MXPA99005191A - Method for correcting errors from a fading signal in a frequency hopped spread spectrum communication system - Google Patents

Abstract

A selective call communication system (10) transmits a frequency hopped spread spectrum signal as a Reed Solomon code word modulated as a four-level frequency shift keying (4FSK) signal. A base site transceiver (150) receives the signal on a plurality of narrow band channels and a DSP (152) performs an FFT on the signal. The DSP (152) has a comparator (342) that computes energy of a 4FSK symbol. The comparator (342) establishes a ratio of a maximum energy and a next largest energy to an eye-opening threshold to indicate a probability of error. A determinator (344) in response to the probability of error determines when there is a fade, an erasure marker (346) marks a position of a Reed Solomon symbol in the fade as an erasure and error correcting code (348) corrects errors in the Reed Solomon code word with the marked erasure.

Description

METHOD TO CORRECT THE ERRORS OF A WAVE SIGNAL IN AN EXTENDED SPECTRUM COMMUNICATIONS SYSTEM WITH FREQUENCY JUMP FIELD OF THE INVENTION This invention relates in general to communication systems and, more particularly, to an extended spectrum communication system and to a method for determining channels with fading interference and channels jammed by interference.

BACKGROUND OF THE INVENTION The idea of the extended spectrum was first used during World War II to combat intentional clogging and to exchange information in a secure manner. Extended spectrum systems must meet at least two criteria. First, the transmitted bandwidth must be much greater than the bandwidth or speed of the information that will be sent and second, some other function is used other than the information that will be sent to determine the radio frequency bandwidth (RF) ) modulated resulting. . In this way, the essence of extended spectrum communications includes expanding the bandwidth of a signal, transmitting that expanded signal and recovering the desired signal upon return to P1309 / 99MX ß "functionally correlate or regroup the extended spectrum received in the bandwidth of the original information for the purpose of providing error-free or error-free information in a noisy signal environment." Numerous schemes have been developed to satisfy these two requirements but, normally, excessive interference, which includes interference from narrow-band interferers, invalidates or spoils these systems. The extended spectrum consists of two different modulation schemes, namely the Extended Direct Sequence Spectrum (DSSS) and the Extended Spectrum Frequency Skip (FHSS). The DSSS modulation uses a high-speed code to extend the data over a large bandwidth, while the FHSS modulation includes rapidly changing the carrier frequency of the narrow-band data signal. The present invention is applies particularly to the FHSS systems but, other applications could be contemplated within the scope of the claims. In 1985, the Federal Communication Commission (FCC) assigned the Industrial, Scientific and Medical bands (ISM) (three unauthorized bands: 902-928 MHz, 2.4- 2.4835 GHz and 5.725 -5.870 GHz) for general purpose communication while spectrum communication Extended P1309 / 99MX is used in accordance with the regulations in part 15. The FCC's deon has had a tremendous impact on the commercial world. Thousands of devices for the ISM band are currently available in the market. Because the spectrum is not authorized, hundreds of other wireless applications are about to arrive. This will cause a drastic increase in the level of interfering noise and may cause the whole band to collapse. The present invention proposes a way to survive in this noisy environment which additionally serves as an effective method to cancel the interference. Thus, what is needed is a method and apparatus for lowering or minimizing interference in an extended spectrum communication system.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is an electrical block diagram of an extended spectrum selective call communication system with frequency hopping, in accordance with the preferred embodiment of the present invention. Figure 2 is a graphical representation illustrating a particular frequency skip pattern in the time-frequency plane, in accordance with the P1309 / 99MX preferred embodiment of the present invention. Figure 3 is a block diagram of function of a selective calling device and the terminal transceiver, in accordance with Figure 1. Figure 4 is a matrix representation of the code words or keywords to be interleaved in accordance with the preferred embodiment of the present invention. Figure 5 is a frequency output that illustrates the frequency offset of level 4 FSK symbols, in accordance with the preferred embodiment of the present invention. Figures 6 and 7 represent the frequency domain output of the 4FSK frequency deviations, in accordance with the preferred embodiment of the present invention. Figure 8 is a flow chart illustrating the operation of the terminal controller that determines when a channel is stuck, in accordance with the preferred embodiment of the present invention. Figure 9 is a matrix representation of the erasure marks of the error corrections when the FH signal is stuck, in accordance with the preferred operation of the invention. Figure 10 is a flow diagram that P1309 / 99MX illustrates the operation of the terminal controller that determines at what time a channel is fading in accordance with the preferred embodiment of the present invention. Figure 11 is a matrix representation of the erasure marks of the error corrections when the FH signal is fading, in accordance with the preferred operation of the invention. Figure 12 is a flowchart illustrating the operation of the terminal controller to correct errors resulting from jammed and / or fading signal interference, in accordance with the preferred embodiment of the present invention. Figure 13 is a matrix representation of the erasure marks of the error corrections when the FH signal is fading and / or when the FH signal is jammed in accordance with the preferred operation of the invention.

