FR2694147A1 - Evaluation of channel quality for radio telephone system - measuring various signal parameters both at base station and mobile station, and accumulating data at base station regarding availability for communication and hand-over - Google Patents

Abstract

The method for evaluating transmission channels between a base station and mobile stations involves measuring the noise (B) on the inactive channels at the base station. At the base station and at the mobile stations one measures the field level (and eventually the error rate) on the active channels. Measurement effects at the mobile stations are transmitted to the base station. Finally measurements are taken, at the base station and the mobile stations, of the noise and interference during time slices of active receiving channels when there is no activity and no carrier signal. USE - Evaluation of channels for time division multiple access (TDMA) radio telephone system.

Description

METHOD FOR EVALUATING RADIO CHANNELS OF
TRANSMISSION
The subject of the invention is a method for evaluating the radio transmission channels between a base station and mobile stations of a communication installation, in particular voice, with time division multiple access, called TDMA or TDMA, with a view to respond to a request for channel allocation for establishing a call, by allocating the free channel which will give the best connection conditions.

It finds a particularly important application in installations consisting of cellular digital radiotelephony as described in the GSM standard to which reference will be made above all; but it can also be used in other installations
AMRT provided that they use a discontinuous broadcast.

The latter term means that the transmitting station ceases to apply the carrier to the antenna when there is no useful data to transmit - for example no voice activity -, that is to say during intervals of silence.

Methods have already been proposed for classifying the various available channels according to their ability to support voice communication. In particular, it has already been proposed to carry out
- at the base station, interference level measurements on the inactive channels, with the aim of preferably allocating, in response to a request, a channel with a low interference level,
- field level measurements at reception and quality on the active channels.

 However, existing solutions have limitations and shortcomings, which can be revealed, for example, by considering the specific case of GSM.

 1 ") According to the GSM standard, each base station measures, on the inactive channels, the radio frequency field level, and considers that it is representative of the interference level on this channel. No interference measurement is in revenge performed on active channels, so allocation is based on measures that may be old (for example when allocating a channel that has just been released).

 As a result, the measurements carried out are not really significant under high load conditions, since there is then little information available. In order for them to be usable, each channel, when it becomes available, should be left inactive for the time necessary to obtain information on the current level of interference. This is in contradiction with an efficient use of resources, since it further increases the blocking rate of the channels.

 This channel classification criterion can even have a perverse effect. Some inactive channels may have a low level of interference not because there is little interference, but due to a malfunction of the reception stages (for example receiver not connected or badly connected to the antenna) .

2 ") Still following the GSM standard, each receiving station (base station or mobile station) measures, during a communication on a channel with a transmitting station:
- the received field level RXLEV, with an accuracy of around + 2 dB,
- RXQUAL quality, made up of the error rate observed.

 The purpose of these measurements is in particular to provide the data making it possible to implement (a) the power control algorithms, using the measurements carried out by the receiver to control the power of the transmitting station and (b) the transfer algorithms or " hand over "from one base station to another.

But the first measurement does not give a directly usable indication on the useful signal level S, because it is representative of the sum of the signal and the noise
S + B. An increase in the level of the measured signal may be due to an increase in noise, not the wanted signal. A command to decrease the power of the transmitter will then cause the opposite effect to that desired, since the signal used will be reduced, therefore the signal to noise ratio. Likewise, a reduction in the measured signal may be due to a reduction in noise, and a command to increase the power is therefore not justified.

 The simultaneous taking into account of the received field level RXLEV and of the error rate RXQUAL makes it possible in principle to evaluate the S + B / B ratio, therefore to reliably decide on the advisability of ordering a power modification, or to decide "hand over" to another channel, usually in a different cell. However, the information on this other channel is limited to a field measurement, which in the case of GSM, is carried out by the mobile on a particular frequency permanently emitted by each cell and which is specific to it. The measurements made by the mobile on these “beacon” channels represent S + B for the adjacent cells.

 This mode of selection proves difficult to implement. It imposes constraints that are difficult to comply with on the "hand over" algorithms, which must be fast because the properties of a channel vary quickly, but not too much so as to not take into account erratic variations. Finally, it is performed without knowing the rate of errors that the new channel will achieve.

