GB2361789A - Mud-pulse telemetry receiver - Google Patents

Abstract

The receiver includes one or more instruments 212,214 for detecting and generating signals in response to a telemetry wave and a noise wave e.g. from pump 306 traveling opposite to the telemetry wave. A filter 220 receives and combines the signals generated by the instruments to produce an output signal in which the noise wave signal is filtered out. An equalizer reduces distortion of the telemetry wave signal caused by transmission through the fluid in the pipeline 206.

Description

2361789 DIGITAL SIGNAL RECEIVER FOR MWD TELEMETRY SYSTEMS

BACKGROUND OF TBE INVENTION 1. Technical Field

The invention relates generally to measurement-while-drilling (MWD) and logging-while-drilling (LWD) telemetry systems. More particularly, the invention relates to a method and apparatus for improving signal detection in the mud-pulse telemetry systems used with MWD/LWD instruments.

2. Background Art

NIWD and LWD technologies provide drilling operators greater control over the construction of a well by providing real-time information about conditions at the bottom of the well. The information that is of interest to the drilling operators, and which must be obtained from the bottom of the well, include directional drilling variables, such as inclination and direction of the well, and geological formation data, such as natural gamma ray levels and electrical resistivity of the rock formation. Typically, the MWD tools make the directional and drilling measurements, and the LWD tools make the geological formation measurements. Often MWID and LWD tools are integrated into a single package and are called MV%rD/LWD tools. In this disclosure, the term MWD system will be used to collectively refer to MVVD, LWD, and MWD/LWD tools. The term MWD system will also be understood to encompass data transmission from the bottom of the well to the surface.

MWD systems measure parameters at the bottom of the well and can transmit the acquired data to the surface. There are several different methods for transmitting data to the surface, but the most widely used method in commercial MWD systems is mud-pulse telemetry. In mud-pulse telemetry, data is transmitted from the bottom of the well to the surface by means of pressure waves in the drilling fluid or "mud" that is pumped down the drill string by pumps on the surface. Figure I illustrates a drilling system 100 that is equipped for N4VVD operation using mud-pulse telemetry. As shown in Figure 1, the drilling system 100 includes a drill string 102 hanging from a derrick 106. The drill string 102 extends through a rotary table 108 into the well 110. A drill bit 112 is attached to the end of the drill string 102, and drilling is accomplished by rotating and pressing down on the drill bit 112. The drill bit 112 may be rotated by rotating the entire drill string 102 from the surface using the rotary table 108 and the kelly 114 or by operating a downhole mud motor 116 above the drill bit 112.

While drilling, mud is pumped from mud pumps 118 on the surface 120 through the standpipe 122, flexible hose 123, and swivel 121 and down the drill string 102. Pulsation dampeners 125, also known as desurgers or accumulators, are located on the outputs of the mud pumps 118 to smooth transients in the mud waves coming from the mud pumps. The mud in the drill string 102 is forced out through jet nozzles (not shown) in the face of the drill bit 112 and returned to the surface through the well annulus 124, i.e., the space between the well 110 and the drill string 102. One or more sensors or transducers 126 are located in a measurement module 127 in the bottomhole assembly of the drill string 102 to measure desired downhole conditions. For example, the transducer 126 may be a strain gage that measures weight-on-bit or a thermocouple that measures temperature at the bottom of the well 110. Additional sensors may be provided as necessary to measure other drilling and formation parameters such as those previously described.

The measurements made by the transducers 126 are transmitted to the surface through the drilling mud in the drill string 102. First, the transducers 126 send signals that are representative of the measured downhole condition to a downhole electronics unit 128. The signals from the transducers 126 may be digitized in an analog-to-digital converter. The downhole electronics unit 128 collects the binary digits, or bits, from the measurements from the transducers 126 and arranges them into data frames. Extra bits for synchronization and error detection and correction may be added to the data frames. The bits ftom the data frames are then passed to a modulator 129, which groups bits in symbols and then uses a process called modulation to impress the symbols onto a baseband or carrier waveform that can be transmitted through the mud in the drill string 102. A symbol consists of a group of one or more bits. The modulated signals serve as input to an acoustic transmitter 130 that generates the telemetry pressure wave that carries data to the surface. A pressure transducer 132 on the standpipe 122 detects changes in mud pressure and generates signals that are representative of these changes. The output of the pressure transducer 132 is digitized in an analog-todigital converter and processed by a signal processor 134, which recovers the symbols from the received waveform and then 2 sends the data to a computer 138 where the transmitted information can be accessed by the drilling operators.

There are several mud-pulse telemetry systems known in the art. These include positive-pulse, negative-pulse, and continuous-wave systems. In a positive-pulse system, the transmitter 130 creates a pressure pulse at higher pressure than that of the mud volume by momentarily restricting flow in the drill string 102. In a negative mud-pulse telemetry system, the transmitter 130 creates a pressure pulse at lower pressure than that of the mud volume by venting a small amount of the high-pressure mud in the drill string 102 to the well annulus 124. In both the positive-pulse and negative-pulse systems, the pressure pulses propagate to the surface through the drilling mud in the drill string 102 and are detected by the pressure transducer 132. To send a stream of data, a series of pressure pulses are generated in a pattern that is recognizable by the signal processor 134.

The pressure pulses generated by positive-pulse and negative-pulse systems are discrete pressure waves. Continuous pressure waves can be generated with a rotary valve or "mud siren." In a continuous-wave system, the transmitter 130 repeatedly interrupts the flow of the drilling mud in the drill string 102, causing a periodic pressure wave to be generated at a rate that is proportional to the rate of interruption. Information is then transmitted by modulating the phase, frequency, or amplitude of the periodic wave as a function of the downhole measured data.

The telemetry pressure wave that carries information from the transmitter 130 to the pressure transducer 132 is subjected to attenuation, reflections, and noise as it moves through the drilling mud. The signal attenuation as it passes through the mud channel may not be constant across the range of frequencies present in the telemetry pressure wave. Typically, lower frequencies are subject to less attenuation than higher frequencies. The pressure wave is reflected off the bottom of the well and at any acoustic impedance mismatches in the drill string 102 and the surface mud system, i.e., the mud pumps 118, standpipe 122, flexible hose 123, swivel 121, and pulsation dampeners 125. The result is that the pressure wave arriving at the pressure transducer 132 on the standpipe 122 is the superposition of the main wave &om the transmitter 130 and multiple reflected waves. The result of the reflections and frequency dependent attenuation is that each of the transmitted symbols become spread out in time and interfere with symbols preceding and following those transmitted symbols. This is known in the art as intersymbol interference (ISI).