DESCRIPTION OF A PREFERRED MODE Referring to Figure 1, a selective calling system (selective calling communications system) is shown as an extended-spectrum selective hopping system (FH) 10, which illustrates a terminal 150 of base site and a selective call device 100 in accordance with P1309 / 99MX the preferred embodiment of the present invention. The selective call terminal 150 comprises a terminal controller 152 coupled to a transceiver 156 of the base site. The terminal controller 152 receives messages from an input device, for example, from a telephone, a computer or an alpha input device or the like (not shown) and by the transceiver 156 and the antenna 154 transmits the messages to the selective calling device 100. The received message (or information) is processed by encoding the message with an address designating the selective calling device 10Q, as is well known in the art. The coded message then passes to the antenna 154 that transmits (or receives) the radio frequency (RF) message to the selective calling device 100 (or thereafter). The transceiver 156 is coupled to, or comprises one of these, well-known in the art, a mixer or a frequency converter 158. The mixer 158 is coupled to a frequency synthesizer 160 to allow the transceiver 156 to tune or select a plurality of frequencies to receive (or transmit) a selective call message over the plurality of frequencies. As is well known, an extended spectrum communication system with frequency hop jumps (or switches) to the plurality of frequencies when the portion of the message or information P1309 / 99MX is transmitted and / or received. A pseudo-random sequence (PN) generator 162 generates a sequence of numbers under the control of a time synchronizer 164 which is synchronized with the PN generator of a transmitter in the selective calling device. The sequence values that are generated by the PN sequence generator 162 are used by the frequency synthesizer 160 and the mixer or by a frequency converter 158 to tune the transceiver to the different frequencies (or jumps) of the signal extended spectrum with frequency hopping, when different portions of the message are transmitted in each of the plurality of jumps. In an extended spectrum communication system with frequency hopping (FH), the information is transmitted and received over a broadband frequency, for example, the Industrial, Scientific and Medical (ISM) band. There are three ISM bands, 902-928 MHz, 2400-2483.5 MHz and 5725-5850 (MHz). For the reverse or descending selective call channel it is preferred to operate at 902-928. However, the applications of the invention are not limited to this band. As is well known, a frequency hopping communication system uses an extended spectrum technique, because the modulated data carrier jumps from one narrowband frequency channel to another frequency randomly within a frequency of P1309 / 99MX specified broadband. These frequency jumps are controlled by a random sequence generator to allow frequency hopping over a plurality of narrowband channels. A corresponding receiver has the ability to duplicate the same random sequence as the random carriers. In the extended spectrum system with FH, the early correction of the error (FEC) is always important. With the FEC, the message that can be retrieved by the FEC when some of the random channels (or carriers) are jammed by interference and / or fading. Two basic characters of the systems with FH are the slow frequency jump (SFH) and the fast frequency jump (FFH). The SFH system transmits several symbols in each frequency hop and the FFH system will jump several times during the transmission of a symbol. Referring to Figure 2, a graphical representation is shown illustrating an example of a particular frequency skip pattern in the time-frequency plane. The graph is an example of the frequency jumps and time slots of each jump. The broadband frequency is subdivided into a plurality of contiguous frequency slots (1-6) over the time intervals (Tc-to-6Tc). For example, in a signaling interval, the transmitted signal occupies one or more of the signal slots.

