides the base station 1 and the time measurement stations 3 with a common time measurement clock frequency via lines 9. Alternatively, the common timing pulse frequency can also be transmitted wirelessly. Finally, a mobile terminal 5 is shown, which also belongs to the Piconet as a slave and whose location is to be determined.

The fact that the mobile terminal 5 is shown as the slave of that piconet with the base station 1 as the master is assumed here only to simplify the illustration. Instead, the mobile terminal 5 can also be the master or slave of another piconet to which the base station 1 also belongs as a slave. Base station 1 then belongs to two different piconets, one as a master and the other as a slave. Piconets connected in this way are called Scatternet. The base station 1 and the mobile terminal 5 only have to belong to one common piconet during the location determination.

The process of determining the location is carried out by the mobile terminal 5 with a

Request to base station 1 initiated for location-specific information. Upon request, the base station 1 sends a signal that the original jump signal, to the mobile terminal 5 and to the time measuring stations 3 and stores the original signal. The original signal activates the time measurement in the time measuring stations 3, which also store the original signal. The mobile terminal 5 then sends a copy of the received original signal as a response signal to the base station 1 and the time measuring stations 3. The base station 1 and the time measuring stations 3 each determine the time of the arrival of the response signal (arrival time) on a common time scale, which is determined by the common Clock 7 is conveyed.

To determine the times at which the response signal arrives at the base station 1 and the time measuring stations 3, the stations each determine the point in time of the bit-wise correspondence between the stored original signal and the received response signal. This method will be described in detail later with reference to Fig. 2a. If the incoming response signal coincides bit by bit with the stored original signal, the response signal has arrived. The point in time at which the match exists provides the arrival time of the response signal at the respective station.

A processor in an evaluation unit, to which the determined arrival times are transmitted by cable or wirelessly, calculates the location of the mobile terminal 5 from the determined arrival times. In the exemplary embodiment, the evaluation unit is integrated in the base station 1, but it can also be an independent unit, e.g. an evaluation computer.

The location is calculated on the basis of a method which makes it possible to determine the location of the mobile terminal 5 on the basis of the differences in the arrival times at the base station 1 or the time measuring stations 3. Such methods are generally known and are called TDOA methods, where TDOA stands for "Time Difference of Arrival". If the time of sending the response signal by the mobile terminal 5 is known with sufficient accuracy or can be estimated, instead of the TDOA method for determining the location, it is also possible to use a TOA method (time of arrival) in which the location is determined based on the arrival times and the transmission time of the response signal is carried out.

Instead of a base station 1 and two time measuring stations 3, as described, more than two time measuring stations 3 can also be present. In this case, the base station 1 does not have to take over the function of a time measuring station and therefore does not have to be equipped with a time recording unit for recording the arrival time of the response signal. This offers the advantage that the method according to the invention for determining the location does not require a base station set up specifically for determining the location. Existing systems can therefore be retrofitted with time measuring stations in order to be able to carry out the method according to the invention.

The method for determining the location of a mobile terminal is described in detail below with reference to FIGS. 2a and 2b. It is assumed that during the request from the mobile terminal 5 to the base station 1, there is already a piconet of base stations 1 and at least two time measuring stations 3, and the mobile terminal 5 also belongs to the same piconet as a slave. If such a piconet does not exist at the time of the request, a piconet or scatternet, which comprises the base station 1 and the mobile terminal 5, exists in any case at the time of the request. The request then triggers the formation of a piconet or, if appropriate, a scatternet, to which both the base station 1 and the time measuring stations 3 belong, the base station 1 forming a piconet with the time measuring stations 3, in which the base station 1 is the master.

After the request, and possibly after the formation of a piconet or a scatternet, the base station 1 forms a Bluetooth group with the time measuring stations 3 and sends an echo request packet as a Bluetooth packet (Data packet in a Bluetooth connection) to the mobile terminal 5. With an echo request packet, the receiving station, in the exemplary embodiment the mobile terminal 5, is asked to send back a copy of the received signal as a response.