3 Pressure waves from the surface mud pumps 118 contribute considerable amounts of noise. The pump noise is the result of the periodic motion of the mud pump pistons and, hence, is harmonic in nature. The pressure waves from the mud pumps 118 travel in the opposite direction from the main information carrying wave, namely from the surface down the drill string 102 to the drill bit 112. Components of the noise waves from the surface mud pumps 118 may be present in the frequency range used for transmission of the telemetry wave. In some cases the components of the mud pump 118 noise waves may have considerably greater power than the received telemetry wave, making correct detection of the received symbols very difficult. Additional sources of noise include the drilling motor 116 and drill bit 112 interaction with the formation. All these factors degrade the quality of the received pressure wave and make it difficult to recover the transmitted information.

Attempts to find solutions for reducing interfering effects in MVVD telemetry signals are not new and many techniques have been proposed. For example, U.S. Patent No. 3,302,457 issued to Mayes proposes a scheme for reducing mud pump noise based on combining the outputs of a static pressure sensor and a differential pressure sensor. U.S. Patent No. 3,555, 504 issued to Fields proposes a method using two pressure taps at spaced points on the surface piping. The pressure taps are connected to flow lines which delay the pressure wave from one tap relative to the other so that the pump noise components from both taps would be in phase at a differential pressure meter, thus canceling the pump noise. U.S. Patent No. 3,488,629 issued to Claycomb extends Fields' invention by including check valves in the flow lines to reduce reflected waves in the flow lines.

U,S. Patent No. 3,747,059 issued to Garcia discloses an electronic noise filter system that eliminates spurious detection caused by mud pump noise waves reflecting back off the flexible hose. The electronic noise filter system is coupled to at least two pressure-sensitive transducers located at spaced points on the mud pump side of the flexible hose. Electronic circuits in the electronic noise filter system introduce relative delays as well as amplitude and phase adjustments to the signals detected by the transducers. After the delays and the amplitude and phase adjustments, the mud pump noise components of the signals are aligned in phase and can be subtracted off, leaving only the signal from downhole. U.S. Patent No. 3,716,830 issued to Garcia proposes an alternative system that eliminates spurious detection caused by mud pump noise waves reflecting back off the flexible hose by placing one of the transducers after the flexible 4 hose on the side furthest away from the mud pumps. The systems proposed by Garcia only reduce the effect of mud pump noise wave reflecting off the flexible hose; other reflections or distortions of the noise or signal waves are not addressed.

U.S. Patent No. 3,742,443 issued to Foster et al. proposes a noise reduction system that uses two pressure sensors at spaced points. The optimum spacing of the sensors is one-quarter wavelength at the frequency of the telemetry signal carrier. The signal from the sensor closer to the mud pumps is passed through a filter having characteristics related to the amplitude and phase distortion encountered by the mud pump noise component as it travels between the two spaced points. The filtered signal is delayed and then subtracted from the signal derived from the sensor further away ftom the mud pumps. The combining function leads to destructive interference of the mud pump noise and constructive interference of the telemetry signal wave, because of the one-quarter wavelength separation between the sensors. The combined output is then passed through another filter to reduce distortion introduced by the signal processing and combining operation. The system does not account for distortion introduced in the telemetry signal wave as it travels through the mud column from the downhole transmitter to the surface sensors. The filter on the combined output also assumes that the mud pump noise wave traveling from the mud pumps between the two sensors encounters the same distortion mechanisms as the telemetry signal wave traveling in the opposite direction between the same pair of sensors. This assumption does not, however, always hold true in actual MVVD systems, as will be shown later.

U.S. Patent No. 4,215,425 issued to Waggener discloses a coherent phase shift keying (PSK) demodulation system that includes a differential filtering o I peration for mud pump noise cancellation using two sensors separated by one-quarter wavelength. U.S.

Patent No. 4,262,343 issued to Claycomb discloses a system in which signals from a pressure sensor and a fluid velocity detector are combined to cancel mud pump noise and enhance the signal from downhole. U.S. Patent No. 4,590,593 issued to Rodney discloses a two sensor noise canceling system similar to those of Garcia and Foster et al., but with a variable delay. The delay is determined using a least mean squares algorithm during the absence of downhole data transmission.

U.S. Patent No. 4,642,800 issued to Umeda discloses a noise-reduction scheme that includes obtaining an "average pump signature" by averaging over a certain number of pump cycles. The assumption is that the telemetry signal is not periodic with the same period as the pump noise and, hence, will average to zero. The pump signature is then subtracted from the incoming signal to leave a residual that should contain mostly telemetry signal. U.S. Patent No. 5,146,433 issued to Kosmala et al. uses signals from position sensors on the mud pumps as inputs to a system that relates the mud pump pressure to the position of the pump pistons. Thus, the mud pump noise signature is predicted from the positions of the pump pistons. The predicted pump noise signature is subtracted from the received signal to cancel the pump noise component of the received signal.

U.S. Patent No. 4,715,022 issued to Yeo discloses a signal detection method for mud pulse telemetry systems using a pressure transducer on the gas filled side of the pulsation dampener to improve detection of the telemetry wave in the presence of mud pump noise. One of the claims includes a second pressure transducer on the surface pipes between the dampener and the drill string and a signal conditioner to combine the signals from the two transducers. Yeo does not describe how the two signals may be combined to improve signal detection.

U.S. Patent No. 4,692,911 issued to Scherbatskoy discloses a scheme for reducing mud pump noise by subtracting from the received signal, the signal that was received T seconds previously, where T is the period of the pump strokes. The received signal comes from a single transducer. A d7elay line is used to store the previous noise pulse from the mud pumps and this is then subtracted from the current mud pump noise pulse. This forms a comb filter with notches at integer multiples of the pump stroke rate. The period T of the mud pumps may be determined from the harmonics of the mud pump noise, or from sensors placed on or near the mud pumps. The telemetry signal then needs to be recovered from the output of the subtraction operation, that includes the telemetry signal plus delayed copies of the telemetry signal. U.S. Patent No. 4,730,281 issued to Rodney proposes an adaptive bucket brigade filter with a feedback loop in place of the delay used by Scherbatskoy to produce a comb filter response that removes a periodic noise and its harmonics while also reducing the delayed copies of the telemetry signal caused by the comb filter response.

