GB2283597A - Multi-lingual communications system - Google Patents

Abstract

A multi-lingual communications system is able to transmit messages comprising alphanumeric symbols 29, ideographic symbols 30, so-called canned phrases 31, and combinations thereof. The message data is broken down into data packets at boxes 32, 33 and 34 respectively. At boxes 35, 36 and 37, a language mode code identifying the language or a canned phrase is added to each data packet to form a codeword. The codewords are grouped in pairs known as long codewords at boxes 38, 39 and 40 and are all combined at box 41 for transmission. A corresponding decoding process (Fig. 5) at the receiver recognises each language mode code, thus avoiding error propagation. The system may be a paging network. <IMAGE>

Description

MULTI-LINGUAL COMMUNICATIONS SYSTEM The present invention relates to a multi-lingual communications system. More particularly, it relates to a paging system, capable of transmitting and receiving multilingual messages as well as canned phrases i.e. commonly used phrases stored in the pagers for retrieval, and encoded for transmission by the transmitting device to save air time, to individuals or groups of users.

Modern paging systems are capable of transmitting data messages from a paging centre to the desired pager(s). A caller makes a phone call to the paging centre and leaves a message which is encoded and transmitted to the selected receiver or group of pagers. On receiving the paging call, a pager decodes and displays the message using a microcomputer which controls a display device and one or more alerting devices.

In locations where more than one language is in general use, a message coding scheme must be devised so that messages containing multi-lingual symbols can be transmitted and rec-ived. The symbols may be used in alphabetic languages such as English, French, Russian or the like, and ideographic languages such as Chinese or Japanese.

Prior art multi-lingual message coding schemes employ storage means in the pager for the storage of control characters received from the transmitting device, which only sends the appropriate control character when the transmission changes from one "active" language currently in use to another. At any one time only one language is in use which is defined by the control character currently stored and this control character is used to switch between different sets of symbols used in different languages.

Thus the control character is a language mode signal that selects the active language. For each language, the pager has a dedicated symbol area that provides a set of symbols to be displayed. The paging centre inserts the control characters as needed and the pager detects them for language decoding.

While such a prior art system may be acceptable, errors occur during actual transmissions due to interference or signal fading. If data corruption occurs at the crucial position of a control character while a multi-lingual message is being received, such that the language mode signal is lost, the part of the message following the control character becomes completely unreadable. This considerable disadvantage of the prior art system is known as the error accumulation effect.

Canned phrase transmission is a standard technique used in the paging industry. A set of commonly used phrases such as "Please Call Me Back" is defined and stored in each of the receivers. Each phrase is assigned a code, which merely represents the address of the phrase, and is much shorter than the phrase itself, the number of phrases stored being limited. The transmitting device uses the code to signal to the pager(s) that a particular phrase is being transmitted.

Upon receiving this code, the pager(s) retrieve the whole phrase from canned phrase memory for display, thus improving transmission efficiency.

The error accumulation effect also occurs in this prior art canned phrase transmission system. If data corruption occurs during transmission of a control code indicating the start of canned phrase(s) transmission, not only the canned phrase but also subsequently transmitted messages may be lost.

It is an object of the invention to provide a communications system for multi-lingual and canned messages which substantially overcomes the error accumulation effect.

Accordingly, the invention consists in a communications system comprising means for encoding data to be transmitted, and means for receiving the transmitted data, the receiving means including means for decoding the received data, the encoding means being operable to encode the data into segments each representing a symbol or a canned phrase belonging to a natural language, and the decoding means being arranged to use information from each of said segments to determine to which language the symbol represented thereby belongs or whether it represents a canned phrase, thus reducing the propagation of an error occurring in one part of the data to other parts thereof.

The symbols may be alphanumeric and/or ideographic characters and preferably the receiving means includes means for displaying the symbols.

The receiving means may be arranged to store at least one canned phrase of a natural language, the or each canned phrase having a corresponding address code associated therewith. The decoding means may be arranged, upon receipt by the receiving means of transmitted data containing at least one address code, to identify the or each address code from information contained in said transmitted data. The receiving means may be operable to retrieve the stored canned phrase associated with the or each address code identified by the decoding means and to display said phrase on said display means. The or each address code and said information identifying the or each address code may each comprise at least one segment.

Said segments may consist of codewords of a standard number of bits, a fixed part (fixed number of bits) of each codeword constituting said information from each of said segments which determines to which language the symbol represented thereby belongs or whether it represents a canned phrase.

The system may be a paging network, the receiving means comprising at least one pager and the network including a paging centre operable to transmit messages each comprising a plurality of said symbols via at least one transmitter to the or each pager.

The receiving means may include a plurality of message memories for storing different types of messages, each message comprising a plurality of said symbols. Said different types of messages may include user messages sent from a caller to the or each pager, personal file messages which are protected against being overwritten by a newly received message and/or databank messages sent from a database to the or each pager.

An embodiment of the invention will now be described by way of example with reference to the accompanying drawings, in which: Figure 1 is a schematic block diagram of a typical paging system which can be used as an embodiment of the invention; Figure 2 is an electronic block diagram of a pager for use with the system of figure 1; Figure 3 is an electronic block diagram of an alternative pager to that shown in figure 2; Figures 4 and 5 are flowcharts showing respectively a data encoding and a data decoding process according to the preferred embodiment of the invention, assuming one set of ideographic characters, one set of alphanumeric characters and one set of canned phrases; Figure 6 is a flowchart showing the flow of data within the message encoder; ; Figures 7A & 7B and 8A & 8B are flowcharts showing the processes of figures 4 and 5 respectively in greater detail and assuming three codewords represent one canned phrase address; Figure 9 schematically shows the encoding of alphanumeric characters; Figure 10 schematically shows the encoding of an ideographic character; Figures 11 and 12 schematically show alternative methods for the encoding of a canned phrase; Figure 13 shows a pager display; Figure 14 shows top, front and side views of a pager; and Figure 15 is a timing diagram showing operations of a pager.

