GB2259226A - Communications systems - Google Patents

Abstract

A communications system of high resistance to interference and jamming is of a frequency hopping type. In each hop period portions of data are simultaneously transmitted on 16 respective pseudo-randomly selected frequency channels. Preferably the data is first encoded using concatenated Reed-Solomon and Reed-Muller encoders (42, 46) the data to be transmitted is also symbol interleaved and then expanded by Direct Sequence Spread Spectrum. Resulting subsets of data from the said portions are thus encoded in both time and frequency domains. <IMAGE>

Description

COMMUNICA2IO,NS SyS2E-MS The present invention relates to communications systems and, more particularly to such systems which are resistant to various types of jamming systems, both existing types of jammers and those which may not yet have been produced but which may be developed to fulfill certain defined requirements in the future.

Frequency hopping radios are resistant to various jammers as they hop their carrier frequency periodically during the course of a transmission over a predetermined number of channels in accordance with a pseudo-random sequence. Some of the channels used during a transmission may be effectively "blocked " by jammers or for other reasons. However it is usually possible with the use of suitable coding techniques to transmit information with reasonable accuracy against the effects of static jammers or noise/chirp jammers provided only a relatively small proportion of the channels are blocked.

However it may be possible to design a "follow jammer" which would be capable of following the pseudo-random sequence of channels being used by a frequency hopping radio system and transmitting a noise signal on the channel in use. Present frequency hopping radios would be incapable of effective communications in the presence of such a follow jammer.

Coding techniques have also been proposed in various forms for use with digital communications systems in general. Some of these codes act directly on the data bits and others fist divide up the data into symbols.

A symbol is a group of bits, for example 5. The nature of these codes is to add a certain amount of data which is redundant with respect to the information to be transmitted but which enables the receiving system to detect the presence of errors in the received transmission and, in some circumstances, correct some or all of them. The greater the redundancy, i.e. the number of extra bits or symbols added to a number of data bits or symbols, the more errors it is possible to detect and/or correct. The latter is true in general but various codes have different properties and are specifically adapted to perform well in certain circumstances. These codes are well described and their detailed structure will not be described in detail herein, although some of their properties will be referred to.Such codes include Hamming codes, Reed Solomon (RS) and Reed-Muller (RM) codes. Descriptions of these and similar suitable codes may be found in the following references: a) Macwilliams F.J. and Sloane N.J.A. "The Thepryof Error Correcting Codes" published by North-Holland 1977, ISBN 0444 84009 0 1977; b) Reed I.S. amd Solomon G. "Polynomial Codes over Certain Finite Fields" J. Soc.Ind. Appl. Math. Vol.8 pp.300-304 June 1960; c) G.D.Forney Jr. "Concatenated Codes" M.I.T.Research Monograph No.37 1966; d) Chien R.T. "Burst-Correcting Codes With High-Speed Decoding" Bell System Technical Journal, Vol.27, pp.379-423, 623-656; and e) Berlekamp E.R., "The Technology of Error-Correcting Codes", Proc.lEEE, Vol.68, No.5 May 1980 pp.564-593.

The likelihood of an error being introduced in a transmitted digital bit in a data transmission is known as the bit error rate (BER). For a channel blocked by a noise jammer the BER will be 0.5. With such an error rate no certainity can be attached to the value of any bit received. Typical noise conditions produce much lower BER, for example of the order of 0.001 for a relatively good channel to 0.1 for a poor channel.

These types of noise conditions produce errors randomly distributed throughout the received data, whereas a blocked channel will produce a burst of errors. Many coding techniques can deal fairly well with randomly occurring errors but are less efficient at dealing with a burst of errors. Symbol codes are generally superior at coping with burst errors.

In order to convert the errors of a burst \ into a relatively random distribution at the decoder the bits or symbols to be transmitted may be interleaved prior to transmission and de-interleaved before decoding takes place. Interleaving is a known technique and can be carried out to various "depths". The "depth" of interleaving is the spacing of two bits or symbols after interleaving which were adjacent before the interleaving process was carried out.

In the following discussion the following terms will be used. A bit or symbol will be referred to as "erased" if it is known to be unreliable. For example if a channel is known to be blocked by a jammer then all bits or symbols received over that channel may be declared "erasures". If a bit or symbol is received, believed to be correct but is actually not correct then it is referred to as "erroneous". If a bit or symbol is received, believed to be correct and is actually correct then it is a. "correct" bit or symbol.

