WO1996037977A1

WO1996037977A1 - Method and apparatus in a communication system for receiving a distorted signal

Info

Publication number: WO1996037977A1
Application number: PCT/US1996/004075
Authority: WO
Inventors: Leo George Dehner, Jr.; Stephen Rocco Carsello
Original assignee: Motorola Inc.
Priority date: 1995-05-22
Filing date: 1996-03-25
Publication date: 1996-11-28
Also published as: TW318297B

Abstract

A method is described for receiving a distorted frequency modulated (FM) signal from a portable transceiver (122) with minimum splatter and for recovering the same with a conventional FM receiver (304) of the base station (116) to obtain data having minimum intersymbol interference using an equalization filter function (306). The method utilizes an optimized prototype pulse which has been predistorted to match the inverse characteristics of the pre-modulation low pass filter (210) of the FM transmitter portion (212) of the portable transceiver (122). When the distorted FM signal is received the equalization filter function (306) processes the distorted FM signal thereby producing electric pulse signals exhibiting minimal intersymbol interference. The electric pulse signals are converted to a sequence of digital data symbols by a clock recovery element (306).

Description

METHOD AND APPARATUS IN A COMMUNICATION SYSTEM FOR RECEIVING A DISTORTED SIGNAL

Field of the Invention

This invention relates generally to the field of communications systems, and more particularly to a receiver which can remove distortion in a distorted carrier signal transmitted by a portable transceiver.

Background of the Invention

A number of problems exist with digital transmission systems. One such problem relates to splatter and adjacent channel interference created when rectangular pulses, such as found in strings of digital information, are transmitted. It is well known that this splatter is reduced by low pass filtering the baseband modulating signal. Many wave shaping functions are known which minimize the spectrum occupied by the baseband modulating signal. Unfortunately, in a frequency modulation (FM) system, minimal occupied baseband spectrum does not imply minimal occupied radio frequency (RF) spectrum. Many of the well known wave shaping functions produce high modulation overshoots which result in excessive splatter, even though the baseband frequency spectrum is well contained. A means of reliably minimizing the splatter of the transmitted RF signal, considering both baseband spectrum and modulation overshoots is required.

A second problem occurring in digital FM modulated transmission systems is the generation of intersymbol interference in the received signal. In general, this is caused by the wave shaping which is required to reduce splatter. When wave shaping is employed, each transmitted bit becomes spread out in time over several adjacent bits resulting in interference during the detection of these bits. Intersymbol interference results in a sensitivity loss in the receiver. Only certain classes of shaped signals are known to exhibit no intersymbol interference. However, most of these shaped signals are extremely complex in structure. A means of adapting one or more of these shaped signals is required if intersymbol interference is to be minimized. Using such complex shaped signals for transmission in portable transceivers is expensive and impractical. Thus, an apparatus and method is needed that allows for the transmission of a distorted signal with intersymbol interference to a receiver, preferably a base station receiver, capable of removing the intersymbol interference inherent in the distorted signal.

Summary of the Invention

One aspect of the present invention is a method of creating a signal exhibiting minimal intersymbol interference for digital communications through a transmission media. The method comprising at a transmitter the steps of generating a multilevel digital baseband signal, filtering the multilevel digital baseband signal using a filter that introduces intersymbol interference and minimizes adjacent channel interference, creating a distorted signal, and transmitting the distorted signal. The method further comprising at a receiver the steps of receiving the distorted signal, and filtering the distorted signal using a filter that removes intersymbol interference and bandlimits any noise introduced by the transmission media.

Another aspect of the present invention is a communication system, comprising a transmitter for transmitting a digital information stream using a distorted carrier signal. The communication system further comprises a receiver for receiving the distorted carrier signal and demodulating the distorted carrier signal to an analog baseband signal s(t). The communication system also includes an extractor for extracting the digital information stream from the analog baseband signal s(t) using an equalization filter function Heq(f), which thereby removes intersymbol interference and bandlimits any noise introduced by a transmission medium. The communication system further includes a processor for controlling operation of the receiver.

