JP2003532311A - Spread spectrum communication system using differential code shift keying - Google Patents

Abstract

(57) [Summary] A spread-spectrum data communication system using a modulation technique called differential code shift keying (DCSK) transmits data in the form of a time shift between continuous cyclic rotating waveforms of length T known as spreading waveforms. I do. A plurality of bits are transmitted during each symbol period, and each symbol period is divided into a plurality of shift indexes, each shift index representing a particular bit pattern. The spread waveform rotates by an amount depending on the data being transmitted, or carries the difference between two consecutive rotations. A correlator (42) using a matched filter having a template of a chirp waveform pattern is used to detect the amount of chirp rotation in the received signal for each symbol. The received data is sent to the shift register (38) and rotated cyclically. For each rotation shift, the matched filter produces a correlation sum. The shift index selected for each symbol corresponds to the shift index that produces the largest correlation sum (44). The difference shift index is generated by subtracting the current reception shift index from the previous reception shift index. Next, the difference shift index is decoded to generate original transmission data.

Description

Detailed Description of the Invention

TECHNICAL FIELD The present invention relates generally to data communication systems, and more particularly to spread spectrum communication systems that transmit and receive data using differential code shift keying.

BACKGROUND ART The use of spread spectrum communication techniques to improve the reliability and security of communications is well known and becoming more and more common. Spread spectrum communication utilizes a much larger spectrum bandwidth to transmit data than the bandwidth of the transmitted data. This leads to more reliable communication in the presence of high narrowband noise, spectral distortion, and pulse noise, among other advantages. Spread spectrum communication systems generally utilize correlation techniques to identify incoming signals.

Spread spectrum communication systems are commonly used in military environments to overcome high energy narrowband hostile jamming. In a commercial or home environment, it can be used to achieve reliable communication over noisy media such as AC power lines.
In particular, certain home appliances and devices can potentially be very disruptive to communication signals carried on power lines. For example, electronic dimmers typically use triacs or silicon controlled rectifiers (SCRs) to control the AC waveforms when implementing the dimming function, so these devices can generate a large amount of noise on the power line. .

Communication media such as AC power lines can be corrupted by fast fading, unpredictable amplitude and phase distortion, and additive noise. In addition, communication channels may be subject to unpredictable time-varying interference and narrowband interference. In order to transmit digital data on such channels, it is preferable to use as wide a bandwidth as possible for the transmission of data. This can be accomplished using spread spectrum technology.

One common type of spread spectrum communication, called direct spread spectrum, is produced by first modulating digital data and then multiplying the result by a signal having particularly desirable spectral characteristics, such as a PN sequence. A PN sequence is a periodic series of bits with period N. Each bit in the sequence is called a chip. This sequence has the property of having a very low autocorrelation for delays greater than one chip. In some systems, the PN sequence is replaced with a chip waveform.

A spread spectrum receiver is required to perform the synchronization typically achieved using acquisition methods in combination with a tracking loop or other tracking mechanism. In noisy and unpredictable environments such as AC power lines, tracking loops typically fail frequently, causing loss of information. Communication systems that overcome these problems are large,
Complex and expensive. In addition, these systems are generally only successful at transmitting 1 or 2 bits per symbol.

SUMMARY OF THE INVENTION The present invention is a spread spectrum technique that utilizes a modulation technique called Differential Code Shift Keying (DCSK) to increase the number of transmitted bits per symbol, reduce synchronization requirements, and improve performance. It is a data communication system. The data is a continuous circular rotation waveform (consecutive circu) of length T, which is called a diffusion waveform.
It is transmitted in the form of a time shift between slightly rotated waves. The spreading waveform can include any type of waveform that has suitable autocorrelation properties. In the example presented here, the standard CEBus chirp can be used as the spreading waveform to allow coexistence of standard CEBus transmissions and transmissions generated by the communication system of the present invention.

During each symbol period, also called the unit symbol time (UST), multiple bits are transmitted. The symbol period is divided into a plurality of shift indexes, each shift index representing a particular bit pattern. The waveform rotates a certain amount depending on the data transmitted. The data is carried in the amount of rotation that is added to the chirp before it is transmitted. Alternatively, the data can be carried with shift differences between consecutive symbols. A correlator is used to decode the received waveform. The correlator uses a matched filter with a chirp waveform pattern template to detect the amount of chirp rotation in the received signal for each symbol. The received data is sent to the shift register and rotated circularly. The matched filter produces a sum of correlations for each rotational shift. The shift index selected for each UST corresponds to the shift index that produces the maximum (or minimum) correlation sum. The differential shift index is generated by subtracting the current receive shift index from the previous receive shift index. The differential shift index is then decoded to generate the original transmission data.

The transmitter sends the data in packets to the receiver. The start of the packet field is placed at the beginning of the packet. The receiver searches for correlation peaks at every possible shift of each symbol received using linear correlation. When the start of a packet field is detected, the receiver searches for two consecutive zeros. Synchronization is achieved when two consecutive zeros are received within the beginning of the packet field. Once synchronization is achieved, the rest of the packet is received using circular correlation. The differential data transmitted by the transmitter is encoded with the shift distance calculated between two consecutive symbols. A differencer in the receiver produces difference data. Then integrating the difference data helps prevent double error effects.

The present invention discloses two embodiments, referred to as a fast embodiment and a reliable embodiment. When using the standard 100 microsecond CEBus chirp, the high speed embodiment is capable of data rates up to about 50 Kbps. Reliable embodiments utilize longer waveform lengths to increase transmission reliability. In addition, the receiver splits the input signal into two or more distinct frequency bands. Each band receives and produces a sum of correlations. The correlation results for each band are combined to determine the maximum correlation shift index. In the example presented here, the trusted embodiment builds a symbol called a super-chirp of 800 microseconds in length from eight individual 100 microseconds of chirp. The entire superchirp is then cyclically shifted depending on the data to be transmitted. A correlator in the receiver with the template of the superchirp pattern is used to detect and decode the superchirp from the received signal.

The spread spectrum communication system of the present invention has the advantages of more reliable transmission, simple and fast synchronization, and immediate recovery from severe fading. In addition, multiple bits can be transmitted per symbol, resulting in longer duration for each symbol, or higher data throughput using the same symbol duration as in a typical direct spread spectrum communication system. You get the rate. Another advantage of the system is that it provides robustness to channels characterized by a frequency varying signal to noise ratio. Furthermore, the present invention is a single V
It can be realized at low cost, such as in an LSI integrated circuit.

Therefore, the present invention is a method of communicating on a communication channel from a transmitter, both of which are connected to the communication channel, to a receiver, which cyclically shifts by an amount depending on the data transmitted in a symbol. Generating a plurality of symbols, each of which is configured using the spread waveform, at the transmitter, placing the plurality of symbols on the communication channel, receiving the signal from the communication channel at the receiver, Decoding the plurality of symbols at the receiver by correlating the signal with a template corresponding to the spreading waveform.

The present invention is also a method of communicating on a communication channel from a transmitter, both of which are connected to the communication channel, to a receiver, which cyclically shifts by a certain amount depending on the data transmitted in a symbol. A step of generating a plurality of symbols configured by using a spread waveform at a transmitter, a step of placing a plurality of symbols on a communication channel, a step of receiving a signal from the communication channel at a receiver, and a received signal With a template corresponding to the spreading waveform to generate a differential shift index that represents a time shift between successive cyclic shifts of the spreading waveform, and decoding the plurality of symbols at the receiver. .

The spreading waveform can include a chirp waveform or a superchirp waveform constructed from multiple individual chirps. Furthermore, the decoding step comprises the step of cyclically shifting each received symbol by a total amount equal to the length of one symbol, and the step of correlating the received symbol with a template corresponding to the spreading waveform for each cyclic shift of the received symbol. , Generating a shift index corresponding to the maximum sum of correlations, and decoding the shift index to generate original transmission data.

The decoding step also cyclically shifts each received symbol by a total amount equal to the length of one symbol, and for each cyclic shift of the received symbol, correlates the received symbol with a template corresponding to the spreading waveform. A step of generating a first shift index and a second shift index respectively corresponding to the positive correlation sum and the negative correlation sum, and decoding the first shift index and the second shift index to decode the first data output and the second shift index. It may also include the steps of respectively generating two data outputs and outputting either the first shift index or the second shift index based on the maximum positive and negative correlation sums.

In addition, the present invention is a transmitter connected to a communication channel, each of which is configured using a spreading waveform that is cyclically shifted by an amount according to the data transmitted in a symbol. A transmitter for generating a plurality of symbols and a receiver coupled to the communication channel, wherein the plurality of symbols are received by receiving a signal from the communication channel and correlating the received signal with a template corresponding to a spread waveform. A spread spectrum communication system for communicating on a communication channel including a receiver for decoding

The present invention also provides a method for generating a signal for transmission on a communication channel using a spread waveform from an input bitstream, the method comprising forming a serial stream of a shift index from the input bitstream; Determining an initial index according to each shift index in the serial stream of indices, cyclically shifting the spreading waveform according to the initial index, and transmitting the cyclically shifted spreading waveform to a communication channel. A method is also provided.

Further, the present invention is a method for generating a spread spectrum signal for transmission on a communication channel using a spread waveform from an input bit stream, the method comprising forming a shift index from the input bit stream, and Formula: initial index = [length of spreading waveform / total number of symbols] -determining an initial index according to the shift index, cyclically shifting the spreading waveform according to the initial index, and cyclically shifting the spreading waveform. Sending to a communication channel. The method further comprises differentiating the input bitstream to generate a differential shift index.