The echo request packet includes an access code that contains a 64-bit pseudo-random sequence. Like all Bluetooth packets, the Echo Request packet has a three-part structure consisting of access code for detecting and addressing the packet, header with control information and Payloader with the actual data. The access code has a length of at least 68 and at most 72 bits, of which four bits are used for a preamble and a (not always available) preview (trailer) of the following data (payload). The remaining 64 bits contain the pseudo-random sequence mentioned above. In addition to the pseudo-random sequence in the access code, other, longer pseudo-random sequences can also be contained in the data part (payload). The data part can contain a maximum of one pseudo-random sequence of 2712 bits in length, so that a maximum of one pseudo-random sequence of length of 2776 bits can be present in a Bluetooth packet. Instead of the pseudo-random sequence, a random sequence or a combination of both can also be used.

In addition to the mobile terminal 5, the echo request packet is also received by the time measuring stations 3, demodulated there and stored in a random access memory (RAM, Random Access Memory). That the time measuring stations 3 are set to receive instead of sending while the response signal is being sent, even though they are slaves of the piconet and should therefore actually be set to sending after receiving a signal from the master (here the base station 1). B. achieve that the timing stations 3 are wired to the base station 1 (master) and receive a signal via cable that sets them to receive.

Alternatively, after sending the echo request packet, the base station 1 can also send a further signal which the mobile terminal 5 sends. let swap the role with the base station 1, so that the base station 1 becomes the slave and the mobile terminal 5 becomes the master. The mobile terminal 5 then sends the copy of the echo request signal to the base station 1, which, like the time measuring stations 3, is now set to receive as a slave. Since the location determination in the exemplary embodiment is based only on the differences in the arrival times at the time measuring stations, the duration of this role change has no influence on the result of the location determination.

As a response signal to the echo request packet sent by the base station 1, the mobile terminal 5 sends a copy of the echo request packet and thus the pseudo-random sequence. The response signal is received by base station 1 and time measuring stations 3 and demodulated in order to extract the copy of the pseudo-random sequence.

The copy of the pseudo-random sequence is correlated with the pseudo-random sequence stored in RAM and obtained from the original signal. To do this, the copy of the pseudo-random sequence is entered into a shift register.

The pseudo-random sequence and its copy are made up of zeros and ones (bits), a one being represented by a high signal level and a zero by a low signal level before modulation and after demodulation. A high level or a low level in the data each takes 1 μs. This means that Bluetooth is able to transmit 1 Mbit of data per second, i.e. one million high or low levels per second. This is the bit rate of the pseudo-random sequence transmitted or its copy. With regard to the bit rate of the pseudo-random sequence or its copy, this results in a clock frequency, hereinafter referred to as the signal clock frequency, of 1 MHz.

The copy of the pseudo-random sequence is entered into the shift register at a clock frequency which is greater than or equal to the signal clock frequency. In the exemplary embodiment, the clock frequency with which the calculation is carried out, hereinafter referred to as the time measurement clock frequency, is 110 MHz. that is 110 times the signal clock frequency. Compared to the signal clock frequency, this time measurement frequency means oversampling (oversampling) by a factor of 110. This means that each high level and each low level of the pseudo-random sequence with a respective duration of 1 μs in 110 "short" high levels or 110 "short" low level with a respective time period, which corresponds to the 110th part of a μs (approx. 9.09 ns), is converted. In this way, the 64 bits of the pseudo-random sequence become 7040 bits, which form 64 groups, each comprising 110 bits, of high and low levels. The 64-bit pseudo-random sequence of the original signal is also broken down in the same way into 7040 bits and the now 7040-bit long pseudo-random sequence of the original signal is stored in RAM.

The copy of the pseudo-random sequence is entered into a shift register in time with the time measurement frequency. For each clock cycle of the timing cycle, hereinafter referred to as the timing cycle, the 7040-bit copy of the pseudo-random sequence is shifted by one of the 7040 bits in the shift register with a length of 7040 bits.

