HK1066864B - Apparatus and method of velocity estimation - Google Patents

Description

Apparatus and method for rate estimation

Background

I. Field of the invention

The present invention relates to wireless communications, and more particularly, to a novel and improved apparatus and method for rate estimation using AGC information.

Description of the related Art

Wireless devices utilize radio waves to provide long range communications without the physical constraints of a wired-based system. Information is provided to the device using radio waves transmitted over a predetermined frequency band. The allocation of the available spectrum is adjusted to enable many users to enter into communication without excessive interference.

A remote receiver tuned to a carrier frequency needs to receive and demodulate signals transmitted from the respective transmitters on the same carrier frequency. The remote receiver recovers a baseband signal from the modulated carrier. The baseband signal may be presented directly to the user or further processed before being presented to the user.

Portable wireless communication devices incorporating both a transmitter and a receiver are used to provide two-way communication. An example of a portable wireless communication device, commonly referred to as a mobile unit, is a radiotelephone. Radiotelephones may form part of a wireless communication system, such as those defined by the Telecommunication Industry Association (TIA)/Electronic Industry Association (EIA) IS-95-B, Mobile STATION-BASE STATION COMPATIBILITY request FOR Dual-mode communication System, and the Wireless STATION-BASE STATION COMPATIBILITY request FOR 1.8 TO 2.0 GHZ DIVIDE MULTIPLE Access System (CDMA), as defined by the American National STANDARDs Institute (ANSI) J-STD-008. The wireless telephones used in the two systems described above must meet the STANDARDS TIA/EIA IS-98-B, RECOMMENDMENDDIMINUM PERFORCE STANDARDS FOR DUAL-MODE SPREAD SPECTRUM CELLULAR MOBILE STATIONS, and ANSI J-STD-018, RECOMMENDED MINUM PERFORCE REQUERNITS SFOR 1.8 TO 2.0 GHZ CODE DIVISION MULTIPLE ACCESS (CDMA) PERSONAL STATIONS, respectively.

Radio receivers operate in relatively hostile environments. A radio signal propagating from a transmitter to a corresponding receiver is scattered and reflected by obstacles and structures surrounding the transmitter and receiver. Structures, such as buildings, and surrounding terrain, such as walls and hills, cause scattering and reflection of the transmitted signal. Scattering and reflection of the transmitted signal results in multiple signal paths from the transmitter to the receiver. The obstacles that affect multiple signal paths are centered on the receiver with a radius proportional to the received signal wavelength. The cause for multiple signal paths changes as the receiver moves.

The incident signal at the receiver antenna is the sum of all multipath signals resulting from the scattering and reflection of the signal from the transmitter to the receiver. The resultant received signal can be modeled as having two components.

The first component is referred to as shadowing, slow fading, log normal fading, or long-term fading. Slow fading is caused by the terrain profile between the transmitter and receiver or as a result of the receiver passing inside a tunnel, under a bridge, or behind a building. The received power measured at any particular location varies over time due to slow fading effects. The received power measured due to the slow fading component is lognormally distributed.

The second signal component is referred to as fast fading, multipath fading, short term fading, or rayleigh fading. Fast fading is caused by reflections and scattering of the transmitted signal caused by obstacles in the transmission path, such as trees, buildings, vehicles, and other structures. Fast fading results in fading of the entire reception bandwidth, where the signals arriving at the receiver combine destructively.

The incident signal at the receiver is synthesized by superimposing a fast fading signal on a slow fading signal. As a result, mobile radio receivers can vary greatly in received signal strength. In addition, the moving receiver is also subject to frequency drift in the received signal. One cause of frequency drift is doppler shift, which causes the frequency offset of the received signal to be proportional to the velocity of the receiver relative to the transmitter.

Mobile radio receivers, such as mobile telephones operating in IS-95 or J-STD-008 communication systems, experience signal fading and frequency doppler shift as a conventional part of their operating environment. Mobile receivers incorporate various techniques to compensate for amplitude and frequency variations of incoming signals.

