WO1998059515A1

WO1998059515A1 - Improvements in, or relating to, mobile radio telephony

Info

Publication number: WO1998059515A1
Application number: PCT/SE1998/001157
Authority: WO
Inventors: Michael Andersin
Original assignee: Telia Ab (Publ)
Priority date: 1997-06-23
Filing date: 1998-06-16
Publication date: 1998-12-30
Also published as: SE9702390D0; SE9702390L; SE517660C2

Abstract

The present invention provides, among other things, a method and algorithm for measuring a mobile transceiver's speed, based on the effects Rayleigh fading on a received radio signal transmitted between a base station and a mobile transceiver. When a mobile transceiver moves rapidly, the received signal will vary rapidly. When a mobile transceiver moves slowly, the received signal will vary slowly. By measuring the fluctuations in the received signal intensity, the speed of a mobile unit can be measured. The algorithm used in the present invention is based on a short term correlation between the fast multi-path fading and the speed of a mobile transceiver relative to its current base station. The present invention has particular application in cellular environments where the fast fading can be modelled according to a Rayleigh distribution. It is possible to exploit the mobile's speed for resource management decision.

Description

Improvements in, or Relating to, Mobile Radio Telephony

The present invention relates to measurement of the speed of a mobile transceiver, mobile radio systems adapted to measure the speed of a mobile transceiver, base stations for use in such mobile radio systems, and transceivers for use with such mobile radio systems

In mobile radio systems, such as NMT and GSM, it is advantageous for the system operator to know a mobile transceiver's instantaneous speed, since the speed can be used in various system applications, for example, in the determination of decision variables for handover functions and transmitter power surveillance

The present invention provides, among other things, a method and algorithm for measuring a mobile transceiver's speed, based on the effects of Rayleigh fading on a received radio signal transmitted between a base station and a mobile transceiver When a mobile transceiver moves rapidly, the received signal will vary rapidly When a mobile transceiver moves slowly, the received signal will vary slowly By measuring the fluctuations in the received signal intensity, the speed of a mobile unit can be measured

The present invention has particular application to modem cellular radiotelephony systems, e g GSM.

In contrast with known techniques for deducing the speed of a mobile transceiver from the properties of a received radio signal, the present invention employs an algorithm which works directly with the received signal power This implies that the algorithm is both simpler and of lower complexity than previously available algorithms used for this purpose

Future mobile radio cellular telephony systems, e g the 3rd generation UMTS systems, will depend on fast and powerful resource allocation to guarantee an efficient usage of the scarce radio spectrum The speed of mobile transceivers, or mobiles, is a useful parameter for decision criteria associated with resource management functions, e g handover and channel allocat'on For example, the deployment of Hierarchical Cellular Systems (HCS) will be a key design feature for enhancing system capacity and improving communication quality In order to reduce the signalling load while maintaining a high capacity in such HCS systems, it is important to assign highly mobile terminals to wide coverage macro-cells and slower moving users (terminals), like pedestrians, to micro-cells The mobile speed is a natural selection criterion in this case Furthermore, several power control algorithms have been deπved, see, for example

M Andersin and Z Rosberg "Time Variant Power Control in Cellular Networks" Technical Report, TRITA-83-RST-9603, ISSN 1400-9137, ISRN KTH/RST/R-96/03- SE, Radio Communication Systems, Department of Signals, Sensors and Systems, Royal Institute of Technology, Stockholm, Sweden, March 1996, and

Z Rosberg "Fast Power Control in Cellular Networks Based on Short-Term Correlation of Rayleigh Fading" Technical Report, TRITA-S3-RST-9607, ISSN 1400-9137, ISRN KTH/RST/R-96/07- SE, Radio Communication Systems, Department of Signals, Sensors and Systems, Royal Institute of Technology, Stockholm, Sweden, Sept 1996,

in which it has been assumed that the mobile speed is known for the power update commands

It is, therefore, important to find techniques that can be used to estimate a mobile's speed accurately and quickly, bearing in mind the rapidly fluctuating radio propagation conditions Some studies have considered this problem