The invention aims in particular to provide a method for evaluating the quality of the active as well as inactive channels, allowing in particular
- to have interference level information both in the base station to mobile direction, for active channels, and in the mobile station to base station direction
- to classify the active and inactive channels in order to be able to allocate them on the basis of the criterion of best signal to noise ratio for communication.

To this end, the invention proposes in particular a method for evaluating transmission radio channels between a base station and mobile stations of a communication installation, in particular voice, TDMA, according to which
- at the base station, noise B is measured on the inactive reception channels by this station,
- at the base station and at the mobile stations, the field level (and possibly the error rate) on the active channels is measured, the measurements made at the mobile stations being transmitted to the base station,
characterized in that the noise is also measured at the base station and at the mobile stations during the time slots of the active reception channels which are without voice activity and without signal carrier.

 In particular, we can then use the measurements to determine the actual signal level and the actual noise and interference level, therefore the actual signal-to-noise ratio. And we can allocate channels, as and when they are available, control the transmit power at the base station and at mobile stations, and carry out transfers or "hand over" according to the results of use of said measurements .

It should be noted in passing that the GSM standard systematically provides for two types of measurement
- FULL measurements made on all time slots of the channel without exception
- SUB measurements carried out only on certain time slots for which the signal cannot be deleted so that, if there was a discontinuous transmission phase in the measurement period, only these SUB measurements must be used since they are representative of
S + B on the active instants, the FULL measurements not being representative since they mix periods of emission and periods of silence.

 Because this method provides signal-to-noise ratio information, it is no longer necessary to use quality measurements and to implement complex algorithms to control the power. The "hand over" algorithms, intended to decide on the appropriateness of the "hand over" and to choose the new channel, can be simplified and more precise and it becomes practically possible to carry out an intra-cellular "hand over" since the You can compare the interference levels of an active channel to those of the inactive channels.

 According to whether one favors the search for modifications as small as possible of the current procedures or a maximum efficiency of implementation of the invention, the mode of feedback of the measurements carried out in the mobile stations will be different, as well as the choice of the criteria. channel allocation.

The invention will be better understood on reading the following description of particular embodiments given by way of nonlimiting example. These embodiments correspond to the implementation in a digital cellular radiotelephone installation of the type
GSM, implementing the recommendations of series 6 of
GSM, which we can refer to.

The description refers to the accompanying drawings, in which
- Figure 1 is a block diagram showing a possible distribution of the transmission slots from three frame manager equipment, each frame having P = 8 slots;
- Figure 2 shows the principle architecture of the part of a base station corresponding to a single frame manager equipment.

Before describing embodiments of the invention, we will recall the characters provided for by the standard
GSM and which influence the implementation.

 A channel made up of successive slots of the same serial number (from 1 to 8) is allocated to each communication to be carried out, in the base station to mobile direction (as in the mobile to base station direction). successive frames, all of the same duration, of a sequence to which are assigned n = 4 frequencies. To balance, as far as possible, the quality of reception, the bursts of information are transmitted in successive slots at different radio frequencies, that is to say by implementing a frequency hopping technique. 1, which corresponds to a base station transmitting three sequences under the control of three different frame manager devices 101, 102, 103, the frequency distribution is indicated by rectangles having different fillings. If for example the white slots are those where there is no emission, the measurements which will be carried out there will represent the interference levels. In FIG. 1, the frame sequences from the same equipment 10 are always applied to the same transmitter TX. However, it is also possible to use architectures making it possible to distribute the frame sequences to any one of the transmitters, for example those described in French patent applications Nos. 91 11434 and 92 04569 of the applicant.

 The basic constitution of the components of the base station can be that given in FIG. 2.

 The transmission channel can then be regarded as comprising, for each physical channel, a set 12 of block or convolutional coding, of interleaving and possibly of encryption. These sets, eight in number for example, attack a time multiplexer 14 intended to constitute the sequences of frames of the kind shown in FIG. 1. The multiplex obtained is applied to a modulator 16. The output signal of the modulator is applied to a circuit frequency agility 18, intended to carry out the transmission with diversity of frequencies such as that shown in FIG. 1.