Figure 1 shows a typical paging system or network in which a call distributer 1 at a paging centre 2 takes incoming calls from callers via the public telephone network 3. The call distributor 1 then distributes the calls to operators at 4, and for each call, the account number of the subscriber being paged is obtained and passed on to a system processor 5 either by the system or the operator. The system processor 5 then retrieves information relating to the subscriber from its database and shows the information on a terminal 6 of the operator. The terminal 6 acts as an input device allowing multi-lingual messages to be taken and stored by the system processor 5 for data transmission.

A message encoder 7 then encodes each multi-lingual message using one of the methods described later with reference to figures 4, 6, 7A and 7B. The encoded data is queued inside the buffer of the message encoder 7 and then broadcast via wired or wireless synchronised data modems 8 to radio frequency transmitters 9. For the purpose of transmission using an international standard, the encoded data is further formatted according to CCIR Radio Paging Code No. 1 (POCSAG code) in the message encoder 7 before transmission. The system includes multi-lingual display pagers 10, of which the one being paged receives and decodes the message.

Figure 2 shows a typical multi-lingual pager comprising a double conversion RF module 11, a digital module 12 and a power supply module 13. An antenna 14 of the RF module 11 picks up the relatively weak RF signal from the air. The signal is amplified by a low noise amplifier 15 and then filtered by a wide band filter 16 in order to suppress the image frequency power and prevent interference. After filtering, the signal is mixed down twice in the RF module 11 to the second IF frequency for signal demodulation. The commonly used double conversion technique is used to demodulate the signal, after which it is wave shaped and passed to the digital module 12 for further processing.

A POCSAG decoder 17 deformats the POCSAG encoded digital data and can detect and match an identity address with an internal preset address. If a match is found, an interrupt is generated to a microcomputer 18, to which the data is passed for handling and storage. The microcomputer 18 looks at the multi-lingual data received, decodes it as described later with reference to figures 5, 8A and 8B, stores it in SRAM 19 and displays it on a pager display which is an LCD panel 20, after retrieving the display pattern for each character from a character mask ROM 21. As a signal to the user that a message has been received, the microcomputer 18 also activates alarm devices, namely an LED 22, a beeper 23 and/or a vibrator 24.

The power supply module 13 of the pager generates all the voltage supplies to the various components. The supply module 13 contains voltage outputs supplying the circuitry of the RF module 11 and the devices in the digital module 12 including an LCD segment and common driver system 25 as well as the LCD panel 20 itself. The supply module 13 also contains a backup power supply 26 to supplement the main battery supply 27, which consists of one "AAA" alkaline type battery, during periods when the battery voltage is low or the battery is removed.

Figure 3 shows an alternative pager using the single conversion technique instead of the double conversion technique used by the pager of figure 2. An alternative RF module 28 carries out only one frequency down mixing process to an IF frequency low enough for demodulation, for example 455kHz. The demodulated signal is processed by the digital module 12 in the manner described above.

In the pagers shown in figures 2 and 3, the electronic components are mounted on both sides of three PCB's, a main PCB housing the LCD panel 20 and the circuitry of the digital module 12, a keyboard PCB on which the power supply module 13 is mounted, and an RF PCB for the RF and IF circuitry of the RF module 11. The keyboard PCB is soldered vertically to the main PCB while the RF PCB is connected via two low profile connectors to the main PCB.

Figure 4 summarises the data encoding process carried out by the message encoder 7. Data representing alphanumeric symbols is at box 29, data representing ideographic language symbols at box 30, and canned phrase addresses at box 31.

The first step of the process is to break down the message data at boxes 29, 30 and 31 into basic data elements called data packets and this is carried out at boxes 32, 33 and 34 respectively. Each data packet is of a fixed word length which is a convenient number of bits such that the alphanumeric, ideographic and canned messages are represented by an integral number of packets. In order to represent Chinese, English, and some canned phrases, a word length of 7 bits means that Chinese symbols are each represented by two packets or 14 bits (allowing a set of 16,384 symbols), English symbols by one packet (using the ASCII coding scheme), and canned phrases by a sufficient number of packets to cover all combinations.

Next, to implement multi-lingual coding, a fixed number of bits, which in this example is a single bit forming the MSB, is added to each packet to form a data codeword.

In general, the fixed number of bits is more than one so that canned phrases can be treated as a separate language. More than one set of canned phrases can be treated in the same way by assigning a different language mode code to each set. This alternative to the described example avoids error propagation.

The process of adding a single bit is carried out for the respective message types at boxes 35, 36 and 37. The added bit is a language mode code identifying the language (or a canned phrase) for that particular data packet and hence that codeword. All codewords are of the same word length, 8 bits, regardless of message type, and each character or canned phrase becomes an integral number of codewords, the number depending on its type.

There may be two sets of canned phrases, one for each language, and a valid data codeword is assigned for each language as a control character. In normal transmission the control character is followed by codewords representing an addressed canned phrase. Each codeword, by definition, carries the language mode code indicating which of the two types of canned phrase is being selected.

The use of this control character to represent the start of a canned phrase code means that errors occurring in the received control character could cause error propagation if the pager fails to switch to canned phrase mode. An error would occur in the count of codewords for the proper decoding of each character which would only be corrected on switching to a different language.