U.S. Patent No. 5,490,121 issued to Gardner et al. discloses a non-linear adaptive equalizer for reducing non-linear distortion of the telemetry signal. The nonlinear equalizer receives an input signal from a pressure transducer and passes the signal through a bank of non-linear function elements. The signal is then processed by a parallel set of 6 linear, or decision feedback, equalizers. One linear equalizer receives the unmodified input signal as its input, and the other linear equalizers receive their inputs from the output of a nonlinear function element. The output signals of the linear equalizers are summed together to provide the nonlinear equalizer's output signal.

U.S. Patent No. 5,969,638 issued to Chin discloses a signal processor for use with MWD systems. The signal processor combines signals from a plurality of signal receivers on the standpipe, spaced less than onequarter wavelength apart to reduce mud pump noise and reflections traveling in a downhole direction. The signal processor isolates the derivative of the forward traveling wave, i.e., the wave traveling up the drill string, by taking time and spatial derivatives of the wave equation. Demodulation is then based on the derivative of the forward traveling wave. The signal processor requires that the signal receivers be spaced a distance of five to fifteen percent of a typical wavelength apart.

All the aforementioned prior art systems are attempting to find a successful solution that would eliminate a substantial portion or all of the mud pump noise measured by transducers at the surface and, in so doing, improve reception of telemetry signals transmitted ftom downhole. Some of these systems also attempt to account for reflected waves traveling back in the direction of the source of the original waves. However, none provide means for substantially reducing mud pump noise while also dealing with distortion caused by the mud channel and reflected waves.

SUMMARY OF THE MWNTION

One aspect of the invention is a receiver for use with a mud-pulse telemetry system. The receiver comprises at least one instrument for detecting and generating signals in response to a telemetry wave and a noise wave traveling opposite the telemetry wave, the generated signals each having a telemetry wave component and a noise wave component. A filter receives and combines the signals generated by the instruments to produce an output signal in which the noise wave component is filtered out. An equalizer reduces distortion of the telemetry wave component of the signals.

Other aspects and advantages of the invention will be apparent from the following description and the appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure I illustrates a drilling system equipped for MWD operation.

Figure 2 is a block diagram of a mud-pulse telemetry system.

7 Figure 3 is a schematic of a drilling system embodying the mud-pulse telemetry system shown in Figure 2.

Figure 4 is a block diagram of the receiver shown in the mud-pulse telemetry system of Figure 2.

Figure 5 illustrates the wave path for reflected mud pump noise and mudpulse telemetry signal.

Figure 6 is a block diagram of one embodiment of the filtering module and equalizer shown in Figure 4.

Figure 7 is a block diagram of a decision feedback equalizer.

Figure 8 is a block diagram of a comb filter.

Figure 9 is a frequency response diagram for the comb filter shown in Figure 8.

Figure 10 i s a b lo ck di agram o f a notch fi lter.

Figure 11 is a frequency response diagram for the notch filter shown in Figure 10.

DETAILED DESCRIPTION OF THE INVENTION

Figure 2 shows a block diagram for a mud-pulse telemetry system 200 that can be employed by a drilling system such as the drilling system 100 illustrated in Figure 1. The mud-pulse telemetry system 200 includes a measurement module 201, an electronics unit 203, a modulator 205, a transmitter 204, a mud channel 206, a receiver 208, and a computer 210. The measurement module 201, electronics unit 203, modulator 205, and transmitter 204 correspond to the downhole portion of the telemetry system, and the receiver 208 and computer 210 correspond to the surface portion of the telemetry system. The mud channel 206 connects the downhole portion of the telemetry system to the surface portion of the telemetry system. The measurement module 201 represents messages that can be transmitted to the computer 210. These messages would be information that is of interest to the drilling operators, e.g., directional and drilling data and geological formation data. The measurement module 201 includes one or more transducers (or sensors) which measure pertinent bottomhole parameters and generate electrical signals related to the parameters measured.

The measurements made by the transducers in the measurement module 201 may be digitized by passing them through analog-to-digital converters (not shown). The binary digits, or bits, representing the measurements are transferred to the electronics unit 203 as message signals M. In the electronics unit 203, extra bits may be added to the message 8 signals M. The extra bits can be used for error detection and correction or for identification. The message signals M may also be filtered or compressed to improve bandwidth efficiency. The electronic module 203 may group the messages M into data frames. Extra bits for frame synchronization, channel identification, equalizer training, or error detection and correction may be included in the data frames.

The output of the electronics unit 203 is a bit stream that is the input to the modulator 205. The modulator 205 groups the bits from the output of the electronics unit 203 into symbols and then impresses these symbols onto a waveform that is suitable for propagation over the mud channel 206. The size of a symbol may be one or more bits. The output of the modulator 205 is transferred to the transmitter 204, which produces the pressure pulses or waves that propagate through the mud channel 206. The telemetry waveform may be a baseband waveform. In which case, symbols are transmitted using a technique called line coding. Examples of line codes that can be used to impress the information are non-retum-to-zero (NRZ), Manchester codes, Nfiller code, time analog, and pulse position modulation. Line codes for mud pulse telemetry are known in the art, see for example "Applying Digital Data-Encoding Techniques to Mud Pulse Telemetry" by S. P. Monroe, SPE 20326, Proceedings of the Petroleum Computer Conference, Denver, June 25-28, 1990, pages 7-16.

Instead of using line -coding, the modulator 205 and the transmitter 204 may impress the symbols onto a suitable carrier by varying the amplitude, phase, or frequency of a carrier, usually a sinusoidal signal, in accordance with the value of a single bit, or group of bits, making up a symbol. This process is called modulation. For example, in binary phase shift keying (BPSK) modulation, the phase of a constant amplitude carrier signal is switched between two values according to the two possible values of a binary digit corresponding to binary I and 0, respectively. Examples of other modulation types include amplitude modulation (AM), frequency modulation (FM), minimum shift keying (MSK), frequency shift keying (FSK), phase shift keying (PSK), phase modulation (PM), continuous phase modulation (CPM), quadrature amplitude modulation (QAM), and trellis code modulation. These modulation types and the aforementioned line codes are known in the art, see, for example, "Digital Communications" by John G. Proakis, 3 rd Edition, McGraw-Hill, Inc., 1995 and "Wireless Communications" by Theodore S. Rappaport, Prentice Hall, Inc., 1996, pages 197-294.