P1309 / 99MX frequency available, preferably a frequency slot. The selection of the frequency slot corresponds to the signaling intervals (Tc-6Tc), and they are carried out pseudo-randomly, in accordance with the output of the sequence generator PN 162. The frequency synthesizer 160 generates a number corresponding to one of the frequency jumps during an appropriate time interval, for example, Te, which causes the mixer a frequency converter 158 to tune to the narrow band frequency slot, eg, the frequency or hop slot 1, to receive the information that is transmitted. After the information is received during the frequency hop 1, the generator PN 162 generates the next number which causes the frequency synthesizer 160 and the mixer a frequency converter 158 to tune the transceiver 156 to the next frequency hop 2 during the interval of time, 2Tc. The generator PN 162 under the control of the time synchronizer 164 continues to generate numbers to cause the frequency synthesizer 160 and the mixer a frequency converter 158 to tune to, for example, the other frequency slots 3-6 corresponding to the intervals 3Tc-6Tc. The transceiver 156 receives the information transmitted during the plurality of frequency hopping of the spectrum system with frequency hopping to receive the information or the P1309 / 99MX message that is transmitted by the selective calling device. The message or information is stored in the memory 166. The selective call device 100 (eg, a selective call receiver with an acknowledgment or acknowledgment return transmitter) transmits an input signal as a signal with a jump of frequency (FH) in response to receipt of a message from terminal 150 of the base site. The output signal from the base site terminal can be received in any other signaling protocol, preferably the protocol FLEX TM. The bandwidth of the input signal of each hop is equal to the bandwidth of the signal with M-th FSK modulation. The input signal FH, according to the preferred embodiment, is modulated as a signal with four level frequency offset (4FSK) and coded as a Reed Solomon code word or key "RS (15.5)". The designed system uses the 4FSK at 400 symbols per second (800 bits per second). Preferably, the signal bandwidth of each hop is 3.125 KHz. The preferred message length of the downlink or reverse selective call channel is 80 bits. Every 20 bits are encoded in a code word Reed-Solomon (15, 5), RS (15.5). There is a total of 4 code words per message. The Reed-Solomon code is P1309 / 99MX well known in the art as a Bose Chaudhuri-Hocquenghem code (BCH) non-binary. It is well known that other code words could also be used. The RS (N, K) is a code that codes K information symbols in a code word with N symbols. The N symbols are equal to (N) = 2 m-l, where m is the number of bits in each symbol. The word code RS can correct a total number of errors (t), and a number of erasures (e), if 2t + e is less than (N-K) + l. An error is defined as a transmission error of which both the location and the value are unknown and an erasure is defined as an error whose location is known but whose value is unknown. For example, a code word RS (15, 5) has a total of 15 symbols, in which 5 are information symbols. Each symbol has 4 bits. The word code RS (15,5) can correct 5 symbol errors without any deletion. However, you can correct up to 10 transmission errors if they are correctly marked as erasures. An increase in capacity is obtained, because the location of the errors is known. The selective calling device 100 comprises an antenna 102 that provides an RF carrier signal to the receiver 104. The receiver 104 generates a recovered signal suitable for processing by a signal processor Digital P1309 / 99MX ("DSP") 106 in a manner well known to a person of ordinary skill in the art. The DSP 106 performs functions such as message coding and decoding and control of the operation of the selective call device 100 well known to those skilled in the art. The DSP 106 processes the received signal to decode the address and compares the decoded address with one or more predetermined addresses contained in a memory, for example, a code plug 118. When the addresses are very similar, the user is alerted that a signal has been received either through an audible or audio alert (eg, a horn or a transducer) 112 or a tactile alert (eg, a vibrator) 114. The received signal may also include optional message data directed towards some selective calling device. Also, if the selective calling device 100 includes an optional voice output, the audio components recovered from the received RF signal may be presented. For a message selective calling device, the retrieved message is stored in a memory 128 for subsequent presentation by an output device, which is, for example, a screen 108. The output device automatically or, when manually selected through the switches 116, will present the message, such as, for example, P1309 / 99MX display or display the message on the screen 108. For a selective callback receiver 100 recognition or acknowledgment (or two-way selective calling device), a signal is transmitted, either automatically or manually, in response to the reception of the selective call message (or signal) by the selective call device 100. The user, in the case of a manual response, uses the switches or switches 116 to select a message from, for example, one of several messages pre-stored in the memory 128 that will be sent to the originator. The message is then encoded