Such error propagation is limited by making C, the number of codewords for each canned phrase address (including the control character) a multiple of I, the number of codewords for an ideographic symbol and also of A, the number of codewords for an alphanumeric symbol. Thus errors occurring in a control character only affect C codewords, after which the system regains synchronisation.

To limit such error propagation still further, a technique is used which introduces a further data structure known as a long codeword. Each long codeword contains a fixed number of codewords; in this example, two. Thus each long codeword contains an integral number of characters for each language and an integral number of canned phrases. The technique makes use of the property that characters of the same language or canned phrase usually come in batches, each of which can be transmitted or received before transmission is switched to a different language or canned phrase type.

In figure 4, the codewords are grouped in pairs at boxes 38, 39 and 40. In cases where there are an odd number of codewords, a filling character codeword is inserted to make up a pair. The transmission data comprising an integral number of long codewords is then entered into a buffer stream at 41 for POCSAG encoding and transmission.

Since each transmission must contain an integral number of long codewords, error accumulation due to one error is limited to one long codeword only. Therefore the long codeword defines a synchronisation frame.

Figure 5 shows how the transmission data is processed by the receiving pager. The whole received message is first captured and buffered for analysis and at box 42, one long codeword is taken from the buffer. This long codeword is evaluated at 43 to determine its language type, and the data packet portion of the first codeword is evaluated to determine if it is a control character representing the start of a canned message. If there is any contradiction in the language mode codes within the long codeword, a simple algorithm such as majority voting, based on the present pair of codewords and the previous language type, is applied to set the language mode.

Alphanumeric type long codewords are passed to box 44, ideographic type to box 45 and canned phrase type to box 46.

Then at boxes 47, 48 and 49 the long codewords are broken down into codewords and then into data packets, by removing the MSB language mode codes. At box 47, two alphanumeric characters are formed from each long codeword, one of which will be discarded if it is a filling character. At box 48, one ideographic character, and at box 49, a canned phrase address is formed. This process is repeated for each long codeword until the entire message is recovered.

The microcomputer 18 then retrieves the display pattern for each character as has been described above in order to display the message. In the case of a canned phrase, the words comprising the canned phrase are first retrieved from an internal canned phrase table area of the microcomputer system inside the pager.

Figure 6 shows how data flows from the system processor 5 to the message encoder 7 and within the message encoder 7.

The system processor 5, also shown in figure 1, takes caller messages from the terminals 6 and converts them into 8-bit bytes, the MSB carrying no information, before storing them in a message memory 118.

One byte of data at a time is read from the message memory 118 to a character buffer 119. If the character in the character buffer 119 is alphanumeric, an alphanumeric character count 120 is incremented.

After having been evaluated and assigned an MSB giving its language mode code, the 8-bit character is read into a buffer stream 121 at the next available location 122 which is indicated by a buffer pointer 123. The buffer pointer is then incremented by one. At box 124, the previous character in the buffer stream 121 is evaluated to determine its type.

When the entire message has been stored in the buffer stream 121, the data is sent to a POCSAG formatter 125 and thence to the transmitters 9.

Figure 7A and figure 7B which is a continuation thereof show an algorithm for the multi-lingual encoding process of figure 4. For this algorithm, each canned phrase is assumed to be represented by two long codewords (including the control character).

After the start of the process at 50, at box 126 all the buffers and registers are cleared. One byte of data is read from the message memory 118 at box 51 and saved into the character buffer 119. At lozenge 52, this character is evaluated to determine if it is alphanumeric, and on a positive result, at box 53, the MSB of the character buffer 119 (the language mode code) is reset to zero and saved into the buffer stream 121. The buffer pointer 123 is incremented by one, and at box 54 the alphanumeric character count 120 is also incremented by one.

If the character evaluated at box 52 is not alphanumeric, at lozenge 55, which corresponds to box 124 in figure 6, the previous character is evaluated to determine if it is alphanumeric. A positive result means that a mode change from alphanumeric to another mode has occurred. On a positive result, at lozenge 56, the alphanumeric character count 120 is evaluated to determine if it is a multiple of two. A negative result means that an odd number of alphanumeric characters has been placed into the buffer stream 121 just before the current character. At box 57, the filling character which in this example is "EOB" is inserted into the buffer stream 121 in order to give an even number of codewords, the buffer pointer 123 is incremented by one, and the alphanumeric character count 120 is reset to zero at box 58.

The character buffer 119 is now evaluated at lozenge 59 to see if it contains the control character, which in this example is "ESC". A positive result means that a canned phrase is commencing, and at box 60, the MSB of the character buffer 119 is reset to zero and saved into the buffer stream 121, the buffer pointer 123 being incremented by one. Then, at box 61, three bytes of data are read from the message memory 118 and the three MSB's are reset to zero and saved into the buffer stream 121, the buffer pointer 123 being incremented by three. The three bytes of data are required for the canned phrase address which, in this example, comprises three data packets.

If the character evaluated at box 59 is not the control character, then it is an ideographic character and is processed at box 62 where the MSB of the character buffer is set to one (the language mode code for ideographic characters) and saved into the buffer stream 121 and the buffer pointer 123 is incremented by one. Next, at box 63, one further byte of data including the second seven bits of the ideographic character is read from the message memory 118, the MSB is set to one and saved into the buffer stream 121 and the buffer pointer 123 is again incremented by one.

In all cases, at lozenge 64 the message is evaluated to determine if its end has been reached. On a positive result, at box 65, the buffer data is sent to the POCSAG formatter 125 for data transmission and the process terminates at 127. Otherwise the process loops back to box 51 for the next byte of data from the message memory 118.