In each timing clock cycle, the content of the shift register is bit-by-bit correlated with the content of the 7040 memory cells in RAM which contain the pseudo-random sequence of the original signal. In a bit-wise correlation, each bit of a register, here the shift register, is multiplied by the corresponding bit of another register, here the RAM. The resulting products are added up. The result of this correlation has a greater value, the better the content of the 7040 bit long shift register matches bit by bit with the content of those memory cells of the RAM in which the 7040 bit long pseudo-random sequence obtained from the original signal is stored. The maximum value occurs when the content of the shift register matches that of the RAM bit by bit. Before the result of the correlation takes its maximum value, the value increases linearly with each cycle, and then decreases linearly again with each cycle. The occurrence of the maximum value of the correlation therefore indicates that the copy of the pseudo-random sequence, and thus the response signal, has been completely is hit. DH also the last bit of the copy of the pseudo-random sequence is entered in the shift register so that the shift register contains the entire pseudo-random sequence.

The described bit-wise correlation of the 7040-bit copy of the pseudo-random sequence with the 7040-bit long pseudo-random sequence of the original signal stored in the RAM provides a comparison of the original signal sent by the base station 1 with the response signal sent by the mobile terminal 5.

The value of the bit-wise correlation is calculated once per clock cycle. Then, with the following timing cycle, the 7040 bit long pseudo-random sequence in the shift register is shifted by a single one of the 7040 bits and the value of the correlation is calculated again. The process of bit-wise reading of the incoming copy of the pseudo-random sequence 20 into the shift register 62 and the correlation of the oversampled signal with the pseudo-random sequence 22 stored in the RAM 66 is shown in FIG. 3a by means of a four-fold oversampling (oversampling with a factor of 4). shown as an example.

During the entire process, a counter of the respective station counts the time measuring cycle from a zero point until the maximum of the correlation function is reached. In the exemplary embodiment, the zero point is formed by the time at which base station 1 sends the echo request packet. Timing station 3 reveals the running time of the echo request packet from base station 1 to time measuring station 3 and thus the zero point through the arrival of the original echo -Request packet and the known distance of the time measuring station 3 from the base station 1. With the arrival of the echo request packet, the counting of the time measuring clock cycles begins. The number of clock pulse cycles times the length of a clock pulse cycle with the duration of 9.09 ns (exemplary embodiment) gives the time between the arrival of the original echo request packet by base station 1 and the occurrence of the maximum value of the correlation, ie the complete arrival of the copy of the echo request packet by which the respective time measuring station 3 or the base station 1 has passed. The occurrence of the maximum value of the correlation thus indicates the point in time at which the response signal has arrived completely. Since the copy of the pseudo-random sequence is shifted by one memory location in the shift register every 9.09 ns and the value of the correlation is determined each time, the arrival of the response signal can be determined with an accuracy of 9.09 ns, i.e. to the 110th part of a μs ,

A processor then forms the differences from the arrival times determined by the respective time measuring stations 3 or by the base station 1 and calculates the location of the mobile terminal 5 from the differences.

Since the arrival times can be determined with an accuracy of approx. 10 ns, the differences in the transit times of the echo request package to the individual time measuring stations can be determined with this method to an accuracy of approx. 20 ns. With this accuracy, location determination of a few meters is possible.

So that the arrival times in the different time measuring stations 3 are comparable, the time measuring clock for the base station 1 and the time measuring stations 3 is made available by a common time clock. This ensures that the time measuring cycle cycles each have exactly the same length.