However, the demodulation algorithms of many mobile receivers can be improved if the mobile receivers know their rate. Furthermore, the mobile receiver-aware rate may be used with a position decision algorithm. In addition, the rate of the mobile receiver may be provided as telemetry data to be transmitted to a remote site, or as data available to the user. It is desirable to be able to determine the velocity of a mobile receiver using the incident signal on the receiver. The rate measurement of the mobile receiver needs to be performed without burdening the communication system.

Summary of The Invention

The present embodiments disclose a novel and improved velocity estimator having: a signal processor for extracting a multipath from a received signal; a signal scaler and multiplier for scaling the received signal by a scaling factor which is the inverse of the AGC gain; an instantaneous envelope calculator; a mobile RMS calculator; a level crossing counter for counting the number of times that the instantaneous envelope value passes through a level crossing threshold; and a look-up table mapping the evaluation cross number to a rate estimate. The rate estimator may further incorporate a FIFO for storing a predetermined number of instantaneous envelope values.

When the velocity estimator is implemented in a CDMA wireless communication device, the received signal is a composite CDMA signal, and individual multipaths can be obtained by despreading and accumulating the pilot signal over a predetermined number of chips.

Level crossing counters may introduce hysteresis in the level crossing count by combining a high level threshold and a low level threshold. The high level threshold may be generated as a first predetermined level M dB above half the moving RMS value. The low level threshold may be generated as a second predetermined level N dB below half the moving RMS value. The values M and N may be three in particular embodiments.

The velocity estimator may use a normalization multiplier to generate a normalized value by multiplying the instantaneous envelope value by a normalization factor. In one embodiment, the normalization factor may be 2/(moving RMS value). When using a normalization multiplier, the level crossing counter uses a predetermined level crossing threshold.

Brief Description of Drawings

The features, nature, and advantages of the present invention will become more apparent from the detailed description set forth below when taken in conjunction with the drawings in which like elements have like numerals wherein:

FIG. 1 is a diagram of a wireless communication system;

FIG. 2 is a block diagram of a wireless receiver;

FIG. 3 is a received signal diagram;

FIG. 4 is a block diagram of received signal estimation;

FIG. 5 is a level hysteresis graph;

FIG. 6 is a block diagram of a rate estimation implementation;

FIG. 7 is a block diagram of a rate estimation implementation;

FIG. 8 is a flow chart diagram of a rate estimation method; and

fig. 9 is a flow chart of another embodiment of a method of rate estimation.

Description of The Preferred Embodiment

Fig. 1 shows a block diagram of a wireless communication system in which a mobile receiver implements rate estimation. The radiotelephone system is provided as an exemplary embodiment only. The mobile receiver rate estimation disclosed herein is not limited to implementation in wireless telephone systems or even wireless communication systems. It will be apparent to those skilled in the art that the rate estimation disclosed herein may be implemented within a mobile radio receiver in the field of receiver design.

A mobile telephone 110 operating in a wireless communication system such as the IS-95 or J-STD-008 system communicates with a base station 120 using radio waves. The base station 120 is identified by an antenna, although in reality the base station hardware would not be located directly at the antenna. The base station 120 antenna may be located on a building 122 or on an antenna tower. Although only one base station 120 is shown, the mobile telephone 110 may communicate with more than one base station 120 at the same time. The transmission from the base station 120 to the mobile telephone 110 ideally traverses a single path, but in reality traverses multiple paths.

The terrain or structure 130 may obstruct the signal path from the base station 120 to the mobile phone 110. Shadowing of the structure 130 of the mobile phone 110 results in slow fading variations in the received signal power. Multiple signal paths from the base station 120 to the mobile station 110 occur due to reflections and scattering of the transmitted signal. Other signal paths may occur due to reflections from structures 142, trees 144, and vehicles 146 that are sufficiently close to the mobile phone 110. The object that results in multiple signal paths is centered on the mobile phone 110 with a radius proportional to the received signal wavelength. The local source of the reflected transmitted signal results in a received signal that experiences fast fading. The moving mobile telephone 110 is subject to fast fading signals caused by constant variations in the multiple signal paths. As described below, the mobile telephone 110 incorporates rate estimation from fast fading signals. The velocity estimate of the mobile phone 110 may be used to generate an error correction factor that is used to demodulate the received signal or retransmitted or translated into a display value for the user.