Figure 8A and figure 8B which is a continuation thereof show an algorithm for multi-lingual decoding process of figure 5.

After the start at 66, two bytes of data are read from the message memory of the microcomputer 18 at box 67 and the bytes are separately saved into two character buffers known as X and Y. At lozenge 68, the data in both buffers is evaluated to determine if it contains errors which the standard POCSAG coding error correction scheme has been unable to correct. On a positive result, at box 69, a black block replacing the entire long codeword is designated to be displayed, indicating an error.

If no errors are detected at lozenge 68, the MSB's of the character buffers X and Y are evaluated at lozenge 70 to see if they are zero. On a negative result, an ideographic character has been transmitted and at box 71 the relative address of the ideographic character in the character mask ROM 21 is calculated. Then, at box 72, that character is designated for display.

At box 73 which is encountered after either of boxes 69 or 72, the microcomputer calls a subroutine to display the designated character on the LCD panel 20.

If at lozenge 70 the MSB's of the buffers X and Y are determined to be zero, then at lozenge 74, the character buffer X is evaluated to determine if it contains the control character. A negative result means that two alphanumeric characters have been transmitted, and at box 75, the two alphanumeric characters are designated for display.

Next, at boxes 76 and 77, a subroutine is called to display the characters in buffers X and Y respectively on the LCD panel 20.

A positive result at lozenge 74 means that a canned phrase is commencing. At box 78, a further two bytes of data are read from the message memory and separately saved into two further character buffers known as M and N. Then the relative address of the canned phrase is calculated at box 79 from the information in character buffers Y, M and N.

The microcomputer 18 now carries out a looped process to display the canned phrase. At box 80, a multi- lingual symbol of the canned phrase is read from the mask ROM 21.

At lozenge 81, this symbol is evaluated to determine if it is an alphanumeric character, and if so, at box 82 it is designated for display. Otherwise, at box 83 an ideographic character is designated and in either case at box 84 a subroutine is called to display the designated character.

Then the canned phrase is evaluated at lozenge 85 to determine if it is at an end. On a negative result the microcomputer loops back to box 80 to continue displaying the canned phrase.

When the canned phrase has been displayed in its entirety, and also after the display of other characters at boxes 73 and 77, the message memory is evaluated at lozenge 86 to determine if the data therein has all been processed.

If so, the microcomputer 18 exits the process at box 87 and otherwise loops back to box 67 to read the next two bytes of data.

As each codeword contains its own language mode code, errors affecting the language selection can often be corrected at the receiver side by checking the language mode code of the codewords preceding and following the erroneous codeword. The decoding process verifies the consistency of the language mode codes and invokes an error handling routine if necessary, such as confirming the language mode code by majority voting or displaying a black block for an error.

If a canned phrase can be represented by a long codeword, an error in the control character does not propagate beyond the one long codeword affected, since the next long codeword always commences with either another control character or a symbol other than a canned phrase. In general, an error in the control character does not propagate beyond the number of long codewords required to represent a canned phrase.

Figures 9-12 show general examples of the use of the coding scheme. The paging system is required to transmit and receive ASCII characters as the alphanumeric characters and Chinese characters coded in GB 2312-80, the national standard defined by the Standard Bureau of the People's Republic of China, as the ideographic characters. A set of at least 128 canned phrases, each having at least a 7-bit address is also required, and the message transmitted may contain a mix of all three data types. Each of the following general examples is applied to one of the data types, the resulting encoded data then being recombined in the order it appears in the message before transmission.

Figure 9 shows the encoding of two ASCII characters each of which is represented by the standard 7-bit ASCII code as its data packet. At box 88, a header of "0" is placed in front of the MSB's of both ASCII codes. This header becomes the MSB of an 8-bit codeword with "0" as its language mode code indicating English. Next, at box 89, the two 8-bit codewords are combined to produce a 16-bit long codeword at box 90. If the ASCII portion of the message to be processed contains an odd number of data codewords, then the filling character "EOB" is inserted as the final character so that the entire message can be encoded as an integral number of long codewords.

Figure 10 shows the encoding of a Chinese character, represented by the standard GB 2312-80 code which has 14 bits, and is broken down into two data packets of 7 bits each. At box 91, a header of "1" is placed in front of the MSB's of both data packets. This header become the MSB's and language mode codes of two 8-bit codewords indicating Chinese. The two codewords are combined to give, at box 92, a 16-bit long codeword representing the Chinese character.

Figure 11 shows how a 7-bit canned phrase address can be encoded, the canned phrase containing alphanumeric and/or ideographic characters. The address is preceded by a control character which is the ASCII code for "ESC". At box 93, headers of "0" are placed in front of the MSB's of both the "ESC" ASCII code and the address. These headers become the MSB's of two 8-bit codewords which are combined at box 94 to give, at box 95, a long codeword of 16 bits. Thus, after encoding, the canned phrase looks like two ASCII characters until the "ESC" control character is detected.

An error occurring in a codeword which can not be corrected results only in the incorrect reception or loss of that long codeword. The error does not propagate to the next long codeword.

In the preferred embodiment of the invention, however, many more than 128 canned phrases are provided in the pager.

For this purpose each canned phrase address consists of three 7-bit data packets rather than one. Figure 12 shows how this address is encoded in a similar manner to the 7-bit address. At box 96 the "0" headers are added to the control character and the three address data packets. At box 97 the four 8-bit codewords are combined to give, at box 98, two long codewords of 16 bits each.

In this embodiment, an error occurring in a codeword will result in the loss of information in a maximum of two long codewords.

Specific examples of message encoding will now be described in order to illustrate the preferred embodiment of the invention.

