ing Systems:
`itectures and Interfcxces
`
`Yi-Bing Lin, National Chiao Tung University
`
`Abstract
`
`In the past few years, the penetration rate of p a g i n g service has significantly
`increased. Paging service is competitive because paging networks offer inexpensive
`service in large service coverage areas, and pagers are cheap and easy to carry.
`This article provides an overview of the one-way paging network architecture and
`the interfaces among the paging network elements. The author also briefly describes
`two-way paging services and uses GSM short message service as an example to illus-
`trate how two-way messaging can be offered under the cellular platform.
`
`its original definition, paging is a one-way, personal
`lective wireless calling system. The first paging system
`was developed by Charles F. Neergard, a hospitalized
`radio engineer who could not tolerate the constant, loud
`voice paging of doctors in the hospital [l]. At the end of 1995,
`there are approximately 80 million paging subscribers world-
`wide [2]. The largest paging industry in the world is the one in
`the Asia-Pacific region. Consider Taiwan as an example: the
`subscriber population has increased by at least 20 every year
`since 1986 (Fig. la). The subscriber density in Taiwan has
`increased rapidly (Fig. lb) and is ranked fifth in the world. In
`the United States, the paging service penetration rate
`increased from 13 to 15.5 in 1996.
`Although paging service does not provide real-time interac-
`tive communications between the calling and the called parties,
`it has several advantages over other personal communications
`services (PCS), such as cellular telephony. For example, the pag-
`ing system requires less radio bandwidth (thus, the service is inex-
`pensive), the paging service coverage area is larger, the pager is
`cheaper, lighter, and smaller, and the battery lasts longer. As long
`as paging continues to keep these advantages, paging service
`will be a strong competitive PCS service in the future.
`This article describes paging systems with the emphasis on
`the network architectures and interfaces. Paging air protocol
`issues such as power saving, repeat intervals, and out-of-band
`confirmations are not covered in this article; these issues are
`discussed in 13, 41. The reader is encouraged to visit Bradley
`Dye's home page http://idt.net/-braddye for information on
`advanced paging technologies.
`fuging Messuge Types
`A paging message can be one of the following four types:
`
`An Alert Tone
`The receiver is a tone pager. A tone pager has a dedicated tele-
`phone number; when the number is dialed, the pager is trig-
`gered. The advantage of tone paging is that it utilizes a small
`amount of air time. However, there are only a few kinds (typi-
`cally four) of alert tones that can be used to identify callers.
`A Voice Message
`In some systems, a voice message may be transmitted after the
`beep. The receiver is a voice pager [5] that can 'receive up to 12
`min of voice messages. However, speech consumes a lot of air
`time, and may not be cost-effective for most public paging systems.
`Some voice paging systems utilize the spare capacity of the
`existing cellular networks using non-real-time transmission.
`Even in non-real time, messages are typically sent within min-
`utes of receipt.
`A Digit Siring
`The receiver is a numeric pager. The pager has a small LCD
`screen to display the digit string. Typically, the string is the
`telephone number of the caller. In Hong Koiig and Singapore,
`the string can be a coded message. The coded message is gen-
`erated by the paging center at the request of the caller, and
`the message is decoded by a code book (which may be built
`into the pager). This type of paging is efficient in air time
`usage just like tone paging. Numeric paging was introduced in
`the early 1980s and is the most popular form of paging.
`Some Texi Strings
`The receiver is an alphanumeric pager, which has a large
`screen to display several text strings. Alphanumeric paging
`was introduced in the late 1980s. Since the caller needs special
`
`56
`
`0890-8044/97/$10.00 0 1997 IEEE
`
`IEEE Network July/August 1997
`
`Authorized licensed use limited to: United States Patent and Trademark Office. Downloaded on June 28,2022 at 16:15:26 UTC from IEEE Xplore. Restrictions apply.
`
`Petitioner Hyundai Ex-1028, 0001
`
`

`

`
`
`.
`
`.
`
`I
`
`.. . .. . .....