In the present invention, a mobile speed estimation algorithm is used which is based on a short term correlation between the fast multi-path fading and the speed of a mobile transceiver relative to its current base station The present invention has particular application in cellular environments where the fast fading _> - can be modelled according to a Rayleigh distribution. Numerical results for the present invention have shown that the mobile's speed can be estimated fairly accurately in a very short period of time. It is, therefore, possible to exploit the mobile's speed for resource management decisions. Furthermore, it can be demonstrated that the results obtained from the present invention are superior to the results obtained by using the algorithms described in A. Sampath and J. M. Holtzman. "Estimation of Maximum Doppler Frequency for Handoff Decisions" Proceedings. IEEE 43rd Vehicular Technology Conference, VTC-93, pp. 859-862, Secaucus, New Jersey, May 1993.

According to a first aspect of the present invention, there is provided a mobile radio system having at least one base station and at least one mobile transceiver characterised in that means are provided for estimating the speed of said mobile transceiver, said means including measurement means for measuring fluctuations in signal power received at said base stations and/or said mobile transceivers, and processor means for deriving a short term correlation of fast multi-path fading and thence deriving an estimation of said mobile transceiver's speed.

Said fast multi-path fading may be Rayleigh fading.

A link gain for a radio signal transmitted between a base station and a mobile transceiver may be modelled as a product of a distance dependent propagation loss for a slow fading component and a fast multi-path fading component.

Said link gains may be normalized over a time interval of a duration sufficiently short to ensure that said slow fading component remains substantially constant.

An estimation of a correlation function dependent on a mobile transceiver's speed may be derived.

Said processor means may operate on an algorithm comprising the following steps:

sampling the power of a received signal over a time interval T, such that the samples r(t_k), k = 1 , 2V, fulfill the relation t_k-t_k_ =t, k=2,4 ,2N

normalising the sampled signal over the time interval T, forming the normalized samples rχt_k),according to the formula

forming N numbers from the measurements according to:

X( =W -^ .)² . *=2, 4, ... , 2/V

estimating the correlation function from the formula

and estimating the mobile transceiver's speed v as a value that yields p.

A mobile's speed may be derived by determining a root v to the equation p = p(t;v) by a numerical method.

A mobile's speed may be derived by determining a root v to the equation p = p(t;v) by using the Newton-Raphsons method.

Precalculated values of p(t;v) may be stored in a look-up table held in said processor means and said mobile transceiver's speed may be determined by comparing a value of p{t;v) derived from measurement with a stored value p(t;v). A mobile transceiver's speed may be estimated by an interpolation process between closest fit comparisons between measured and stored values of p{t;v).

Said mobile radio system may be a GSM system.

Said mobile radio system may be an NMT system.

Said mobile radio system may be a UMTS system.

Said mobile radio system may employ a Hierarchical Cellular System in which mobiles travelling at a substantial speed are assigned to wide coverage macro-cells and slower moving mobiles are assigned to micro-cells.

Said derived mobile speed may be used as a decision parameter in channel allocation.

Said derived mobile speed may be used as a decision parameter in handover procedures.

Said derived mobile speed may be used as a decision parameter in transmitter power surveillance.

According to a second aspect of the present invention, there is provided, in a mobile radio system having at least one base station and at least one mobile transceiver, a method of estimating a mobile transceiver's speed, characterised by measuring fluctuations in signal power received at said base stations and/or said mobile transceivers and deriving a short term correlation of fast multi-path fading and thence deriving an estimation of said mobile transceiver's speed.

Said fast multi-path fading being Rayleigh fading.

A link gain for a radio signal transmitted between a base station and a mobile transceiver may be modelled as a product of a distance dependent propagation loss for a slow fading component and a fεst multi-path fading component.

Said link gains may be normalized over a time interval of a duration sufficiently short so to ensure said slow fading component remains substantially constant.