Reed-Solomon (RS) codes are symbol codes which can work on combinations of erased and erroneous symbols.

Such codes are capable of correcting erasures more efficiently than errors as each error has first to be located and then corrected. The (31,15) RS code has 16 redundant symbols in each 31 symbol code word. Suppose a code word has S symbol errors and T erasures then the decoder can correct the code word completely if and only if 2*S + T is less than or equal to 16.

The technical problem posed in the present application is to provide an efficient communications system which may be used in conjunction with known coding techniques to provide high resistance to a variety of jamming techniques.

This problem is solved in accordance with the invention by providing a communications system comprising a transmitter adapted to transmit simultaneously information modulated onto a plurality of different frequency channels pseudo-randomly selected from a predetermined set of frequency channels for a predetermined hop period, the plurality of frequency channels being reselected for each hop period.

This system is advantageous in that it is extremely resistant-to any known or expected jamming system. When such a system is used with existing coding techniques, various novel coding strategies can be adopted which provide coding in both the time and frequency domains and result in impressive communications reliability even in circumstances where a number of the predetermined set of channels are blocked by jammers and the BER on other channels is relatively high. When the existing codes are combined in the manner to be described in more detail below the improvement exceeds that which might be expected having regard to the performance of the individual codes.

Embodiments of the invention will now be described, by way of example only, with reference to the accompanying diagrammatic drawings, in which : Figure 1 is a block diagram of a transmitter for a communications system in accordance with the invention; Figure 2 is a block diagram of a receiver for a communications system including the transmitter of Figure 1; and Figure 3 is a detail of a channel decoding unit for use in the receiver of Figure 2.

The communications system to be described primarily comprises à pseudo-random frequency selection device 2 which is controlled by an output from a control clock 4. The frequency selection device 2 periodically outputs a pseudo-randomly selected set of sixteen frequency channels which are fed to an MSK modulator 6.

The sixteen selected frequency channels are used during a succeeding hop period. These frequency channels are selected from a predetermined set of channels, for example 256 channels spaced from one another by, for example, 25 kHz.

The MSK modulator 6 is also supplied with data which has preferably been encoded in encoding apparatus 8 to be described in more detail hereinafter. A selected subset of the data bits to be transmitted during each hop period are modulated onto each of the sixteen selected frequency channels to be used during that hop period and simultaneously output along a line 10.

Synchronisation bits may be added to the data to be transmitttd on each channel. The means for doing this are indicated by a synchronisation bit generator 12 which is connected to the output of the MSK modulator 6 by means of a device 14. The addition of the synchronisation bits is carried out under the control of control clock 4. Alternative synchronisation strategies 'are possible, but this implementation is conceptually simpler.

Prior to transmission the sixteen channels are mixed in mixer 16 with an intermediate frequency signal output from oscillator 18. The intermediate frequency may either be fixed or may be changed for each hop period.

The output of the mixer 16 is then fed to a transmitter 20 for transmission to a receiver of the communications system. Alternative frequency translation strategies are possible where fewer mixers are used.

At the receiver of the communications system which is illustrated diagrammatically in Figure 2 the received data are mixed with an intermediate frequency signal corresponding to that produced by the oscillator 18 in the transmitter. The mixer and oscillator at the receiver are not illustrated in the drawing. The received data are then fed to a channel separator which divides the incoming signals into the sixteen frequency channels osed for transmission. Each frequency channel having its own data modulated thereon is fed to an individual channel decoding unit 32a, 32b...32p and 32p. The channel decoding unit 32 is shown in more detail in Figure 3 and will be described later.The data recovered from each channel is then assembled in a data assembler 34 and preferably fed through decoding apparatus 36 which recovers the original data as encoded by the encoding apparatus 8 in the transmitter.

This data is output along line 38. The decoding apparatus 36 may also produce on line' 34 a variety of information indicative of the number and positions of errors located in the received data.

It will be appreciated that an important aspect of the communications system so far described is the transmission of data simultaneously: on a number of channels which are pseudo-randomly selected; i.e. bear no obvious relationship with one another that is carried forward from hop period to hop period. Thus, whereas with a frequency hopping radio system, a follow jammer could effectively prevent communications taking place, the described communications system when operating in the presence of a single follow jammer would only have one of its sixteen channels impaired during any single hop period. Therefore, if the data transmittMd on the remaining fifteen channels was correctly received and even a simple coding technique was used, communications could still be reliably achieved with such a system.In fact it is believed that the described communications system would be capable of combatting the presence of several ideal follow jammers operating in its vicinity.