`
`200 -
`
`160-
`
`80 -
`
`40 -
`
`(1 0.000s)
`
`0 . 7 I
`
`/
`
`I
`
`I
`
`I
`
`8
`
`I
`
`I
`
`I
`
`I
`
`Year
`(a) Paging subscriber population
`
`2;FJ,,
`
`,
`,
`,
`'93'94'95'96
`
`,
`0
`'85'86 '87'88'89'90'91'92
`I
`Year
`I
`(b) Paging subscriber density
`. I
`Figure 1 . Paging subscriberpopulatiorz in Taiwan.
`
`I
`
`!
`
`i
`
`.~
`
`input devices or a special input procedure with a phone set,
`alphanumeric paging is not as popular as numeric paging.
`Depending on the paging system, the paging process can
`be done either manually or automatically. In a manual paging
`system, the caller sends a message to the paging operator by
`a telephone call. The operator then delivers the message to
`the pager through the paging network. In an automatic pag-
`ing system, a paging terminal automatically processes caller
`requests and delivers messages to the pagers [6]. This article
`will focus on automatic paging.
`The Paging Network Architecture
`F tecture. The network consists of six basic elements.
`igure 2 illustrates an example of the paging network archi-
`
`User Terminal Equipment (Input Device)
`The caller sends messages by using user terminal equipment
`(common telephone handsets, computer with modem, or spe-
`cific input devices). The alert tone and numeric messages are
`typically generated from telephone handsets and delivered
`through the public switched telephone network (PSTN), while
`alphanumeric messages are generated by computers or tele-
`phone handsets (through the operator of a paging center)
`and delivered to the paging network through the public
`switched data network (PSDN).
`Paging Terminal
`Through the user access interface, messages are sent to the
`paging terminal. A high-capacity paging terminal may support
`more than 1 million pagers, and is typically connected to a
`central office of the PSTN through T1 (up to 24 subscriber
`lines) or El (up to 30 subscriber lines) trunks. A high-capacity
`paging terminal can support more than 1900 input lines. It
`may also connect to other paging terminals, and the connect-
`ed paging terminals communicate with each other via an
`internetwork interface. The paging terminal maintains a cus-
`tomer database that contains information necessary to alert
`the individual pager (e.g., pager number, pager code, and
`types of message) and for billing. Based on the alerted pager's
`record in the customer database, the paging terminal may
`deliver the message to its' base station controllers or forward
`the message to other paging terminals. A voice message can
`be converted into text for an alphanumeric pager. The paging
`terminal may store the message in a mailbox (with capacity
`for several hundred hours of voice storage).
`The paging terminal can be maintained locally (e.g., removing
`or installing cards) or remotely when the system is operational.
`
`I
`I
`I equipment
`
`User
`
`I
`
`I
`telephone
`I network
`j
`
`or
`
`'I
`Paging
`terminal
`
`-
`
`Base
`station +
`
`The Operation and Maintenance Center
`The operation, administration, and maintenance functions of
`a paging network are conducted by the operation and mainte-
`nance center (OMC). The OMC accesses the customer
`database of the paging terminal to add new customer records,
`delete terminated customer records, collect billing informa-
`tion, and so on. The OMC may also page the customers
`through the paging terminal. Typical OMC commands include
`read (from the paging terminal database), write (to the paging
`terminal database), and alert (a specific pager).
`The Base Station Controller
`A base station consumes high power when transmitting a paging
`signal, and is in low-power mode when it is idle. Since the ser-
`vices of an alerted pager may be limited to specific geographical
`areas smaller than the area covered by the paging terminal, a
`base station controller is used to link the paging terminal to the
`base stations to control the base stations to be powered up.
`If the radio channel is shared by different paging service
`uroviders. the base station controllers must be coordinated so that
`at most one paging signal is transmitted in
`the air at a time. The paging signal may
`be sent to multiple paging transmitters
`simultaneously (simulcast paging) or
`sequentially (transmitter sequencing).
`Although transmitter sequencing does
`not use air time efficiently (it takes
`longer to communicate with a pager), it
`may be required when simulcasting from
`two transmitters creates a region where
`the pager cannot receive reliable data
`from either transmitter.
`The Base Station
`Base stations may be designed for two-
`way voice, although most are for one-
`way paging. Three
`transmission
`technologies are possible for delivering
`
`.