by the DSP 106 and passed to a transmitter 126 for transmission by the antenna 102. Preferably, the message is modulated by the DSP 106 and is coded in the code word RS (15.5) in a form well known in the art. The DSP modulates in the word code RS to the signal 4FSK for the transmission. After the signal is encoded and modulated and, in accordance with the preferred embodiment of the present invention, a PN sequence generator 120 generates a sequence that is used by the frequency synthesizer 122 to generate a frequency output. The frequency output is supplied to a mixer by a frequency converter 124 which sets the frequency of the transmitter 126 in, for example, the frequency slots 1 to 6 during the P1309 / 99 X time intervals Tc-6Tc. The transmitter 126 transmits different portions of the message during the frequency slots 1-6, until the entire message is received at the terminal 150 of the base site. It is understood that the PN sequence generator 162 in the controller 152 of the base site is identical to the generator PN 120 in the selective call device 100. The time synchronizer 164 synchronizes the transceiver at the terminal 150 of the base site with the transmitter 126 in the selective call device 100 in a manner well known in the art. Referring to Figure 3, there is shown, in accordance with Figure 1, a functional block diagram of the receiver FH (receiver at the terminal 150 of the base site) and the transmitter FH (the transmitter in the selective call device 100). An encoder 302 for the advance correction of the error (FEC) encodes the information or message of the selective call device 100 with the code word RS (15,5). The resulting RS code word is interleaved by interleaver 304 which generates the spread spectrum signal with interleaved frequency hopping. Referring to Figure 4, interleaver 304 forms a fifteen-by-four matrix. Each column of the matrix represents an RS code word. There are four columns that represent four RS code words. The four words RS code are P1309 / 99MX stored in columns and transmitted in rows. The matrix is stored in memory 128 (Figure 1), before transmission. Each block represents a bit 402, two blocks (or two bits) represent the symbol 4FSK 404, and four blocks (or four bits) represent the Reed Solomon symbol (or RS symbol) 406. The memory 128 has the ability to store the four RS symbols and other output messages received. In accordance with Figure 3, the 4FSK 306 modulator modulates each bit in the FSK signaling scheme of level four (4FSK) followed. Referring to Figure 5, the index k = 0, 1, 2 and 3, marks the deviation of the frequency of the 4FSK signal. Each index corresponds to one of the four frequency shifts of the 4FSK signal. An FH 308 modulator modulates portions of the 4FSK signal, e.g., each row or row of the matrix (Figure 4), in one of the frequency slots of the spread spectrum signal with frequency hopping. The PN sequence generates a number that is used by the frequency synthesizer 310 to generate the frequencies corresponding to the frequency slots of Figure 2. During each hop, the transmitter 312 transmits a row or row of the matrix of Figure 4. Each row corresponds to four RS symbols or eight 4FSK symbols. For example, during the time period Te, the row or row 1 of the P1309 / 99MX matrix and during the 2Tc time period, line 2 of the matrix is transmitted, etc. The data or signal FH 320 is transmitted as a radio frequency signal on a wireless channel that is susceptible to interference 322 in the form of other signals that clog the signal of the narrow bands. Interference 322 also includes fading, multipath fading, or random-type errors well known to any person of ordinary skill in the art. A receiver 330, for example, in the base site controller, receives the FH 320 signal plus the interference. An FH demodulator 332 is coupled to a frequency synthesizer 334, which receives a PN sequence similar to and which is synchronized with the PN sequence of the transmitter (selective call device 100). The PN sequence is shown connected by a dotted line to the PN sequence of the transmitter to illustrate that the PN sequence is synchronized with the transmitter. In order to modulate the 4FSK signal, a fast Fourier transform (FFT) 336 is applied to each 4FSK symbol. As is well known, the sampling frequency of the signal must be fast enough to satisfy the Nyquist velocity well known to those of ordinary skill in the art. For example, an FFT of L points is required, where L is equal to the sampling frequency P1309 / 99 X multiplied by the duration of the sample. Preferably, the symbol rate is 100 4FSK symbols per second at 800 4FSK symbols per second, a higher symbol rate requires a higher computational speed of the DSP. A sampling rate of four times the symbol rate is preferred to ensure that the 4FSK symbols are decoded correctly. As is well known, a sampling and holding circuit (not shown) samples the time domain signal and an analog-to-digital converter (ADC) converts the time sample into digital values before the FFT transforms the domain signal from digital time in a frequency domain signal suitable for use by a 4FSK demodulator 338 to demodulate 4FSK signals. The FFT samples comprise a plurality of frequency samples. Referring to Figures 6 and 7, the output of the frequency domain is shown as a representation of the deviations or displacements 4FSK indicated by the indices k = 0, 1, 2 and 3, which correspond to the center of each of the four deviations frequency