The communications system will now be described in more detail using, as an example, a specific coding strategy. It will be appreciated that alternative coding stratagies may be employed. The use of the described communications system together with suitable coding strategies produces an even more reliable communications system due to the interaction between the various properties of the code and the overall system.

in the illustrated embodiment the coding scheme used by the encoding apparatus 8 will now be described in more detail. A message to be transmitted may be taken to be 375 bits long. If the message is longer than this it can be divided up into 375 bit sections. Each section can then be transmitted as described. The 375 message bits are divided into 75 5-bit symbols. The symbols are separated into five groups of 15 symbols each.

Each group of 15 symbols is encoded in a Reed-Solomon encoder 42 using a (31,15) RS code. Each group of 15 symbols then becomes a code word of 31 symbols. At the output of the encoder 42 after all 375 bits of the message have been encoded with the (31,15) RS code there will be five coded words each of 31 symbols.

These five coded words of 31 symbols are then symbol interleaved to depth 5 so that the first symbol from each of the five code words of 31 symbols are adjacent each other in sequence after the interleaving has been carried out, and are followed by the second symbols from each code word and so on until all 155 symbols have been interleaved. The interleaving process is carried out in an interleaver matrix former 44 which outputs a series of 31 initial data matrices (IDMs).

Each IDM consists of one symbol from each of the 31 RS encoded code words in each row. Each symbol consists of 5 data bits so each IDM is a 5 x 5 matrix. The data to be transmitted during each hop period will be derived from the data contained within a single initial data matrix. Therefore as the 375 bit message consists of 31 IDMs it will be necessary to transmit the message over 31 hop periods. Each IDM is passed to a (16,5) RM encoder 46. The encoder 46 applies a (16,5) Reed Muller code to each of the five columns of the IDM.

The output of the encoder 46 may be regarded as an extended data matrix (EDM) having a 16 x 5 matrix structure.

Each 5 bit row of the EDM is encoded in a 5 bit/32 chip cyclic encoder 48. The encoder 48 uses each 5 bit row to determine the number of cyclic shifts applied to a fixed 32 chip long sequence possessing a good autocorrelation property. The output of the encoder 48 therefore consists of a spread data matrix (SDM) having a 16 x 32 matrix structure. The chip rate of the 32 chip long sequence is chosen to be 6.4 times as fast as the' bit rate of one row of the EDM. Thus the time to transmit 32 chips is identical to the time to transmit the 5 bits.

Each of the sixteen rows of the spread data matrix is then modulated in the MSK modulator 6 onto a respective one of the 16 channels pseudo-randomly selected by the frequency selection device 2. The sixteen modulated channel frequencies output along line 10 from modulator 6 are then simultaneously transmitted after passage through device 14, mixer 16 and transmitter 20.

The same process is repeated during each of the next 30 successive hop periods in order to complete transmisston of the 31 IDMs representing the whole of the 375 bit message to be transmitted. If there are further message sections then these are encoded in a similar manner and separately transmitted.

At the receiver 16 the channels transmitted during each hop period are separated by channel separation device 30. Each separate channel is then fed to an individual channel decoding unit 32. In this decoding unit, which is illustrated in more detail in Figure 3, each 5 bit row of the ED as fed to encoder 48 is restored.

In the decoding unit 32 the data is first fed to a hard limiter 50 which shapes the incoming waveform so that it represents a 32 chip sequence. Thus the limiter 50 reduces to some extent the effects of random noise introduced during transmission. The output of the hard limiter is then fed to an analogue correlator 52 in which the received 32 chip sequence is correlated with the fixed chip sequence used in the encoder 40. This fixed chip sequence is fed to the correlator 52 from a reference signal source 54.As the fixed chip sequence has a good auto-correlation property, the output of the analogue correlator 52 peaks at the time when the reference signal is subject to the same number of cyclic shifts as were applied to the original chip sequence 'in encoder 48. As the received 32 chip sequence may have suffered distortion during transmission it is necessary to feed the output of the correlator 52 through a likelihood processing circuit 56 in order to determine the most likely value of the applied number of cyclic shifts and therefore the most likely value of the 5 bits of the EDM which were transmitted on that channel.This data is fed out from the channel decoding unit along line 58 to the data assembler 34. The information generated in the likelihood processing circuit 56 may be applied in the later decoding stages to determine the probabilities and therefore the confidence to be attached to each decoded item of data.