`
`I
`I
`,., !,.,=
`
`I
`I
`
`data network
`
`Internetwork interface
`
`User access interface
`
`to other paging
`terminals
`
`.. . . -
`W Figure 2. An example of the paging network architecture.
`
`_. . .. . .
`
`.
`
`.
`
`.
`
`... .
`
`Air interface
`
`........
`
`.
`
`.
`
`IEEE Network July/August 1997
`
`57
`
`Authorized licensed use limited to: United States Patent and Trademark Office. Downloaded on June 28,2022 at 16:15:26 UTC from IEEE Xplore. Restrictions apply.
`
`Petitioner Hyundai Ex-1028, 0002
`
`

`

`the message from the paging terminal to the base stations:
`leased telephone lines, PSDN (X.25), or satellite-based net-
`works. Satellite communications technology is typically used
`between the paging terminal and the dispersed base stations.
`In some advanced paging services, satellites can serve as base
`stations to broadcast messages to the pagers [7].
`Base stations broadcast the message to the pager by radio
`signal. The radio path between a base station and the pagers
`can be dedicated paging channels, FM radio, or subcarriers of
`TV stations [8]. (In the last case, a signal is superimposed on
`the normal TV or radio channel, which does not interfere
`with the normal TV broadcast. The signal is extracted by the
`pager to obtain the paging message.) The base stations have
`high transmitter power (hundreds of watts to kilowatts to pen-
`etrate walls of buildings) and high antennas (for a large cover-
`age area). Strong one-way radio transmission allows
`low-complexity, low-power paging receivers.
`In simulcasting, messages from the paging terminal may
`arrive at the base stations asynchronously (because the dis-
`tances from the paging terminal to the base stations may not
`be the same). Different time lag values or audio equalizers are
`used at the base stations to synchronize the message receipt of
`the base stations.
`The Pager
`A pager consists of four basic components: a receiver,
`decoder, controller logic, and display. The receiver is tuned to
`the same radio frequency (RF) as the base station to receive
`and modulate the paging signals. The decoder decodes the
`binary information and identifies the code for the pager (and
`rejects messages for other pagers). A pager may share the
`same code with other pagers for group paging, or may be
`assigned multiple page codes (typically up to four) for differ-
`ent paging functions. The control logic provides service fea-
`tures and operation functions such as duplicate message
`detection (repeated messages will not be stored in memory),
`message locking (selected messages are not overwritten), mes-
`sage freeze (the message is kept on the screen when reading),
`multiple alerting modes (e.g., audible tone, visual flashing,
`and silent vibration), and so on.
`A pager is typically powered by a single AA battery. Battery-
`saving techniques are used to conserve power by periodically
`switching the pager into low-power mode. When the pager is
`off, stored messages are retained in nonvolatile memory.
`
`User Access Interface
`e caller sends a paging request to the paging terminal
`
`r” through the user access interface. Protocols for this interface
`
`include analog trunk protocols and digital protocols. We describe
`two digital protocols, Telocator Alphanumeric Input Protocol
`(TAP) [9] and Telocator Message Entry Protocol (TME) [lo].
`Telocator Alphanumeric Input Protocol
`In TAP, the caller prepares the pages to be sent in advance
`(thus, text creation does not consume any air time). To send
`the message, the input device dials the paging terminal’s num-
`ber. After the line is connected, the caller (input device) and
`paging terminal exchange messages as described below.
`Msg I [Caller + PT) - The input device repeats the carriage
`return
`Msgl = <CR>
`at 2-s intervals until the paging terminal replies. If the paging
`terminal does not reply after three repetitions, the paging
`request fails.
`
`Msg2 ( P I + Caller) - Within 1 s of receipt of Msgl, the pag-
`ing terminal requests the input device to provide the pager ID
`by sending the sequence
`Msg2 = ID =<CR><LF>
`where <LF> is the line feed symbol.