Figure 3 shows a simplified diagram of a drilling system 300 embodying the mud- 9 pulse telemetry system 200 previously illustrated in Figure 2. The mud channel 206 of the telemetry system 200 corresponds to the drilling mud in the drill string 302. The transmitter 204 uses the telemetry waveform signal generated by the modulator 205 to control a mechanism 213, which alters the flow of mud in the drill string 302 to generate a pressure wave. In one embodiment, the mechanism 213 is a rotary valve or "mud siren" that generates periodic waveforms, in fluid. An example of a mud siren is disclosed in U.S. Patent 5,375,098 issued to Malone el al., assigned to Schlumberger Technology Corporation.

The mud siren includes a slotted rotor and a slotted stator. The slotted rotor periodically interrupts the flow of fluid through the slots in the stator to generate the carrier wave. A phase change in the carrier wave can be produced by applying a brake to the rotor, and thus slowing its rate of rotation. The brake is released when the phase of the carrier wave is the required number of degrees from what it was before the brake was applied. It should be clear that the mechanism 213 is not limited to changing the phase of a carrier wave, but, depending on the modulation scheme, may also change other properties, e.g., amplitude and frequency, of the carrier wave. The mechanism 213 also does not have to be a mud siren, but may be a type that generates positive pressure pulses or negative pressure pulses, as well known in the art.

The signal wave genefated by the transmitter 204 propagates to the receiver 208 through the mud channel 206, which is the drilling mud in the drill string 302. One or more transducers (or sensors), e.g., transducers 212 and 214, in the receiver 208 detect changes in the signal wave and generate electrical signals 216 and 218, respectively, that are representative of the changes. In the example embodiment, the transducers 212 and 214 measure changes in mud pressure. Other types of measurement, e.g., flow measurements, may also be used. The transducers 212 and 214 are spatially separated so that they receive the telemetry signal from the downhole transmitter 204 at different times related to the speed of sound in the drilling mud. For example, the transducers 212 and 214 may be mounted on the surface piping 304 that connects the mud pumps 306 to the flexible hose 308. Alternatively, at least one transducer may be placed in the well annulus 310 or on the gas filled side of the pulsation dampener 312.

Figure 4 shows a block diagram for the receiver 208. The receiver 208 includes a filtering module 220, which receives samples of the transducer signals 216 and 218 as input. The samples may be generated by passing the signals 216 and 218 through analog- to-digital converters. The primary purpose of the filtering module 220 is to cancel mud pump noise from the received signal. The output signal 222 of the filtering module 220 is transferred to a noise filter 224, which filters frequencies outside of the band occupied by the telemetry signal. In a system using carrier modulation, the output signal 226 of the noise filter 224 is demodulated by a demodulator 228. The demodulator 228 basically reverses the modulation process carried out by the modulator 205 and the transmitter 204, i.e., the demodulator 228 recovers from the output signal 226 the baseband waveforms that were modulated onto a carrier waveform. In a carrier modulated system, the demodulator 228 may also track and compensate for gain changes, carrier frequency offsets, phase offsets, and sample timing offsets so that the bits can be successfully demodulated. Techniques that can be used to track and compensate for frequency, phase, and timing offsets are known in the art, see, for example, "Synchronization Techniques for Digital Receivers" by Umberto Mengali and Aldo N. D'Andrea, Plenum Publishing Corporation, 1997.

The demodulated signal 230 is transferred to an equalizer 232 and detector 236. The function of the equalizer 232 is to compensate for, or reduce, distortion in the telemetry component of the signal 230 prior to the detector 236. The detector 236 is a decision device that makes decisions about the received symbols. The detector 236 outputs a stream of bits, which are passed on to the computer 210. The computer 210 recovers the data frames and then the original messages M from the bit stream. The received messages may then be stored in a database, analyzed and/or displayed, The equalizer 232 may be a linear transversal equalizer, with symbol- spaced taps or fractional ly-spaced taps or a decision feedback equalizer. These types of equalizers reduce distortion in the signal prior to the detector, and symbol detection is usually on a symbol-by-symbol basis. There are also equal izers/detectors that make symbol decisions based on received samples spanning a group of symbols or by determining the most likely sequence of symbols. These usually include some means of estimating the impulse response of the channel that causes the signal distortion. The maximum likelihood sequence estimator (MLSE) and maximum a-posteriori (") receivers are examples of equalizer/detectorsthat do not make symbol decisions on a symbol-by-symbol basis. All these equalizers as well as other types of equalizers are known in the art, see, for example, "Digital Communications" by John G. Proakis, McGraw-Hill, Inc., 1995, pages 583-679.

Equalizers based on neural networks, such as multi-layer perceptions or a radial I I basis function networks, are alternative types of equalizers. Examples of these equalizers are described in "Adaptive Filter Theory" by S. Haykin, P Edition, Prentice Hall International, Inc. 1996, pages 817-874. The equalizer 232 may also be a hybrid of the previously mentioned equalizers. For example, the feedback section of a decision feedback equalizer may be used as part of a maximum-likelihood sequence estimator, as described in "Developments of the conventional nonlinear equalizer" by A. P. Clark and R. S. Marshall, IEE Proceedings, Vol. 129, Pff, No.2, April 1982, pages 82-94. Examples of hybrid equalizers that use radial basis function components are described in "Applying Radial Basis Functions" by Bernard Mulgrew, IEEE Signal Processing Magazine, March 1996, pages 50-65. The equalizer 232 may also help to reduce residual pump noise components present in the signal 230. Certain types of equalizers, e.g., Volterra filters and Neural network filters, may help to reduce non-linear distortion in the signal. Examples of publications describing these equalizers are "Adaptive Filter Theory" by S. Haykin, 3rd Edition, Prentice Hall Intemational, Inc. 1996, pages 16-18, and "Channel Equalization Using Adaptive Complex Radial Basis Function Networks" by Inhyok Cha and Saleern A. Kassam, IEEE Journal on Selected Areas in Communications, Vol. 13, No. 1, January 1995.

Referring back to Figure 3, both the telemetry signal wave from the downhole transmitter 204 and the noise wave from the mud pumps 306 encounter numerous acoustic reflectors as they propagate through the mud system. For the purpose of illustration, consider the surface system as being comprised of two acoustic reflectors, e g., reflector R, representing the flexible hose 308 and swivel (not shown) on the downhole side of the surface piping.304 and reflector R2 representing the combination of the pulsation dampener 312 and mud pumps 306. Reflector R, has a reflection coefficient pl, and reflector R2 has a reflection coefficient P2. Transducer 214 is located at position A, closer to the mud pumps, and transducer 212 is located at position B, closer to the downhole side.