Fig. 2 is a block diagram of a conventional heterodyne receiver 200. The antenna 210 is used to interface the radio receiver 200 with incoming radio waves. The receiver 200 is implemented in conjunction with a corresponding transmitter in the radiotelephone, with which the antenna 210 is also shared. In a radiotelephone, a duplexer is used to couple signals from the antenna 210 to the rest of the receiver, and from the transmitter to the antenna 210. For the sake of simplicity, the duplexer and the corresponding transmitter are not shown in fig. 2.

The output of antenna 210 is coupled to a Low Noise Amplifier (LNA). The LNA 220 behind the antenna 210 is used to amplify the received signal. The LNA 220 is also the primary cause of the receiver noise figure. The noise figure of the LNA 220 adds directly to the noise figure of the receiver, while the noise figure of the subsequent stages decreases in proportion to the LNA 220 gain. Thus, the LNA 220 is selected to provide the minimum noise figure in the receive band while amplifying the receive signal with sufficient gain to minimize the noise figure from subsequent stages.

The signal amplified in the LNA 220 is coupled to an RF filter 224. The RF filter 224 is used to provide rejection of signals outside the receive band. To reduce the effect of the filter on the overall receiver noise figure, an RF filter 224 is used after the LNA 220 stage. The output of the RF filter 224 is coupled to the input of the RF mixer 230.

The RF mixer 230 mixes the amplified received signal with a locally generated frequency signal to down-convert the signal to an Intermediate Frequency (IF). The IF output of the RF mixer 230 is coupled to an IF filter 232. The IF filter 232 is used to pass only IF generated from a single received signal. The IF filter 232 is used to reject signals outside the IF band, particularly adjacent channel signals and undesired mixer products. The IF filter 232 has a much narrower frequency response than the RF filter 224. Since the RF mixer 230 down-converts the desired RF channel to the same IF regardless of the frequency of the RF channel, the IF filter 232 may have a much narrower band. Conversely, since any channel within the receive band can be allocated for the communication link, the RF filter 224 must pass the entire receive band. The output of the IF filter 232 is coupled to a Variable Gain Amplifier (VGA)236 that is used to increase the signal level.

VGA236 is used as part of an Automatic Gain Control (AGC) loop to maintain a constant amplitude within the received signal for subsequent stages. The gain of VGA236 is varied with a control loop that senses the amplitude of the amplifier output. The output from VGA236 is coupled to the input of IF mixer 240.

The IF mixer 240 down-converts the IF signal to a baseband signal. The Local Oscillator (LO) used with the IF mixer 240 is separate and distinct from the first LO. The baseband output of the IF mixer 240 is coupled to a baseband filter 242. Baseband filter 242 is used to low pass the down-converted baseband signal to remove any undesired mixer products and to provide further rejection of any adjacent channel signals not previously rejected by IF filter 232. The output of the baseband filter 242 is coupled to an analog-to-digital converter (ADC)250 where the analog baseband signal is converted to digital samples. When receiver 200 IS a CDMA receiver, such as used in a communication system defined in IS-95 or J-STD-008, the baseband signal IS comprised of an in-phase component and a quadrature component. The in-phase and quadrature baseband components may be separated by downconverting the IF signal in two separate mixers, where the LO signal to each IF mixer is in quadrature. The ADC 250 samples the in-phase and quadrature baseband signals to produce in-phase and quadrature digital samples, respectively. The output of the ADC 250 is coupled to a baseband processor 270.

The baseband processor 270 stage represents all subsequent processing performed on the baseband signal. Examples of subsequent processing include, but are not limited to: despreading, deinterleaving, error correction, filtering, and amplification. For example, a baseband processor within a CDMA telephone incorporates a plurality of demodulator fingers arranged as a rake receiver. The processed information is then routed to the appropriate destination. The processed information may be used as control signals within the wireless device or routed to a user interface such as a display, speaker, or data port.