`Msg3 (Ca//er -+ PI) - The input device sends a string
`Msg3 = <ESC> <SST> <PPPPPP>
`to the paging terminal. The escape character <ESC> indi-
`cates that the paging information will be sent in automatic
`dump mode. The first two alphanumeric characters SS repre-
`sent the type of service. For example, the combination SS =
`PG represents that the information will be sent as a pair of
`two fields where field 1 contains the pager ID and field 2 con-
`tains the message. The last character, T, represents the type of
`input device. For example, 1x0 or PET devices arc in the cat-
`egory T = 1. The (optional) six digits <PPPPPP> are the
`password of the input device, which may be interpreted as a
`caller ID or system entry key.
`Msg4 ( P I + Caller) - When the paging terminal receives
`Msg3, one of the following three situations occurs:
`e If the request is accepted, an acknowledgment is sent to the
`input device:
`Msg4 = < C R > d C K > < C R >
`*If the format of Msg3 is not correct, the paging terminal will
`request retransmission by sending a nonacknowledgment mes-
`sage:
`Msg4 = ‘‘ < CR > <NAK> < CR > “
`If the paging terminal is not available to handle the paging
`request, the request is rejected by a forced disconnect message:
`Msg4 = ‘‘ < CR > <ESC> <EOT> < CR > “
`where <EOT> means “end of transmission.”
`Msg5 ( P I + Caller) - If Msg4 is an accept acknowledgment,
`the paging terminal will prepare for message input, and when
`ready it sends a “go ahead” sequence to the input device:
`Msg5 = <CR> [p <CR>
`Msg6 (Caller -+ PI/ - The input device starts to send the
`paging information to the paging terminal. The information is
`partitioned into data blocks of 256 bytes. Every block consists
`of three control characters, a message text of length up to 250
`characters, and a three-character checksum. For illustration
`purposes, we assume that the service type is PG1, and the
`paging information is partitioned into two data blocks (i.e.,
`two transmissions are required to deliver the message text).
`Then the first block is delivered in the following format:
`Msg6 = <STX>Pager_ID<CR>Text<CR>Cksum<ETB>
`where <STX> means “start transmission,” and <ETB> is
`used as a block terminator if the transaction is continued into
`the next block. Pager-ID is a string of ASCII digits to identify
`the destination pager, Text is a part of the paging message,
`and Cksum is the three-character checksum.
`Msg7 (PT + Caller) - After receiving Msg6, one of the fol-
`lowing occurs:
`0 If the transmission is correct, an acknowledgment is sent
`back to the input device:
`Msgl = <Message-Sequence > < CR > <ACK> < CR >
`
`58
`IEEE Network * July/August 1997
`Authorized licensed use limited to: United States Patent and Trademark Office. Downloaded on June 28,2022 at 16:15:26 UTC from IEEE Xplore. Restrictions apply.
`
`Petitioner Hyundai Ex-1028, 0003
`
`

`

`If a transmission or checksum error occurs, the following
`message is sent to request the input device to resend Msg6:
`Msg7 = <MessageSequence > < CR > <NAK> < CR >
`Note that Message-Sequence is optional, and in some systems
`may only be included in the last response message (i.e.,
`MsglO) before disconnect.
`Msg8 (Caller + PT) - Suppose that the input device is to
`send the last block. The message is of the form
`Msg8 = <SIx>Pager-ID<CR> Text<CR>Cksum < E m >
`where <EXT> indicates that Msg8 delivers the last piece of
`the paging information.
`Msg9 (PT -+ Ca//er) -When
`the paging terminal receives
`Msg8, one of three situations may occur. The first two situa-
`tions are the same as those for Msg7; the third occurs if the
`transmission violates a system rule (e.g., the destination pager
`cannot be paged). Then Msg8 is abandoned by the following
`replied message:
`Msg9 = <Message-Sequence > < CR > <RS > < CR >
`The reason for abandonment is given in MessageSequence.
`Msg 10 (Ca//er + PT, - If Msg9 is of type d C K > or <RS>,
`the input device completes the transmission by sending
`MsglO = <EOT><CR>
`Msg / 7 (PT + Caller) - The paging terminal breaks the con-
`nection by sending
`Msgll = <Message-Sequence> <CR> <ESC> <EOT> <CR>
`where MessageSequence reports the degree of acceptability of
`information on this service.