Another aspect of the present invention is a receiver for receiving a digital information stream from a distorted carrier signal. The receiver comprising a receiver module for receiving the distorted carrier signal and demodulating the distorted carrier signal into an analog baseband signal. The receiver further comprises an extractor for extracting the digital information stream from the analog baseband signal s(t) using an equalization filter function Heq(f), which thereby removes intersymbol interference and bandlimits any noise introduced by a transmission medium. The receiver further includes a processor for controlling operation of the receiver module.

Brief Description of the Drawings

FIG. 1 is an electrical block diagram of a selective call communication system in accordance with the present invention. FIG. 2 is a representation of a rectangular digital prototype pulse or signaling waveform which results in consumption of relatively large amounts of frequency spectrum when modulated on a radio frequency carrier signal in accordance with the present invention.

FIG. 3 is a graphic representation of a reference pulse exhibiting minimum intersymbol interference in accordance with the present invention.

FIG. 4 is a representation of a window function applied to the reference pulse to minimize energy outside the window in accordance with the present invention. FIG. 5 is a representation of the spectrum derived by taking the

Fourier transform of the windowed reference pulse in accordance with the present invention.

FIG. 6 is a representation of the spectrum of the reciprocal of the transfer function of the transmitter portion of the portable transceiver in accordance with the present invention.

FIG. 7 is a representation of the spectrum derived by applying the transformed pulse of FIG. 5 to the response derived in FIG. 6 in accordance with the present invention.

FIG. 8 is a representation of a pre-distorted pulse matching the characteristics of the transmitter pre-modulation filter, derived by taking the inverse Fourier transform of the waveform of FIG. 7 in accordance with the present invention.

FIG. 9 is a presentation of the final prototype pulse derived by windowing the pre-distorted pulse of FIG. 8 in accordance with the present invention.

FIG. 10 is an electrical block diagram of the portable transceiver in accordance with the present invention. FIG. 11 is an electrical block diagram of the base station transceiver employed to receive transmissions from the portable transceiver of FIG. 10 in accordance with the present invention.

FIG. 12 is a flowchart of the equalization filter function method in accordance with the present invention.

FIG. 13 is a flow chart of the method used by the base station receiver portion of FIG. 11 to remove intersymbol interference generated by the portable transceiver of FIG. 10 in accordance with the present invention.

Description of the Preferred Embodiment

Referring to FIG. 1, an electrical block diagram of a selective call communication system in accordance with the preferred embodiment of the present invention comprises a fixed portion 102 and a portable portion 104. The fixed portion 102 comprises a plurality of base transceivers which are base stations 116 coupled by communication links 114 to a controller 112 for controlling the base stations 116. Optionally, the fixed portion 102 may also comprise a plurality of base receivers 117 likewise coupled by communication links 114 to the controller 112. The hardware of the controller 112 is preferably the combination of the Wireless Messaging Gateway (WMG™) Administrator! paging terminal and the RF-Conductor!™ controller manufactured by Motorola, Inc. The hardware of the base stations 116 is preferably a combination of the Nucleus® Orchestra! transmitter and RF-Audience!™ receivers manufactured by Motorola, Inc. The hardware of the base receivers 117 is preferably similar to the RF-Audience!™ receivers manufactured by Motorola, Inc. Other similar controller hardware can be utilized as well for the controller 112 and base stations 116. Each of the base stations 116 transmits radio signals to the portable portion 104 comprising a plurality of portable transceivers 122 preferably via a transmitting antenna 120. The base stations 116 and base receiver 117 each receive radio signals from the plurality of portable transceivers 122 preferably via a receiving antenna 118. The radio signals comprise selective call addresses and messages transmitted to the portable transceivers 122 and acknowledgments received from the portable transceivers 122. It will be appreciated that the portable transceivers 122 can also originate messages other than acknowledgments. The radio signals received by the base stations 116 and the base receivers 117 from the portable transceivers 122 are in the form of distorted carrier signals, preferably a distorted FM signal, which are transformed by the base station 116 receivers and base receivers 117. It will be appreciated that other modulation schemes such as amplitude modulation (AM) and FM modulation may be used for modulating a distorted carrier signal. Examples of AM and FM modulation can include quadrature amplitude modulation (QAM) and frequency shift keyed (FSK) modulation, respectively. The controller 112 preferably is coupled by telephone links 101 to the public switched telephone network (PSTN) 110 for receiving selective call originations therefrom. Selective call originations comprising voice and data messages from the PSTN 110 can be generated, for example, from a conventional telephone 124 coupled to the PSTN 110 in a manner that is well known in the art.