The method may include the following steps:

sampling the power of a received signal over a time interval T, such that the samples r(t_k), k = 1 2V, fulfill the relation

Wι=^f. fr⁼²»⁴ >^2N

normalising the sampled signal over the time interval T, forming the normalized samples f(t_k), according to the formula

^)-~^L- ^• *"¹ - ^2N

forming N numbers from the measurements according to:

estimating the correlation function from the formula

and estimating the mobile transceiver's speed v as the value that yields p. A mobile's speed may be derived by determining a root v to the equation p = p(t;v) by a numerical method

A mobile's speed may be derived by determining a root v to the equation p = p(t;v) using the Newton-Raphsons method

Precalculated values of p(t;v) may be stored in a look-up table and determining said mobile transceiver's speed by comparing a value of p(t;v) derived from measurement with a stored value p(t;v)

A mobile's speed may be estimated from an interpolation process between closest fit comparisons between measured and stored values of p(t;v)

According to a third aspect of the present invention, there is provided a base station for use in a system as set forth in any preceding paragraph, characterised in that said base station includes means for estimating the speed of a mobile transceiver, said means including measurement means for measuπng fluctuations in signal power received at said base stations and processor means operating an algoπthm for deπving a short term correlation of fast multi-path fading and thence deriving an estimation of said mobile transceiver's speed

Said base station may be adapted to operate a method as set forth in any preceding paragraph

According to a fourth aspect of the present invention, there is provided a mobile transceiver for use with a system as set forth in any preceding paragraph, characterised in that said mobile transceiver includes means for estimating its speed relative to a base station, said means including measurement means for measuring fluctuations in signal power received at said mobile transceiver and processor means operating an algorithm for deriving a short term correlation of fast multi-path fading and thence deriving an estimation of said mobile transceiver's speed

Said mobile transceiver may be adapted to operate a method as set forth in any preceding paragraph.

Embodiments of the present invention will now be described, by way of example, with reference to the accompanying drawings in which:

Figure 1 is a schematic illustration of a mobile radio system of the type in which the present invention can be implemented.

Figure 2a is a diagram illustrating fluctuations in radio power received by a base station and/or a mobile transceiver resulting from Rayleigh fading where the relative speed between base station and mobile transceiver is high.

Figure 2b is a diagram illustrating fluctuations in radio power received by a base station and/or a mobile transceiver resulting from Rayleigh fading where the relative speed between base station and mobile transceiver is low.

Figure 3 is a diagram schematically illustrating the circuitry used for estimating a mobile transceivers speed according to the present invention

In order to appreciate the operation of the present invention, it is necessary to discuss the system model for radio propagation on which it is based. The system model, described herein, is based on the system model in Z. Rosberg. "Fast Power Control in Cellular Networks Based on Short-Term Correlation of Rayleigh Fading". Technical Report, TRITA-S3-RST-9607, ISSN 1400-9137, ISRN KTH/RST/R-96/07-SE, Radio Communication Systems, Department of Signals, Sensors and Systems, Royal Institute of Technology, Stockholm, Sweden, Sept.1996.

Consider a generic radio link in a cellular system with an arbitrary transmitter and receiver. For the uplink case, the transmitter is a mobile transceiver and the receiver is its corresponding base station. For the downlink case, the roles of mobile transceiver and base station are reversed. When the transmitter is transmitting at time t, it uses a power p(t). Given that at time t, the link gain between transmitter and receiver is g(t), the received power r(t) equals r t)=g(t).p(t) (1)

Note that the receiver noise and the interference is omitted in this analysis. This limitation is discussed, in more detail, later in this specitication.

A conventional model is used for the link gain g(t), see for example, the following references:

W. C. Jakes, Jr. "Microwave Mobile Communications". John Wiley & Sons, New York, 1974.

W. C. Y Lee. "Mobile Communications Engineering". McGraw-Hill, New York, 1982.

Z. Rosberg. "Fast Power Control in Cellular Networks Based on Short-Term Correlation of Rayleigh Fading". Technical Report, TRITA-S3-RST-9607, ISSN 1400-9137, ISRN KTH/RST/R-96/07-SE, Radio Communication Systems, Department of Signals, Sensors and Systems, Royal Institute of Technology, Stockholm, Sweden, Sept.1996.