`After the paging terminal sends a message, a timer of at
`least 4 s is set (for Msg2, the timer is 8 s). Similarly, a timer of
`at least 10 s is set for the input device after it sends a mes-
`sage. The connection is disconnected after timer expiration.
`Telocator Message Entry Protocol ITME)
`TME is the data input protocol of the Telocator Data Protocol
`(TDP) [lo]. TME is considered the successor to the TAP pro-
`tocol. TME relaxes the 7-bit, even-parity TAP coding format
`by allowing unrestricted 8-bit data transfer. It also relaxes the
`conventional TAP short ASCII text messages by providing
`entry for long messages of text or binary data (e.g., e-mail
`with attached files, spreadsheets, and database information).
`TME extends TAP one-way paging to two-way communica-
`tions (to be addressed in a forthcoming revision). It also pro-
`vides new service features such as priority paging (to indicate
`that a request is a priority or emergency page to be sent
`immediately), deferred paging (to send a page at a particular
`time of day), periodic paging (to send a page periodically until
`a cancel message is received), message forwarding, message
`deletion (e.g., to cancel periodic paging), and so on.
`TME follows the open systems interconnection model
`except that no presentation and session layers are required.
`Transmission Control Protocol/Internet Protocol (TCP/IP) is
`recommended to serve as the lower-layer (the network layer
`and below) protocol of TME, although other protocols such
`as X.25 [ l l ] can also be used to support the TME application
`layer. When the input device requests connection to the pag-
`ing terminal in the TCP/IP solution, it must include the TCP
`port number 4076 and an available unregistered port number
`to uniquely identify itself as the client application. More
`details of the TME lower-layer protocols can be found in [lo].
`
`The TME application layer is specified by the remote opera-
`tions service element (ROSE) [12, 131, and all TME operations
`are defined by using the Abstract Syntax Notation 1 (ASN.l)
`notation and Packed Encoding Rules (PER) [14]. There are ten
`basic TME operations; three of them are listed below. The read-
`er is referred to [lo] for details of the other operations.
`The login operation establishes a session connection between
`the input device and the paging terminal. Password is an
`optional argument of the login operation. Note that the
`caller password is required if operations such as deleteOp
`(to delete a paging request) or d i r (to list the messages cur-
`rently in the system, which have been sent to or from the
`subscriber’s account) are to be performed in the session.
`The send operation sends a message from the input device to
`the paging terminal. The message consists of an envelope and
`the message content to be delivered. The paging terminal
`only forwards the message content to the pager, which can
`be transparent data, tone only, numeric and alphanumeric
`data, or 8-bit wireless message format (WMF) [lo] data.
`The logout operation terminates a communication session.
`This operation may be issued under the request of the
`input device or system operator, or due to the expiration of
`the inactivity timer or session timer. The session can also be
`terminated if situations such as security violation or
`resource shortage occur.
`
`1
`
`The hersystem Inferface
`o extend paging service coverage, paging terminals can be con-
`nected to form paging terminal networks. In order to receive
`intersystem paging services, the pager should keep the paging ter-
`minal updated about its locations. This information specifies the
`group of paging terminals to receive the paging requests and is
`stored in the customer database of the paging terminal. Several
`proprietary protocols such as the Glenayre Data Link Module
`(DLM) protocol and Spectrum Data Link Handler (DLH) pro-
`tocol have been developed for paging terminal networks. Note
`that paging terminals using different protocols cannot talk to
`each other, and network paging gateways are needed to per-
`form protocol conversion. Several hundred types of gateways
`are used to connect paging terminal networks.
`An industry standard protocol, Telocator Network Paging
`Protocol (TNPP) [ E , 161, has been developed for paging ter-
`minal connection. TNPP is a point-to-point communication
`protocol which can be built on top of TCP/IP networks [17].
`To move paging request data from the source to the destina-
`tion in a TNPP network, a routing algorithm is required,
`although routing of paging requests is not covered in the
`TNPP specifications. The DLH network follows the same
`point-to-point philosophy. On the other hand, the DLM net-
`work uses a token-passing protocol; that is, a token is passed
`around the DLMs (every paging terminal is connected to a
`DLM). The DLM who grasps the token will transmit the pag-
`ing request data. All other DLMs will listen, and the destina-
`tion DLMs will read the data from the network and pass the
`data to their paging terminals.