The time of transmission of the echo request packet does not necessarily have to be selected as the zero point of the counting of the clock cycles. Other zero points are also possible. There is also no need for a common time clock for the time measurement cycle. In this case, however, the time measuring stations must each have their own time clock. The phase deviations occurring over the measurement time in the time measurement cycle of the individual time generators of the time measurement stations 3 must not exceed a certain value. Before a location is determined, the location system is calibrated. The calibration is carried out in order to determine the transit time of the original signal from the base station 1 to the time measurement stations 3, so that the time measurement stations 3 can determine the zero point of the count (time of transmission of the original signal) from the arrival time of the original signal when the original signal is received. Although the transit time of the original signal and thus the zero point can in principle also be calculated from the distance of the time measuring stations 3 from the base station 1, the calibration measurement is the more suitable means, in particular if the location determination system system is changed frequently, for example by changing it frequently dismantled and rebuilt. Another advantage of the calibration measurement is that it also takes into account the transit time differences of the original signal that occur within the receiving time measurement stations 3. Such runtime differences cannot be calculated from the distance of the time measuring stations 3 from the base station 1.

Instead of the difference in arrival times, the location can also be calculated from the arrival times themselves if the time of sending the copy of the response signal in relation to the time measurement clock is known or can be estimated with sufficient accuracy.

As described above, the correlation is carried out with the clock frequency of 110 MHz. If the correlation were only performed at a clock frequency of 1 MHz (the clock frequency resulting from the bit rate of the Bluetooth signal), this means that a correlation is not carried out every 9.09 ns, but only every microsecond. The maximum value of the correlation - and thus the arrival time of the copy of the echo request packet - can then only be determined with an accuracy of 1 μs. However, this means that the location can only be determined to within a few 100 meters. This is too imprecise for determining the location of a mobile terminal 5 in a piconet or a scatternet. The above-described method with oversampling is therefore used to improve the accuracy of the location determination. applies. In networks other than Bluetooth-based, the clock frequency of the bit rate of the transmitted data can be higher than 1 MHz. Oversampling may then not be necessary. Even with a clock frequency of 1 MHz, oversampling is not necessary if a spatial resolution of a few 100 m is sufficient.

An improvement in the accuracy in the determination of the arrival time of the response signal can be achieved in addition to a shortening of the timing cycle by interpolating between the correlation values determined for each timing cycle. This is done by laying a curve, which is known from preliminary considerations, through the correlation values and adapting the parameter or parameters of the curve until it maximally matches the correlation values (fitting process). For adjustment, for example, the sum of the squares of the deviations of the values of the curve from the correlation values can be calculated. The parameters of the curve are then varied until this sum assumes a minimum. With this method, an interpolation, the maximum can be determined with an accuracy that is greater than the accuracy that is determined by the length of the timing pulse cycles.

A simple case of interpolation is obtained for exactly rectangular pulses of the pseudo-random sequence or random sequence. If in this case the values are plotted before the occurrence of the maximum value depending on the state of the counting of the time measuring cycle cycles, then an ascending straight line is obtained, whereas after the maximum value occurs a descending straight line is obtained. The intersection of this straight line then shows the maximum value of the correlation.

Instead of the described 110-fold oversampling, however, oversampling with a different factor can also be selected. The factor of oversampling depends on the desired spatial resolution. When choosing the factor, the quality of the fitting process can also be taken into account. ever the more precisely the maximum can be determined with the fitting process, the lower the factor needs to be.

It is also advantageous if the selected factor of oversampling, i.e. the ratio of the timing pulse frequency and the signal timing frequency, does not result in the length of a high or low level being an integral multiple of the duration of a timing pulse cycle. Otherwise, the time of occurrence of the maximum can only be determined with an accuracy of one time measuring cycle in the case of an integer ratio. The selected factor for oversampling should not result in the duration of the level being an integral multiple of the duration of a time measurement cycle for any of these possible time periods of a high or low level. It is therefore particularly advantageous if the relationship between the timing pulse frequency and the signal timing frequency is not rational, since this ensures in any case that the duration of a level is not an integral multiple of the duration of a timing pulse cycle.