In the first example, the ASCII character message "TEL: 7128903" is to be transmitted. The message is encoded into the following hexadecimal ASCII sequence of long codewords (16 bits each): 54S45S, 4C,3AH, 37B31Hl 32B38H, 39,30,, 33R17E- As shown above, an "EOB" character, 17z, is concatenated to the end of the sequence to give an integral number of long codewords, since there are an odd number of ASCII characters in the message. For each 8-bit byte, the LSB is transmitted first.

The GB2312-80 standard defines a set of 7,445 characters including commonly used Chinese characters as well as ASCII characters, although in this paging system, GB2312-80 is not used for alphanumeric characters. The character set is divided into 94 sections, each holding a maximum of 94 characters. Each character is represented by two hexadecimal numbers, a section number (01 to 5E) and a column number (01 to 5E). For computer storage purposes, the numbers are converted into internal section and column numbers by adding the hexadecimal constant AOH to each number, and then into internal codes by combining the two numbers.For example, a Chinese character having a section number of 13E and a column number of 56E is converted into the internal code B3F6E since 13H + AOH = B3H and 56E + AOH = F6a. By definition the internal section number and internal column number must both have MSB's of "1". The internal code forms the long codeword for transmission in the paging system, the internal section number being transmitted LSB first and then the internal column number, also LSB first.

In the second example, a multi-lingual message consisting of Chinese and ASCII characters is to be transmitted. Translated into English, the message reads "Mr Chan Tai Man please call 7881256". The long codewords for the Chinese words are: B3HC2H, B4HF3H, CEHC4H, CFHC8H, C9HFAH, C7HEBH, B5HE7H; and the long codewords for the telephone number in ASCII characters are: 37 38z 38H31al 32,35H, 36H17H For the POCSAG transmission of the encoded message, first of all a preamble is transmitted to permit the pagers to attain bit synchronisation and to prepare them to acquire word synchronisation. The preamble is a bit pattern of reversals, 101010... repeated for a period of at least 576 bits. The POCSAG standard has 32-bit data words and after the preamble, a specific 32-bit synchronisation word is transmitted.

The 32-bit words are transmitted in batches. Each batch is composed of a synchronisation word followed by eight frames. Each frame contains two 32-bit words so that one batch contains 17 32-bit words (including the synchronisation word). The frames are numbered 0 to 7 and the pager population is similarly divided into eight groups. Each pager is assigned a 21-bit identity and is dedicated to one of the eight frames according to the three least significant bits (LSBs) of its 21-bit identity (e.g. "000"= frame 0, "111"= frame 7). These three bits will not be transmitted.

Each pager only examines 32-bit words within its assigned frame. Hence, the 32-bit address word (including the 18 most significant bits of the 21-bit identity as well as function bits and check bits) must be transmitted in the frame that is allocated to the pager. The 32-bit message words follow directly after the 32-bit address word for that pager.

In the second example, these 32-bit message words are transmitted as shown in the following tables:

Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 1100 1101 0100 0011 0010 x y Hex. 3H BH 2H CH 4H Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 1101 1100 1111 0111 0011 x y Hex. BH 3H FH EH CH Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 0010 0011 1111 0011 0001 x y Hex. 4H CH FH CH 8H Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 0011 1001 0011 0101 1111 x y Hex. CH 9H CH AH FH Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 1110 0011 1101 0111 1010 x y Hex. 7H CH BH EH 5H Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 1101 0001 0111 1110 1100 x y Hex. BH 7H EH 7H 3H Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 0001 1100 0001 1100 1000 x y Hex. 8H 3H 8H 3H 1H Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 1100 0100 1100 1010 1100 x y Hex. 3H 2H 3H 5H 3H

Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P. Bits 1 0110 1100 1110 1000 1111 x y Bits in 6H 3H 7H 1H FH Hex. (U@@@@@ed @@@@ will z@@ be 1)

x - BcH(31,21) code caIaiIa by bits 2-21 y - Even parity bit calculated by bits 1-31 Each unused part of the last 32-bit data word of the message is filled with a "1".

As stated previously, each canned phrase is represented by four 8-bit codewords, the first of which is the control character "ESC", or 1 BH. The other three codewords contain three ASCII characters representing the address of the canned phrase. The set of valid ASCII characters is: 0,1,2,3,4,5,6,7,8,9,A,B,C,D,E For example, the canned phrase address for the Chinese character phrase meaning "I am occupied today, can't go to your home" is 3,5,A and the 16-bit long codewords are: 1BH33H, 35H 41H To transmit the canned message, each 8-bit byte, starting with the "ESC" character, is transmitted LSB first.

In the final example the three canned phrases consisting of Chinese characters which when combined mean "Mr Wong, Beijing Railway Station" are to be transmitted.

The multi-lingual long codewords for these canned phrases are: 1BH30H, 33H41H, 1BH30H. 45H42H, 1BH34H, 31H31H.

Again, the POCSAG preamble, synchronisation word and address word are transmitted. 32-bit POCSAG words are then transmitted as follows:

Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 1101 1000 0000 1100 1100 x Hex. BH 1H 0H 3H 3H Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 1100 1000 0010 1101 - 1000 x Y Hex. 3H 1H 4H BH 1H Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 0000 1100 1010 0010 1000 x y Hex. 0H 3H 5H 4H 1H Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 0010 1101 1000 0010 1100 x y Hex. 4H BH 1H 4H 3H Bit 1 2-5 6-9 10-13 14-17 18-21 22-31 32 Number BCH Bit Even P.

Bits 1 1000 1100 1000 1100 1111 x y Bits in 1H 3H 1H 3H FH Hex. (Uafilled blts will all be 1) x - BCH(31,21) code calculated by bits 2-21 y - Even parity bit calculated by bits 1-31 As before, the unused parts of the last 32-bit data word are each filled with a "1".