To facilitate an understanding of the present invention, a discussion of the method of generating a digital pulse which is optimized both in terms of the time domain (bit rate, pulse duration) and frequency domain (frequency consumed) ensues. For purposes of this example, it is assumed that digital pulses are normalized to 1 bit/sec such as the pulse illustrated in FIG. 2. If a series of such nearly square wave or rectangular pulses were modulated on an electromagnetic carrier wave, the sharp transitions in such pulses would result in the consumption of an unduly large amount of frequency spectrum. This is not acceptable given the bandwidth constraints which are imposed by modern communication systems. For this reason, a pre-modulation low pass filter is used with the transmitter of the portable transceiver 122 to minimize adjacent channel interference at the expense of creating intersymbol interference at the base station 116 receiver or base receiver 117. It is desirable to reproduce a prototype pulse which, when received by the base stations 116 or base receivers 117, minimizes intersymbol interference introduced by the spectral transfer function of the transmitter portion of the portable transceiver 122. The prototype pulse is preferably compatible with data transmissions over a broad range of data bit rates, such as 800 bit per second to data bit rates of 9600 bits per second, and higher. The prototype pulse should be generated by a FM receiver using an equalization filter Heq(f) which can remove intersymbol interference from a distorted FM signal transmitted by a portable transceiver 122 using a pre-modulation low pass filter Hipf(f) for removing adjacent channel interference. The derivation of one embodiment of the equalization filter H_eq(f) which generates the prototype pulse of the present invention is shown in FIGs. 3-5 and 6-8.

FIG. 3 is a graphic representation of the reference pulse Po(t) used as the starting point. The reference pulse used is the well known sync function (sin πt)/πt which exhibits the desirable characteristic of providing no intersymbol interference. In an alternative embodiment in accordance with the present invention, Po(t) may also use the well known raised cosine function ((cos2πβt) / (1 - (4fit)2)) * ((sin πrbt) / πrfc_>t) wherein 0 < β < rb/2 and wherein rb is a transmitting data rate in bits per second. The reference pulse represented by the sync function and the raised cosine function are too complex to use directly and thus are preferably transformed as explained below.

As shown in FIG. 4, a window function K(t) is applied to the reference pulse Po(t) truncating the pulse to a length of from ±1 second to ± 3.5 seconds. The optimum value selected depends on the frequency deviation used in the system. The preferred window function K(t) is a Kaiser window which truncates the reference pulse Po(t) to a length of ± 2 seconds. A description of the Kaiser window function may be found in a textbook by Childers and Durling, entitled "Digital Filtering and Signal Processing", published 1975 by West Publishing Company of St. Paul, Minnesota on pages 437 to 440 the inclusion of which is incorporated by reference herein. In an alternative embodiment in accordance with the present invention, the window function K(t) is the well known rectangular window function. The resultant window reference pulse Pι(t) is obtained by multiplying the reference pulse Po(t) by the window function K(t). Once the windowed reference pulse has been determined, as described, the spectrum Pι(f) is determined by taking the Fourier transform of the windowed reference pulse Pi (t). A graphic representation of the spectrum Pi(f) derived is shown in FIG. 5.