The advantages of the described method are summarized again below: The method makes it possible, in particular, to dispense with the synchronization of the time clocks in the time measuring stations and in the master with a common system time. Furthermore, it does not require a defined time delay in the mobile terminal between the reception of the echo request signal and the transmission of the echo response signal. The position can be determined with an accuracy that is several times better than the distance corresponding to the duration of a bit. Finally, the location of the mobile terminal can be determined from the one-time reception of the signal code.

These advantages are achieved by the following technical measures: the received signal is sampled before it has been synchronized to the clock of the baseband processor. The received signal is still oversampled at a frequency that is many times higher than the bit rate. Furthermore, the value of the correlation function between received and stored code in each sampling interval, the time of the maximum of the correlation function is determined from the one-time reception of the signal code and finally. The accuracy of the localization can be increased by repeating the measurement several times and averaging.

The determination of the position of a mobile terminal by the TDOA method according to the invention in one exemplary embodiment is explained in more detail below with reference to FIG. 2b. In this context, the mobile terminal 5 is also marked with "M" for a better understanding of the following equations. In the present exemplary embodiment there are three time measuring stations 3, which are additionally marked with "Z1", "Z2" or "Z3", and one Base station 1 is provided, which is identified by “B”. Furthermore, there is a time clock 7 (also “T”) to which the time measuring stations are connected.

The signal sequence explained with reference to FIG. 2a is symbolized in FIG. 2b by arrows of different line types. The temporal course of the signals, which cannot be represented here, corresponds to that explained with reference to FIG. 2a. The additional third time measuring station carries out the same steps as the time measuring stations shown in FIG. 2a.

In addition to the signals explained in FIG. 2a, an activation signal originating from the base station is provided in the present exemplary embodiment, which is shown in the form of arrows with a dashed line. The activation signal is sent to all timing stations as well as the mobile terminal. The echo request packet is then sent to the time measuring stations and the mobile terminal. This is shown by arrows with a dotted line. The copy of the echo request signal sent by the mobile terminal is sent as an echo response signal from the mobile terminal to the time measuring stations. This is shown in the form of arrows with a solid line. The mathematical side of determining the time difference is explained in detail below. First, the arrival time of the echo request signal at the time measuring stations Z | certainly. It is given by

tlZ, i ⁼ tstart ⁺ tBZ, l (1)

Here, t _start denotes the point in time at which the base station transmits the echo request signal, t _B z, ι signal propagation time from the base station B to the i-th time measuring station Z | (i = 1, 2, 3) and t _1Zl ι arrival time of the echo request signal at the i-th time measuring station Z _\ .

For the mobile terminal M, the arrival time tι _{M of} the echo request signal is given by

where t _BM denotes signal propagation time from base station B to the mobile terminal.

The arrival time of the echo response signal at the time measuring stations is

tstart ⁺ tB ⁺ tv ⁺ t Z, i (3)

Here, t _v denotes the delay time in the mobile station between the reception of the echo request signal and the transmission of the echo response signal, t _M z, signal runtime from the mobile station M to the i-th time measuring station Zj and t ₂ z, ι Arrival time of the echo response signal at the i-th time measuring station.

The time difference between receipt of the echo request and echo response at the i-th time measuring station thus results from equations (1) and (3)

Δtj = t ₂ z, i

tstart ⁺ tBM ⁺ tv ⁺ Z, i "tstarrtBZ.i ⁼

(4) Finally, the difference between the signal arrival times at the time measuring stations i and j is Δtdoa.u

Δtdoa, i, j ⁼ Δtj- Δtj = t _B M ⁺ tv ⁺ tMZ, i-tBZ, i "tß - v-tMZj + tBZj

that is, in the result

Δtdoa, i, j ⁼ tMZ, i-tBZ, rtMZ, j ⁺ tBZ, j (5)

The signal propagation times t _B z, ι from the base station B to the i-th time measuring station Z _\ are known or can be determined in a calibration measurement.