of the 4FSK symbols. The 4FSK demodulator 338 demodulates the 4FSK symbols after the FFT transforms the time domain signal into the frequency domain representation of the same to demodulate the 4FSK signal in the domain of P1309 / 99MX frequency of a subset of frequency samples. The frequency derivation of the 4FSK signal is represented as Fj, k, where the jth index corresponds to each frequency hop containing the information shown in the rows or rows of the matrix of Figure 4. The 4FSK demodulator 338 generates a matrix of four times J (4xJ) that is stored in memory 166 (Figure 1). A comparator 342 measures the energy, magnitude or value of the frequency offset or frequency deviation of the 4FSK symbols corresponding to the intervals k = 0-3 for all J values, where J = 0 to 7, the number of 4FSK symbols transmitted in each frequency hop. The comparator 342 compares the measured frequency offset with a predetermined threshold, the details for determining the predetermined threshold will be discussed below. A determiner 344 determines that the 4FSK symbols get stuck when there is no variation or there is very little variation in the magnitude or energy of the FFT output from symbol to symbol to a frequency offset or shift. As is well known, a narrowband interference signal could block or jam the entire signal contained in a frequency hop. Referring to Figure 6, when a frequency slot or jump jams, the energy or magnitude P1309 / 99MX of the displacement frequency 4FSK, for example, the frequency offset corresponding to the interval k = 0, each of the eight 4FSK symbols indicates a magnitude or energy practically equal to all eight 4FSK symbols of the interval k = 0 and the energy of the interval k = 0 is substantially greater than the energy at the other frequency shifts of k = the 3. Referring to Figure 7, in the normal environment when the signal is not clogged, the magnitude or energy of the output FFT will vary from frequency shift to frequency shift through all eight 4FSK symbols transmitted in the frequency hop. As illustrated, the energy or magnitude of the displacements 4FSK k = 0 to 3 of all eight 4FSK symbols will vary substantially between the indices k = 0 to 3 and the indices J = 0 to 7. Returning to Figure 3, a marker of erasure 346 marks all location of the bits that have a higher probability of errors, the erasure mark will be discussed in detail later. A deinterleaver 340 deinterleaves the data and the results of the erasure marker 346 and deinterleaver 340 are passed to a FEC decoder 348 which performs error correction to correct errors based on interference by a clog and fading signal. The correction of P1-309 / 99 X error will be discussed in more detail later. Referring to Figure 8, a flow diagram is shown illustrating the operation of the terminal controller that determines when a channel is jammed by an interference. When a signal is received by the base site receiver, the demodulator FH demodulates the FH data of the plurality of frequency jumps, step 800. As is well known, the demodulated data is sampled and then passed to the FFT to convert the time domain signal in a frequency domain signal, step 802. Of the magnitude or energy values that correspond to the FFT samples, the displacements corresponding to the symbol 4FSK, k = 0-3, are selected, they are selected based on the deviation of the frequency of the 4FSK, step 804. A matrix representation of the data is generated and stored as a matrix, step 806. The matrix has eight rows or rows and four columns, the eight lines correspond to the number of bits of each row or row of the interleaved matrix of Figure 4 and the four columns correspond to the number of displacements or frequency deviations 4FSK, k = 0 to 3. The difference between adjacent lines d e the matrix is calculated for each of the indices k = 0-3 [dj (= abs (fj, - fj-ijk,) -r step 808. The differences are added for the indices k = 0-3 together with the lines P1309 / 99MX of the matrix [Uk =? dj / k for j = 1 to 8 in k = 0-3], step 810. The total energy of each column of the matrix is calculated by adding or adding the magnitude or energy of each of the indices, [Vk =? fjík for j = 0 to 7 in k = 0-3], step 812. The column that has the maximum energy value is identified or determined as the column with the maximum energy, kmax, step 814. The sum of the difference is divide by the maximum energy of the column, [W = Ukmax / Vkmax], step 816. A predetermined threshold value is set as the clogging threshold, T a 0.5, step 818. If the sum of the difference divided by the maximum value of column energy, W, is greater than the jam threshold Tj, then the line has not experienced a jamming signal during transmission, step 820. The corresponding value is set to zero, step 824, and that value is written or registers in the clogging matrix, step 826. When the sum of the difference is divided by the maximum energy W, it is greater than or equal to the clogging threshold Tj, step 820, the value is set to one and the result is recorded or recorded in the binding matrix, step 820. The result is repeated until they have been all the data is processed and a matrix is generated in which the RS symbols are marked as having a high probability of having an error. Referring to Figure 9, a matrix representation of the 4FSK symbols is generated P1309 / 99MX decoded to determine which of the hops of the FH signal was jammed by an interference signal during transmission. When it is determined that a column