In addition, the present invention is a transmitter for generating a signal for transmission on a communication channel using a spread waveform from an input bitstream, shifting from each group of n bits in the input bitstream. Means for forming an index,
A transmitter is provided that includes means for determining an initial index according to a shift index, means for cyclically shifting a spreading waveform according to the initial index, and means for providing a cyclically shifted spreading waveform to a communication channel.

The means for cyclically shifting the spreading waveform comprises counting means adapted to receive the initial index and lookup table means for outputting sample points of the spreading waveform corresponding to the output of the counting means. . The transmitter further includes a differentiator that subtracts the input bitstream to generate a differential shift index.

The present invention is also coupled to a communication channel for receiving a plurality of symbols each comprised of a spreading waveform that cyclically rotates by an amount depending on the data transmitted during a particular symbol time. A receiver having sampling means for sampling a received input signal and a plurality of taps, shift means for cyclically rotating the output of the sampling means, and a plurality of taps of the shift means And a correlation means for generating a plurality of correlation sums for each cyclic shift of the shift means using a template corresponding to the spread waveform, and a maximum correlation determination for determining a maximum correlation sum from the plurality of correlation sums. Also provided is a receiver including means and a data decoder for decoding the shift index associated with the maximum correlation sum and producing an output therefrom.

Further, according to the present invention, in a receiver connected to a communication channel to receive data encoded as a plurality of symbols, each of which is transmitted by using a spread waveform, a reception input signal is transmitted at a plurality of frequencies. A signal dividing means for dividing into bands, a signal dividing means for outputting a plurality of band pass signals respectively associated with a single frequency band, a sampling means for sampling the plurality of band pass signals, Correlation means for correlating the output of the sampling means for each frequency band, and a correlation means for generating a plurality of band correlation sums for each frequency band, and a plurality of correlations by adding each of the plurality of band correlation sums Summing means for generating the sum, maximum correlation determining means for determining the maximum correlation sum from a plurality of correlation sums, and decoding the received symbol using the maximum correlation sum and outputting from it And a data decoding means for generating the.

In addition, the present invention receives data encoded as a plurality of symbols, each of which is composed of a spread waveform that cyclically rotates by a certain amount according to the data transmitted during a specific symbol time. In the receiver connected to the communication channel, the signal dividing means for dividing the received input signal into a plurality of frequency bands, the signal dividing means outputting a plurality of band-pass signals respectively associated with a single frequency band. A plurality of sampling means for sampling a plurality of bandpass signals, and a plurality of shifting means for cyclically rotating the output of each sampling means associated with each frequency band, A shift means having taps and a plurality of correlation means each connected to one output of the shift means, the shift means using a template corresponding to a spread waveform. Correlation means for each cyclic shift of, to generate a plurality of band correlation sums for each received symbol, and the sum of the plurality of band correlation sums output by each correlation means Addition means for generating a plurality of correlation sums, maximum correlation determining means for determining a maximum correlation sum from the plurality of correlation sums, and data decoding for decoding a shift index associated with the maximum correlation sum and generating an output therefrom And a receiver including a receiver.

The receiver further includes a differentiator coupled to the output of the maximum correlation detecting means, the differentiator producing a differential shift index corresponding to the time difference between the two continuous cyclic rotational spread waveforms.

The receiver further includes an integrator coupled to the output of the data decoder, the integrator for integrating the output of the data decoder. The signal dividing means includes a plurality of band pass filters each having a bandwidth and a center frequency depending on its frequency band. The sampling means comprises a 1-bit analog-digital converter or comparator and a sampling circuit.
The sampling means comprises means for producing both an I or in-phase data stream and a Q or quadrature data stream, where the Q data stream is delayed in time with respect to the I data stream by a predetermined amount.

The correlation means includes complex correlation means. The complex correlation means includes means for applying a non-linear function to the result of the complex correlation. The non-linear function includes a square function.

The present invention is a method for receiving data encoded as a plurality of symbols each transmitted using a spread waveform and transmitted on a communication channel, in which a received input signal is divided into a plurality of frequency bands. Correlating the sample streams associated with each frequency band, generating a plurality of bandpass signals each associated with a single frequency band, sampling the plurality of bandpass signals to generate a sample stream, and Generating a plurality of band correlation sums therefrom, adding a plurality of band correlation sums of the respective bands to generate a plurality of correlation sums, determining a maximum correlation sum from the plurality of correlation sums, and Utilizing the sum of correlations to decode the shift index of each received symbol and generate an output therefrom.

Further, according to the present invention, a plurality of symbols each composed of a spread waveform that cyclically rotates by a certain amount according to the data transmitted during a specific symbol time are encoded and transmitted through a communication channel. A method of receiving data, the method comprising: dividing a received input signal into a plurality of frequency bands; generating a plurality of bandpass signals each associated with a single frequency band; and sampling the plurality of bandpass signals. To generate a sample stream, cyclically rotate the sample stream of each frequency band, and correlate the cyclically rotated sample stream to each frequency band using a template corresponding to the spread waveform. And a step of generating a band correlation sum for each cyclic rotation to generate a plurality of band correlation sums for each symbol A step of adding a plurality of band correlation sums to each frequency band to generate a plurality of correlation sums, a step of determining a maximum correlation sum from the plurality of correlation sums, and a shift index associated with the maximum correlation sum. Decoding and producing an output therefrom.

In addition, the invention provides a method of synchronizing receivers in a spread spectrum communication system for communicating on a communication channel, the method comprising: a transmitter and a receiver both connected to the communication channel, Transmitting a plurality of spread waveforms having a zero differential shift with respect to each other, receiving and decoding the plurality of spread waveforms,
Achieving synchronization by receiving a minimum predetermined number of spreading waveforms with a zero differential shift between them. Each spreading waveform can have a zero rotation shift.

BRIEF DESCRIPTION OF THE DRAWINGS The invention will now be described, by way of example only, with reference to the accompanying drawings. FIG. 1 shows a chirp waveform suitable for use in the spread spectrum communication system of the present invention. FIG. 2 shows the waveform of a sample symbol stream produced by rotating each chirp pattern by an amount representative of the data to be transmitted. FIG. 3 shows a packet structure of the data communication protocol of the present invention. FIG. 4 is a high level block diagram showing the transmitter portion of the first embodiment of the present invention. 5A and 5B are high-level block diagrams illustrating the receiver portion of the first embodiment of the present invention. FIG. 6 is a high level block diagram showing the transmitter portion of the first embodiment of the present invention suitable for transmitting differential data or absolute data with special non-data symbols. 7A and 7B are high-level block diagrams showing the receiver portion of the first embodiment of the present invention in further detail. FIG. 8 is a high-level flowchart showing the preamble and synchronous reception method of the first embodiment of the present invention. FIG. 9 is a high-level flowchart showing the cyclic mode reception method according to the first embodiment of the present invention. FIG. 10 is a high level flow chart illustrating the linear tracking correction method of the present invention. FIG. 11 shows a waveform of a super chirp generated from a plurality of single chirps and including one super UST. FIG. 12 is a high level block diagram showing the transmitter portion of the second embodiment of the present invention. FIG. 13 is a high level block diagram showing the transmitter portion of a second embodiment of the invention suitable for transmitting differential data or absolute data with special non-data symbols. 14A, 14B and 14C are high level block diagrams illustrating the receiver portion of the second embodiment of the present invention. FIG. 15 is a high-level flow chart showing the preamble and synchronous reception method of the second embodiment of the present invention. FIG. 16 is a high-level flow chart showing the cyclic mode reception method of the second embodiment of the present invention.

DETAILED DESCRIPTION OF THE INVENTION The present invention is a spread spectrum data communication system applicable to relatively noisy environments such as AC power lines. Two different embodiments of the spread spectrum data communication system of the present invention are disclosed. The first embodiment, also referred to as the high speed embodiment, includes a transmitter and receiver pair capable of relatively high speed data communication. The second embodiment, also referred to as the trusted embodiment, includes a transmitter and receiver pair that communicates at a lower data rate than the first embodiment, but achieves a higher reliability level. Both embodiments of the present invention are particularly suitable for use in an environment that includes a modem operating according to the CEBus communication standard. In the first embodiment, communication is possible at a higher data rate than the current data communicable rate of the CEBus standard. The CEBus standard is based on the Electronics Industries Association.
on) and is known as the EIA-600 standard.

The spread spectrum system of the present invention transmits data in the form of time shifts between continuous circulating rotation waveforms of length T, called spread waveforms. The spreading waveform can include any type of waveform with suitable autocorrelation properties. The spreading waveform preferably comprises a chirp waveform. An illustration of a chirp waveform, a spread waveform suitable for use in the spread spectrum communication system of the present invention, is shown in FIG. The spreading waveform shown in FIG. 1 is
It spans a duration called the unit symbol time (UST). Multiple bits are transmitted during each symbol period or UST. The symbol period is divided into a plurality of shift indexes, each shift index representing a particular bit pattern.
The spread waveform rotates by an amount that depends on the data to be transmitted. The data is carried in the amount of rotation that is added to the chirp before it is transmitted. Alternatively, the data can be carried in shift differences between consecutive symbols. A chirp typically contains a swept frequency signal. For example, the frequency sweep can range from 200 to 400 KHz, similar to the chirp specified in the CEBus standard, and then 100 KHz to 200 KHz. Alternatively, the chirp can include the swept frequency waveform shown in FIG.

The spread spectrum communication system of the present invention uses differential code shift keying (DC
Send data using the technique known as SK). Using this technique
The data is transmitted in the form of a time shift between continuous circular rotating diffusion waveforms, or in absolute shift itself. The spreading waveform can include standard CEBus chirps to prevent contention in the CEBus modem environment. Other spreading waveforms can be utilized in the present invention in a non-CEBus environment, or when CEBus device interoperability is not critical.