In the method described here, the delay time t _v between reception of the echo request signal and transmission of the echo response signal need not be known in the mobile station. The time t _Sta n of transmission of the echo request signal may also be arbitrary. The delay times between reception and transmission of a signal in the time measuring stations and in the base station are not required at all for this method. The method described here differs from known methods which measure the signal _{arrival times} t _2Zι j directly at the base stations and, for this purpose, require the values of the time delay t _v .

In order to measure the time differences between the arrival of the echo request signal from the master and the arrival of the echo response signal from the mobile terminal, each time measuring station can use its own, non-synchronized time clock. The time errors resulting from frequency deviations of the time clocks must be smaller than a clock time of the shift register of the correlator via the time difference Δtj. This can be achieved with quartz-stabilized oscillators.

3a shows schematically the reading in of the response signal into a shift register, the signal being oversampled. The incoming copy 20 of the pseudo-random sequence 22 is read into the shift register 62 bit by bit. The signal is oversampled four times. 3b and 3c show an error source in the time determination of the assignment of a pulse to the shift register cells to explain the inaccuracy of the position determination at a low sampling rate. In a first case (Fig. 3b) an incoming pulse (solid line) generates a "1" in four cells of the shift register due to the given timing (vertical lines). In a second case (Fig. 3c) generates an equally long pulse due to a "1" of a slightly shifted clock in only three cells of the shift register. In this way, the correlation result will be different depending on the position of the timing relative to the signal. This causes an inaccuracy when determining the time of arrival of a signal and thus the position determination. The figures accordingly make it clear that by increasing the sampling rate via the time resolution, the spatial resolution of the position determination can be increased. Because the higher the sampling rate, the smaller the error in the correlation.

The structure of a receiver suitable for Bluetooth, which enables the use of the above-described method, is described below with reference to FIG.

Figure 4 shows schematically the receiver as a receiving part of a Bluetooth transceiver. The receiving part comprises a high-frequency antenna 50, a low-noise high-frequency amplifier 52 for amplifying the received signal, at the input of which the high-frequency signal received by the high-frequency antenna 50 is present, a demodulator 54 for demodulating the amplified signal, at the input of which the amplified signal is present, and a pulse shaper 56 for bringing the pulses of the demodulated signal into the desired form, at the input of which the demodulated signal is present. The output of pulse shaper 56 is connected to shift register 62 and baseband processor 64.

The receiver shown is a superimposed receiver, also called a superheterodyne receiver or a superheterodyne receiver. This is amplified between the low-noise high-frequency amplifier 52 and the demodulator 54 Signal mixed in a mixer 58 with a frequency signal z, which is generated by a voltage-controlled oscillator 60. The result of the mixing is a sum frequency signal from the sum of the two mixed frequencies and a difference signal (intermediate frequency signal) of 111 MHz, which results from the absolute value of the difference between the frequencies. The intermediate frequency signal is taken from the mixer 58 (down mixer) and fed to the demodulator 54.

The demodulator 54 is designed in such a way that it delivers the largest possible output signal for the frequency swing customary in Buetooth. For this purpose, the demodulator 54 has a resonance curve with a steepest possible slope. On the other hand, the resonance curve must be wide enough not to allow drift. The signal tapped at the output of the demodulator 54 is applied as an input signal to the pulse sharpener 56, the output signal of which is present at the input of the shift register 62 and the baseband processor 64.

The length of the shift register, as well as the number of memory locations in RAM, which are provided for the pseudo-random sequence extracted from the original signal, depend on the choice of the oversampling factor, since this specifies in how many bits a single bit of the pseudo-random sequence is split up. As described above, oversampling with a factor of 110 generates 110 "short" bits with a duration of 9.09 ns (110th part of a μs) from a single bit with a duration of 1 μs. For each "short" bit the shift register must or its own memory space can be provided in RAM.