The data assembler 34 receives an input from each of the 16 channel decoding u-nits 32a, 32b 32O and 32p and is able to restore the 16 rows of the EDM and teed these to a Reed-Muller decoder 60. The decoder 60 acts on each of the five columns of 16 bits of the EDM applying the inverse of the (16,5) RM code used in encoder 46. The output of the decoder 60 is therefore the 5 x 5 IDM.

During each hop period a single IDM is received. The IDMs received over 31 consecutive hop periods are accumulattd in the symbol de-interleaver 62 and after receipt of the 31st IDM each 31 symbol code word can be de-interleaved and fed to the Reed-Solomon decoder 64.

The Reed-Solomon decoder decodes each code word of 31 symbols applying the inverse of the (31,15) RS code used in encoder 42. The decoder 64 output the 375 bit message originally transmitted on line 38 together with errata information on line 40.

The encoder and decoder used in the transmitter and receiver described may be of various types commercially available. They may also be implemented using the known encoding and decoding algorithms using suitable microprocessors, or mini computers.

It will be appreciated that the described coding scheme not only takes advantage of the concatenation of Reed Solomon and Reed-Muller codes, but also has the associated advantages provided by the spread spectrum techniques introduced by the 5 bit/32 chip encoder and the use of 16 pseudo randomly selected channels during each hop period. Thus the data is effectively encoded in the time domain by the interleaved Reed Solomon code and in the frequency domain by the frequency selection device 2, the Reed-Muller code and the 5 bit/32 chip spreading. It is believed that such a transmission technique would be virtually intractable to known or contemplated interference devices and jammers apart from broad band, noise-like jammers.

It is to be understood that the Reed-Muller decoder may utilise information on the reliability of data provided by the channel decoding units from their likelihood processing circuits. In turn the Reed-Solomon decoder will utilise information provided by the Reed-Muller decoder relating to the presence of erasures.

It will be appreciated that the described coding scheme is only one of many that can be used with the described transmission method of simultaneously transmitting data on a plurality of parallel channels, the frequencies of which are each independently hopped to provide for encoding of the data in both time and frequency domains thereby ensuring a high degree of immunity to selective jamming techniques.

Claims (14)

1. A communications system comprising a transmitter adapted to transmit simultaneously selected subsets of information, each subset being modulated onto a respective one of a plurality of different frequency channels pseudo-randomly selected from a predetermined set of frequency channels for a predetermined hop period, the plurality of frequency channels being reselected for each hop period.

2. A system as claimed in claim 1, further comprising means for encoding data to be transmitted to provide said information.

3. A system as claimed in claim 2, wherein said encoding means comprises means for employing concatenated codes.

4. A system as claimed in claim 2 or 3, wherein the code or codes employed are selected from Reed-Solomon codes, Reed-Muller codes and Hamming codes.

5. A system as claimed in any one of the preceding claims, further comprising spectrum spreading means operative on data to be transmitted to provide said information.

6. A system as claimed in any one of the preceding claims, comprising an MSK modulator for modulating said information onto said frequency channels.

7. A system as claimed in any one of the preceding claims, wherein said subset of information transmitted on each channel includes synchronisation information to allow a receiver to identify the channel and the beginning and end of the subset.

8. A communications system substantially as herein described with reference to the accompanying drawings.

9. A method of transmitting data comprising the steps of encoding the data, simultaneously transmitting subsets of the encoded data on respective ones of a plurality of channels for a predetermined time interval, and reselecting the frequenc(es of said channels after each predetermined time interval before transmitting further subsets of the encoded data.

10. A method as claimed in claim 9, wherein said encoding step comprises the steps of encoding the data with a first code, and encoding the data with a further code.

11. A method as claimed in claim 10, wherein the first code is a Reed-Solomon code and the further code is a Reed-Muller code.

12. A method as claimed in claim 10 or 11, wherein the encoding step further comprises interleaving the data encoded by the first code before encoding with the further code.

13. A method as claimed in any one of claims 9 to 12, further comprising spreading the spectrum of the encoded data prior to transmission.

14. A method of transmitting data substantially as herein described with reference to the accompanying drawings.