An illustration of the waveform of the sample symbol stream generated by rotating each chirp pattern by an amount representative of the data to be transmitted is shown in FIG. DCS of the present invention
The K modulation scheme transmits data by rotating the chirp waveform by a specific amount depending on the transmitted data. Thus, during each UST, the chirp starts from a point in the chirp waveform that corresponds to the data transmitted during that particular UST. Referring to FIG. 2, the four USTs that make up the sample symbol stream are shown. The data transmitted within each UST is carried by the amount of rotation applied to each chirp waveform. For example, in the first UST, the chirp waveform rotates by a certain amount indicated by the length of the horizontal arrow. The vertically downward arrow indicates the start of the original chirp waveform with no rotation applied. Within each UST, the data transmitted determines the amount of rotation applied to the chirp before transmission.

The DCSK modulation method of the present invention has the advantage of being robust to synchronization errors, relatively easy to implement, and producing performance close to that of error correction codes in the presence of white Gaussian noise. In operation, each UST is divided into a predetermined number of shift indexes or shift positions. In the example presented here, each UST is divided into 32 shift indexes. However, each UST can be divided into a higher or lower number of shift indexes than 32. Dividing each UST into 32 shift indexes means, in other words, a transmission rate of 5 bits per symbol. The high speed embodiment will now be described, followed by the high reliability embodiment.

FIG. 3 shows the packet structure of the data communication protocol portion of the present invention. Typically,
The packet structure of the present invention includes a preamble, a start of packet (SOP) field, an L-byte data field, and a cyclic redundancy check (CRC) field, similar to the standard CEBus packet structure. However, the packet structure of the present invention includes additional fields. The preamble part 450 is similar to the preamble field defined in the CEBus standard. The start-of-packet (SOP) field 452 contains the symbol "1111", which is recognized by the receiver of the invention as three consecutive zeros. It is important to note that the "1" in the packet start field represents an absolute shift of zero, while the term "zero" represents a differential shift of zero.

The term DCSK derives from the fact that the receiver detects the difference in rotation between consecutive symbols received. When the last two zeros in the packet start field are correctly detected, the receiver is synchronized and can continue to receive. Protocol version 45
4 is a 3-bit field containing the protocol version used for that particular packet. The protocol version field enables various types of transmission protocols ranging from, for example, high data rate transmission to low data rate transmission. This field can also implement any user protocol. The protocol version field is transmitted using one symbol and the receiver must have a shift value other than zero in order to detect the end of the packet start field and the start of the protocol version field. In addition, the protocol version field must be transmitted with a fixed number of bits per symbol known a priori to the receiver. This is to ensure that the receiver can receive and decode the protocol version field.
Once decoded, other receive modes can be set. The structure and decoding of the rest of the packet is also determined by the protocol version field.
This includes the number of bits per symbol, ie the number of chirp shifts per symbol time. Generally, 5 bits are transmitted for each symbol constituting one chirp.

The packet length 456 is a 7-bit field indicating the size of the packet in the number of bytes. Generally, the packet size is limited to a specific number such as 128 bytes or 1024 bits. The header error detection code (HEDC) 458 is an 8-bit field containing an error detection code for the protocol version field and the packet length field. The data field 460 contains a series of DCSK data chirps. The start of this field is aligned with the chirp boundary. The length L in bytes of this field is determined by the packet length field. Cyclic Redundancy Check (CRC) field 462 contains a 16-bit error detection field. This field follows the DCSK data chirp continuously without alignment of chirp boundaries. If unused bits remain after the CRC field, they are padded with zeros to the end of the last chirp.

The transmitter portion of the communication system of the present invention will now be described in more detail. A high level block diagram illustrating the transmitter portion of the high speed embodiment of the present invention is shown in FIG. The transmitter and receiver parts of the invention are described for the case where the chirp or symbol time is divided into 32 shift indexes. Therefore, 5 bits are transmitted at a time during each symbol. However, one skilled in the art will be able to modify the receiver and transmitter of the present invention so that the number of bits transmitted per symbol time is higher or lower. Referring to FIG. 4, the host provides the data to be transmitted to the transmitter, generally referenced 12. The host populates the data to be transmitted with a header and a CRC field already generated. The host data used to form the shift index is input to the initial index portion 14 of the transmitter. The shift index includes a number in the range 0 to 2 ⁿ -1, where "n" represents the number of bits transmitted per symbol time.
In the example described here, “n” is 5. Therefore, the shift index is 0
It is a number in the range from 1 to 31. The initial index of the chirp is calculated by dividing the length of the chirp by the total number of symbols in the coding set, eg 2 ⁿ , and multiplying by the shift index, as shown below. Initial index = [chirp length / total number of symbols] · shift index In this example, the chirp length is set to 512. So each chirp is 32
Are divided into 16 indices, each spaced at 16 intervals from each other, ie 0, 1
6, 32 and so on. The initial index is then input to a counter 16 which counts the modulo-chirp length or modular 512. The sync signal serves to clear the counter first. The output of the counter is provided to the address input of the chirp sample read only memory (ROM) 18. This ROM contains a digitized representation of the chirp frequency waveform. The output of the ROM is input to the D / A converter 20, and its analog output is first filtered by a bandpass filter (BPF) 21 having an appropriate passband depending on the signal width. The output of the BPF is then amplified by the output amplifier 22. The output of the amplifier includes the transmit output signal.

The receiver portion of the communication system of the present invention will now be discussed in more detail. A high level block diagram illustrating the receiver portion of the high speed embodiment of the present invention is shown in FIG. The received signal is input to a bandpass filter (BPF) 32, which has a bandwidth wide enough to receive the frequency range transmitted within the chirp. The output of the bandpass filter is input to the 1-bit A / D converter 34. The 1-bit A / D converter can consist of a comparator in combination with a sampling device that is clocked at the appropriate sampling frequency. The output of the A / D converter is input to the shift register # 1 35 and one input of a 2-input multiplexer (mux) 40. The output of the multiplexer is input to the second shift register # 2 38. For purposes of illustration, both shift registers # 1 and # 2 are each 256 bits long. Each of the 256-bit outputs from shift register # 1 is input to shift register # 2. The output of the shift register # 2 is input to the correlator 42. The correlator is implemented with a matched filter that functions to recognize the chirp pattern. The chirp pattern is stored as a template in the correlator and is used to detect the presence of chirp in the received input signal. The serial output of shift register # 2 is circulated to the second input of multiplexer 40. The multiplexer select output is controlled by the linear / circular control signal.

The receiver, generally referred to as 30, is capable of operating in either linear mode or cyclic mode. For linear mode operation, the multiplexer is an A / D
It is set to select the output of the converter as the output to shift register # 2.
Before synchronization occurs, the correlator is set to operate in linear mode. As each bit is received, it is clocked into shift register # 2. The output of the shift register # 2 is input to the correlator 42. Each bit input to the correlator is multiplied by the corresponding bit from the template. The products of all 256 are then added to form the output of the correlator. The multiplication of each input bit and the template bit in the correlator can be realized by using an XOR function.

The sum output of the correlators is input to the maximum correlation detector 44. For each symbol period,
The maximum correlation detector functions to determine the maximum of all 256 sums output by the correlator. The maximum correlation detector outputs two values after detecting the maximum correlation sum. The first value N _MAX indicates the position associated with the maximum correlation sum within the 256 possible shifts. In addition, the value S _MAX represents the particular sum of correlations found to be maximum, which is associated with the position marker N _MAX . Therefore, the value N _MAX can take any value in the range of 0 to 255. The shift index that produces the maximum correlation is then input to the differentiator. The differencer is an adder 50,
A delay unit 52 and a second adder 54 are included. The delay unit 52 functions to delay the input value by one unit symbol time. The output of the delay unit is
Subtracted from the current index value modulo the chirp length via adder 54. The output of adder 54 represents the difference or delta shift between the two shifts detected corresponding to two consecutive symbol times.

The output of the adder and the correlation maximum value S _MAX are input to the differential data decoder 56. The differential data decoder functions to map the shift index, which may be in the range 0 to 255, to a value in the range 0 to 31 depending on the original data transmitted.

In the operation in the circulation mode, the output of the 1-bit A / D converter 34 is the shift register #
136 is input. The received data is clocked in shift register # 1 until it is full. At that point, the shift register contains data representing the complete symbol time. Once full, the contents of shift register # 1 are loaded into shift register # 2 in parallel. Multiplexer 40 is selected to cycle serial data from shift register # 2 to its serial data input. Counter 46 functions to count the length of a chirp, which is typically 1UST wide. The initial sync signal is used to initially reset the counter. The load signal output from the counter is input to shift register # 2 which functions to provide timing for dumping the contents of shift register # 1 to shift register # 2. Shift register # 2 is clocked as many times as there are bits that make up the length of the shift register. For each rotation of the shift register, the correlator produces a sum, which is input to the maximum correlation detector. For every 256 rotations of the shift register, the maximum correlation detector outputs N _MAX and S _MAX values corresponding to the maximum correlation sum and the index that produces the maximum sum itself. The counter provides a counting index, which is input to the maximum correlation detector. This index is
It provides a counter value for each rotation of shift register # 2.

The tracking correction circuit 48 finely adjusts the value of the counter for each symbol. The tracking correction circuit functions to finely adjust the counter value for each symbol. The small difference between the received index and the ideal index is used as an input to the tracking correction circuit 48. A positive or negative error signal is output by the tracking correction circuit and input to the counter. This error signal serves to fine tune the counter value to better track chirp reception and correlation within each symbol time.