Figure 5 shows the mud pump noise wave path and the telemetry signal wave path. As shown, a telemetry signal from the transmitter 204 first reaches transducer 212 and then at a later time reaches transducer 214. A pressure wave traveling from the mud pumps 306 down the drill string 302 will first reach the transducer 214 and then at a later time reach the transducer 212. Let -i represent the time it takes for a wave to travel from position A and B, or vise versa. Let '12 represent the time it takes for a wave to travel from 12 position A to reflector R2 and back to position A. Let T3 represent the time it takes for a wave to travel from position B to reflector R, and back to position B. Let s(t) represent the telemetry signal wave that arrives at position B from the downhole transmitter 204 at time t, and let n(t) represent the noise wave that arrives at position A from the direction of the mud pumps 306 at time t.

Transducer 212 at position B will, at time t, detect telemetry signal, s(t), from the downhole transmitter 204 plus a first reflected wave, P2s(t-(2T+T2)), a second reflected wave, P]P2s(t-(2T-'-T2+'C3)), and subsequent reflections, where p, and P2 are reflection coefficients. The first reflected wave is caused by the telemetry wave, s(t), reflecting back off reflector R2. The second reflected wave is caused by the first reflected wave being reflected by reflector & The magnitude and phase of the reflected waves depend on the reflection coefficients p, and P2. The telemetry signal component, SB(t), detected by transducer 212 at position B may then be expressed as a convolution of s(t) with h2(t), where h2(t) is the sum of the direct and reflected waves detected at position B. That is, SB (1):__ SW 0 k (1) (1) = s(f) 0 [8(t) + p28(1 - (2.c + T2)) + P1P28(t - (2.c + '12 + TM+ 1 Taking into consideration a delay of -c seconds in the time it takes the telemetry signal, s(t), to reach position A, the telemetry signal component, SA(t), detected by transducer 214 at position A can be expressed as a convolution of s(t-,c) with h3(t), where h3(t) is the sum of the direct and reflected waves detected at position A. That is, SA1) = S(/ - 7) 0 MI) (2) = S (t - 1) 018(t) + P245(t -'r2) +A P245(t -(2z-+ 1'2 + 1,3D + The mud pump noise component, nA(t), detected by the transducer 214 at position A will include the direct wave from the mud pumps 306 and reflected waves from reflector R, and reflector R2. That is, n., (t) = n(t) 0 h4(t) = n(t) 0 [8(t) + p, 8(t - (2r +CA) + PI P28(t (2 r + Z2 + Z3))+ ' 1 The mud pump noise component, nB(t), detected by the transducer 212 at position B is delayed -c seconds as the mud pump noise wave travels from position A to position B, hence 13 % (1) = n(t - z.) 0 h, (t) (4) = n(t - z.) 0 [8(t) + pl.5(t -T3) + pIp28(t - (2.c +r2 + Z3))+ 1 The transducer 214 at position A detects a combination of telemetry signal, sA(t), and mud pump noise, nA(t). The combined signal, CA(t), at position A may be expressed as:

CA (1) = SA (t) +nA (t) (5) = s(t - z) 0 k (t) + n(f) 0 h4(1) The combined signal, CB(t), at position B may be expressed as:

CB (1)::- SB (t) + n. (1) (6) = s(t) 0 h2(t) + n(t - z.) 0 hs (1) Using Laplace Transforms, equations (5) and (6) can be transformed into the following algebraic expressions:

C, (s) = S(s)e H, (s) + N(s)H, (s) (7) C,Js) = S(s)H,(s)+N(s)eH,(s) (8) Mud pump noise component, N(s), may then be eliminated by multiplying equation (7) by ' (S) e' end subtracting the resulting expression from equation (8). H 4(S) The result is:

CB (S) - CA (S) H, (s) e' = S(s) HAS)_ HAs)H5W e (9) H,, (S) 1 H, (S) The telemetry signal, s(t), which has itself passed through the mud column in the drill string as it propagated to the surface, may be expressed as a convolution of the telemetry signal transmitted by the transmitter 204 with the distortion caused by the mud channel from the transmitter 204 to position B- Let d(t) be the telemetry signal transmitted by the transmitter 204 and hi(t) be the effects of the mud channel from the transmitter 204 to position B, then s(f) = d(t) 0h, (t) (10) Taking Laplace Transform of equation (10) and substituting the result into equation (9), the following expression is obtained:

14 C,, (S) - C, (S) H, (S) e " = D(s)H, (s) HAS)_ HAs)H_, (s) e -2sr H4W 1 H4W = D(s)M(s) Let s = jeo. The frequency response of the system is then given by:

D(jco)M(jw) = D(jco)H, (jco)H, (jo)) 1 - H3 UCOW5 U0)) e (12) 1 H2 U0)W4 U0j) The mud pump noise cancellation operation explained above is carried out in the filtering module 220, as illustrated in Figure 6. The output of the filtering module 220 is a distorted signal, Doa))Moco). The signal distortion causes intersymbol interference (ISI) that makes correct detection of the received symbols difficult and increases the probability of errors in the detector 236. The equalizer 232 compensates for the distortion, Moco), of the telemetry signal, Doco), caused by the mud channel in the drill string, Hioco), the surface reflections, H20o)), H30o)), H40o)), and H50o)), observed at positions A and B and the pump noise filtering module 220. The simple two-reflector model described above shows how distortion of the telemetry signal and mud pump noise arises between positions A and B. On a real world drilling rig, the acoustic system may be considerably more complex with multiple reflectors at different distances from positions A and B. In the example embodiment, the filtering module 220 performs a differential filtering operation. Differential filtering exploits the fact that the noise signal from the mud pumps 306 and the signal from the transmitter 204 travel in opposite directions through the surface piping 304. To cancel the mud pump noise in the received signal, the signal 218 is passed through a filter 215 that accounts for amplitude and phase differences in the mud pump noise components of signals 216 and 218, caused by reflected waves represented in the previous discussion by Mco) and Hsoo)). The output 217 of the filter 215 is then passed through a delay element 219, where it is delayed by the time z that it takes the mud waves to travel from position A to position B. After the filter and delay operations, the mud pump noise component of signal 221 should be in phase and have the same amplitude as the mud pump noise component of the signal 216. The phase-aligned and scaled signal 221 is then subtracted from signal 216 to cancel the mud pump noise. The filtering operation by the filter 215 and the time delay by the delay element 219 may be performed in any order. The delay may even be incorporated into the filter 215. In a practical system, the filter 215 would need a causal impulse response and hence will is introduce some delay of its own. The signal 216 from transducer 212 will need to be delayed by an amount equivalent to the group delay of filter 215.