A. Sampath and J. M. Holtzman. "Estimation of Maximum Doppler Frequency for Handoff Decisions". In Proceedings. IEEE 43rd Vehicular Technology Conference, VTC-93, pp. 859-862, Secaucus, New Jersey, May 1993.

For every time instant t, the link gain g(t) is modelled as a product of a distance dependent propagation loss for a slow fading component and a fast multi- path fading component. That is, g(t)=L(t).S(t).R²(t) (2)

Assume that L(t) = D^"°(t), where D(t) is the distance between the transmitter and the receiver at time t, and α is a propagation constant. The shadow, or slow, fading component S(t) is assumed to be log-normally distributed with a log-mean of 0 dB, and a log-variance of or², dB. The third factor R²(t) is the power attenuation due to fast multi-path fading. The envelope of R(t) usually follows a Rayleigh distribution, implying many scattered signals of comparable strengths, see:

W. C. Jakes, Jr. "Microwave Mobile Communications". John Wiley & Sons, New York, 1974; and

W. C. Y Lee. "Mobile Communications Engineerinc". McGraw-Hill, New York, 1982.

This also holds in the case of non-line of sight propagation in a micro- cellular environment. However, whenever the user is in-line of sight with the base station, the direct path will tend to dominate in strength, and the envelope distribution .will be more typically Rican, see R. J. C. Bultitude and G. K. Bedal. "Propagation Characteristics on Micro-cellular Urban Mobile Radio Channels at 910 MHZ" IEEE Journal on Selected Areas in Communications, Vol.7, No.1 , Jan.1989. The variable R(t) is assumed to have a standard deviation, o„ = 1/Y2 that is, E(R²(t)) = 1.

Let v be the mobile speed, which is to be estimated, and t₀ be an arbitrary time reference. For the time variant Rayleigh fading process, the Jake's model is used, see:

W. C. Jakes, Jr. "Microwave Mobile Communications". John Wiley & Sons, New York, 1974;

R. H. Clarke, "A Statistical Theory of Mobile Radio Reception". Bell Systems Technical Journal, Vol.47, pp.957-1000, July 1968; and

J. D. Parsons and A. M.D. Turkmani. "Characterization of Mobile Radio signals: Model Description". IEE Proceedings-I, Vol. 138, pp.549-556, Dec.1991 The Jake's model has been shown to agree well with field experimental data. Let {Z^(k)(t)}, where k = 1, or 2, be two independent Gaussian processes with stationary means and variances, 0 and f_R respectively. Each process evolves independently of the rest of the system according to the following process - for every time reference to and time distance t>0,

z < +0 =P(* ; ).z<*>(y ^1 -P²(t; v).≠^k)(t), *=1,2 (3)

where 0^k t)} are independent normal random variables with mean 0 and variance 0 and p(t;v) is the zeroth order Bessel function of the first kind evaluated at (2π.v.f.t)/c. That is,

P(';v)=J₀( ) (4)

Here, f is the carrier frequency and c is the speed of light. The quantity f_d = (v.f)/c is termed Maximum Doppler Frequency, see A. Sampath and J. M. Holtzman. "Estimation of Maximum Doppler Frequency for Handoff Decisions", in Proceedings. IEEE 43rd Vehicular Technology Conference, V7C-93, pp. 859-862, Secaucus, New Jersey, May 1993.

It should be observed that the Rayleigh process {R(t)| ; 0} is the envelope of the complex correlated Gaussian process {Z^t) + i.Z²|t-tO] whose evolution is modulated by the correlated process defined in (3).

From the Rayleigh power definition and (3) we have,

₊2p²(f;v)_v ^/1 -p²(t;v) (Z'¹>(f₀) ₊Z<²>(f₀).Z<²⁾(t)) (5)

where {R⁽²⁾(t)} are independent exponential random variables with mean 1. This expression is used to derive the mobile speed estimator. The speed estimation algorithm, of the present invention, is based on the system model described above and the findings set out in Z. Rosberg. "Fast Power Control in Cellular Networks Based on Short-Term Correlation of Rayleigh Fading". Technical Report, TRITA-S3-RST-9607, ISSN 1400-9137, ISRN KTH/RST/R- 96/07-SE, Radio Communication Systems, Department of Signals, Sensors and Systems, Royal Institute of Technology, Stockholm, Sweden, Sept.1996.