The transmitter 12 shown in FIG. 4 transmits data using the absolute transmission mode. In this mode, all 2 ⁿ symbols are transmitted directly, without difference or integration.
To directly determine the rotation shift index for each chirp in UST,
Each 5-bit symbol is used. Therefore, the receiver 30 shown in FIG.
It includes an integrator 62 that functions to integrate the delta shift produced by the differential data decoder. Alternatively, additional transmit and receive modes are possible. For example, the transmitter can be used in differential mode so that the transmitter modulo the chirp length before transmitting the data. Therefore, the receiver must difference the received data in order to receive properly, as it does in the receiver shown in FIG. However, in this case no integrator is needed. In another alternative, the data is first subtracted modulo the number of transmitter shift indices, encoded, and then integrated before being applied to the chirp sample ROM. Therefore, the receiver first subtracts, decodes the output of the differencer, and finally integrates the output of the decoder to form the output of the receiver. The encoder in the transmitter functions to encode all symbols, including both data symbols and non-data symbols.

This last alternative embodiment can be used to encode data symbols (2 ⁿ or any other number) in addition to the extra symbols not in the set of data symbols. In the 5-bit example presented here, this allows a total of more than 32 symbols to be transmitted, some of which are non-data symbols. To achieve this, the chirp symbol time is divided into a number of shift indices greater than 32 to accommodate the extra non-data symbols.

A high level block diagram illustrating the transmitter portion of a high speed embodiment of the invention suitable for transmitting differential or absolute data with extra non-data symbols is shown in FIG. An optional differencer 72 for transmitting data using the difference transmission mode.
Is not required. The host provides the computational initial index unit 80 with data that acts as a shift index. The host calculates the host index within the chirp symbol and inputs it to an integrator 85 that includes an adder 82 and a delay unit 84. The adder 82 adds modulo 2 ⁿ , that is, modulo 32. The output of the adder is delayed and added to the output of the calculation initial index unit 80.
The output of the integrator is a counter 86 that functions similarly to the counter of the transmitter shown in FIG.
Entered in. The chirp sample ROM 88, the D / A converter 90, the bandpass filter (BPF) 91, and the output amplifier 92 function in the same manner as their counterparts of the transmitter shown in FIG.

The receiver portion of the communication system of the present invention will now be discussed in more detail. A high level block diagram showing the receiver portion of the high speed embodiment of the present invention in more detail is shown in FIGS. 7A and 7B. The analog received data is input to a bandpass filter (BPF) 102 having a bandwidth set so as to span the frequency range of the chirp waveform. The output of the BPF filter is input to a 1-bit A / D converter 104, which can be implemented using a comparator in combination with a sampling circuit. The output of the A / D converter is input to the shift register # 1 106 and the multiplexer (mux) 114. The output of the multiplexer 114 is input to the shift register # 2 108,
Its serial output circulates to the second input of the multiplexer. As with the receiver of FIG. 5, the multiplexer is selected by the linear / cyclic control signal.

Linear mode reception is utilized during the preamble for synchronization. Once synchronization is achieved, the cyclic receive mode is utilized to demodulate the rest of the packet. During linear mode operation, the multiplexer is selected to input data from A / D converter 104 to shift register # 2. Each bit is shift register # 2
Each time it is shifted through, correlator 110 produces a sum output, which is input to maximum correlation detector 112. Once synchronization is achieved, the receiver switches to the circular mode of operation. In this mode, the data of the entire UST is shifted into the shift register # 1 and the load command from the counter 116 causes the shift register # 1 to be shifted.
The entire contents of 1 are transferred to shift register # 2. Next, the contents of shift register # 2 are rotated bit by bit through multiplexer 114. The output of shift register # 2 is input to correlator 110, which acts as a matched filter. Each of the 256 bits of shift register # 2 is multiplied by the corresponding bit of the template stored in correlator 110. Then all 256 products are added to produce the output of the correlator. The multiplication can be realized using an XOR gate. Note that after proper conversion, the correlator produces a sum that can take either a positive or negative sign.

Alternatively, the correlator can use less than 256 taps to implement a matched filter. The number of taps used by the correlator is actually shift register # 2.
Can be reduced to almost one-third of the total number of taps corresponding to This is accomplished by sampling the template against both positive and negative threshold values of the template. Positive value templates are compared against a positive threshold and values below this threshold are discarded and not used. Similarly, negative value templates are compared against a negative threshold, values below the negative threshold are discarded and not used. In this way, the number of taps can be reduced by almost two thirds. This effectively improves performance because no noise is introduced from the removed taps and they do not contribute to the correlator output sum.

The correlator output sums are input through a maximum correlation detector that serves to find the maximum correlation sum for both positive and negative sums within each UST. The maximum correlation detector outputs two shift indexes that represent the shift value that achieves the maximum correlation for both positive and negative correlation sums N _POS and N _NEG, respectively. In addition, the corresponding absolute values S _POS , _SNEG of the correlation sums of both positive and negative correlation maxima are also output respectively. The sum S _POS and S _NEG are low pass filters (LPF) 150.
, 152 respectively. The correlation sum is smoothed before being input to the comparator 154. Comparator 154 functions to determine the maximum value between the positive and negative correlation sums. The output of the comparator forms the basis for selecting the positive index N _POS or the negative index N _NEG .

The positive index is input to the positive receive logic circuit 144 and the negative index is input to the negative receive logic 146. Both positive and negative receive logic functions similarly, and only positive receive logic is shown for clarity. The indices output by the maximum correlation detector are first subtracted. The differencer is an adder 126,
A delay unit 128 and an adder 130 are included. The differencer produces the shift delta between each shift index found by the maximum correlation detector. This delta shift value is then rounded to the nearest shift value. In the example presented here, 2
Using a shift register with 56 bits and a symbol with 5 data bits in other words means that the shift indices are 8 bits apart from each other. Therefore, the delta shift index output by the differentiator is rounded to the nearest multiple of 8 bits, ie 0, 8, 16, 24, etc.

The rounded delta shift index is then input to a differential data decoder 136 that functions to decode the shift index into a value between 0 and 31 representing the transmitted data. When the transmitter is set to absolute transmission mode, i.e., linearly coding the data bits into symbols without difference or integration, the output of the difference data decoder represents the difference between the symbols and the transmitted original data In order to restore, we need to integrate. The integrator 148 includes an adder 138 and a delay unit 1
Including 40. The current shift index, which includes values in the range 0 to 31, is 32
Is added modulo the previous output of the adder. This value forms the output data of the receiver and represents the 5 bits originally transmitted.

The output of integrator 148 is input to multiplexer 142, along with the output of the corresponding integrator and negative receive logic 146. The output of comparator 154 serves as the select input to multiplexer 142. Therefore, the index that yields the larger correlation sum is used to determine the received output data.

The receiver also includes a linear mode tracking correction circuit 118 that functions to check for shift indexes near the upper edge of the UST. It is undesirable for a correlation peak to occur near the upper edge of the UST during receiver operation. At very high correlation peaks, the correlation peaks may span the boundary between two UST periods. Therefore, if it is detected that the peak occurs near the UST boundary, the linear mode tracking correction circuit functions to adjust the counter value by about 10% to move the correlation peak away from the UST boundary.

The value output by the linear mode tracking correction circuit is subtracted from the counter.
The modification of the counter value is effective to readjust the symbol reference point of the receiver so that the correlation peak does not cross the boundary between symbols. Note that the linear mode tracking correction circuit operates during the linear mode of operation used to receive the start-of-packet field. Upon completion of tracking and synchronization, the receiver switches to circular mode or reception of the rest of the packet.

In addition to providing counter correction, the linear mode tracking correction circuit also provides a correction signal to the adder 126, which is part of the differencer of the positive receive logic portion 144.
Similarly, the correction signal is also applied to the corresponding adder of the negative reception logic circuit 146. A correction signal to the adder is required to keep the counter value synchronized with the differencer.

In addition, receiver 100 operates to correct clock drift in both linear and cyclic modes of operation. The correction signal is generated based on the difference between the differential shift index both before and after being rounded to the nearest integer shift value. The truncated shift value is input to the adder 134. The rounded shift value is then subtracted from the unrounded shift value, the difference low pass filtered, and counter 1
Used to adjust 16 values. Depending on whether comparator 154 chooses to use a positive or negative shift index, multiplexer 124 functions to pass either value from positive receive logic or negative receive logic to low pass filter 122. . The output of the low pass filter serves to smooth the truncation correction before it is input to summer 120. When the output of adder 120 overflows, the counter is reloaded with zeros for a new count. The correction signal from the low pass filter is subtracted from the current value of the counter by the adder. The clock drift correction signals output from both the positive receive logic adder 134 and the negative receive logic corresponding adder can have either positive or negative signs. Using this technique, synchronization of the counter with the symbol period is maintained.

A high level flow chart illustrating the preamble and synchronous receive method of the high speed embodiment of the present invention is shown in FIG. First, all flags and counters are reset (
Step 160). The linear operating mode of correlation is then set (step 16).
2). The received data bits are shifted into shift register # 2 until the maximum correlation is found (step 164). Once the maximum correlation peak is found, the receiver searches for the next maximum correlation peak. When zero difference is detected (step 166)
, Zero counter is incremented (step 172). A zero difference is detected if the absolute value of the difference between consecutive correlation peaks is less than one half of the minimum delta shift, ie less than 1/2 (1/2 ⁿ ) UST. Where n is the number of bits per initial symbol before the protocol version field is read, eg 3 bits. It is also checked whether the peak value of the correlation sum is larger than a predetermined threshold value (step 174). If the peak correlation value is greater than the threshold value, a "carrier detect" signal is reported (step 176). It then serves to count the number of zero deltas received that are above the threshold, ie the number of deltas equal to zero.
The "high level zero" counter is incremented (step 178). The receiver is considered to be synchronized if a minimum of two high level zeros are received (step 180). Then the time base is corrected according to the value of _{N MAX} (step 182). After the receivers are synchronized, this shift index value represents the offset of the current counter value from the symbol boundary. The counter is adjusted using the value of N _MAX such that each symbol is properly framed, that is, the counter starts counting at the beginning of each symbol. Receiving the rest of the packet then continues using the circular receive mode (step 188).