In Figure 6, the equalizer 232 is shown as the optimum linear equalizer for this system. Other types of equalizer may, in practice, produce better results. It may be very difficult in practice to predict or estimate the reflection profiles for the signal and mud pump waves at each sensor location. Hence, an adaptive filter 215, 219 and an adaptive equalizer 232 may be used. In such a case, the adaptation of adaptive filter 215 and equalizer 232 may be carried out separately or even jointly. In addition, a simpler type of filtering module which comprises a single gain factor G may be used in place of the filter 215, followed by the time delay 219. To perform this simpler filtering operation, the signal 218, which is the signal from the transducer 214 closer to the mud pumps, is shifted in time according to the distance between the transducers 212 and 214. A positive gain is applied to the delayed signal 218 so that the mud pump noise detected by the transducers 212 and 214 have approximately the same amplitude. The signals 216 and 218 are compared to determine how much gain and delay to apply to the signal 218. After applying the gain and delay, the signal 218 is subtracted from the signal 216. This operation is equivalent to passing the signal 218 through a filter that has constant gain across all frequencies and linear phase, meaning that all frequencies are delayed by the same amount. If the group delay through the channel 206 is approximately constant for all frequencies in the range occupied by the telemetry signal, applying the correct gain G and delay T values will result in most of the pump noise interference being eliminated irrespective of the pump noise frequency.

Because the telemetry signal wave from the transmitter 204 travels in the opposite direction to the noise wave from the mud pumps 306, the time shifting operation on the signal 218 will not align the telemetry signal components in the signals 216 and 218. Hence, telemetry signal cancellation does not occur when signal 221 is subtracted from signal 216, unless the spacing of the transducers is such that they are separated by n/2 wavelength, where n E=- {0,1, 2, 3,...}, at the telemetry signal carrier frequency. Performing the gain and delay filtering on signals from transducers that are separated by less than 1/8 wavelength will result in a reduction in the amplitude of the telemetry signal, because the telemetry signal components will destructively interfere. However, if the mud pump noise component is reduced by a greater amount than the telemetry signal component, this process may still work for systems where the sensors are spaced less than 16 118 wavelength apart.

The transducers 212 and 214 are preferably spaced between 118 and 3/8 of a wavelength apart at the frequency of the carrier wave so that the carrier signal components of the signals 216 and 218 add constructively in the differential filter 220. The preferred spacing between the transducers 212 and 214 is 1/4 wavelength apart at the frequency of the carrier wave. At this preferred spacing, the time delay used to align the mud pump noise detected by the transducers 212 and 214 will move the wanted telemetry signal component from the transducer 214 by 1/2 of a carrier wavelength relative to the signal from transducer 212. After applying gain and a delay to signal 218 and subtracting the resulting signal from signal 216, the telemetry signal components from the transducers 212 and 214 will be in phase. The telemetry signal components of the signals 216 and 218 thus add constructively to increase the signal strength.

The differential filter 220 cancels out most of the mud pump noise in the received signals 216, 218 but does not account for signal distortion from reflected waves or non linear impairments in the signals received at the different transducer locations. Whenever a change in acoustic impedance occurs, part of the pressure waves, from both the transmitter 204 and the mud pumps 306, are reflected in the direction in which they were traveling. Examples of locations where acoustic impedance changes occur are at changes in pipe diameter, pipe junctions, the flexible hose 308 connecting the drill string 302 to the standpipe 304 on the surface and at pulsation dampeners 312 on the outputs of mud pumps 306. The reflected waves interfere with subsequent waves and may result in destructive interference at some frequencies and constructive interference at other frequencies. As the distance between the transducers increases, the telemetry signals at the transducers 212 and 214 become less correlated with each other. This means that the telemetry signal distortion will differ at each transducer. As a result, transducers at different locations may detect different signal fading patterns. Because the two signals 216, 218 are combined in the differential filter 220, the output of which is the sum of the two distorted signals, the output of the filter 220 will itself be a distorted signal.

Part of the waves from the mud pumps may also be reflected back towards their source and cause different fading patterns on the pump noise at the two transducers 212, 214. The relative amplitudes and phases of pump noise components at the two transducers 212, 214 may thus be different from that expected flom a simple time shift. This will result in imperfect noise cancellation from the simple gain and delay differential filter 220.

17 The delay value may be determined from knowledge of the spacing and estimates of the speed of sound in the mud, or, alternatively, if the spacing is not known, the delay can be estimated from the signals received at the surface transducers 212, 214 before data transmission from downhole occurs or even during data transmission. Those skilled in signal processing methods will appreciate that a variety of methods exist for estimating the time difference of arrival of a signal at multiple sensors.

The equalizer 232 in the receiver 208 compensates for signal distortion present in the output signal 222 of the filtering module 220. The equalizer 232 may also help reduce residual mud pump noise components, especially if the equalizer taps are fractionally spaced. Figure 7 illustrates one embodiment of the equalizer 232 as a decision-feedback equalizer which includes a fractional I y- spaced linear feedforward transversal filter 234, the detector 236, and a linear feedback transversal filter 238. The forward filter has Nf taps, WFo to WF(Nf-1), and the feedback filter has Nb taps, WBO to WB<Nb-1). For illustration purposes in Figure 7, Nf is set to 4 and Nb is set to 3. The fractionallysp aced linear feedforward transversal filter 234 is made up of a shift register 235 that stores samples of the input signal 237. The input signal 237 is sampled from the output 230 of the demodulator 228 (shown in Figure 4.) A symbol timing circuit (not shown) ensures that the samples are taken at the optimum point during each symbol interval.

Delay elements 239-241 represent the time between samples of the input signal. The input signal 237 is sampled at a fast enough rate that no aliasing of the input signal occurs in the sampling process. For example, if the received signal is bandlimited to a frequency range less than the symbol rate, two samples per received symbol period may be used. The samples in the shift register 235 that align with the taps 242-245 of the forward filter 234 are then multiplied by the tap weights, WFo to WF(Nf1), and summed to produce the output 246 of the feedforward filter 234. The output 246 of the feedforward filter 234 is determined once per symbol period. The fractional ly-spaced nature of the feedforward filter 234 allows the filter to adaptively synthesize a matched filter for the received signals. In the presence of residual mud pump noise, the fractional ly-spaced feedforward filter 234 may adaptively synthesize notch filter characteristics for reducing residual mud pump noise present in the signal 237 entering the forward filter 234.