Finally, a power transmitter 20 attacks the transmission antenna 22.

 The reception channel has a symmetrical constitution of the previous one. The receiver 24 drives a selector 26 which retains the frequencies assigned to the successive slots and supplies a demodulator 28.

 The implementation of the invention requires that the receiver 24 be provided with a field measurement circuit 30 making it possible, depending on whether the slot on which the measurement is carried out is active or not, to provide a representative indication either of the sum of the signal or noise, i.e. noise, i.e. the interference level.

 The signal thus obtained is transmitted by a line 32 to the channel allocation equipment 10.

Thanks to the implementation of the invention, as defined above, the channel allocator equipment will not only have the information provided by the GSM standard, that is to say
- the noise (or interference level) B on the inactive channels, in the mobile direction towards the base station,
- the field level S + B (given by RXLEV) in both directions, for the only channels which are allocated and active, one directly supplied by the circuit 30, the other included in the reports of ascent to the station basic.

 It also has, by the invention, information on the noise B, in both directions, for the active channels.

 The base station thus has data enabling it to carry out various operations reliably.

 1 ") The channel allocator equipment 10 can permanently classify the channels, active or not, by interference level. As soon as an active channel is released, it can be allocated if it has a better signal to noise ratio since the equipment 10 knows the current noise level of each of the channels and the capacities (in terms of transmission power) of the mobiles.

The channel allocator may therefore allocate a channel whose interference level is high (or respectively low) when the transmission capacity of the mobile is high (or respectively low), so as to carry out a channel allocation having a report
S + B / B substantially constant or optimum.

2 ") The base station is able to control the power of each mobile taking into account the ratio
S + B / B (and not just S + B), which eliminates problems due to a lack of knowledge of interference levels.

 3) Since the noise level information B is available for all the channels of the cell in communication with the mobile as well as adjacent cells, and this for the active and inactive channels, the base station also has the value of S + B / B in both directions of communication for all channels and it is possible to implement a simpler "hand over" algorithm than those currently planned, aiming to obtain an optimal value of S + B / B, at the network level, for each communication.

 We will not describe here the modes of voice activity detection, discontinuous transmission and insertion of comfort noise at reception, as they can be those described in the GSM series 7 recommendations. It is sufficient to recall that each station transmitter, as soon as it stops transmitting due to the absence of voice activity, transmits, with a recurrence of 480 ms (i.e. 104 frames) 1 silence indication frame, called SID. The reception of this frame allows the receiving station to update the comfort noise generated towards the user and indicates to him that the transmission is discontinuous.

 Various embodiments of the invention will now be described; they use information already provided for by the GSM standard and the fact that, as long as SID frames are received, the measurements carried out in the time slots constituting the speech or data blocks are representative of the noise or interference level B.

According to the GSM standard, the receiving station supplies the channel allocator equipment 10 and the equipment responsible for the power control and "hand over" algorithms with information representative of the radiofrequency field. This information is provided directly for the mobile direction to base station, transmitted by messages in the opposite direction. This information is repeated every 104 frames and named
RXLEV-FULL-SERVING-CELL. The information is constituted by the average of the measurements made over all the time slots of the channel considered, including the slots for which there is no transmission.

When the base station uses discontinuous transmission, this measurement is not significant, as we do not know the proportion of time slots for which there was no transmission. The GSM standard therefore plans to systematically insert, in the measurement message, an additional field measurement called RXLEV-SUB
SERVING-CELL, consisting of the average of the measurements carried out only on the time slots of the channel for which the transmitting station is prohibited from implementing discontinuous transmission. This partition of time slots is predetermined and immutable. It is known by the base station and by all mobiles.

 In short, only the RXLEV-SUB value (and possibly the corresponding RXQUAL-SUB value if a quality criterion is used) can be used. It comes from a reduced number of samples.

 Two specific, non-limiting solutions will now be described.

 The solution (1) is particularly effective; it requires modifying the structure of the measurement feedback messages from the mobile to the base station.

The solution (2) has the advantage of not modifying in any way the messages, which can remain those defined by the standard
GSM current.