The output of the ADC 250 is also coupled to an Automatic Gain Control (AGC)260 stage. AGC 260 measures the energy of the incoming signal and provides a control signal to LNA 220 and VGA236 to scale the received analog signal so that the downconverted baseband signal remains within a predetermined dynamic range of ADC 250. Since the LNA 220 is generally not a variable gain amplifier, the control signal coupled to the LNA 220 is generally used to control the amplifier DC bias.

The dynamic range of the received analog signal can be as high as 80dB, which typically requires an ADC bit width of about 13-14 bits to quantize it without causing saturation or truncation loss. However, due to the desire to minimize external leads, data path hardware, and power consumption, IQ sampling bit widths on commercially available Application Specific Integrated Circuits (ASICs) are typically limited to 4-6 bits. The AGC loop captures a large dynamic range with a small ADC bit width. The AGC measures the energy of the incoming signal and controls the LNA and VGA to scale the analog signal so that it stays within the dynamic range supported by the ADC bit width.

As discussed above, the physical topography and structure surrounding a mobile telephone creates multiple signal paths from the transmitter to the mobile telephone receiver. The composite received signal may be modeled with a plurality of signals having slow fading components and fast fading components. Fig. 3A shows an example of one of the multipath signals of a moving mobile phone, showing the variation of the received signal power over time due to slow fading and fast fading components. The slow fading component of the composite received signal is shown in fig. 3B. The fast fading, or rayleigh fading, component of the composite received signal is shown in fig. 3C. The AGC loop in the receiver can compensate for almost all the effects of slow fading and can compensate for some of the effects of rayleigh fading.

The rate estimate can be made with power measurements in individual multipaths over time. The speed of a mobile unit may be estimated from the number of times that power has passed at half its RMS power level over a given time period, which is referred to as the level crossing rate. In this way, the number of times fast fading results in a crossing with half the RMS power level threshold can be used to estimate the mobile unit's rate. The threshold of half the RMS power level is not the only threshold that can be used for rate estimation. Other fractions or multiples of any RMS power level may be selected as the threshold level. However, using half the RMS power level as a threshold results in a maximum level crossing rate for a given rate.

One challenge in implementing such a velocity estimator is to establish an accurate meter within the mobile unit to measure the power of a multipath. Since the level crossing rate algorithm of the rate estimation needs to know the variation of the received power in the individual multipaths with time, the power in the individual multipaths has to be separated from the total received power. Even if the power in individual multipaths is separated, a second problem arises due to the effect of AGC. Since the AGC attempts to keep the received signal envelope approximately constant, the information needed by the mobile unit to estimate the received power is corrupted. Rate estimation with level crossing does not work properly if the received signal power cannot be measured accurately.

In order to use the level crossing rate of the received power as an estimator of the receiver velocity, the receiver must separate a single multipath, correct for the AGC effect on the signal amplitude, and count the number of times the received signal passes a threshold based on RMS received power.

In a first embodiment, a CDMA receiver uses characteristics of a pilot signal to assist rate estimation. In a CDMA system, a base station sends out a pilot signal that can be used by a mobile receiver as a phase reference for coherent demodulation. Knowledge of the pilot power of a multipath is sufficient to determine the level crossing rate. The value of the pilot power in one multipath is normalized by its RMS power level, which is equivalent to the normalized total power received in one multipath. The equivalence of the two ratios assumes that the ratio of the pilot energy to the total energy transmitted by the base station (Ecp/Ior) is constant. A constant ratio of pilot signal to total transmitted energy is true for CDMA systems. In other words, since fading affects all Walsh-coded channels equally, normalized power information is available after using the power of one channel (pilot). The power in one multipath is separated from the total received power by coherently integrating the despread pilots over a sufficient amount of time. Coherent integration separates the pilot power from the power in the orthogonal data channels. A CDMA receiver may assign a finger of the rake receiver to an identified multipath. However, since the signal may be dynamically scaled by the AGC circuit, the separate despread, integrated pilots are not accurate estimates of the power of their respective paths at the receiver.