The microcomputer 18 has four separate message memories. A user message memory can store 25 incoming user messages each of a maximum length of 80 Chinese characters, 160 ASCII characters, or an equivalent length for a mixed Chinese/ASCII message, longer messages being truncated at that length. Each user message has an associated time/date stamp in the format HR:MN DT/MH/YR indicating the 24-hour time and the date when the message is received, a memory number and a function code. Each message occupies approximately 196 bytes of memory, making a total of 4,900 bytes.

When the user message memory is full, the oldest message, Message No. 1, is deleted to accomodate a new message being received which becomes Message No. 25, or No. 26 if a tone only message, described below is present.

For secure storage, important user messages may be copied from the user message memory to a protected personal file memory which can also store 25 messages, of the same maximum length, some of which may be manually entered personal records. The personal file memory cannot be deleted by the receipt of an incoming message but can be deleted manually. After having been copied into personal file memory, original messages remain intact in the user message memory.

The pager can receive a tone-only call which has no message content other than the paging address and is treated in a manner completely transparent to the user. A tone-only message call is stored separately in a tone-only memory location to conserve overall memory. Only the time/ date stamp, memory number and function code of the most recently received tone-only call is stored regardless of its function code, and when it is reviewed, the display also shows a Chinese message meaning "Please call paging station".

During memory review and deletion, the tone-only message is treated logically as one of the user messages and can be copied and deleted in the same manner.

There are also two databank memories which together can receive and store up to 100 databank messages, each databank memory being able to store up to 80 databank messages. The databank messages also contain up to 80 Chinese characters, 160 ASCII characters, or an equivalent combination, as well as a memory number and a time/date stamp, the message received comprising database information such as stock prices, forex data, traffic, weather etc. If a databank message is received after that databank memory is full, the oldest message is deleted to accommodate the new message.

The function code of a databank message is ignored and is not displayed.

Figure 13 shows a preferred format for the LCD display panel 20 of the pager. Two dot matrix display areas 99, each of 16 rows and 112 columns, can display two lines, each of either 7 Chinese characters, of 16 x 16 pixels each including an inter-character space, 14 ASCII characters, or an equivalent combination. In addition, the areas 99 are able to display graphic characters contained in the GB231280 set. Any group of characters suspected of containing errors is displayed in reverse video format and any character containing error(s) that can not be corrected is displayed as a black block.

If databank messages are to be assumed to consist only of ASCII characters, a small display option can be selected as a 2-bit EEPROM factory option for one or both databank memories, causing the display panel 20 to show four lines, each of up to 18 ASCII characters, in a small font. By default the small display option is disabled.

Function code icons A-G indicating the function code of a message and other icons e.g. 100 can also be displayed.

The function code G indicates that a message just received or stored is a group address message. The alert icon 100 indicates that the switch 101, described below, is in position N. A page icon 112 indicates that a message displayed on the screen has more than one page and a service area icon 113 may indicate that the pager is within the effective paging service area. This latter is an EEPROM factory option which is disabled by default. If displayed, the service area icon 113 is switched off four minutes after the pager leaves the service area. When the personal file memory or one of the databank memories is being reviewed, a personal file icon 114 or one of two databank icons 115 respectively is displayed.

A real-time clock is displayed in the standby mode, described below, and during the storage of time/date stamps.

The real-time clock function is supported by a built-in calendar.

The dimensions of the LCD panel may be 24.5 x 57.5 x 2.3mm.

Figure 14 shows external views of the pager, which may have the dimensions 85.5 x 54 x 17mm.

A sliding switch 101 on the side of the pager has three possible positions, OFF, M and N. At OFF, all the functions of the pager except the real time clock are turned off. No paging or databank messages are received and a minimum power consumption mode operates, however the message memories are saved.

Position M is for mute operation, cutting off the beeper 23 of the pager completely through an electronic circuit. If the vibrator 24 is fitted, silent vibration replaces the beeper alert. In position N the pager operates to receive paging and databank messages normally.

A reset button 102 lying directly above the switch 101 is used to re-order the pager to the standby mode in which an incoming paging call is awaited. The reset button 102 can be used as a fast and convenient method of leaving a mode, especially when the user forgets the operating procedure.

There are also five control buttons 103-107 on the top of the pager. A "down" or "set" button 103 is used to review messages in descending order of their memory numbers, the highest number being displayed after No. 1, either stepping through them or scrolling through at the rate of two seconds per message when the button is depressed for more than two seconds. When pressed in the standby mode, the "down" button causes the most recently received user message to be displayed for 12 seconds, after which the pager reverts to the standby mode. The "down" button 103 is also used to enter the time set mode, by holding it down while moving the switch 101 from OFF, and, during setting of the clock, alarms and calendar, to decrement each digit.

A "page" or "del" button 104 is used to review subsequent pages of a message of any type which is too long to be displayed on one screen, i.e. when the page icon 112 is displayed. The "page" button 104 is also used in the time set mode to select the next digit.

An "up" or "dex*" button 105 operates in the same manner as the "down" button 103, but in reverse, stepping or scrolling up, rather than down, through messages of any type being reviewed. The "up" button 105 is also used to increment each digit in the time set mode.

A "mode" or "protect" button 106 is used to select whether messages are to be reviewed from the user, personal file, databank one or databank two message memories, by repeated pressing.