In connection with step 166, the zero counter is cleared (step 170) if the continuous correlation peaks are not within one half of the minimum delta shift of each other. in addition,
The high level zero counter is also cleared (step 168). The receiver then continues in the linear receive mode, trying to find the maximum correlation peak during the next UST period.

If the peak value is less than the predetermined threshold (step 174), then it is checked whether the zero counter value is greater than 5 (step 184). If the value of the counter is less than 5, the high level zero counter is cleared (step 168).
), The receiver continues to search for the next maximum correlation peak. If the zero counter value is greater than 5, this indicates that a standard CEBus packet is being received, and the receiver switches to standard CEBus reception using the receiver's linear receive mode (step 186).

A high level flow chart illustrating the circular receive mode of the high speed embodiment of the present invention is shown in FIG. Circular receive mode is commonly used to receive a portion of a packet after synchronization. The first step is to find the correlation peak for each of all bits in the UST, 256 shifts of shift register # 2 108 (FIG. 7) (step 190). When a zero difference is detected (step 192), the zero counter is incremented by 1 (step 204). The absolute value of the previous maximum correlation peak position subtracted from the current correlation peak position is less than one half of the minimum delta shift, that is, 1 /
If it is less than 2 (1/2 ⁿ ) UST, a zero difference is detected. Here, n is the initial number of bits per symbol before the protocol version field is read, for example, 3 bits. If the value of the zero counter is less than or equal to 5, control returns to step 190 and the receiver searches for the next maximum correlation peak (step 20).
6). If the value of the zero counter is greater than 5, then a standard CEBus bus packet has been received and the receiver is switched to a linear mode of operation to perform standard CEBus reception (step 208).

If the absolute value of the difference between the two correlation peaks is not within one half of the minimum delta shift, the protocol version field of the packet is decoded (step 1
94). As mentioned above, the Start of Packet (SOP) field contains four symbols with a zero rotation shift. The receiver differentially decodes these symbols as zero. The protocol version field is a single symbol with a non-zero shift, so
Detection of a non-zero delta shift indicates the start of the protocol version field.

Next, the packet length and the header error detection code (HEDC) are read (step 196). If the header error detection code is correct (step 198), the rest of the packet is read (step 200). If the header error detection code is incorrect, the packet is ignored (step 210). When a complete packet is received, an indication of the end of packet and CRC check status is provided (step 2).
02).

A high level flow chart illustrating the linear tracking correction method of the present invention is shown in FIG. As mentioned above, the tracking correction is performed by the linear mode tracking correction circuit 118 (FIG. 7). If the chirp length minus the current receive shift index (represented by N) is less than a predetermined threshold (step 220), the time value ΔT is preferably set to 10% of the chirp length. (Step 222). The counter 116 (FIG. 7), which counts the modulo-chirp length, is modified with this ΔT value (step 226). In particular, the upper limit of the counter is adjusted according to ΔT. In addition, the last positive and negative maximum correlation shift positions are also corrected according to the ΔT value (step 228). If the chirp length subtracted from the current shift index is greater than or equal to the predetermined threshold, then ΔT is set to zero and the counter is unmodified (step 224).

The second or reliable embodiment of the spread spectrum communication system of the present invention will now be described in more detail. The trusted embodiment achieves a higher level of reliability by combining multiple single UST chirps to produce a single superchirp. For example, eight 100 microsecond UST periods can be combined to form an 800 microsecond super UST period. Then, in this super chirp, as for each individual symbol in the fast embodiment described above,
Differential code shift keying (DCSK) is applied.

In a reliable embodiment, the data is transmitted in the form of time shifts between super-chirps that rotate in a circular fashion. In the example presented here, each superchirp has a length of 800
It contains eight standard CEBus chirps that form a microsecond superchirp.
Each individual chirp within the superchirp shifts cyclically by a specific amount. The individual shift amount of each chirp within the superchirp is constant for all transmitted superchirps. The amount of shift or rotation of each chirp is chosen such that the superchirp autocorrelation spurious peaks are relatively low. In addition, the shift of each individual chirp is such that the shift between successive chirps is from zero so that the superchirp is not perceived as a packet start (SOP) of a standard CEBus packet or a packet sent using the fast embodiment of the invention. Selected to be far enough apart.

As with the fast embodiment, the number of bits transmitted within each superchirp determines the number of shift indexes required. FIG. 11 shows an example of the shift index in the case of a specific data sample. In this case, the super chirp starts to be sent from the point of the downward arrow. The transmission is cycled at the end of the waveform, returning to the point of the down arrow.

The packet structure of the communication system of the high reliability embodiment of the present invention is similar to that of the high speed embodiment shown in FIG. The packet structure of the reliable embodiment is a preamble,
It includes a packet start (SOP) field, a protocol version field, a packet length field, a header error detection code (HEDC), a differential code shift data field, and a CRC field. The packet structure of the trusted embodiment is a fast embodiment in that the packet start field of the trusted embodiment contains four superchirp symbols with a zero rotation shift, rather than four normal chirp symbols with a zero shift. Different from that. The four superchirp symbols are perceived by the receiver as three zero differential shifts. Receiving is possible if at least two or more of the last transmitted zeros are correctly detected. At this point, the receiver is synchronized with the received symbol stream. The remaining fields are similar to the corresponding fields in the packet structure of the fast embodiment.

The additional reliability in the second embodiment is achieved by dividing the reception band into two or more equally sized subbands. In the example presented here, the reception band is divided into three equally sized subbands. 4 from 100 KHz
In the case of a chirp waveform with a frequency up to 00 KHz, the three bands are, for example, 10
It can be 0 to 200 KHz, 200 to 300 KHz, and 300 to 400 KHz. The receiver therefore includes three bandpass input filters. The output of each bandpass filter is input to a 1-bit A / D converter that converts the output of each bandpass filter into a binary value. The 1-bit A / D converter may include a comparator followed by a sampling device clocked at the appropriate sampling frequency. 5.12M
Assuming a clock rate of Hz, each band is sampled at a frequency of 320 KHz to form an I or in-phase data stream. The output of the 1-bit A / D converter is also input to a delay unit which delays the signal by an amount equal to 1 / 4f _c . The value f _c represents the demodulation frequency of each passband. The output of the delay unit is sampled at the sampling rate f _s , forming a Q or quadrature data stream. Therefore, the demodulation frequency of the Q sample is delayed by 90 degrees with respect to the I sample in each pass band. The I samples of the three bands are aligned after sampling, but the Q samples are not aligned due to the band dependent delay.

The demodulation frequency f _c of each band is preferably a multiple of ½ of the sampling frequency f _s . If the sampling frequency is 320 KHz, then it is preferable to have the center frequency be a multiple of 160 KHz. The demodulation frequency can be selected to be a multiple of f _s greater than 2 that is closest to the middle of a particular frequency band. In the example presented here, band # 1 is in the range 100-200 KHz, and f _c is 1.
60K is selected. Band # 2 is in the range of 200-300 KHz and f _c is 3
Selected to 20 KHz. Band # 3 is in the range of 300~400KHz, _{f c} is selected to be 320 kHz. Delay between each of the I and Q data streams is equal to 1 / 4f _c, which is at 90 degrees of the particular carrier. The I and Q data streams are then complex-correlated with the complex template to give the real and imaginary correlation sums. These sums are then squared, summed and input to the maximum correlation detector. The maximum correlation sum from all three passbands is determined and used to generate the receiver output for a particular symbol.

A high level block diagram illustrating the transmitter portion of a trusted embodiment of the present invention is shown in FIG. The transmitter, generally referred to at 30, is suitable for generating absolute transmitted data as opposed to differential transmitted data. If it is preferable to send difference data,
An integration step is required. With reference to FIG. 12, data is received from the host. The host generates and adds header and CRC total check in advance. The data is input to the calculation initial index unit 232. The data from the host forms the shift index that is used to compute the index super-chirp. This index is calculated in a similar manner to that of the transmitter of FIG.
The length of the superchirp is divided by the number of possible shifts for each superchirp symbol and then multiplied by the shift index. If each superchirp sends 5 bits, the shift index can include values from 0 to 31. In addition, for the example presented here, the superchirp length is taken to be 2048 samples. Therefore, the initial index includes the numbers 0 to 2047. This initial index is then input to the counter 234. The counter is an 11-bit modulo 2048 counter divided into two parts, a 3-bit length part and an M-bit length part. M is equal to 8 in this case. The 3-bit portion corresponds to 8 chirp periods forming a super chirp symbol. The upper 3 bits are input to the index ROM 36 which functions to output the M-bit value corresponding to the starting point or initial shift index of each individual chirp in the superchirp. These eight initial shift indices are selected a priori for all transmitted symbols and selected to maximize the superchirp autocorrelation. The M bits output by the index ROM are added by adder 237 with the M significant bits from counter 234. The adder 237 adds these two values together, modulo the chirp length of each individual chirp comprising the superchirp, in this case 256. The output of the adder enters the chirp sample ROM 238 which is addressed using the 8 bits output by adder 237. The output of the chirp sample ROM is converted to analog by a D / A converter 240, filtered by a bandpass filter (BPF) 241, and amplified by an output amplifier 242. The output of the amplifier forms the transmit output signal.