The effect of AGC can be removed by scaling the received signal by the inverse of the AGC gain. Fig. 4 shows a block diagram of an implementation that scales an input signal to remove the effects of an AGC loop. The signal is input to a circuit using AGC. This is shown in fig. 4 as an input signal to VGA 436. When the circuit of fig. 4 is implemented in a CDMA receiver, the input to VGA 436 may be a down-converted IF signal. The variable gain loop includes an RF stage or may only include a baseband stage. The distribution of the variable gain is not a limitation of the scaling circuit.

The signal on control line 438 determines the amplification level of VGA 436. In a CDMA receiver, the output of VGA 440 is an IF signal that is coupled to a signal processing stage that performs quadrature mixing and filtering 440. The quadrature baseband output consists of in-phase and quadrature signal components. The in-phase signal is coupled to the first ADC 452 and the quadrature signal is coupled to the second ADC 454. The sampled output signals from the two ADCs 452 and 454 are coupled to a power estimator 462 which forms part of the AGC loop. The power estimator 462 makes an estimate of the instantaneous received power from the ADC output. The output of the power estimator is subtracted from the predetermined set point with an inverting adder 464. The predetermined set point is selected to represent the signal power value near the upper boundary of ADCs 452 and 454. Power estimates that exceed the set point result in a signal on control line 438 that decreases the gain of VGA 436. Power estimates below the set point result in a signal on control line 438 that increases the gain of VGA 436.

The output of the inverting adder 464 is coupled to a low pass filter 466. The output of the low pass filter 466 is coupled to a digital-to-analog converter (DAC)468 that generates an analog control signal that is applied to the VGA 436. DAC 468 is not required when the VGA can accept digital control signals. The VGA 436 changes its gain in accordance with the control signal.

The output of the low pass filter 466 is also coupled to an inverter, here shown as multiplier 482 having a multiplication factor of-1. The output of multiplier 482 is coupled to a log-to-linear converter 484. The log to linear converter 484 is configured with a transfer function that is the inverse of the AGC control signal to signal gain transfer function. The transfer function of the log-to-linear converter 484 is the inverse of the VGA response, where the VGA 436 is the only variable gain element. When there is more than one variable gain element within the control of the AGC circuit, the log to linear converter 484 must compensate for the variable gain of the additional element. The output of the log-to-linear converter 484 is an estimate of the received signal voltage, in which any effect of the AGC circuitry is removed.

The signal scaler provides sufficient signal processing to allow the level crossing rate to be accurately determined when the received signal-to-noise ratio (SNR) is high. This is because the noise component within the received signal is insignificant relative to the received signal power and thus does not adversely affect the determination of the level crossing rate. However, when the noise component has a significant effect on the total received power, the noise component adversely affects the determination of the level crossing rate when a single threshold is used.

As described above, the pilot signal is integrated over a period of time to separate the power in one multipath from the total received power. This coherent integration provides an instantaneous estimate of the pilot signal power. The pilot integration time is determined based on the highest frequency component for which fast fading is desired. The frequency components of the fast fading can be estimated with the expected range of rates that the receiver will experience. In order to reliably detect level crossings, the pilot integration time must be sufficiently below the period of fast fading. Since this amount of time is limited when the channel has fading, the measured pilot power has a certain amount of noise associated with it. SNR and Iorhat of power of integrated pilot₀(Ecp/Ior)/(Ioc + Nt) is directly proportional, wherein iormat₀Is the amount of signal power at the mobile unit in path 0, (Ecp/Ior) represents the ratio of the pilot energy to the total energy transmitted at the base station, and (Ioc + Nt) represents the total interference caused by the adjacent reference and thermal noise. Thus, the measurement of the pilot power is noisy when the total received power is low or when the interference level and noise are high. This noise causes the estimate of the pilot power to cross the threshold many times, while the actual received pilot measured in a noiseless system only passes once.

To reduce the effect of the SNR of the signal on the level crossing rate, level hysteresis may be used. Hysteresis means the use of a high threshold and a low threshold. The signal amplitude should not be considered to pass the threshold level unless it starts with a lower hysteresis threshold (set N dB below the threshold level) and then passes a higher hysteresis threshold (set M dB above the threshold level) and vice versa. The threshold level that results in the maximum number of level crossings is half the RMS signal power. However, the actual threshold used is not limiting in design and may be selected to be any level relative to the RMS level. In an exemplary implementation, M and N are set to 3. The values of M and N need not be the same and may not depend on the predetermined threshold level. Thus, small variations in signal amplitude measurement below (N + M) dB are not counted in the level crossing calculation when hysteresis is implemented. The level-hysteresis algorithm may be represented by a pseudo-random code, where s (n) is the symbol amplitude at time n, T_HIs a high hysteresis threshold level, and T_LIs a low hysteresis threshold level.