1 ") Process with modification of messages
When the method must be implemented both when the receiving station uses discontinuous transmission and otherwise, the method implies that the receiving station indicates, in the measurement message used by the base station
- instead of RXLEV-FULL, the value of the field measured over all the time slots where there was emission, representative of S + B,
- instead of RXLEV-SUB, the value of the field measured over all the time slots where there was no transmission, representative of B.

 This solution has the advantage of being usable whether the discontinuous DTX transmission is used or not, with simply in the latter case the constraint that B (interference level) is not known.

It is also possible to deviate from the usual constitution only if the DTX is used. When it is not, the message remains unchanged. When the DTX is used, we insert, instead of RXLEV-FULL, the value of the field measured over all the time slots where there was no transmission, representative of the interference level B. RXLEV
SUB remains as defined by GSM and represents S + B.

2 ") Process without modification of messages. With addition of information
The method then simply involves the base station adding information on the measurement feedback messages so that the channel allocator has, for both directions of communication
- RXLEV-FULL (average of the measurements over all the time slots of the channel, including those for which there is no transmission),
- from RXLEV-SUB, which represents S + B.

 We then have S + B and B for each direction.

Base station to mobile.

 The base station indicates, at each measurement message feedback period, the number of time slots R for which it has not transmitted.

The number of time slots in each period being equal to 104 (or 52 for half-speed channels), RXLEV
FULL, measured by the mobile represents
RXLEV-FULL = (104-R) * (S + B) + R * B (1)
104
RXLEV-SUB is measured over a predetermined number of time slots Q where there is necessarily transmission by the base station; So we have
RXLEV-SUB = 9 * (S + B) (2)
Q
From the values 1) and 2) the noise level B is determined; since S + B is available we know how to determine the S + B / B ratio in the base to mobile direction.

Mobile to base station
The base station counts the number of time slots R 'for which the mobile does not transmit, using the SID information. We can then proceed as in the mobile direction towards base station and write
RXLEV-FULL = (104-R ') * (S + B) + R' * B (lbis)
104
Since we know RXLEV-SUB, measured on a predetermined number Q '(Q' being equal to Q in the case of GSM) of time slots where there is necessarily transmission by the mobile station, we have
RXLEV-SUB = Q '* (S + B) (2bis)
Q '
From (Ibis) and (2bis) we know how to determine the noise level B; since we have S + B, we know how to determine the S + B / B ratio in the moving direction towards base.

 From this information, the classification of the channels can be carried out in a reliable manner, the assignment being then carried out in a known manner.

Claims (5)

 1. Method for evaluating the radio transmission channels between a base station and mobile stations of a time division multiple access communication installation, according to which  - at the base station, noise and interference B are measured on the inactive channels of reception by this station,  - at the base station and at the mobile stations, the field level (and possibly the error rate) on the active channels is measured, the measurements made at the mobile stations being transmitted to the base station,  characterized in that noise and interference are also measured at the base station and at the mobile stations during the time slots of the active reception channels which are inactive and without signal carrier.  2. Method according to claim 1, characterized in that the measurements are used to determine the actual signal level, the actual noise and interference level, therefore the actual signal-to-noise ratio and in that: the channels, as and when they are available, i transmit the transmit power to the base station and to the mobile stations, and the transfers are carried out according to the results of use of said measurements.  3. Method according to claim 1 or 2, characterized in that, the installation being in accordance with the GSM standard, and upon cessation of transmission in the absence of activity, the representative field level on the active channels is defined by the average of measurements carried out only on predetermined and immutable time slots of the channel for which the transmitting station is prohibited from implementing discontinuous transmission.  4. Method according to claim 3, characterized in that, the receiving station (base station or mobile) using discontinuous transmission, said receiving station communicates, in measurement messages used by the base station  - the value of the field measured over all the time slots in which there was emission,  - the value of the field measured over all the time slots where there was no transmission.  5. Method according to claim 4, characterized in that the base station is intended to provide cyclically over a known period  - the average of the field measurements over all the time slots of the channel, including those for which there is no transmission for a determined period;  - the average of the field measurements on the only time slots where there is emission;  - the number of time slots R for which there was no transmission, for each direction and to determine the noise level B and the signal-to-noise ratio from this information.