The next step in determining the equalization transfer function Heq(f) is to determine a filter function FR(f) which is the reciprocal of the transfer function of the transmitter portion Htχ(f) of the portable transceiver 122 of the preferred embodiment of the present invention. That is, FR(f) = 1/Htχ(f) for all values of f, where Htχ(f) is the transfer function of the transmitter pre-modulation low pass filter Hlpf(f) and the digital to analog converter HD/A(Γ)/ i-e., Hipf(f) * HD/Arø- For optimum minimization of adjacent channel interference in the signal transmitted by the transmitter portion of the portable transceiver 122, the pre- modulation low pass filter Hlpf(f) used in the transmitter portion is preferably a second order Butterworth filter with a .6 Hz cutoff frequency. A graphic representation of the spectrum of the reciprocal filter function FR(f) is shown in FIG. 6. From the reciprocal filter function FR(f), the spectrum P2(f) of the desired prototype pulse P2(t) can be determined by multiplying the spectrum of the filter function FR(f) by the spectrum Pl(f) of the windowed reference pulse. A graphic representation of the spectrum P2(f) of the desired prototype pulse is shown in FIG. 7. The desired prototype pulse P2(t) is next determined by taking the inverse Fourier transform of the spectrum P2(f) of the desired prototype pulse P2(t) obtained in FIG. 7. A graphic representation of the desired prototype pulse P2(t) is shown in FIG. 8. It should be noted that the desired prototype pulse P2(t) has been distorted from an ideal sync function by the process described, resulting in zero crossings to occur at intervals along the time axis which are other than at integer increments, as with the sync function. The distortion imposed is due to using the reciprocal of the transfer function of the transmitter portion Htχ(f) of the portable transceiver 122 in deriving P2(t). The P2(t) distortion is removed by the cancellation effect of the filter function FR(f) with the transfer function of the transmitter portion Htχ(f). The desired prototype pulse P2(t) as shown in FIG. 8 is too complex for utilization. The final step in deriving a final prototype pulse providing the original goals of minimum intersymbol distortion, is shown by the graphic representation illustrated in FIG. 9. The final prototype pulse P3(t) is determined by applying a second predetermined window function W(t) to truncate the time span of the desired prototype pulse P2(t) to a finite length while preserving virtually all of the energy of the pulse. The preferred method is to apply a window equal to unity in the region from ±4 seconds, having cosine squared shaping, which is well known to one of ordinary skill in the art, attenuating P2(t) in the regions -5 to -4 and +4 to +5 seconds, and thereafter being zero outside of ±5 seconds. A graphic representation of the final prototype pulse P3(t) is shown in FIG. 9, and was obtained by multiplying the desired prototype pulse P2(t) by the second window function W(t). In an alternative embodiment in accordance with the present invention, the window function W(t) can be the well known rectangular window function. The Fourier transform of the final prototype pulse P3(t) gives the spectrum of the equalization filter function H_eq(f).

To summarize, a method of mathematically deriving the equalization filter H_eq(f) has been illustrated in FIGs. 3-5, 6-8, and 9. The equalization filter H_eq(f) has the characteristic of providing minimum intersymbol distortion when being detected in the receiver. As previously described, the derivation obtained was for a pulse normalized for 1 symbol/second. It will be appreciated that the result obtained may be scaled for a wide range of data bit rates. When scaling the final prototype pulse to other data bit rates, it will be appreciated that the time indicated scales down and frequency scales up as the data bit rate is increased.

Figure 10 is an electrical block diagram of the portable transceiver 122 of the present invention. The heart of the transmitting apparatus is a microprocessor 202, such as a MC6805 microprocessor integrated circuit manufactured by Motorola, Inc. Coupled to the microprocessor 202, is a random access memory (RAM) 204 and a read only memory (ROM) 206. The RAM 204 preferably provides temporary storage for the microprocessor 202. The data output is a stream of multilevel signals such as binary information corresponding to the portable transceiver 122 address and messages desired to be transmitted. The transmitted symbols are multiplied by a constant which depends on the data to be transmitted. The constants are, for example, +1 to send a logic one, and -1 to send a logic zero. The method is easily extended to other multilevel transmission such as four level. For example, constants of -1, -.333, +.333, and +1 could be used for four-level transmission. These constants are scaled to digital values which are sent by the microprocessor 202 to a D/A converter 208 utilizing conventional means well known in the art. The D/A converter 208 converts the digital symbol levels to analog symbols levels. The analog symbols levels have fast transitions edges which result in adjacent channel interference during modulation. For this reason, a pre-modulation low pass filter 210 is used for pulse shaping thereby removing the high harmonics which cause adjacent channel interference. Although pulse shaping of the analog symbols removes adjacent channel interference, it adds intersymbol interference. The intersymbol interference is removed by the base station 116 receiver by using the equalization filter H_eq(f). The analog symbols levels which have been pulse shaped by the pre-modulation low pass filter 210 are then sent to a FM modulator 212 which modulates the symbols to a carrier, preferably using the well known technique of FSK modulation, and thereafter amplifies the signal for transmission. The amplified carrier generated by the FM modulator 212 is then transmitted by a transmitter antenna 214. The portable transceiver 122 also has a receive portion comprising a receiver antenna 216 and a FM receiver 218 utilizing conventional circuits well known in the art. The receiver portion of the portable transceiver 122 is used for receiving preferably FSK signals from the base stations 116, which represent selective call messages. Note that one physical antenna could be used instead of antennas 214 and 216 in the portable transceiver 122 using convention duplex or switching techniques.