If (s (n) < T_L){

If (threshold value mark ═ 0) retaining tone

Level crossing counter + +;

}

the threshold flag is 1;

else if (s (n) > T_H){

If (threshold value mark ═ 1) retaining tone

Level crossing counter + +;

}

the threshold flag is 0;

}

FIG. 5 shows a strip in a noisy environmentThe signal in multipath estimates 530 the variation over time. Typically, fast fading results in a signal as shown in fig. 3C. However, the effect of the noise component may result in a noisy estimate of the multipath signal. The high hysteresis threshold is denoted T_H510, low hysteresis threshold denoted T_L520. The predetermined threshold level occurs at a power level between the high and low hysteresis thresholds and is not shown in fig. 5. The operation of the pseudo random code provided above results in the level crossing counts being counted only at those points indicated by X.

Fig. 6 shows a first embodiment of a velocity estimator 600. The velocity estimator 600 is shown for a receiver such as a CDMA radiotelephone that utilizes orthogonal signals. However, the operation of the velocity estimator 600 does not require the use of orthogonal signals.

In a first embodiment, the velocity estimator 600 IS implemented in a CDMA radiotelephone, such as a radiotelephone operating in a communication system defined in IS-95 or J-STD-008. The sampled in-phase 602 and quadrature 604 signals are provided to a signal processor 610, which PN despreads and accumulates the pilot signal over a predetermined N-chip period. The signal processor 610 operates as a multipath extractor. The signal processor 610 may be a microprocessor, a digital signal processor, an ASIC, or any combination of signal processors capable of performing this function. The signal processor 610 may be capable of performing the functions of a stand-alone unit or may be capable of performing the functions in conjunction with memory through the use of instructions stored in the memory. Also, the signal processor 610 may be dedicated to the above-described functions or may perform additional functions. The output of the signal processor 610 is in-phase pilot symbols 612 and quadrature pilot symbols 614. The in-phase and quadrature pilot signals 612 and 614 are coupled to respective multipliers 622 and 624.

The same sampled in-phase 602 and quadrature 604 signals provided to the signal processor 610 are also provided to the signal sealer 608. The signal sealer 608 may be integrated with the AGC circuit as shown in fig. 4, or may be a circuit that performs an equivalent function. The output of the signal sealer 608 is a scaling factor that represents the inverse of the AGC gain. The scaling factor is coupled to each of multipliers 622 and 624.

The outputs of multipliers 622 and 624 are signals representing in-phase and quadrature pilot signals in which the effects of the AGC circuitry are removed. The scaled signal is coupled to a low pass filter 630. The output of the low pass filter is coupled to a power calculation stage 640 which calculates the square root of the signal energy, denoted as s (n), by taking the square root of the sum of the squares of the scaled in-phase and quadrature pilot symbols.

The calculated values s (n) are coupled to a mobile RMS calculator 650. The moving RMS calculator 650 calculates a moving RMS value using a predetermined number of consecutive values of s (n). The calculated moving RMS value is coupled to a threshold calculation stage 652. The threshold calculation stage 652 calculates the upper and lower level crossing thresholds with predetermined hysteresis values M and N. The calculated upper and lower thresholds are coupled to a level crossing counter 660.

The calculated value s (n) is also coupled to a first-in-first-out (FIFO) buffer 642. Determining the depth of the FIFO corresponds to the number of symbols used in the moving RMS calculation. The symbol output of FIFO 642 is coupled to level crossing counter 660. Level crossing counter 660 counts the number of level crossings by the calculated upper and lower thresholds to achieve hysteresis in the count.