A "light" or "hold" button 107 operates an LCD backlight when depressed and then released. The backlight remains lit for 12 seconds in the standby mode or for the length of the timeout period in other modes. In the memory review mode, however, for as long as it is depressed, the "light" button 107 operates to "freeze" whatever information is on screen against such a timeout which would otherwise occur to save power. While the "light" button 107 is depressed in this manner, the buttons 104-106 assume different functions, the "light" button 107 operating as a "shift" or "alternative function" key. The page button 104 is used to delete the message on screen from the memory, the "up" button 105 clears an entire message memory and the "mode" button 106 copies a message from the user message memory to the personal file memory. When the "light" button 107 is operated as a "shift" key the LCD backlight is not lit.

If a paging call is received while the user is operating the pager, any incomplete operation will be aborted.

Figure 14 also shows a beeper grill 108 and a chain ring 109 on the side of the pager, an LED lens 110 for the LED 22 in the top corner and a battery door 111 on the back.

The pager according to the preferred embodiment of the invention includes a built-in canned phrase table with over 1000 frequently used phrases, each of which may contain a number of GB2312-80 characters. The canned phrase table can be customised to the user's requirements. The GB2312-80 character set contains approximately 7700 Chinese, English and graphic characters.

Regardless of message type, the time/date stamp is always displayed in the lower right hand corner of the -last page of the message during reception or memory review.

If a normal user or group message, but not a databank message, is received which is identical to an earlier message stored in memory, within 15 minutes of reception of the earlier message and having the same date stamp, then a duplicate message display function operates. A "D" sign is displayed at the top left hand corner of the first page of the duplicate message, above the memory number. The "D" sign will not appear in a duplicate message that has been copied into the personal file memory. A duplicate message can have a different function code from the earlier message.

When a paging call is received, the function code, memory number, service area icon 113, alert icon 100, "D" sign and page icon 112, the last three if appropriate, are displayed. The message itself also appears in the dot matrix areas 99. One of the function code icons A-D will be displayed, and the function code icon G in the case of a group call.

Unless the user intervenes and assuming the sliding switch 101 is in the N position, the receipt of a paging call will cause a paging alert, in which the beeper 23 sounds for eight seconds. The display remains on screen for 12 seconds before standby mode recommences. The LED 22 will flash regardless of the position of the switch 101, flashing in accordance with the beeping in the case of the N position. If the switch 101 is in the M position, the vibrator 24 will operate for eight seconds unless the reset button 102 or another button is pressed. Figure 15 shows how the beeper 23 sounds when messages having each of the eight possible function codes A-D and GA-GD are received.

The figure shows the operation over a two second period, which is repeated over the eight seconds. If the beeping alert sound is stopped by the user pressing any button except the reset button 102, the display remains on screen for a further 12 seconds with the pager in user message memory review mode. If the reset button 102 is pressed the beeping stops and the pager returns to standby mode immediately.

If a second paging call is received during the paging alert operation described above, the pager immediately performs a paging alert for the second call.

The display format for the review of user messages, personal file messages and databank messages is the same as that for received paging calls except that in the latter two review cases one of the icons 114 or 115 is displayed.

The pager supports four addresses, namely for normal user paging calls, group calls, and calls to the two databank memories which have separate addresses to receive databank messages directly. Thus databank messages containing different types of information can be segregated between the two databank memories. The normal user address contains one of the function codes A, B, C or D, and the group address, one of the function codes GA, GB, GC or GD.

On receipt of a databank message, no alert is raised and the message is stored directly in the appropriate databank memory for later retrieval.

The databank message reception function of the pager can be disabled or enabled at any time by means of on-air programming. Upon reception of a control message in the normal user address consisting of the hexadecimal ASCII version of the pattern "99999991" or "99999992", the message reception function is disabled for databank one or databank two respectively. The function is enabled upon receiving the hexadecimal ASCII version of the pattern "88888881" or "88888882" for databanks one or two respectively.

For the on-air programming to be available the dedicated EEPROM programmer must first be used to set te appropriate internal EEPROM option at the factory. The default option is for this entire function to be disabled, and is preferred if only normal user messages are to be received. This is because more power is consumed if on-air programming is enabled, regardless of whether databank message reception is enabled.

An alerting melody output may be provided which the user can assign to one of the four function codes for normal user messages, normally that of the first paging call, the other three actuating a beep alert. This is an EEPROM factory option which is enabled by default, the disabled alternative actuating the beep alert in all four function code cases.

The data transmission rate is either 512 or 1200 bps (bits per second).

An EEPROM factory option allows dates to be displayed in either day/month/year or month/day/year format, with the former being selected by default.

In the standby mode, the alert icon 100, optionally the service area icon 113, the real time clock and, if appropriate, a low battery message are all displayed. The low battery message, which is in Chinese and means "Battery Low!!!" is displayed on the LCD panel 20 if the battery voltage falls below a threshold of 1.15V. Also, if the sliding switch 101 is in the N position, the beeper 23 will produce a long tone, represented at 116 in figure 15 and the LED 22 will light simultaneously for 4 seconds every 30 minutes, the beeping being replaced by vibration if the switch 101 is in the M position.

During power up, when the sliding switch 101 is moved from OFF to the N position, the beeper produces a long alert tone, to confirm proper operation, of approximately four seconds, as shown at 117. Moving the switch 101 from OFF to the M position causes a four second vibration alert. The LED 22 flashes and all pixels of the LCD panel 20 are switched on for the four second period, after which the pager enters the standby mode. Any button can be pressed to stop the alert tone or vibration, immediately initiating the standby mode.

Chinese words are displayed to inform the user how the pager is operating. If an empty message memory is to be reviewed, the word displayed means "no message".