Figure 11 is an electrical block diagram of the base station 116 transceiver employed to receive the transmissions from the portable transceiver 122 of FIG. 10. The transmitted stream of FM signals from the portable transceiver 122 is intercepted by an antenna 118 and received by FM receiver 304. The FM receiver 304 is a conventional FM receiver which is well known in the art, and utilizes any of a number of well known demodulator circuits, such as pulse count discriminators and peak and valley detectors, for the detection of the received FM signals. The output from FM receiver 304 is an analog baseband signal s(t) which depicts a waveform stream characteristic of the distorted FM signal transmitted by the portable transceiver 122. The distorted FM signal transmitted by the portable transceiver 122 has minimal adjacent channel interference, at the expense of having intersymbol interference, as described above. The intersymbol interference significantly reduces the sensitivity of the base station 116 receiver thereby increasing the difficulty of properly sampling the binary information patterns depicted by the waveform stream. To remove the intersymbol distortion the output of the FM receiver 304 is fed to an extractor 306 which removes the intersymbol interference and bandlimits any noise introduced by the transmission medium by using the equalization filter H_eq(f) derived above. In one embodiment in accordance with the present invention, the extractor 306 can be a pulse generator. The pulse generator generates a sequence of optimized electric pulse signals, with minimal intersymbol interference, by taking a convolution of the analog baseband signal s(t) with an impulse response h_eq(t) of the equalization filter function H_eq(f). The pulse generator thereafter converts the sequence of optimized electric pulse signals into a sequence of digital data symbols representative of the digital information stream received. The digital data symbols are delivered to the processor 308 for processing. In order to convert the sequence of optimized electric pulse signals into a sequence of digital data symbols the pulse generator further comprises a clock recovery circuit which obtains bit synchronization from the recovered data signal in a manner well known to one of ordinary skill in the art. The clock recovery circuit controls sampling of the output of the equalization filter H_eq(f). The sampling function is accomplished by a conventional A/D converter well known in the art. Bit decisions are made by a comparator which uses the amplitude of each sampled bit. When the recovered signal is positive, a positive optimized prototype pulse signal, a logical one is generated at the output of comparator. When the recovered signal is negative, indicating a negative optimized prototype pulse signal, a logical zero is generated at the output of comparator. For multilevel symbols, for example, four level data streams, the comparator generates two binary bits per sample. There is no intersymbol distortion in this decision process, except that which is unavoidable due to the IF filtering and that caused by multipath signal reception. It will be appreciated that variations of the above circuit can be used to receive multilevel data symbols at a range of baud rates such as 800 to 9600 and higher. It will be further appreciated that the clock recovery process can also be implemented by the processor 308 coupled to the extractor 306. The processor 308 is one of the family of DSP56000 digital signal processors manufactured by Motorola, Inc. It will be appreciated that other well known processors employing complex instruction set computer (CISC) architecture and reduced instruction set computer (RISC) architecture may be used. The processor 308 is further coupled to a RAM 310 for temporary storage and for calculation processing. In an alternative embodiment in accordance with the present invention, the extractor 306 comprises a sampler and a convolution element 316 stored in a ROM 312 coupled to the processor 308. The sampler samples the analog baseband signal s(t) into a baseband digital sequence s(n). The sampler is preferably an A/D converter utilizing techniques well known in the art. In addition, an equalization filter element 314 is stored in ROM 312. The equalization filter element 314 stores a sequence of numerical values which represent a unit pulse response h_eq(n) of the equalization filter function Heq(f). The convolution element 316 generates a discrete time sequence of optimized digital pulse signals, with minimal intersymbol interference and bandlimits any noise introduced by the transmission medium and spurious signals which arise in the digital to analog conversion process, by taking a convolution of the baseband digital sequence s(n) with a unit pulse response h_eq(n). The convolution element 316 thereafter converts the discrete time sequence of optimized digital pulse signals into a sequence of digital data symbols representative of the digital information stream received. The conversion of the discrete time sequence of optimized digital pulse signals into a sequence of digital data symbols is derived by a clock recovery element which controls sampling of the optimized digital pulse signals utilizing techniques well known in the art. It will be appreciated by one of ordinary skill in the art that the convolution element 316 can also be implemented as a separate hardware module using digital circuits which use digital signal processing techniques which perform discrete time convolution and clock recovery techniques for properly sampling the digital data stream as described above.