The super chirp is constructed with 8 chirps for compatibility with the standard CEBus system. However, the only requirement is that the symbol length be longer than that used in the fast embodiment. Thus, the superchirp may alternatively include a single chirp over the entire symbol length. The use of longer length chirps provides higher reliability due to the increased symbol length and correspondingly more accurate correlation. In addition, higher reliability is also achieved by utilizing multiple bandwidths at the receiver. To increase the reliability of the transmission, the receiver can be configured to utilize only one of these techniques, or a combination.

A high-level block diagram showing the transmitter portion of a reliable embodiment of the invention suitable for transmitting differential or absolute data with extra non-data symbols is shown in FIG.
3 shows. The transmitter of FIG. 13, generally referenced 250, is similar to that shown in FIG. 12, except for the addition of a differencer 252 and an integrator 262.

The transmitter 250 shown in FIG. 13 transmits data using the absolute transmission mode.
In this mode, all symbols, eg 2 ⁿ symbols, are transmitted directly without difference or integration. Each 5-bit symbol is used to directly determine the rotational shift index for each superchirp in the UST. Therefore,
The receiver includes an integrator that functions to integrate the data shift produced by the differential data decoder. Alternatively, additional transmit and receive modes are possible. For example, the transmitter can be used in differential mode, whereby the transmitter integrates the data modulo the superchirp length before transmitting the data. Therefore, the receiver has to differ the received data to receive properly. However, in this case no integrator is needed. In another alternative, the data is first subtracted modulo the number of transmitter shift indices, encoded, and then integrated before being applied to the chirp sample ROM. Therefore, the receiver first subtracts the data, decodes the differencer output, and finally integrates the decoder output to form the receiver output.

This last alternative can be used to encode data symbols (2 ⁿ or any other number) in addition to the extra symbols not in the set of data symbols. In the 5-bit example presented here, this allows the total number of symbols to exceed 32, some of which are non-data symbols.
To achieve this, the superchirp symbol time is divided into more than 32 shift indexes to accommodate the extra non-data symbols.

An optional differencer 252 for transmitting data using the difference transmission mode.
Is not required. The host provides the computational initial index unit 260 with data that acts as a shift index. The shift index within the superchirp symbol is calculated and includes an integrator 26 including an adder 264 and a delay unit 266.
Entered in 2. The adder 264 adds modulo 2 ⁿ , that is, modulo 32. The output of the adder is delayed and added with the output of the calculation initial index unit 260. The output of the integrator is input to a counter 268 which functions similarly to the transmitter counter shown in FIG. Chirp sample ROM 274, D / A converter 276
, Bandpass filter 277, and output amplifier 278 function similarly to their transmitter counterparts shown in FIG.

If the transmitter and receiver pairs are used in differential mode, the initial index data needs to be integrated by integrator 262 before being applied to counter 268. In differential mode, the receiver only needs to differ the shift index output by the maximum correlation sum detector. If the symbol set contains more than 2 ⁿ symbols, then the differencer 252 is needed. Data from the host is the delay unit 25
4 and is subtracted by adder 256 from the current data received from the host. The output of the adder added modulo the number of data symbols, 32, is
It is output to the data encoder 258. The data encoder maps 2 ⁿ data symbols to a particular set of shift indices. A data encoder is required if the symbol coding set contains non-data symbols. For example, the encoding set is
It may include 2 ⁿ data symbols in addition to the extra non-data symbols used for various purposes.

The shift index is used to calculate the initial index input to the integrator. The output of the integrator is delayed by delay unit 266 and added by adder 2
It is added to the initial index by 64. The adder adds the two quantities modulo the length of the superchirp. Counter 268, index ROM 270, adder 272, chirp sample ROM 274, D / A converter 276, filter 277.
, And the output amplifier operates similarly to the transmitter counterpart shown in FIG.

A high-level block diagram showing the receiver portion of a trusted embodiment of the present invention is shown in FIG. 14A.
, 14B and 14C. As mentioned above, the receiver is divided into three pass bands, labeled band # 1, band # 2, and band # 3. The received signal has a band pass filter (BPF) 282 that covers the frequency range of band # 1, a band pass filter 292 that covers the frequency range of band # 2, and a band pass that covers the frequency range of band # 3. It is input to the filter 296. In the case of the chirp pattern in the range of 100 to 400 KHz, band # 1 has a pass band of 100 to 200 KHz, band # 2 has a pass band of 200 to 300 KHz, and band # 3 is 300 to 4
It has a pass band of 00 KHz.

Alternative embodiments may use different numbers of bands to achieve improved reliability, as long as the number of bands is two or more. For three bands,
Two of the bands may be corrupted by noise and cannot be received, but the logic of the receiver in the remaining bands can still output correct data. In addition, splitting the received signal into three bands has the advantage that the receiver has a higher immunity to phase distortion. Although the amount of distortion each band can handle is unchanged, the working of the three bands together contributes to increasing the amount of phase distortion the receiver can handle and still receive properly.

The output of the band pass filter 282 of band # 1 is input to the 1-bit A / D converter 284. The output of the band pass filter 292 of band # 2 is the 1-bit A / D converter 2
It is input to 94. Similarly, the output of the band pass filter 296 of band # 3 is input to the 1-bit A / D converter 298. The output of each 1-bit A / D converter is input to the receiver subunit. In particular, the output of each 1-bit A / D converter is input to the receiver sub-units or receiver logic 322, 324, 326 for each of band # 1, band # 2, band # 3. The receiver logic for each band includes the same circuitry, and therefore only the receiver logic for band # 1 is shown in FIG. 14A for clarity.

The operation of the receiver of FIGS. 14A, 14B, and 14C is similar to that of FIGS. 7A and 7B, except that additional circuitry has been added to perform complex correlation rather than real correlation.
It is similar to the receiver shown in. The output of A / D converter 284 is sampled at sampling frequency f _S by sampling device 288 to form an I data stream. A / D
The output of the converter 284 is also sampled at the sampling frequency f _S by the sampling device 290 after being delayed by a delay circuit 286 for a period of 1/4 f _c . The output of the sampler forms a 90 degree quadrature Q bit stream.

The I and Q data streams are each input to a separate set of shift registers. The shift register # 1 is used for the I bit stream or the in-phase bit stream.
It is input to one of the two inputs of 300 and multiplexer 308.
The output of the multiplexer 308 is input to the serial input of the shift register # 2 302. The serial output of shift register # 2 forms the second input to multiplexer 308.

Similarly, the Q or out-of-phase bitstream is shifted register # 1 304
And to one of the two inputs of multiplexer 310. The output of the multiplexer is input to the serial input of shift register # 2 306. The serial output of shift register # 2 circulates and forms the second input of multiplexer 310.
The linear / circular control signal forms the select input to multiplexers 308, 310.
Both sets of shift registers for the I and Q channels operate similarly to that of the receiver shown in FIGS. 7A and 7B.

Note that the size of the shift register in this embodiment is 256 bits long. Each shift register holds the equivalent of one super-chirp or a chirp that spans eight USTs. The 256-bit I value and the 256-bit Q value from the shift register # 2 are input to the complex correlator 312. The complex correlator is
It functions to multiply the complex input I + jQ by the complex template M _i + jM _q to produce: Output of complex correlator = (I + jQ) × (M _i −jM _q ) = (I · M _i + Q · M _q ) + j (−I · M _q + Q · M _i ) = (Re) + j (Im) correlator The result of the complex multiplication performed by is the real and imaginary sums, each 9 bits wide. The real correlation sum is then squared by the square function number 314 and the imaginary correlation sum is squared by the square function 316. The square of the real sum and the imaginary sum is then added by the adder 3
18 is added. The output of adder 318 forms the output of the receive logic for band # 1. Similarly, receive logic in bands # 2 and # 3 produce similar outputs. The three correlation sum outputs are then added by adder 320. The output of the adder 320 is then input to the maximum correlation detector 382.

The remainder of the receiver shown in FIG. 14C is the reception described in connection with the fast embodiment of FIGS. 7A and 7B, except that only the positive index and positive correlation sum are utilized in the reliable embodiment. Works like a machine. The index N _{P OS} that produces the maximum correlation is input to a differencer including an adder 342, a delay unit 344, and an adder 346. The output of the differential shift index in the range 0 to 255 is rounded to the nearest shift value by the rounding function 348. The rounded shift value is then
It is input to the differential data decoder 350 which decodes the shift index value into a value within the original data sample range, in this case a number within the range 0 to 31. Since the output of the differential decoder is the differential data value, it must be integrated before it is output by the receiver. An integrator that includes adder 354 and delay unit 356 functions to integrate the difference data values to produce received output data.

Similar to the receiver in the high speed embodiment, the I and Q data values can include either +1 or -1, which is a digital binary single bit or 0 or 1.
Can be expressed as In addition, the components M _i and M _q of the complex template are +
It can be either 1, -1, or 0. The multiplication by +1 or -1 is performed using the XOR function, and the multiplication by 0 is performed by not connecting that particular tap to the shift register. Performing a multiplication by 0 by removing taps in this way not only improves performance, but also saves considerable hardware. With this technique, only about one third of the original shift register taps need to be connected and then summed.

The complex template of the complex correlator is calculated by first passing the superchirp through each of the bandpass filters. The output of each bandpass filter is then sampled to produce I and Q data streams. Each sample is then +
It is quantized into three levels: 1, -1, and 0. If the absolute value of the sample is below a certain threshold, it is quantized to zero. Otherwise, the sign is checked, the positive sign is encoded as "1" and the negative sign is "0". This same technique can be used to generate the template of the receiver shown in Figures 7A and 7B.