of the matrix is stuck, the binding matrix is generated with all the ones in the row corresponding to the location of the matrix in step 806 of Figure 8. Similarly, when it is determined that a jump has not been jammed by interference during transmission, the jam matrix is generated with all zeros in the row corresponding to the position of the jump. Referring to the clogging matrix, it was determined that the second, eighth and ninth lines were jammed by interference. Referring to Figure 10, there is shown a flow chart illustrating the operation of the terminal controller to determine at what time a fading channel is in accordance with the preferred embodiment of the present invention. When the receiver receives a signal at the base site, the demodulator FH demodulates the FH data of the plurality of frequency jumps, step 1000. As is well known, the demodulated data is sampled and then passed to the FFT to convert the signal of time domain in a frequency domain signal, step 1002. Of the magnitude or energy values corresponding to the samples of the FFT, P1309 / 99MX select the displacements corresponding to the symbol 4FSK, k = 0 to 3, based on the frequency deviation of the symbol 4FSK, step 1004. A matrix of representation of the data is generated and stored as a matrix, step 1006. matrix has eight rows and four columns. The eight lines correspond to the number of bits of each row of the interleaved matrix of Figure 4 and the four columns correspond to the number of displacements or frequency deviations 4FSK, k = 0 to 3. The maximum energy or value of magnitude is determined and the next higher energy or magnitude value of the frequency deviations for each row of the array and the ratio of the maximum energy to the next higher energy, Sj, step 1008 is determined. A predetermined threshold is set or set as the threshold value alert equal to 1.2, step 1012. In step 1012, when the proportion is determined to be greater than the alert threshold (1.2), the ratio is set to 1, step 1014. Alternatively, when the proportion is determined to be lower than the alert threshold, step 1012, the ratio is set to zero, step 1016. The total energy is calculated for each of the 4FSK symbols by the formula [Xj =? f / k for j = 0 to 7 in k = 0-3], step 1018. The total energy of the 4FSK symbol is multiplied by the ratio of each 4FSK symbol by the formula [Yj = Xj / k x Sj for j = 0 a P1309 / 99MX 7], step 1020. Each adjacent para of the 4FSK symbols are multiplied to obtain a measure of the energy for the code word RS (15.5), since each RS symbol comprises two 4FSK symbols that appear adjacent to each other another in the matrix. The value is written or recorded in a fade matrix in the corresponding location, step 1022. In step 1024, it is determined if the fade matrix is fully generated and if so, the fade determination programs terminate, otherwise , the process returns to step 1004 to be repeated until the fading matrix is fully generated, step 1004. Referring to Figure 11, a matrix representation of the decoded 4FSK symbols is generated to determine which hop or jumps of the FH signal It was fading during the transmission. The total energy of the RS symbol is measured and weighed or weighted against the proportion of the two highest magnitude values of the 4FSK symbols to determine the probability of a fade during transmission. The lower the values of the fading matrix, the greater the probability that the RS symbol will experience fading during transmission, referring to Figure 12, a flow chart illustrating the operation of the P1309 / 99MX terminal controller to correct errors resulting from interference of the jammed and / or fading signal, in accordance with the preferred embodiment of the present invention. A counter M is initiated in zeros, step 1201 and, in a resulting array, the M-th column of the jamming matrix of Figure 9, step 1202 is received and stored. The values of the binding matrix are verified and the number of erasures are recorded as Lm, step 1204. As discussed above, the RS code words were interleaved to reduce the impact of errors such as high energy and fading clogging signals. Each RS code word is stored in columns but it is transmitted in rows from the matrix, therefore, when a frequency jump experiences an interference signal, the errors will be distributed through the four RS code words. As is well known in the art, when the locations of the errors can be determined, the number of errors that can be corrected increases. This method is called error correction code by soft decoding or error correction code by soft decoding. As stated above, the word code RS (N = 10, K = 5) can correct a total number of errors (t), and a number of erasures (e), if 2t + e is less than (NK) +1 = (15-5) P1309 / 99MX + 1 = 11. An error is defined as the transmission errors whose location and value are unknown, and an erasure is defined as an error whose location is known but whose value is unknown. For example, RS (15.5) has a total of 15 Reed Solomon symbols in which 5 are information symbols. Each symbol has 4 bits. The word code RS (15,5) can correct 5 symbol errors without any deletion. However, in this mode, the codes of the RS code are assigned to the correction of up to 10 transmission errors if they are correctly marked as erasures. The increase in capacity is obtained because the location of the errors is known. The binding matrix in Figure 9 allows the position of the 4FSK symbols that have a high probability