`A TNPP packet has a 4-byte source address and a 4-byte
`destination address. It uses a 2-byte sequence number to dis-
`tinguish between new and retransmitted packets. A 2-byte iner-
`tia counter is used to limit the number of nodes visited by the
`packet so that the packet will not travel in the network forever.
`Like TAP, TNPP is an ASCII-oriented protocol. After a
`packet delivery, the destination node may reply with one of
`four characters to the source node: <ACK> (the delivery is
`successful), <NAK> (an error has occurred, and the packet
`should be resent), <CN> (the destination address is invalid
`and the packet is canceled), and <RS> (the destination node
`
`IEEE Network
`
`July/August 1997
`
`59
`
`Authorized licensed use limited to: United States Patent and Trademark Office. Downloaded on June 28,2022 at 16:15:26 UTC from IEEE Xplore. Restrictions apply.
`
`Petitioner Hyundai Ex-1028, 0004
`
`

`

`1 '-P';';&-pLq;:$
`-- .- - -
`32 bits 64 bits
`
`576 bits
`
`544 bits
`
`544 bits
`
`.. .. - -
`
`_ . . . - - -
`
`64 bits
`
`~
`
`One! (1: m o r e
`
`I batches
`I 8 frames
`
`j 2 codewords
`
`I
`
`Address
`cod CWo rd
`
`or
`
`--- . . L _ .
`- - - _ . ,
`..
`64 bits
`
`The POCSAG coding format (Fig. 3)
`consists of a 576-bit preamble and one or
`more 544-bit batches. The preamble is a
`string of an alternating "1010 ..." pattern,
`which is used to identify the POCSAG
`signal. The decoder of the pager utilizes
`the preamble to synchronize with the data
`stream.
`Every batch consists of one 32-bit frame
`synchronization codeword and ,eight 64-bit
`frames. The frame synchronization code-
`word has a unique bit pattern, which is
`used to identify the beginning of a batch.
`Every frame consists of two 32-bit code-
`words. A codeword can be an address, a
`message, or an idle pattern. An address
`codeword consists of five fields. The first
`field is a bit 0. The second field is an 18-
`bit address. The third field is a 2-bit
`source identifier used to identify four dif-
`ferent paging sources. The fourth field is
`a 10-bit error detection and correction
`code, which can be used to correct single-
`bit errors and detect multiple-bit errors.
`The last field is an even paritv bit. The
`structure of a message codeword is simiiar to the
`address codeword, except the codeword begins with 1,
`and bits 2-21 are occupied by the message text. An idle
`codeword is a unique bit pattern. An address codeword
`is always followed by zero or more message codewords.
`If the second half of a frame is empty, it will be padded
`- - - -
`with the idle codeword to ensure that every frame has
`.'k f. fl,L:~,F/
`m'1TF-, . -:;;,?2'!!>,,:
`. - ._I--.
`17 sPcnnds _I 64 hits.
`~-
`.
`-11 I
`A pager can only be paged at one of the eight frames
`1 G bakhes
`.
`:. , \ .. '
`---
`.___,___ .
`. , ',,,';, .
`(the 3-bit uattern of the frame is stored in the uaner).
`... ... .
`.
`U Figure 4. The ERMESframe structure.
`Thus, the ieceiver of the pager can be turned offdlring
`the other seven frames to reduce power consumption.
`POCSAG is efficient for large volumes of data trans-
`mission. For a lightly loaded mixture of encoding formats, air
`time may be wasted (many frames will be inserted with idle
`codewords).
`ERMES
`ERMES is an open system developed by consensus with oper-
`ators and manufacturers. It is the only International Telecom-
`munications Union (ITU) recommended paging code
`standard. The ERMES air interface I1 operates with 16 fre-
`quencies ifi the radio band 169.4125-169.8125 MHz. It uses a
`4PAM (four-level pulse amplitude modulated) frequency
`modulation (FM) scheme. In this scheme, four frequencies
`are used to represent the binary codes 00, 01, 10, and 11. At
`any moment, only one of the four is used to transmit the sig-
`nal, and two-bit information is delivered at the moment. The
`effective transmission rate of a frequency is at 3750 b/s.