Consider the received signal power in (1). It is assumed that the transmitted power p(t) is known by the receiver. This is a reasonable assumption in the downlink, where mobile terminals know the transmitted power on the beacon pilot signal. This implies that the receiver also knows the link gain g(t).

The object of the algorithm of the present invention is to estimate the function in (4) (which is dependent of the mobile speed) based on measurements of the received signal power. This means that the Rayleigh fading factor {R (t)|t-_0} must be found. This, in turn, requires that the distance dependent and the slow shadow fading component L(t).S(t) be removed from the link gain g(t). The algorithm can, therefore, be divided into two steps

The first step is to remove the component L(t). S(t) in (2). It has been shown in M. Andersin and Z. Rosberg. "Time Variant Power Control in Cellular Networks". Technical Report, TRITA-83-RST-9603, ISSN 1400-9137, ISRN KTH/RST/R-96/03- SE, Radio Communication Systems, Department of Signals, Sensors and Systems, Royal Institute of Technology, Stockholm, Sweden, March 1996, that the distance dependent, and the slow shadow, fading factors have the following asymptotic representations:

S(t.+f) =S(f₀)(1 ₊a.{ut)^V2.N(t))Λ0°^ ) (7)

where a = (σ_s/10)ln10 and o(x) is a function with the property that lim_x__, o(x)/x = 0. The variables {N(t)} are independent standard normal random variables, u = 2v/X is the normalized speed, and the parameter X is the effective correlation distance of the slow shadow fading. From (5), (6) and (7) it foliows that for a short time horizon, T, the distance dependent and the slow shadow fading time variation is negligible compared to the fast Rayleigh fading variation. Therefore, the link gains {g(t)} can be normalized over this time interval, yielding the Rayleigh fading component R²(t). From now on it will be assumed this component is known.

It is necessary to investigate the deviations of the Rayleigh fading component, because it is expected that the size of the deviations will increase as the mobile speed v, or alternatively, the Doppler frequency, increases. Since a deviation between two samples may both be negative and positive, the squared difference between two samples is taken. The samples are separated in time with a distance t. Let {X(t_k)}^N _k=ι be a sequence of such measurements, i.e.

X(t_kHR²(t_k+t)-R²(t_k))², =1, N (8)

Note that the time between the measurements X(t_k) can be chosen independently of the time interval t. From (5) and (8), it follows that

Hence, by forming a sample average over the measurements {X(t_k)}^N _k=1 an estimate of the correlation function p(t;v) can be obtained which is dependent on the mobile speed. Note that this function is expressed in terms of a Bessel function. This implies that for a given time interval t<t', there is a one-to-one relation between the function p(t;v) and the mobile speed v, as long as the speed is below a certain threshold v < V. Consequently, the mobile speed can be uniquely determined.

The time interval T plays an important role in the specification of the algorithm. The larger T is. the more samples can be obtained. On the other hand, the larger T is, the less valid is the assumption of constant distance dependent, and slow shadow, fading components. Hence, there exists a natural trade-off for the time interval T. It can be demonstrated that only a small number of measurements are required to obtain a fairly accurate estimate of the mobile speed.

The mobile Speed Estimation Algorithm (SEA) can now be formulated as follows.

1) Sample the power of the received signal over a time interval T, such that the samples r(t_k), k = 1 2V, fulfill the relation t_k-t_k_ =t, k=2,4,....,2N

2) Normalise the sampled signal over the time interval T, i.e. form the normalized samples f(t_k), according to

3) Form N numbers from the measurements, according to (8), i.e.

Denote {X(t,)}^N _i=1 as the resultant sequence of numbers.