of error to be identified. The maximum number of erasures is preferably set at ten. The number of positions corresponding to I s in each column is calculated or accumulated and step 1206 verifies whether the number of erasures is equal to ten, the maximum number of erasures. If not, the number of erasures is checked to determine if they exceed the predetermined number of erasures, step 1208. The predetermined number of erasures can be any number equal to or less than the error correcting capability of the error correcting code by soft decoding. . If so, all P1309 / 99MX erasures exceeding the maximum number of erasures is ignored, step 1226. When the erasure number does not exceed the maximum erasure number, step 1208, the M-th column is received from the fade matrix of Figure 11, step 1210. An additional number of errors or erasures can be corrected when the maximum number of erasures was not exceeded in the clogging matrix. The addition number of erasures is equal to the maximum number of erasures minus the actual number of erasures accumulated in the corresponding column of the clogging matrix. To determine other erasure candidates in the fade matrix, the smallest values of the fading matrix are located in the M-th column of the fade matrix, step 1212. The smallest values are checked to determine if they are greater than Fade threshold, step 1214. The fade threshold is set equal to 0.05. If not, the value is checked to see if it is equal to one, step 1216. If not, the RS symbol is identified as an erasure and the value is set to one, step 1218. Once the location is determined of the erasure, that position is blocked in the fading matrix to prevent the same position being selected again. The number of erasures, Lm, is increased by one, step 1220. Step 1222 checks to determine whether the number of erasures recorded, P1309 / 99MX Lm, is equal to the maximum number of erasures. If not, the process returns to step 1212 to locate other erasures. If the recorded number of erasures is equal to the maximum number of erasures, step 1222 or if the smallest value of the fading matrix is greater than the fade threshold, step 1214, or if the recorded number of erasures of the clogging matrix is equal to the maximum number of erasures, step 1206 or continuation of step 1226, the counter, M, is increased by one to select the next column of the clogging and / or fading matrices, step 1224. In step 1228, determines if all the columns of the two matrices have been processed. If so, the errors are corrected in accordance with the marked erasures, step 1230. If not, the process repeats itself from step 1202, until all the data has been processed. Referring to Figure 13, a representation of the matrix with the erasure marks for error correction is shown when the FH signal is fading and / or when the FH signal is jammed, in accordance with the preferred operation of the invention. When this matrix is compared to the binding matrix of Figure 9, this matrix will have more positions with l's, because there was fading in the channels. The fade matrix indicates the RS symbols that have a P1309 / 99MX high error probability. The error position of the binding matrix and the error positions of the fading matrix are combined and indicated as l's in this matrix. Therefore, the FEC decoder uses the error indications of the array to determine which RS symbols of the data have a high probability of error and those positions are marked as erasures. In this way, the message is modulated as a level four frequency offset and is encoded as Reed Solomon code words. Several RS code words, preferably four, are stored in a matrix in the direction of the columns and the RS symbols are transmitted in the direction of the rows to achieve data interleaving to minimize the bursts of type bursts. The four RS symbols are transmitted by the spread spectrum with frequency hop in a plurality of jumps. Each jump is received and the 4FSK symbols are verified by interference of the clogging signals. The four RS symbols include eight 4FSK symbols. Each 4FSK symbol has two bits that correspond to the four frequency deviations of the 4FSK modulation. A matrix representation of each jump is generated as a matrix (8x4), where the 8 rows represent the number of 4FSK symbols per jump and the four columns represent the number P1309 / 99MX of bits per 4FSK symbol. An FFT is performed on each of the 4FSK symbols and the results are used to determine whether the signal experienced interference from a jamming signal or a fading signal. Two matrices are generated to represent the RS symbols that are most likely to be found with errors and the results are combined to identify each of these positions as an erasure to allow a correction block in advance of the errors to correct errors in the data . As established, the selective call system's input channel will operate in the ISM band that was approved for the extended spectrum communication systems with frequency hopping. It is anticipated that there will be numerous other systems operating in the ISM band, which may cause interference, therefore, this method and apparatus will correct any errors that result from fading or random-type errors. In summary, a selective call communication system comprises a selective calling device that transmits a modulated signal as a level four frequency shift signal (4FSK) and encoded as a Reed Solomon code word and transmitted as a spectrum signal extended with frequency jump interspersed in a