In an alternative embodiment in accordance with the present invention, the extractor 306 comprises the sampler, described above, and a multiplication element 318 stored in the ROM 312 coupled to the processor 308. In addition, the sequence of numerical values stored in ROM 312 depict a discrete frequency spectrum H_eq(k) of the equalization filter function H_eq(f). The multiplication element 318 generates a discrete time sequence of optimized digital pulse signals, with minimal intersymbol interference and bandlimits any noise introduced by the transmission medium and spurious signals which arise in the digital to analog conversion process. The multiplication element 318 determines a discrete frequency spectrum S(k) by taking a discrete Fourier transform (DFT) of the baseband digital sequence s(n). The multiplication element 318 then multiplies the discrete frequency spectrum S(k) with the discrete frequency spectrum H_eq(k) stored in ROM 312 to generate a discrete frequency spectrum Sι (k). The multiplication element 318 then determines a discrete time sequence sι(n), representative of a discrete time sequence of optimized digital pulse signals, by taking an inverse DFT of the discrete frequency spectrum Sι(k), and thereafter converts the discrete time sequence sι(n) into a sequence of digital data symbols representative of the digital information stream received. The conversion of the discrete time sequence si(n) of optimized digital pulse signals into a sequence of digital data symbols is derived by a clock recovery element which controls sampling of the optimized digital pulse signals utilizing techniques well known in the art. It will be appreciated by one of ordinary skill in the art that the multiplication element 318 can also be implemented as a separate hardware module using digital circuits which use digital signal processing techniques which perform discrete frequency multiplication and clock recovery techniques for properly sampling the digital data stream as described above. The base station 116 also includes a transmitter portion comprising a transmitter antenna 120 and a FM transmitter 320 utilizing conventional circuits well known in the art. The transmitter portion of the base station 116 is used for transmitting preferably FSK signals comprising selective call messages to be received by portable transceivers 122. Also note that a base receiver 117 preferably comprises all the elements of the base station 116 except for the transmitter portion. Additionally, as was noted with the portable transceivers 122, the base stations 116 may also use one antenna in appropriate configurations as is well known in the art. The method of generating an optimized prototype pulse is summarized in the flowchart 400 of FIG. 12. The method starts with a function Po(t) at step 402. In one embodiment in accordance with the present invention, Po(t) is the well known sync function (sin πt)/πt. In another embodiment in accordance with the present invention, Po(t) is the well known raised cosine function ((cos2πfit) / (1 - (4βt)2)) * ((sin πr_Dt) / πrbt) wherein 0 < β < rb/2 and wherein rb is a transmitting data rate in bits per second. The selected function Po(t) is windowed at block 404, by a predetermined window function to obtain a pulse Pi (t) having minimum energy outside the window. The spectrum Pi(f) of the pulse Pl(t) derived at step 404, is preferably determined by taking the Fourier transform of Pi(t) at step 406. The reciprocal transfer function FR(f) of the pre-modulation low pass filter is determined at step 408. The reciprocal transfer function FR(f) is preferably multiplied by the spectrum Pi(f) to determine the spectrum P2(f) at block 410. The desired prototype pulse is determined, at step 412, by preferably taking the inverse Fourier transform of the spectrum P2(f). The optimized prototype pulse P3(t) is determined by windowing the desired prototype pulse with a second predetermined window function, at step 414. Finally the desired equalization filter function H_eq(f) is determined at step 416 by preferably taking the Fourier transform of the optimized prototype pulse P3(t). For the pulse generator, described above, the impulse response h_eq(t) of the equalization filter function H_eq(f) is preferably used for generating optimized prototype pulses. In one embodiment of the present invention, data defining the equalization filter function Heq(f) preferably is stored in the ROM 312 of the base station 116 receiver preferably as a unit pulse response h_eq(n) used by the convolution element 316 described above. In another embodiment of the present invention, data defining the equalization filter function Heq(f) is stored in the ROM 312 of the base station 116 receiver as a discrete spectrum H_eq(k) used by the multiplication element 318 described above. It will also be appreciated, as described above, that the embodiments which include the convolution element 316 and the multiplication element 318 can be implemented as a separate hardware module using digital circuits employing digital signal processing techniques well known in the art. In the above embodiments, the equalization filter function Heq(f) provides a mechanism for removing intersymbol interference and bandlimits any noise introduced by the transmission medium and spurious signals which arise in the digital to analog conversion process.