In addition, the length of the shift register and the number of memory locations provided in RAM for the pseudo-random sequence extracted from the original signal depend on the length of the pseudo-random sequence sent. In the exemplary embodiment described, the length of the pseudo-random sequence is 64 bits, which means that a shift register with 64 times 110, that is 7040, memory locations must be provided. The same applies to the area of RAM in which the pseudo-random sequence of the original signal should be saved. Longer or shorter pseudo-random sequences require longer or shorter shift registers and more or less memory locations in RAM with the same oversampling factor.

It is important that the demodulated response signal sharpened by the pulse shaper 56 is tapped before the baseband processor 64 in order not to lose the time information due to the synchronization carried out there to the clock of the baseband processor 64. In order to compensate for the time deviation of the signal clock resulting from the bit rate of the response signal from the clock of the baseband processor 64, the clock generator of the baseband processor 64 is synchronized with the signal clock. This is done by adding an offset to the clock cycles of the baseband processor, whereby the time information of the received signal is lost.

The response signal can be tapped anywhere in front of the baseband processor 64. However, the earlier the response signal is tapped, the more additional elements must additionally be present between the tapping point and the correlator. If the response signal were tapped behind the mixer 58, for example, the branch to the correlator 62 would have to have its own demodulator and its own pulse shaper. It is therefore best to tap the response signal after the pulse shaper.

Deviations from the exemplary embodiments shown in detail are possible within the scope of the invention.

In the exemplary embodiment, the common timing pulse frequency is conveyed via lines (see FIG. 1 and associated description). Alternatively, however, it can also be transmitted wirelessly.

In addition, the time measuring station 3 and the base station 1 can each have an individual time clock instead of a common one. However, these individual time clocks must each have a time provide a sequence with sufficiently long and stable timing cycles. If this does not happen, the sum of the deviations of the time measurement cycle from each other would become so large during the measurement that the location can no longer be determined with the required accuracy. The required quality of the individual timers is therefore dependent on the desired accuracy of the location determination.

Both the base station 1 and the time measuring stations 3 are designed as stationary stations in the exemplary embodiment, but they can also be mobile stations, provided that their position is known sufficiently well during the location determination for the location determination to be carried out.

The clock when entering the response signal in the shift register does not have to be given by the time measurement clock. It is sufficient if the clock used is synchronized with the time measurement clock.

The described location determination only leads to a clear result if there are three time measuring stations. However, if there are only two time measuring stations and the uniqueness of the result can be established in a different way than by a third time measuring station, two time measuring stations are sufficient for determining the location. E.g. If a procedure based on the arrival times and not on the differences of the arrival times, a so-called ToA procedure (Time of Arrival) is used, a circle is obtained for each time measuring station on which the mobile terminal can be located. The radius of the circle shows the distance of the mobile terminal from the respective time measuring station. With two timing stations, these circles generally have two intersections, unless the mobile terminal is exactly in the middle between the two timing stations. If one of the two intersections can be excluded, for example because the time measuring stations are located on a wall and the mobile terminal can only be located on one side of the wall, two time measuring stations are sufficient for determining the location.

Claims

claims

1. Procedure for determining the location of a mobile terminal with the steps:

Sending an originating signal originating from a base station (1) to the mobile terminal (5) and to at least two time measuring stations (3),

Sending a response signal to the original signal to the time measuring stations (3) by the mobile terminal (5),

Determining the arrival times of the response signal at the time measuring stations (3) and

Determining the location of the mobile terminal (5) on the basis of the arrival times of the response signal at the time measuring stations (3).

2. The method according to claim 1, characterized in that the original signal is sent to at least three time measuring stations.

3. The method according to claim 1 or 2, characterized in that the arrival times of the original signal are determined at the time measuring stations.

4. The method according to any one of the preceding claims, characterized in that the time measuring stations (3) are conveyed a time scale with a common time measuring cycle and a common zero point.

5. The method according to claim 4, characterized in that the arrival times in the time measuring stations (3) are determined by counting the time measuring clock cycles until the arrival of the response signal.