Interpolation can be used to further increase the resolution of the receiver. This can be accomplished by placing an interpolator at the output of complex correlator 312 before the real and imaginary output correlation sums are squared. The interpolator effectively doubles the sampling rate by producing an intermediate value. The interpolated value can be calculated by adding each consecutive pair of numbers and multiplying by a suitable constant, such as 0.5. 0.6
A multiplication of 25 is useful to achieve a sinx / x or synchronous type approximation of the interpolation function. The output of the interpolator is then squared and added.

The use of interpolation allows the use of lower sampling rates. For example, interpolation can be used to reduce the sampling rate from 320 KHz to 160 KHz. Correspondingly, the demodulation frequency f _c for each of the three bands is a multiple of one half of the sampling rate f _s or a more preferred multiple of 80 KHz.
Can be changed to Therefore, the center frequency of band # 1 can be 160 KHz, the center frequency of band # 2 can be 240 KHz,
Band Center frequency of band # 3 can be 320 KHz.

The receiver starts operating in preamble listening mode. The receiver calculates the maximum absolute value of the complex correlation. In addition, the differential shift and correlation sum thresholds are also checked. When two consecutive zero shifts are detected, "carrier detected" is reported. The sampling windows are then synchronized and the rest of the packet is decoded using a circular correlator.

A high level flow chart illustrating the preamble and synchronous reception method of the reliable embodiment of the present invention is shown in FIG. First, all flags and counters are reset (step 360). The linear operating mode of the correlator is then set (step 362). The received data bits are shifted into shift register # 2 (step 364) until the superchirp or 8 UST maximum correlation is found. When the maximum correlation peak is found, the receiver searches for the next maximum correlation peak. When a zero difference is detected (step 368), the zero counter is incremented (step 372).
The absolute value of the difference between consecutive correlation peaks is less than half the minimum delta shift, ie 1
If it is less than / 2 (1/2 ⁿ ) UST, a zero difference is detected. Where n is the initial number of bits per symbol, for example 3 bits, before the protocol version field is read. It is also checked whether the peak value of the correlation sum is larger than a predetermined threshold value (step 374). If the peak value of the correlation sum is greater than the threshold value, then it is checked whether the value of the zero counter is greater than or equal to 2 (step 378). If the zero counter value is greater than or equal to 2, the linear mode tracking correction circuit 336 of FIG. 14C is used to correct the linear mode tracking (step 380). The rest of the packet is then received using the circular receive mode (step 382). If the peak value of the correlation sum is not greater than the threshold value, then it is determined whether the value of the zero counter is greater than 5 (step 3).
76). If not, control is passed to step 360 and the process begins again. If it is greater than 5, the receiver continues to receive zeroes (step 384).

At step 368, if the difference between successive correlation peaks is greater than or equal to one half of a bit time, then the zero counter is cleared (step 370) and control is passed to step 360.

A high level flow chart illustrating a method of receiving in circular mode in a reliable embodiment of the present invention used to receive the rest of the packet is shown in FIG. In the first step, for each bit in the superchirp, namely shift register # 2 302,3
For the 256 shifts of 06, find the correlation peak (FIG. 14C) (step 390). When a zero difference is detected (step 392), the zero counter is incremented by 1 (step 394). If the absolute value obtained by subtracting the previous maximum correlation peak from the current correlation peak is less than half the bit time, that is, less than 1/2 (1/2 ⁿ ) UST, zero difference is detected. Where n is the initial number of bits per symbol before the protocol version field is read, eg 3 bits. If the value of the zero counter is not greater than 5, control returns to step 390 and the receiver searches for the next maximum correlation peak (step 400). If the value of the zero counter is greater than 5, then the receiver waits until no zero is received and control returns to step 390 (step 402).

If the absolute value of the difference between the two correlation peaks is not within one half of the maximum delta time, the protocol version field of the packet is detected (step 39).
6). As mentioned above, the Start of Packet (SOP) field contains four identical symbols, which are preferably non-rotating superchirps with zero shift.
The receiver differentially decodes these symbols as zero. The detection of a non-zero delta shift indicates the start of the protocol version field, since the protocol version field is a single symbol with a non-zero shift. Once the protocol version is decoded, the shift sensitivity is set accordingly (step 398).

Next, the packet length and the header error detection code (HEDC) are read (step 404). If the header error detection code is correct (step 406), the rest of the packet is read (step 408). If the header error detection code is incorrect, the packet is ignored (step 412). When the rest of the packet is read (step 408), the end of packet and CRC check status is signaled (step 412).

Although the present invention has been described with reference to a limited number of embodiments, it will be appreciated that many variations, modifications, and other applications of the invention are possible.

[Brief description of drawings]

FIG. 1 shows a chirp waveform suitable for use in the spread spectrum communication system of the present invention.

FIG. 2 shows a waveform of a sample symbol stream generated by rotating each chirp pattern by an amount representative of the data to be transmitted.

[Figure 3]   1 shows a packet structure of a data communication protocol of the present invention.

[Figure 4]   FIG. 3 is a high level block diagram showing a transmitter portion of the first embodiment of the present invention.

[Figure 5]   FIG. 3 is a high-level block diagram showing a receiver part of the first embodiment of the present invention.

FIG. 6 is a high level block diagram showing the transmitter portion of the first embodiment of the present invention suitable for transmitting differential data or absolute data with special non-data symbols.

FIG. 7 is a high level block diagram showing the receiver portion of the first embodiment of the present invention in further detail.

FIG. 8 is a high-level flowchart showing the preamble and synchronous reception method of the first embodiment of the present invention.

[Figure 9]   3 is a high-level flow chart illustrating the cyclic mode reception method of the first embodiment of the present invention.

[Figure 10]   3 is a high level flow chart illustrating the linear tracking compensation method of the present invention.

FIG. 11 shows a waveform of a super chirp generated from multiple single chirps and containing one super UST.

[Fig. 12]   FIG. 6 is a high level block diagram showing a transmitter portion of the second embodiment of the present invention.

FIG. 13 is a high-level block diagram showing the transmitter portion of the second embodiment of the present invention suitable for transmitting differential data or absolute data with special non-data symbols.

FIG. 14   FIG. 6 is a high-level block diagram showing a receiver part of the second embodiment of the present invention.

FIG. 15 is a high-level flowchart showing a preamble and synchronous reception method according to the second embodiment of the present invention.

FIG. 16   6 is a high-level flow chart illustrating a cyclic mode reception method according to a second embodiment of the present invention.

─────────────────────────────────────────────────── ─── Continued front page    (81) Designated countries EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, I T, LU, MC, NL, PT, SE), OA (BF, BJ , CF, CG, CI, CM, GA, GN, GW, ML, MR, NE, SN, TD, TG), AP (GH, GM, K E, LS, MW, SD, SL, SZ, TZ, UG, ZW ), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), AE, AL, AM, AT, AU, AZ, BA, BB, BG, BR, BY, CA, CH, CN, C R, CU, CZ, DE, DK, DM, DZ, EE, ES , FI, GB, GD, GE, GH, GM, HR, HU, ID, IL, IN, IS, JP, KE, KG, KP, K R, KZ, LC, LK, LR, LS, LT, LU, LV , MA, MD, MG, MK, MN, MW, MX, NO, NZ, PL, PT, RO, RU, SD, SE, SG, S I, SK, SL, TJ, TM, TR, TT, TZ, UA , UG, UZ, VN, YU, ZA, ZW [Continued summary] Box is decoded and original transmission data is generated.

Claims (54)