FIG. 13 is a flowchart 500 summarizing the method used by the base station 116 receiver portion or base receivers 117 of FIG. 11 to remove intersymbol interference generated by the portable transceiver of FIG. 10 in accordance with the preferred embodiment of the present invention. In step 502 the base station 116 receiver portion or base receiver 117 receives a distorted FM signal transmitted by the portable transceiver 122. The transmitter portion of the portable transceiver 122 uses a pre-modulation low pass filter which minimizes adjacent channel interference, but at the same time introduces intersymbol interference at the base station 116 receiver. In step 504 the base station 116 receiver demodulates the distorted FM signal transmitted by the portable transceiver 122 utilizing techniques well known in the art. The output of the base station 116 receiver generates an analog baseband signal which inherently has intersymbol interference. Steps 506, 508 and 510 represents three embodiments of the extractor 306 described above in accordance with the present invention. Step 506 uses a pulse generator which uses the impulse response heq(t) to derive optimized digital pulse signals, with minimal intersymbol interference while preferably bandlimiting any noise introduced by the transmission medium. The digital information stream is preferably derived by a clock recovery circuit which controls sampling of the output of the equalization filter H_eq(f), as described above. In an alternative embodiment, step 508 uses a convolution element 316 which generates a discrete time sequence of optimized digital pulse signals, with minimal intersymbol interference while preferably bandlimiting any noise introduced by the transmission medium and spurious signals which arise in the digital to analog conversion process, as described above. In another embodiment, step 510 uses a multiplication element 318 which generates a discrete time sequence of optimized digital pulse signals, with minimal intersymbol interference and bandlimits any noise introduced by the transmission medium and spurious signals which arise in the digital to analog conversion process, as described above. Finally in step 512 the processor 308 decodes the message included in the digital information streams utilizing techniques well known in the art.

While a specific embodiment of this invention has been shown and described, further modifications and improvements will occur to those skilled in the art. All modifications which retain the basic underlying principles disclosed and claimed herein are within the scope and spirit of the present invention.

What is claimed is:

Claims

1. A method of creating a signal exhibiting minimal intersymbol interference for digital communications through a transmission media, comprising the steps of: at a transmitter generating a multilevel digital baseband signal, filtering the multilevel digital baseband signal using a filter that introduces intersymbol interference and minimizes adjacent channel interference, creating a distorted signal, and transmitting the distorted signal; and at a receiver receiving the distorted signal, and filtering the distorted signal using a filter that removes intersymbol interference and bandlimits any noise introduced by the transmission media.

2. The method of claim 1, wherein the method further includes the step at the transmitter of modulating a carrier signal with the distorted signal thereby creating a distorted carrier signal and the step at the receiver of demodulating the distorted carrier signal.