4) The correlation function in (4) is estimated using (9),

/V/.1

5) The estimated mobile speed v is obtained as the value that yields p according to (4)

To find the mobile speed in step 5) is equivalent to finding the root v to the equation p = p(t;v). This can be solved numerically, for instance using the Newton-Raphsons method.

If the component L(t).S(t) is constant over the time interval T, then v -v as N - ~.

The SEΞA algorithm proposed above, is a very simple algorithm with little complexity. The mobile speed is found with the help of a standard root finding method. These are usually of an iterative nature where the number of steps to converge to a final solution cannot he determined in advance. However, since the time interval t is fixed, this means that this problem can be overcome by simply calculating the correlation function p(t;v) in advance for different values of the mobile speed v. These values can then be stored in a look-up table. The number of elements in the table is dependent on the required granularity of the estimates. For example, if it is sufficient to divide the speed into 5 km/h intervals, it only require 30' elements to span a speed interval from 0 to 150 km/h. It is also possible to make a linear interpolation between two consecutive values of the mobile speed to get a more accurate result. That is, the algorithm can easily be implemented even in handheld mobile units.

The proposed SEΞA algorithm is general, in the sense that it can both be applied in the uplink and the downlink. In the uplink, the base station is required to know the transmission power of the mobile. In the GSM system, the base station is in charge for the power update commands and consequently knows the transmission power of the mobile.

The mobile speed is a useful decision variable for efficient resource management decisions. The present invention provides a fast and simple speed estimator that is based on short term correlation of fast multi-path fading. Numerical results show that it gives fairly accurate estimates of the mobile speed in a very short period of time.

One crucial assumption, in the derivation of the algorithm of the present invention, is that fast multipath fading is modelled according to a Rayleigh fading process In some environments, this may not be true In these cases, the ideas behind the SEA algorithm, can still be used There will not be any significant difference except that some of the expressions will have a sl'ghtly different form In particular, the expression in (9) will be slightly more complex It should be noted, that it is necessary for the required fading characteristics to be known for the environment under consideration.

The analysis presented in this patent specification has neglected the effect of interference and receiver noise However, in contemporary interference limited cellular systems, the received signal from the intended transmitter is much larger than the interference It can, therefore, be expected that the interference, duπng the short estimation interval, will not change dramatically and consequently will be cancelled out through the normalization process The fact that the interference does not degrade the performance of this type of estimator has been verified in A. Sampath and J M Holtzman "Estimation of Maximum Doppler Frequency for Handoff Decisions" In Proceedings IEEE 43rd Vehicular Technology Conference, VTC-93, pp 859-862, Secaucus, New Jersey, May 1993

In the analysis herein set forth, it has been assumed that the received signal is not subject to any severe inter-symbol interference, implying a very small delay spread However, in the case of non-negligible delay spread, the algoπthm can still he used In that case, the estimates of the channel taps from the equalizer are used The individual channel taps are subject to multipath fading and it is only required to take samples from one of them (preferably the strongest)

It is believed that the present application also has application to CDMA and OFDM based cellular systems

Turning now to Figure 1 , there is shown a mobile radio communications system having a plurality of base stations, two of which are illustrated at 1 and 2 A plurality of mobile transceivers, 3 to 7, transmit to the base stations At any given time, each of said mobile transceivers communicates with a single base station The mobile transceivers move at different speeds relative to the base stations, for example the mobile transceiver 7, mounted in a car, will clearly be moving substantially faster than a mobile transceiver 3, carried by a pedestrian. This means that handover procedures must be provided as a mobile transceiver moves from the coverage area of one base station to the coverage area of another base station. As previously stated, the speed of a mobile transceiver relative a base station is an important parameter for handover decisions between base stations. Thus, it is clearly important for the speed of a mobile transceiver, relative to a base station, to be determined. The measurement of speed may be carried out either in the base station, or in the mobile transceiver.