Chinese words meaning "delete the message" are displayed when the "light" and "page" buttons 107, 104 are pressed as described above. The user confirms deletion of the message by pressing the "page" button 104 a second time with the "light" button 107 still held down. The message is erased and Chinese words meaning "message deleted" are displayed until the "light" button 107 is released.

Similarly, when the "light" and "up" buttons 107, 105 are pressed together in memory review mode, Chinese words meaning "delete this file" are displayed. If the user presses the "up" button 105 a second time without releasing the "light" button 107, the entire memory file is erased and the display shows words meaning "file deleted" for as long as the "light" button 107 is held down.

In the time set mode, which is entered when theswitch 101 is moved from OFF with the "down" button 103 held down, the pager prompts the user for entry of the time and date to set the real time clock. The digits are set one by one and each digit is incremented by pressing the "up" button 105 or decremented by pressing the "down" button 103. The next digit is selected by pressing the "page" button 104. The user then confirms and stores the set time and date by pressing the reset button 102 to enter standby mode. If the reset button 102 is pressed when the clock is only partially set, an incorrect time is stored which can then be corrected. In the time set mode, paging call reception is inhibited.

The pager also has two appointment alarms. These are selected for setting by holding down the "mode" button 106 while the sliding switch 101 is moved from OFF. The alarms are set one by one in the same manner as the real time clock. Paging call reception is also inhibited in this mode.

The maximum current consumption is 0.4mA when the pager is switched off, 9mA when able to receive messages and below 2mA on average with no alerts. The battery life is longer than ten days when two calls are received each day and the pager is switched on continuously. The backup power supply 26 allows the real time clock to run for over two minutes during replacement of the "AAA" battery 27. The message data stored in the pager is protected during this time.

When the low battery message is displayed, the pager should be switched off and the battery replaced quickly to avoid losing data and clock settings.

The pager has the following further specifications: RF Frequency Range : 130-170 MHz (other frequencies possible) Channel Spacing : 25 kHz Modulation : NRZ FSK + 4.5 kHz Paging Sensitivity : < 7pV/m (Front, Body) Spurious Image Rejection : > 50 dB Selectivity : > 65 dB at + 25 kHz Frequency Stability : + Sppm at -10 to +55 C Alert Tone Sound : > 80dBA at 0.3m Level Operating Range : 0 to 50'C, 90% R.H., no dew forming The system of the invention places no limitation on the mixing of languages or canned phrases for each message transmission.

Also, as each codeword contains its own language mode code, a random error occurring in the transmission that affects the language selection can often be corrected at the receiver side by checking the language mode code of the codewords that precede and follow the erroneous codeword.

This is because each alphanumeric character and each ideographic character is composed of a known number of codewords, and characters of the same language usually come in batches. The decoding algorithm can make use of these properties to set up error correction rules. Even if the error is not corrected, error accumulation is limited to one or two long codeword(s), depending on the number of long codeword(s) representing a canned phrase.

Whilst a particular embodiment of the invention has been described, modifications may be made without departing from the scope of the invention.

For example, it is not essential to use POCSAG code for transmission purposes, and the Golay Sequential Paging Scheme or any other suitable format may be used.

Instead of using control characters to introduce the canned phrases, the canned phrases could be treated as a separate language with its own language mode code. A canned phrase would be represented by one or more codewords each having that mode code. Different sets of canned phrases could each have a different language mode code. For this modification, the language mode codes would of course consist of more than one bit each. This method of encoding the canned phrases is robust against error.

The invention can be used with other languages than English and Chinese and with other character sets than ASCII and GB2312-80.

Claims (12)

1. A communications system comprising means for encoding data to be transmitted, and means for receiving the transmitted data, the receiving means including means for decoding the received data, the encoding means being operable to encode the data into segments each representing a symbol or a canned phrase belonging to a natural language, and the decoding means being arranged to use information from each of said segments to determine to which language the symbol represented thereby belongs or whether it represents a canned phrase, thus reducing the propagation of an error occurring in one part of the data to other parts thereof.

2. A communications system according to claim 1, wherein the symbols are alphanumeric and/or ideographic characters.

3. A communications system according to claim 1 or 2, wherein the receiving means includes means for displaying the symbols.

4. A communications system as claimed in claim 1, 2 or 3, wherein the receiving means is arranged to store at least one canned phrase of a natural language, the or each canned phrase having a corresponding address code associated therewith.

5. A communications system as claimed in claim 4, wherein the decoding means is arranged, upon receipt by the receiving means of transmitted data containing at least one address code, to identify the or each address code from information contained in the said transmitted data.

6. A communications system as claimed in claim 5, wherein the receiving means is operable to retrieve the stored canned phrase associated with the or each address code identified by the decoding means and to display said phrase on said display means.

7. A communications system as claimed in claim 5 or 6, wherein the or each address code and said information identifying the or each address code each comprise at least one segment.

8. A communications system as claimed in any preceding claim, wherein said segments consist of codewords of a standard number of bits, a fixed part (fixed number of bits) of each codeword constituting said information from each of said segments which determines to which language the symbol represented thereby belongs or whether it represents a canned phrase.

9. A communications system as claimed in any preceding claim, wherein said system is a paging network, the receiving means comprising at least one pager and the network including a paging centre operable to transmit messages each comprising a plurality of said symbols via at least one transmitter to the or each pager.

10. A communications system as claimed in any preceding claim, wherein the receiving means includes a plurality of message memories for storing different types of messages, each message comprising a plurality of said symbols.

11. A communications system as claimed in claim 10, when dependent upon claim 9, wherein said different types of messages include user messages sent from a caller to the or each pager, personal file messages which are protected against being overwritten by a newly received message and/ or databank messages sent from a database to the or each pager.

12. A communications system substantially as described