The output of the level crossing counter 660 is coupled to a look-up table 670 which maps the number of level crossings in a given time period to an estimated rate. The rate estimates are output directly from the look-up table 670. Alternatively, instead of implementing the lookup table 670, the output of the level crossing counter 660 is used directly as the rate estimate. Subsequent stages may be adapted to use the level crossing counts directly, or rate estimates may be calculated from the level crossing counts. When the subsequent stages are adapted to use level crossing counting directly, no intermediate translation provided by the look-up table 670 is necessary. When the level crossing count is used directly by the subsequent stage, the level crossing count itself represents the rate estimation.

Fig. 7 shows an alternative embodiment. The alternative velocity estimator 700 shown in fig. 7 uses many of the same stages as the velocity estimator used in fig. 6. Those stages shown that remain the same have the same index number. The difference between the two embodiments is the calculation of the level crossings. The stages up to the square root of the signal energy s (n) remain the same in both embodiments. In a second embodiment, signal s (n) is coupled to moving RMS calculation 650 and to FIFO 642 as in the first embodiment. However, in the second embodiment, the output of the moving RMS calculation 650 is coupled to a normalization factor stage 752. The normalization factor stage 752 calculates 2/x, where x represents the output of the moving RMS calculation stage 650. The value 2/x represents the normalization factor. The output of the normalization factor stage 752 is coupled to a normalization multiplier 744.

The output of FIFO 642 is coupled to another input of normalization multiplier 744. The output of the normalization multiplier 744 is the FIFO 642 output normalized by half the moving RMS value. The multiplier 744 output is coupled to the level crossing counter 660. The level crossing counter 660 is the same as that used in the first embodiment. However, rather than using a varying hysteresis threshold, the level crossing counter can use a constant hysteresis threshold. Since the input signal is normalized by a value proportional to the moving RMS value, the level crossing counter 660 can use a constant hysteresis threshold. The output of the level crossing counter 660 is again coupled to a look-up table 670 to determine a rate estimate. The implementation of the look-up table 670 is then again optional. The level crossing count may be used directly by subsequent stages or it may be possible to calculate a rate estimate from the level crossing count.

Fig. 8 shows a flow chart of a first embodiment of a rate estimation method. The routine begins by receiving an input signal 802. In CDMA phones, these are down-converted signal samples.

Next, a multipath is extracted 804 from the received signal. This is performed in CDMA phones by integrating one pilot over a certain time period. This is performed within one finger of a multi-finger rake receiver. Instantaneous envelope signal values are calculated from the extracted multipaths 806. In CDMA telephony, the instantaneous value is obtained by adding the squares of the in-phase and quadrature components of the integrated pilot signal. A moving RMS value is calculated 808 from a predetermined number of instantaneous envelope values. The calculated moving RMS value is then used to calculate a level crossing threshold 810. The higher level crossing threshold is calculated by adding a predetermined higher hysteresis value to the calculated moving RMS value. Similarly, a lower level crossing threshold is calculated by subtracting a predetermined lower hysteresis value from the calculated moving RMS value.

Once the level crossing threshold has been calculated, the number of level crossings is counted 812. A counting method incorporating hysteresis may be used to eliminate the effects caused by noise. In one embodiment, the accumulated level crossings are then mapped to rate estimates using a predetermined lookup table. Alternatively, the accumulated number of level crossings itself represents the rate estimate. No look-up table is needed when the accumulated number of level crossings represents a rate estimate.

Fig. 9 shows an alternative rate estimation embodiment. The routine begins 902 by receiving an input signal. The routine then proceeds to block 904 where a multipath is extracted from the received signal. The routine then calculates 906 the instantaneous envelope value for this one multipath. The routine calculates a running RMS value 908 using the plurality of instantaneous envelope values. The routine then calculates a scaling factor 910 equal to 2/(moving RMS value). The routine then scales 912 each instantaneous envelope value by a scaling factor. These normalized values are then used to count 914 the number of level crossings. The number of level crossings is then mapped to a rate estimate in a look-up table 916. Here, the implementation of the look-up table is optional, as discussed in the previous embodiments. The number of level crossings may independently represent the rate estimate.