[Claims] 1. A method of communicating through a communication channel from a transmitter, both of which are connected to the communication channel, to a receiver, the spread waveform being cyclically shifted by an amount depending on the data carried by the symbol. Generating a plurality of symbols respectively configured in the transmitter, generating a transmission signal according to the plurality of symbols, transmitting the transmission symbols to the communication channel, and the communication channel in the receiver Receiving the transmitted signal from the source and generating a received signal therefrom, and decoding the received signal at the receiver by cyclically shifting the received signal for each cyclic shift. To correlate the received signal with a template corresponding to the spread waveform and generate a correlation sum therefrom. Method comprising a decoding step was, and determining the received data in accordance with a shift corresponding to the maximum correlation sum. 2. The method of claim 1, wherein the spreading waveform comprises a chirp waveform. 3. The method of claim 1, wherein the spreading waveform comprises a superchirp waveform composed of a plurality of individual chirps. 4. The decoding step cyclically shifts each received symbol, and for each cyclic shift of the received symbol, correlates the received symbol with a template corresponding to a spreading waveform to generate a correlation sum. The method of claim 1, comprising the steps of: generating a shift index corresponding to the maximum sum of correlations; and decoding the shift index to generate original transmission data. 5. The decoding step cyclically shifts each received symbol, correlates the received symbol with a template corresponding to a spreading waveform for each cyclic shift of the received symbol, and positively correlates. Generating a first shift index and a second shift index respectively corresponding to a sum and a negative correlation sum, and decoding the first shift index and the second shift index,
Generating a first data output and a second data output, respectively, and outputting either the first shift index or the second shift index based on the maximum value of the positive correlation sum and the negative correlation sum. The method of claim 1, comprising the steps of: 6. A spread spectrum communication system for communicating on a communication channel, the transmitter being coupled to the communication channel, wherein the spread waveform is cyclically shifted by an amount corresponding to the data carried by the symbol. A transmitter for generating a plurality of symbols each configured by using a receiver, and a receiver connected to the communication channel, which receives a signal from the communication channel and cyclically receives the received signal. Said plurality of symbols by shifting and correlating the received signal with a template corresponding to the spreading waveform for each cyclic shift to generate a correlation sum therefrom and determining received data according to the shift corresponding to the maximum correlation sum. And a receiver for decoding the. 7. The decoding step cyclically shifts each received symbol by a total amount equal to the length of one symbol, and for each cyclic shift of the received symbol, the received symbol corresponds to a spread waveform. And a step of generating a first shift index and a second shift index corresponding to a positive correlation sum and a negative correlation sum, respectively, and decoding the first shift index and the second shift index,
Generating a first data output and a second data output, respectively, and outputting either the first shift index or the second shift index based on the maximum value of the positive correlation sum and the negative correlation sum. The method of claim 1, comprising the steps of: 8. The system of claim 7, wherein the spreading waveform comprises a chirp waveform. 9. The system of claim 7, wherein the spreading waveform comprises a superchirp waveform composed of a plurality of individual chirps. 10. A method of generating a signal for transmission on a communication channel from an input bitstream using a spread waveform, the method comprising: forming a serial stream of shift indexes from the input bitstream; Determining an initial index according to each shift index in the shift index of the stream, cyclically shifting the spreading waveform according to the initial index, and transmitting the cyclically shifted spreading waveform on the communication channel. Method including and. 11. A method of generating a spread spectrum signal for transmission on a communication channel from an input bitstream using a spread waveform, the method comprising the step of forming a shift index from the input bitstream: Initial index = [spread waveform length / total number of symbols] determining an initial index according to a shift index, cyclically shifting the spread waveform according to the initial index, and transmitting the cyclically shifted spread waveform to the communication channel Transmitting above. 12. The method of claim 11, wherein the spreading waveform comprises a chirp waveform. 13. The method of claim 11, wherein the spreading waveform comprises a superchirp waveform composed of a plurality of individual chirps. 14. The method of claim 11, further comprising the step of differencing the input bitstreams to generate a differential shift index. 15. A transmitter for generating a signal for transmission on a communication channel from an input bitstream using a spread waveform, wherein a shift index is formed from each group of n bits of the input bitstream. Transmission including means, means for determining an initial index according to the shift index, means for cyclically shifting a spreading waveform according to the initial index, and means for transmitting the cyclically shifted spreading waveform on the communication channel Machine. 16. The transmitter of claim 15, wherein the spreading waveform comprises a chirp waveform. 17. The transmitter of claim 15, wherein the spreading waveform comprises a superchirp waveform composed of multiple individual chirps. 18. The means for cyclically shifting the spreading waveform outputs counting means adapted to receive the initial index, and sampling points of the spreading waveform corresponding to the output of the counting means. 16. A transmitter as claimed in claim 15 including lookup table means. 19. The transmitter of claim 15, further comprising a differentiator that subtracts the input bitstream to generate a differential shift index. 20. A receiver coupled to a communication channel, the receiver for receiving data encoded as a plurality of symbols each transmitted using a spread waveform, wherein a receiver input signal is 1 Signal dividing means for dividing into one or more frequency bands and outputting one or more band-pass signals each associated with a single frequency band; and sampling the one or more band-pass signals And a correlating means for correlating the output of the sampling means with respect to each frequency band to generate one or more band correlation sums respectively corresponding to different frequency bands, and the one or more Adding means for adding the above band correlation sums to generate a total correlation sum; maximum correlation detecting means for determining a maximum correlation sum from a plurality of total correlation sums calculated over a certain period; Receiver for decoding the received symbols using the sum of correlations and producing an output therefrom. 21. The receiver of claim 20, wherein the spreading waveform comprises a chirp waveform. 22. The receiver of claim 20, wherein the spreading waveform comprises a superchirp waveform composed of multiple individual chirps. 23. The receiver of claim 20, further comprising means for synchronizing the correlating means with symbol time. 24. The receiver of claim 20, further comprising a differencer coupled to the output of the maximum correlation detection means to generate a difference shift index corresponding to a time difference between two continuous cyclic rotation spread waveforms. 25. The receiver of claim 20, further comprising an integrator coupled to the output of the data decoder for integrating the output of the data decoder. 26. A receiver as claimed in claim 20, wherein said signal splitting means comprises one or more bandpass filters, each bandpass filter having a bandwidth and a center frequency according to its corresponding frequency band. 27. The receiver according to claim 20, wherein said sampling means comprises a 1-bit analog-to-digital converter. 28. The receiver of claim 20, wherein the sampling means comprises a comparator and a sampling circuit. 29. The sampling means is I or an in-phase data stream and Q.
21. The receiver of claim 20, further comprising means for generating both orthogonal data streams, the Q data stream being delayed in time by a predetermined amount with respect to the I data stream. 30. The receiver of claim 20, wherein the correlating means comprises a complex correlating means. 31. The receiver of claim 30, wherein the complex correlation means comprises means for applying a non-linear function to the complex correlation result. 32. The receiver of claim 31, wherein the non-linear means comprises a squared function. 33. The receiver of claim 20, further comprising tracking means for maintaining alignment of the correlating means with a symbol time. 34. A method for receiving data transmitted as a plurality of symbols, each of which is encoded using a spread waveform and transmitted on a communication channel, wherein a received input signal is divided into a plurality of frequency bands. Generating a plurality of bandpass signals each associated with a single frequency band, sampling the plurality of bandpass signals to generate a sample stream, and correlating the sample streams associated with each frequency band. , A step of generating a plurality of band correlation sums therefrom, a step of adding a plurality of band correlation sums to generate a plurality of correlation sums, a step of determining a maximum correlation sum from the plurality of correlation sums, Decoding the shift index for each received symbol utilizing the maximum correlation sum and producing an output therefrom. 35. The method of claim 34, wherein the spreading waveform comprises a chirp waveform. 36. The method of claim 34, wherein the spreading waveform comprises a superchirp waveform composed of a plurality of individual chirps. 37. The method of claim 34, further comprising synchronizing the correlation to symbol time. 38. The method of claim 34, further comprising the step of subtracting the shift index to generate a differential shift index corresponding to a time difference between two continuous circular rotational spreading waveforms. 39. The method of claim 34, further comprising integrating the output of the decoding step. 40. The dividing step comprises bandpass filtering each frequency band having a bandwidth and a center frequency according to each frequency band.
The method according to 4. 41. The method of claim 34, wherein the sampling step comprises providing a 1-bit analog-to-digital converter and applying to each bandpass signal. 42. The method of claim 34, wherein said sampling step comprises the step of providing a comparator and sampling circuit for applying to each bandpass signal. 43. The step of sampling includes the step of generating both an I or in-phase data stream and a Q or quadrature data stream, the Q data stream being timed with respect to the I data stream by a predetermined amount. 35. The method of claim 34, wherein the method is delayed. 44. The method of claim 34, wherein the correlating step comprises applying a complex correlation to the sample stream. 45. The method of claim 44, wherein applying the complex correlation comprises applying a non-linear function to a result of the complex correlation. 46. The method of claim 45, wherein the non-linear function comprises a squared function. 47. The method of claim 34, further comprising maintaining a match of the correlation to symbol time. 48. A spread spectrum communication system for communicating in a communication channel, both of which includes a transmitter and a receiver connected to the communication channel, wherein a plurality of spread waveforms having a differential shift of zero to each other are provided in the communication channel Transmitting above; receiving the plurality of spread waveforms from the communication channel to generate a received signal; decoding the received signal and having a minimum predetermined shift between them. Declaring synchronization when it receives a number of spread waveforms. 49. The method of claim 48, wherein the spreading waveform comprises a chirp waveform. 50. The method of claim 48, wherein the spreading waveform comprises a superchirp waveform composed of a plurality of individual chirps. 51. The method of claim 48, wherein each spreading waveform has a zero rotational shift. 52. A method of communicating on a communication channel from a transmitter, both of which is connected to the communication channel, to a receiver, each comprising a spreading waveform cyclically shifted by an amount dependent on the data carried by the symbol. Generating at the transmitter a plurality of symbols according to the plurality of symbols for transmitting on the communication channel, the receiver receiving the transmission signal from the communication channel, wherein At the receiver by correlating the received signal with a template corresponding to a spreading waveform to generate a differential shift index representative of a time shift between successive cyclic shifts of the spreading waveform. Decoding the received signal. 53. Receiving data transmitted on a communication channel encoded as a plurality of symbols each comprised of a spreading waveform that is cyclically rotated by an amount corresponding to the data transmitted during a particular symbol time. A method of dividing a received input signal into a plurality of frequency bands to generate a plurality of bandpass signals each associated with a single frequency band; and sampling the plurality of bandpass signals to generate a sample stream. And cyclically rotating the sample stream for each frequency band, and correlating the cyclically rotating sample stream for each frequency band using the template corresponding to the spreading waveform. Generating a band correlation sum for rotations to generate a plurality of band correlation sums for each symbol; Adding a number of band correlation sums, generating a plurality of correlation sums, determining a maximum correlation sum from the plurality of correlation sums, decoding a shift index associated with the maximum correlation sums, Generating output from the method. 54. A communication channel for receiving data encoded as a plurality of symbols, each of which is composed of a spreading waveform that is cyclically rotated by an amount corresponding to the data transmitted during a particular symbol time. A combined receiver, which divides a received input signal into a plurality of frequency bands and outputs a plurality of band pass signals each associated with a single frequency band, and a sample of the plurality of band pass signals. A plurality of sampling means for converting the output of each sampling means associated with each frequency band in a cyclic manner, and a plurality of shifting means each having a plurality of taps; For each received symbol, a correlation sum is generated for each cyclic shift of the shift means using a template corresponding to a spread waveform, each of which is connected to one output. A plurality of correlating means for respectively generating a plurality of band correlation sums, a plurality of adding means for respectively adding a plurality of band correlation sums output by the respective correlating means, and a plurality of correlations A receiver comprising a maximum correlation determining means for determining a maximum correlation sum from the sum and a data decoder for decoding a shift index associated with the maximum correlation sum and generating an output therefrom.