The present invention, as described above, uses the effect of Rayleigh fading on the fluctuations of a received signal in order to measure the relative speed between a mobile transceiver and a base station. Figures 2a and 2b illustrate the effect of Rayleigh fading on radio signals received from, or transmitted by, mobile transceivers travelling at different speeds. Figure 2a refers to a mobile transceiver travelling at a speed substantially less than the speed of the mobile, transceiver to which Figure 2b relates. As can be clearly seen, although the fluctuations in both Figure 2a and 2b are essentially random, the rate of fluctuations are greater in Figure 2b than Figure 2a. The present invention makes use of this phenomenon to measure the speed of mobile relative to a base station.

The structure of the apparatus used to measure the speed of a mobile transceiver relative to a base station is schematically illustrated in Figure 3. It should be noted that speed measurements may be made either in the mobile transceiver, or the base station, or both. An incoming radio signal is picked up by the arial and passed to the receiver. The power meter measures the received radio power, which fluctuates as a result of both slow fading and fast fading, as explained in detail above. Power measurements are sampled and quantised and passed to the processor, which implements the algorithm described above. As previously explained, the algorithm of the present invention requires that the root of the equation p = p(t;v) be found. This may be done by storing values of p(t;v), as a function of v in a look-up table and comparing a value for p(t;v) derived by the processor with values in the look-up table and extracting the corresponding values of v held in the look-up table. The results of this comparison can then be passed to an interpolator which estimates the actual speed of a mobile transceiver from the two closest comparisons. The estimated value of the mobile transceivers speed is then passed to a resource management unit responsible for controllong such functions as handover and power surveillance.

Claims

CLAI S

1 A mobile radio system having at least one base station and at least one mobile transceiver characterised in that means are provided for estimating the speed of said mobile transceiver, said means including measurement means for measu╧Çng fluctuations in signal power received at said base stations and/or said mobile transceivers, and processor means for deriving a short term correlation of fast multi-path fading and thence de╧Çving an estimation of said mobile transceiver's speed

2 A mobile radio system, as claimed in claim 1, characterised in that said fast multi-path fading is Rayleigh fading

3 A mobile radio system, as claimed in either claim 1 , oi 2, characte╧Çsed in that a link gain for a radio signal transmitted between a base station and a mobile transceiver is modelled as a product of a distance dependent propagation loss for a slow fading component and a fast multi-path fading component

4 A mobile radio system, as claimed in claim 3, character.sed in that said link gains are normalized over a time interval of a duration sufficiently short to ensure that said slow fading component remains substantially constant

5 A mobile radio system, as claimed in any previous claim, characte╧Çsed in that an estimation of a correlation function dependent on a mobile transceiver's speed is derived

6 A mobile radio system, as claimed in any previous claim, characte╧Çsed in that said processor means operates on an algorithm comprising the following steps

sampling the power of a received signal over a time .nterval T, such that the samples r(t_k), k = 1 2V, fulfill the relation - t_k-t_k_ t, k=2,4, 2/V

forming N numbers from the measurements according to:

X(t_k)=(╬╗t_k)-╬╗t_k.,))² , *=2, 4 2╬¢/

estimating the correlation function from the formula

and estimating the mobile transceiver's speed v as a value that yields p.

7. A mobile radio system, as claimed in claim 6, characterised in that a mobile's speed is derived by determining a root v to the equation p = p(t;v) by a numerical method.

8. A mobile radio system, as claimed in claim 7, characterised in that a mobile's speed is derived by determining a root v to the equation p = p(t;v) by using the Newton-Raphsons method.

9. A mobile radio system, as claimed in claim 7, characterised in that precalculated values of p(t;v) are stored in a look-up table held in said processor means and said mobile transceiver's speed is determined by comparing a value of p(t;v) derived from measurement with a stored value p(t;v}.

10 A mobile radio system, as claimed in claim 9, characterised in that a mobile transceiver's speed is estimated by an interpolation process between closest fit comparisons between measured and stored values of p(t;v}.

11. A mobile radio system, as claimed in any previous claim, characterised in that said mobile radio system is a GSM system.

12. A mobile radio system, as claimed in any of claims 1 to 10, characterised in that said mobile radio system is an NMT system.

13. A mobile radio system, as claimed in any of claims 1 to 10, characterised in that said mobile radio system is a UMTS system.