6. The method according to claim 5, characterized in that the arrival of the original signal in the time measuring stations (3) each activates a counter for counting the time measuring clock cycles.

7. The method according to claim 5, characterized in that the arrival of a signal sent from a known location in the time measuring stations (3) each activates a counter for counting the time measuring clock cycles.

8. The method according to claim 1, characterized in that the transmission of the response signal in the time measuring stations (3) activates a counter for counting the time measuring clock cycles and the arrival of a stop signal sent from a known location stops counting in the time measuring stations (3) ,

9. The method according to claim 7 or 8, characterized in that the respective difference in the arrival times of the original signal and the response signal or the response signal and the stop signal in the

Timing stations is determined.

10. The method according to any one of the preceding claims, characterized in that the time measuring stations (3) store the original signal, that the mobile terminal (5) sends a copy of the original signal as a response signal and that the arrival of the response signal by a

Comparison of the response signal with the stored original signal is determined.

11. The method according to claim 4 and claim 10, characterized in that the comparison of the response signal with the original signal is carried out by the response signal in the time measuring stations (3) in

Time measurement clock is entered in a shift register and in each time measurement cycle the response signal located in the shift register is correlated bit by bit with the stored original signal.

12. The method according to claim 6 and claim 11, characterized in that the arrival of the response signal is determined by the maximum value of the correlation and the counter reading at which the maximum value occurs represents the arrival time.

13. The method according to any one of claims 4 to 12, characterized in that the timing cycle is shorter than the signal clock cycle resulting from the bit rate of the response signal.

14. The method according to claim 13, characterized in that the timing cycle is 10 to 1000 times shorter than the signal cycle.

15. The method according to any one of claims 12 to 14, characterized in that in order to determine the maximum value of the correlation, a parameterized curve or straight line is adapted to the correlation values.

16. The method according to claim 15, characterized in that the adaptation takes place in such a way that the squares of the deviations of the values of the curve or the straight line from the correlation values are determined and the parameters of the curve or the straight line are varied for as long as possible, until the sum of the squares of the deviations takes a minimum.

17. The method according to any one of claims 12 to 14, characterized in that a straight line is laid through the values determined before and after the maximum value in order to determine the maximum value of the correlation.

18. The method according to any one of the preceding claims, characterized in that the original signal comprises a pseudo-random sequence and / or a random sequence.

19. The method according to any one of the preceding claims, wherein the base station (1) also performs the tasks of one of the timing stations (3).

20. The method according to any one of the preceding claims, wherein the communication between the mobile terminal, the base and the time measuring stations takes place according to a Bluetooth protocol.

21. Arrangement for determining the location of a mobile terminal with a base station (1), at least two time measuring stations (3), the time measuring stations (3) each having a time recording unit, with which the arrival time of an outgoing from the mobile terminal (5)

Signal can be detected, and a location determination unit for calculating the location from the arrival times recorded by time recording units, characterized in that the time recording units each comprise a shift register (62) and a correlator.

22. The arrangement according to claim 21 with at least three time measuring stations (3).

23. The arrangement according to claim 21 or 22, characterized in that a common time clock (7) is provided for supplying a common time measurement frequency to all time measuring stations (3).

24. The arrangement according to claim 23, characterized in that the common clock (7) supplies the timing stations (3) with a clock frequency which is 10 to 1000 times higher than the frequency of the signal clock.

25. Arrangement according to one of claims 21 to 24, characterized in that each time measuring station (3) comprises a receiver according to claim 17 or claim 18.

26. Arrangement according to one of claims 21 to 25, characterized in that one of the time measuring stations (3) is integrated in the base station (1).

27. Arrangement according to one of claims 23 to 26, characterized in that the time measuring stations are each connected to a receiver which has a receiving device for receiving a signal and a baseband processor, the receiver being designed such that the received signal is present the baseband processor can be tapped.

28. The arrangement according to claim 27, characterized in that the receiver and time measuring station are integrated in one device.