P1309 / 99MX plurality of narrowband channels. A base site transceiver receives the spread spectrum signal with frequency hop interleaved in the plurality of narrowband channels by skipping the frequency in the plurality of narrowband channels in a pseudorandom sequence. A Digital Signal Processor (DSP) performs the fast Fourier transform on a 4FSK symbol to generate a frequency domain signal. The DSP further comprises a comparator that measures the energy at a frequency deviation of a 4FSK symbol received on a narrowband channel. The comparator establishes a ratio of the maximum value of energy and the next higher value of energy in an alert threshold to indicate the probability of error of a Reed Solomon symbol. A determiner determines that the frequency domain signal was fading in response to the probability of error that is below the fading threshold, a marker of erasures marks the position of the Reed Solomon symbol that was determined to be fading, as an erasure and an error correction code word comprises a soft decoding error correction code to correct errors that have Reed Solomon symbols marked as erasures in the Reed Solomon code word. The position that is marked improves the capacity P1309 / 99MX error correction on the code word Reed Solomon. P1309 / 99MX

Claims (10)

1. NOVELTY OF THE INVENTION Having described the present invention, it is considered as a novelty and, therefore, the content of the following CLAIMS is claimed as property: 1. In an extended spectrum communication system, a method to correct errors when minus one channel is fading, comprising the steps of: receiving a signal transmitted as an extended-spectrum signal over a plurality of narrow-band channels; converting the spread spectrum signal into a frequency domain signal comprising a plurality of frequency samples; calculating the energy value of a frequency sample that will be received in a narrowband channel; establish the proportion of the maximum value of energy and the next higher value of energy as a warning threshold; generate a matrix that indicates the probability of error of the frequency sample; compare the probability of error in the matrix with a fade threshold; mark, in response to the comparison step, the frequency sample that was determined P1309 / 99MX is fading, like an erasure; and correcting the error of the plurality of frequency samples with the erasure that was marked when it is determined that at least one channel is in fade. The method according to claim 1, wherein the establishment step further comprises a step for measuring the energy at a frequency deviation corresponding to the level four frequency shift signal, wherein the energy at the deviation of the frequency. The method according to claim 1, wherein the marking step, marks the position of a frequency sample as an erasure and the error correction step uses an error correction code by soft decoding where the position that was marked improves the error-correcting capability of a Reed Solomon code. 4. The method according to claim 1, wherein the spread spectrum signal comprises an extended spectrum signal with frequency hopping. 5. The method according to claim 1 further comprises the step of deinterleaving the transmitted signal as an extended spectrum signal. 6. A selective call communication system, comprising: a selective calling device for P1309 / 99MX transmitting an extended spectrum signal with frequency hopping, encoded as a Reed Solomon code word and modulated as a level four frequency shift signal (4FSK); a base site transceiver receives the spread spectrum signal with frequency hop on a plurality of narrow band channels in a pseudorandom sequence; a digital signal processor (DSP) for performing the fast Fourier transform on the spread spectrum signal with frequency hopping that generates a plurality of frequency samples; the DSP further comprises: a comparator to calculate the energy of a 4FSK symbol that is received in a narrow band channel, the comparator establishes a proportion of the maximum energy value and the next higher energy value as a warning threshold to indicate the error probability of a Reed Solomon symbol; a determiner, in response to the probability of error, to determine the time at which there is a fade signal in which narrowband channel; an erasure marker to mark the position of the Reed Solomon symbol that was determined, is fading as an erasure; and a bug-correcting code to correct P1309 / 99MX the errors of the code word Reed Solomon transmitted on the narrowband channel with the erasure marked. The selective call communication system according to claim 6, wherein the selective calling device comprises a modulator for modulating the signals as a four level frequency shift signal (4FSK) and a decoder for encoding the 4FSK signal in the code word Reed Solomon. 8. The selective call communication system according to claim 6, where the comparator measures the energy at a frequency deviation corresponding to the symbol 4FSK. 9. The selective call communication system of claim 6, wherein the DSP determines that a narrowband channel is fading when the probability of the error is below the fading threshold. The selective call communication system according to claim 6, wherein the erasure marker marks the position of a Reed Solomon symbol as an erasure and executes the error correcting code by soft decoding, wherein the position that was marked improves the error-correcting ability on the code word Reed Solomon. P1309 / 99MX