14. A mobile radio system, as claimed in any previous claim, characterised in that said mobile radio system employs a Hierarchical Cellular System in which mobiles travelling at a substantial speed are assigned to wide coverage macro- cells and slower moving mobiles are assigned to micro-cells.

15. A mobile radio system, as claimed in any previous claim, characterised in that said derived mobile speed is used as a decision parameter in channel allocation.

16. A mobile radio system, as claimed in any previous claim, characterised in that said derived mobile speed is used as a decision parameter in handover procedures.

17. A mobile radio system, as claimed in any previous claim, characterised in that said derived mobile speed is used as a decision parameter in transmitter power surveillance.

18. In a mobile radio system having at least one base station and at least one mobile transceiver, a method of estimating a mobile transceiver's speed, characterised by measuring fluctuations in signal power received at said base stations and/or said mobile transceivers and deriving a short term correlation of fast multi-path fading and thence deriving an estimation of said mobile transceiver's speed.

19. A method, as claimed in claim 18, characterised by said fast multi-path fading being Rayleigh fading.

20. A method, as claimed in either claim 18, or 19, characterised by modelling a link gain for a radio signal transmitted between a base station and a mobile transceiver as a product of a distance dependent propagation loss for a slow fading component and a fast multi-path fading component.

21. A method, as claimed in claim 20, characterised by normalizing said link gains over a time interval of a duration sufficiently short so to ensure said slow fading component remains substantially constant.

22. A method, as claimed in any of claims 18 to 21, characterised by deriving an estimation of a correlation function dependent on a mobile transceiver's speed.

23. A method, as claimed in any of claims 18 to 21 , characterised by the following steps:

sampling the power of a received signal over a time interval T, such that the samples r(t_k), k = 1, 2V, fulfill the relation

normalising the sampled signal over the time interval T, forming the normalized samples r(t_k), according to the formula

forming N numbers from the measurements according to:

estimating the correlation function from the formula

and estimating the mobile transceiver's speed v as the value that yields p.

24. A method, as claimed in claim 23, characterised by deriving a mobile's speed by determining a root v to the equation p = p(t;v) by a numerical method.

25. A method, as claimed in claim 24, characterised by deriving a mobile's speed by determining a root v to the equation p = p(t;v) using the Newton- Raphsons method.

26. A method, as claimed in claim 24, characterised by storing precalculated values of p(t;v) in a look-up table and determining said mobile transceiver's speed by comparing a value of p(t;v) derived from measurement with a stored value P(t;v).

27. A method radio system, as claimed in claim 26, characterised by estimating a mobile's speed from an interpolation process between closest fit comparisons between measured and stored values of p(t;v).

28. A method, as claimed in any of claims 18 to 27, characterised by said mobile radio system being a GSM system.

29. A method, as claimed in any of claims 18 to 28, characterised by using said derived mobile speed as a decision parameter in channel allocation.

30. A method, as claimed in any of claims 18 to 29, characterised by using said derived mobile speed as a decision parameter in handover procedures.

31. A method, as claimed in any of claims 18 to 30, characterised by using said derived mobile speed as a decision parameter in transmitter power surveillance.

32. A base station for use in a system as claimed in any of claims 1 to 17, characterised in that said base station includes means for estimating the speed of a mobile transceiver, said means including measurement means for measuring fluctuations in signal power received at said base station and processor means for deriving a short term correlation of fast multi-path fading and thence deriving an estimation of said mobile transceiver's speed.

33. A base station, as claimed in claim 32, characterised in that said base station is adapted to operate the method claimed in any of claims 18 to 31.

34. A mobile transceiver for use with a system as claimed in any of claims 1 to 1 , characterised in that said mobile transceiver includes means for estimating its speed relative to a base station, said means including measurement means for measuring fluctuations in signal power received at said mobile transceiver and processor means for deriving a short term correlation of fast multi-path fading and thence deriving an estimation of said mobile transceiver's speed.

35. A mobile transceiver, as claimed in claim 34, characterised in that said mobile transceiver is adapted to operate the method claimed in any of claims 18 to 31.