Previous symbol decisions from the detector 236 are stored in the feedback shift register 252. Delay elements 260-264 represent the time between samples of the input signal to the feedback register 238. The previous symbol decisions 254-258 are multiplied 18 by the feedback filter tap weights, WBo to WB(Nb-1). The outputs of the multiplication operations are then summed to produce the output 248 of the feedback filter 238. The output 248 of the feedback filter 238 is subtracted from the output 246 of the feedforward filter 234, and the result is the input 266 to the detector 236. When the samples from the next symbol are taken, the samples stored in the feedforward shift register 235 are shifted across and the new samples shifted in to the shift register. Similarly for the feedback shift register 252, the previously detected symbols are shifted across and the most recently detected symbol shifted into the feedback shift register 252.

The equalizer 232 may be trained by sending a sequence of symbols, that is known to the receiver 208, through the channel 206. When the sequence of symbols is detected by the receiver 208, the filter coefficients WFo to WF(Nf-1) and WBO to W) for the linear feedforward transversal filter 234 and the linear feedback transversal filter 238, respectively, are determined, usually by an algorithm that minimizes the mean squared error between the samples of the received signal and the ideal reference signal. Once the initial training of equalizer 232 is completed, the decisions at the output of the detector 236 may be used to update the filter coefficients WFr) to WF(N1-1) and WBo to WB(Nb-1). Other criteria for adjusting the tap weights are known in the art, e.g. zero forcing or minimizing peak distortion, and may alternatively be used. Adaptive filter algorithms that are commonly used to determirre the filter coefficients are described, for example, in "Adaptive Filter Theory" by S. Haykin, 3 rd Edition, Prentice Hall International, Inc. 1996, and "Efficient Least Squares Adaptive Algorithms for FIR Transversal Filtering" by G-0 Glentis, K. Berberidis, and S. Theodoridis, IEEE Signal Processing, July 1999, pages 1341. Blind adaptive algorithms are yet another alternative for adapting the tap weights, see for example, pages 772-816 of "Adaptive Filter Theory" by S. Haylcin.

Equalization may be carried out at baseband or on the passband signal prior to conversion to baseband. The linear feedforward transversal filter 234 is implemented in the digital domain. Accordingly, input signal 237 to the linear feedforward transversal filter 234 should be in digital form. This may require using a digital-to-analog converter (not shown) to process input signal 237. This does not mean, however, that the invention is limited to a transversal filter that operates in the digital domain.

In operation, the transmitter 204 sends data symbols in the form of a telemetry signal to the receiver 208. The transducers 212 and 214 detect the transmitted signal and generate signals 216 and 218, respectively, in response. The signals 216 and 218 are 19 transferred to the filtering module 220. The filtering module 220 cancels mud pump noise in the transmitted signal. The output of the filtering module 220 is passed through the noise filter 224 and the demodulator 228. The noise filter 224 attenuates noise at frequencies outside of the band occupied by the carrier signal. The demodulator 228 demodulates the transmitted signal. The noise filter 224 and the demodulator 228 operations are linear and can be performed in any order, even before the filtering module 220. The output of the demodulator 228 is transferred to the equalizer 232 and detector 236. The equalizer 232 compensates for, or reduces, distortion and residual pump noise in the received signal. The detector 236 makes decisions about the received symbols and outputs a stream of bits which is passed on to the computer 210.

The invention has been described with respect to the filtering module 220 using differential filtering to cancel out the mud pump noise component of the transducer signals 216 and 218. Alternatively, the filtering module 220 can filter out the mud pump noise component and retrieve the telemetry signal component of the transducer signals 216 and 218 using a method similar to the one disclosed in U.S. Patent No. 5,969,638 issued to Chin. This alternate filtering method expresses the acoustic pressure in the drill string, p(x,t), as the sum of the upward traveling wave, g, and a downward traveling wave, f That is, P(X, 0 = g(x - CO + f (X + ct) (13) where t is time, x is distance, and c is the speed of sound in drilling mud.

The information transmitted from downhole is contained in the upward traveling wave g(x-ct). Reflected signals and mud piston noise are contained in the downward traveling wave f(x+ct). The information transmitted from downhole can then be obtained by isolating g(x-ct) from p(xt). This is accomplished by taking space and time derivatives of equation (13) and solving for the derivative of g(x-ct). The resulting expression for the derivative of g(x-ct) is:

ICPX -PI 9 2c (14) where p,, is the space derivative of p(x,t) and p, is the time derivative of p(x,t). Because the derivative g' contains the same phase, amplitude, and frequency information, it is not necessary to integrate g' to obtain g. The equalizer 232 can then be used to reduce distortion in g' as previously discussed.

The filtering module 220 may also use other alternative pump noise filtering methods that process input signals from two or more transducers, for example, methods such as those disclosed in U.S. Patent No. 4,262, 343 issued to Claycomb and U.S. Patent No. 5,146,433 issued to Kosmala et aL The invention is, however, not limited to a filtering module 220 that receives input signals from two or more transducers.

In an alternative embodiment, a comb filter that receives input from only one transducer may be used in place of the filtering module 220. The comb filter is designed to have nulls at frequencies corresponding to the harmonics of the mud pump noise wave. An example of a type of adaptive comb filter for MWD applications is disclosed in U.S. Patent No. 4,692, 911 issued to Scherbatskoy. As shown in Figure 8, the comb filter 320 includes a delay element 322 and a module 324 for tracking the pefiod of the noise wave from th e mud pumps 306 (shown in Figure 3). The comb filter 320 receives an input signal 326 from a transducer 328, which would be placed on the surface piping 304 (shown in Figure 3). The comb filter 320 combines the unmodified input signal 326 with a delayed copy 330 of the input signal 326 to produce the output signal 332. The module 324 adjusts the delay so that it corresponds to the period T of the mud pumps 306. The period T of the mud pumps may be determined from the harmonics of the noise wave from the mud pumps 306 or from sensors placed on or near the mud pumps 306- The comb filter is adaptive so that it can take into account variations in the period of the mud pumps 306. The output of the comb filter may be used as a feedback signal to the tracking module 324, to improve tracking and rejection of the harmonics of the mud pump noise wave.