3. A method of creating a signal exhibiting minimal intersymbol interference for digital communications, comprising the steps of: at a transmitter generating a multilevel digital baseband signal, filtering the multilevel digital baseband signal using a filter that introduces intersymbol interference and minimizes adjacent channel interference, creating a distorted signal, modulating a carrier signal with the distorted signal, creating a distorted carrier signal, and transmitting the distorted carrier signal; and at a receiver receiving the distorted carrier signal, demodulating the distorted carrier signal to a baseband signal, and filtering the baseband signal using a filter that removes intersymbol interference and bandlimits any noise introduced by a transmission medium.

4. A method for recovering a digital information stream received from a distorted frequency modulation (FM) signal transmitted with a pre- modulation low pass filter having a predetermined filter transfer function, said method comprising the steps of: at a transmitter generating a multilevel digital baseband signal, filtering the multilevel digital baseband signal with a pre- modulation low pass filter Hlpf (f) having a predetermined filter transfer function which minimizes adjacent channel interference and thereby generates a distorted pulse signal, modulating a carrier signal with the distorted pulse signal, creating the distorted FM signal, and transmitting the distorted FM signal; and at a receiver receiving the distorted FM signal, demodulating the distorted FM signal to an analog baseband signal s(t), and extracting the digital information stream from the analog baseband signal s(t) using an equalization filter function Heq(f), which thereby removes intersymbol interference and bandlimits any noise introduced by a transmission medium.

5. The method of claim 4, wherein said extracting step comprises the steps of: generating a sequence of optimized electric pulse signals, with minimal intersymbol interference, by taking a convolution of the analog baseband signal s(t) with an impulse response h_eq(t) of the equalization filter function H_eq(f); and thereafter converting the sequence of optimized electric pulse signals into a sequence of digital data symbols representative of the digital information stream received.

6. The method of claim 4, further comprising the step of storing a sequence of numerical values depicting the equalization filter function H_eq(f).

7. The method of claim 6, further comprising the step of sampling the analog baseband signal s(t) and generating therefrom a baseband digital sequence s(n).

8. The method of claim 7, wherein said extracting step comprises the steps of: generating a discrete time sequence of optimized digital pulse signals, with minimal intersymbol interference, by taking a convolution of the baseband digital sequence s(n) with a unit pulse response h_eq(n) of the equalization filter function Heq(f); and thereafter converting the discrete time sequence of optimized digital pulse signals into a sequence of digital data symbols representative of the digital information stream received.

9. The method of claim 7, wherein said extracting step comprises the steps of: determining a discrete frequency spectrum S(k) by taking a discrete Fourier transform of the baseband digital sequence s(n); multiplying the discrete frequency spectrum S(k) with a discrete frequency spectrum H_eq(k) of the equalization filter function H_eq(f) to generate a discrete frequency spectrum Sι(k); determining a discrete time sequence sι(n), representative of a discrete time sequence of optimized digital pulse signals, by taking an inverse discrete Fourier transform of the discrete frequency spectrum Sid ); and thereafter converting the discrete time sequence sι (n) into a sequence of digital data symbols representative of the digital information stream received.

10. The method of claim 4, wherein said equalization filter function H_eq(f) comprises the steps of: selecting a first pulse Po(t) which exhibits a property of low intersymbol interference and low adjacent channel interference; multiplying the first pulse Po(t) with a first predetermined window function K(t) to obtain a second pulse Pl(t) minimizing energy density outside the first predetermined window function K(t); determining a spectrum Pi(f) by taking a Fourier transform of the second pulse Pι(t); determining a filter function FR(f) which is a reciprocal of a transfer function of the transmitter; multiplying the spectrum Pi(f) with the reciprocal of the filter function FR(f) to obtain a spectrum P2(f); determining a third pulse P2(t) by taking an inverse Fourier transform of the spectrum P2(f); multiplying the third pulse P2(t) by a second predetermined window function W(t) to obtain an optimized electric pulse P3(t); and determining a spectrum for the equalization filter function Heq(f) by taking an inverse Fourier transform of the optimized electric pulse P3(t).