Those of skill would further appreciate that the various illustrative logical blocks, modules, and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both. To clearly illustrate this interchangeability of hardware and software, various illustrative components, blocks, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application in which the overall system is used and design constraints imposed on the overall system. The skilled person will recognize the interactivity of the hardware and software in these cases and how best to implement the described functionality for each particular application.

The previous description of the preferred embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without the use of the inventive faculty. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims (18)

1. A velocity estimator, comprising:

a signal processor for extracting a multipath from a received signal;

a signal scaler for generating a scaling factor which is the inverse of the AGC gain;

a multiplier for scaling the extracted multipath by a scaling factor;

an instantaneous envelope calculator for calculating an instantaneous envelope value of the extracted multipath;

a moving RMS calculator for calculating a moving RMS value using the plurality of instantaneous envelope values;

a level crossing counter for counting a number of level crossings, wherein the number of level crossings is a number of crossings of the instantaneous envelope value by a level crossing threshold resulting from the running RMS value.

2. The velocity estimator of claim 1 further comprising a look-up table that maps the number of level crossings to a velocity estimate.

3. The velocity estimator of claim 1 wherein said received signal is a composite CDMA signal.

4. The rate estimator of claim 3 wherein the signal processor extracts the multipath by despreading and accumulating the pilot signal from the composite CDMA signal over a predetermined number of chips.

5. The velocity estimator of claim 1 further comprising a FIFO storing a predetermined number of instantaneous envelope values equal to the plurality of RMS values used to calculate the moving RMS value.

6. The velocity estimator of claim 1 wherein the level crossing counter achieves hysteresis in counting level crossings by using a higher level crossing threshold and a lower level crossing threshold resulting from a moving RMS value.

7. The velocity estimator of claim 6 wherein the higher level crossing threshold is calculated as a first predetermined number of dB, M, above a level crossing threshold.

8. The velocity estimator of claim 7 wherein the lower level crossing threshold is calculated as a second predetermined number of dB below the level crossing threshold, N.

9. The velocity estimator of claim 8 wherein the level crossing threshold is one half of the moving RMS value.

10. The velocity estimator of claim 9 wherein M and N are 3.

11. A velocity estimator, comprising:

a signal processor for extracting a multipath from a received signal;

a signal scaler for generating a scaling factor which is the inverse of the AGC gain;

a multiplier for scaling the extracted multipath by a scaling factor;

an instantaneous envelope calculator for calculating an instantaneous envelope value of the extracted multipath;

a moving RMS calculator for calculating a moving RMS value using the plurality of instantaneous envelope values;

a normalization multiplier that generates a normalized RMS value by multiplying the instantaneous envelope value by a normalization factor generated from the moving RMS value;

a level crossing counter for counting a number of level crossings, wherein the number of level crossings is a number of crossings of the instantaneous envelope value with a predetermined level crossing threshold; and

a look-up table for mapping the number of level crossings to a rate estimate.

12. The velocity estimator of claim 11 wherein the normalization factor is 2/(running RMS value).

13. A method of rate estimation, comprising:

receiving a composite input signal;

extracting a single multipath signal from the composite input signal;

calculating the instantaneous envelope value of the extracted multipath signal;

calculating a moving RMS value of the extracted multipath signal;

the number of crossings with the level crossing threshold made by the instantaneous envelope value is counted.

14. The method of claim 13, further comprising mapping the number of crossings to a rate estimate.

15. The method of claim 13 wherein said composite input signal is a CDMA signal.

16. The method of claim 13, wherein extracting the signal of a single multipath is performed by PN despreading and accumulating the pilot signal over a predetermined number of chips.

17. The method of claim 13, wherein the instantaneous envelope value is calculated by taking the square root of the sum of the squares of the inphase signal component and the quadrature signal component.

18. The method of claim 13 wherein the number of crossings is calculated using a high level crossing threshold and a low level crossing threshold, wherein the high level crossing threshold is greater than half the moving RMS value by a first predetermined number M of dB and the low level crossing threshold is less than half the moving RMS value by a second predetermined number N of dB, and wherein a level crossing occurs when an instantaneous envelope value representing an instantaneous envelope value over time passes both the high level crossing threshold and the low level crossing threshold.