Figure 9 shows the frequency response of the comb filter 320. The comb filter attenuates signal components in the region of integer multiples of some base frequency. The base frequency in this case would be the stroke rate or the reciprocal of the period, T, of the mud pumps 306. The adaptive comb filter 320 filters mud pump harmonics, but introduces distortion of the telemetry signal wave. Using the equalizer 232 after the comb filter 320 will help to compensate for, or reduce, distortion of the telemetry wave component caused by the mud channel and comb filter. Other forms of adaptive comb filters are known, see, for example, U.S. Patent No. 4,642,800 issued to Umeda, U.S. Patent No. 4,730,281 issued to Rodney el aL, and "Digital Signal Processing: Principles, Algorithms and Applications" by J.G. Proakis and D.G. Manolakis, 2 nd Edition, Prentice Hall of India Private Limited, 1995, pages 3 58-36 1.

An alternative method for compensating for mud pump noise and telemetry signal 21 distortion is to precede the equalizer 232 with a notch filter. The notch filter passes most frequencies with little attenuation, but has nulls in its frequency response at frequencies corresponding to mud pump noise harmonics that fall with the range of frequencies used for transmission of the telemetry wave. Figure 10 shows an example of a notch filter 340. The notch filter 340 uses a module 342 to track the frequency of the mud pump noise harmonics and then synthesizes a filter 344, which has a frequency response that includes nulls at the frequencies of the mud pump noise harmonics that fall within the range of frequencies used for transmission of the telemetry wave. The notch filter 340 is adaptive so that it can take into account variations in the frequency of the mud pump harmonics. The output of the notch filter 344 may be used as a feedback signal to the tracking module 342 to improve the noise rejection capabilities of the notch filter 344.

The notch filter 340 takes input from a single transducer 346 and can be used in place of the filtering module 220 (shown in Figure 3). Figure I I shows the frequency response of the notch filter 340 with a null at a ftequency f... th. Adaptive notch filters may be implemented in a variety of different ways. For examples of alternative implementations see "Adaptive Noise Cancelling: Principles and Applications" by B. Widrow et aL, Proceedings of the IIEEE, Vol. 63, No. 12, December 1975, pages 16921716, "Adaptive Noise Canceling Applied to Sinusoidal Interferences" by J.R. Glover Jr., IEEE Transactions on Acoustics, Speech and Signal Processing, Vol. ASSP-25, No.6, December 1977, pages 484-491, "Adaptive Harmonic Noise Cancellation with an Application to Distribution Power Line Communications" by J-D. Wang and HT Trussell, IEEE Transactions on Communications, Vol. 36, No.7, July 1988, pages 875-884 and and "Digital Signal Processing: Principles, Algorithms and Applications" by J.G. Proakis and D. G. Manolakis, 2"' Edition, Prentice Hall of India Private Limited, 1995, pages 3 5 5-3 57.

While the invention has been described with respect to a limited number of embodiments, those skilled in the art will appreciate that other embodiments of the invention can be devised which are within the scope of the invention. Accordingly, the scope of the invention should be limited only by the appended claims- 22

Claims (25)

1. A receiver for use with a mud-pulse telemetry system, the receiver comprising:

an instrument for detecting and generating signals in response to a telemetry wave and a noise wave traveling opposite the telemetry wave, the generated signals each having a telemetry wave component and a noise wave component; a filter for receiving and combining the signals generated by the instrument to produce an output signal in which the noise wave component is filtered out; and an equalizer for compensating for or reducing distortion of the telemetry wave component of the output signal.

2. The receiver of claim 1, further comprising:

at least two instruments for detecting and generating signals in response to the telemetry wave and the noise wave traveling opposite the telemetry wave.

3. The receiver of claim I or 2, further comprising a detector for detecting a message transmitted by the telemetry wave from an output of the equalizer.

4. The receiver of claim I or 3, wherein the equalizer further compensates for or reduces residual noise wave in the output signal.

5. The receiver of claim I or 2, wherein the filter is an adaptive filter.

6. The receiver of claim I or 2, wherein the equalizer is an adaptive equalizer.

7. The receiver of claim I or 2, further comprising a demodulator for demodulating the signals generated by the instruments.

8. The receiver of claim 1, wherein the filter attenuates components of the signals generated by the instruments in the region of integer multiples of a frequency of the noise wave.

23

9. The receiver of claim 1, wherein the filter attenuates components of the signals generated by the instruments at frequencies of the noise wave within a range of frequencies of the telemetry wave.

10. The receiver of claim 1, wherein the filter is a comb filter.

11. The receiver of claim 1 or 2, wherein the filter is a notch filter.

12. The receiver of claim 2, wherein the instruments are spatially separated such that they receive the telemetry wave and noise wave at different times related to the speed of sound in drilling fluid.

13. The receiver of claim 2, wherein the filter is a differential filter.

14. The receiver of claim 13, wherein the differential filter has characteristics related to reflections of the telemetry wave and the noise wave in a transmission channel between the instruments.

15. The receiver of claim 12, wherein the two instruments are spaced between 1/8 and 3/8 of a wavelength apart at a frequency of the telemetry wave.

16. The receiver of claim 12, wherein the filter retrieves the telemetry wave component of the signals generated by the instruments.

17- The receiver of claim 2, wherein the equalizer has characteristics related to distortion of a transmission channel of the telemetry wave.

18. The receiver of claim 17, wherein the instruments are spatially separated and the equalizer has characteristics related to reflections of the telemetry wave and the noise wave in a channel between the instruments.

19. The receiver of claim 12, wherein the two instruments are spaced onequarter of a wavelength apart at a frequency of the telemetry wave.

24

20. A method for detecting a telemetry wave being transmitted through a mud channel by a mud-pulse telemetry system, the method comprising: using at least one instrument to detect and generate signals in response to the telemetry wave and a noise wave traveling opposite the telemetry wave-, filtering out a noise wave component of the signals generated in response to the telemetry wave and the noise wave; and compensating for or reducing distortion of a telemetry wave component of the signals generated in response to the telemetry wave and the noise wave.

21. The method of cl aim 20, further comprising detecting a message transmitted by the telemetry wave.

22. The method of claim 20, further comprising demodulating the signals generated in response to the telemetry wave and the noise wave.

23. The method of claim 20, wherein filtering out a noise wave component includes combining the signal generated in response to the telemetry wave and noise wave with a delayed copy of the signal.

24. The method of claim 23, further comprising synchronizing combination of the signals with the period of the noise wave.

25. The method of claim 20, wherein filtering out a noise wave component includes synthesizing a filter characterized by nulls at the ftequencies of the noise wave that fall within the range of frequencies used for transmitting the telemetry wave.