`ERMES has improved the paging capacity over POCSAG by
`a factor of four (in terms of the number of users per hertz).
`Every ERMES pager is identified by a 35-bit radio identity
`code. Thus, the address space is large enough to accommo-
`date hundreds of millions of pagers.
`ERMES partitions every hour into 60 cycles, as shown in
`Fig. 4. Every cycle is partitioned into five subsequences, and
`every subsequence is partitioned into 16 batches. Battery-sav-
`ing mode is implemented such that a pager is programmed to
`be paged on specific subsequences or cycles. Like POCSAG,
`information in a batch is partitioned into codewords. Every
`nine codewords are grouped into a codeblock. Instead of
`transmitting the codewords sequentially, the codewords in a
`
`1 2
`
`SoLirce idmtificatibn bits
`21 22
`
`Even parity bit
`
`31 '\32
`
`Wlessdye
`i
`(otle\wrcl
`I Figure 3. The POCSAG coding format.
`
`I
`
`5 sulxcqucnces
`
`I+.-
`
`I' , ..I.,
`
`I
`
`~
`
`cannot process the packet). Note that the source and destina-
`tion nodes can send packets to each other at the same time.
`TNPP can be used for satellite communications between
`paging terminals. Note that the destination paging terminal
`cannot acknowledge receipt of information in TNPP satellite
`communications. To provide reliable satellite transmission, a
`retransmission technique is used to transmit multiple copies
`of data to the destination.
`The Air Interface
`S
`everal signaling formats have been used in the paging air
`interface. POCSAG (Post Office Code Standardization
`Advisory Group) was initiated by the British Post Office in
`1970. Most paging systems are based on this protocol. In early
`1990, several high-speed protocols became available, including
`ERMES (European Radio Message System) [MI, approved by
`the European Telecommunications Standards Institute
`(ETSI); FLEX [6], developed by Motorola; and APOC
`(Advanced Paging Operations Code), developed by Philips
`Telecom. We describe POCSAG and ERMES in this article.
`POCSAG
`POCSAG is designed for a one-frequency, one-operator pag-
`ing network, and cannot be extended for multinetwork opera-
`tion. The POCSAG coding format can accommodate 2 million
`pagers. The original format was specified to operate at 512 bls.
`Without modifymg the coding format, POCSAG can be oper-
`ated at 1200 bls, and up to 2400 bls for some applications.
`
`69
`
`IEEE Network JulyIAugust 1997
`
`Authorized licensed use limited to: United States Patent and Trademark Office. Downloaded on June 28,2022 at 16:15:26 UTC from IEEE Xplore. Restrictions apply.
`
`Petitioner Hyundai Ex-1028, 0005
`
`

`

`codeblock are transmitted interleaved; that is, the previous
`bits of all nine codewords are transmitted before the next bits
`are transmitted. Due to the bursty error nature of radio
`transmission, codeword interleaving is likely to spread only
`one or two error bits to each of several codewords, which can
`easily be fixed by the error correction mechanism.
`
`Future Trends
`raditional paging systems do not provide confirmation that
`
`T correct messages have been delivered to pager wearers. In
`
`the future, a potential solution is two-way paging systems.
`These systems can be implemented by using cellular phone
`network platforms, such as the short message service of the
`Global System for Mobile Communications (GSM) protocol.
`In GSM, short messages are delivered using the signaling
`channel; the traffic channels that carry voice and bearer ser-
`vices are not affected. Thus, GSM short message service is
`considered a high-value-added network service. The GSM
`short message service architecture is illustrated in Fig. 5 [19].
`In this architecture, a GSM mobile station (MS) serves as a
`pager. The short message sender can be a wireline user using a
`telephone or an input device as we described before (la). The
`short message is sent to a short message service center (SM-SC),
`which is a specific paging terminal tailored to GSM. The mes-
`sage sender can also be another MS (lb). In this case, the mes-
`sage is sent to a special GSM mobile switching center (MSC),
`the interworking MSC (IWMSC). The IWMSC utilizes a net-
`work protocol to communicate with the SM-SC. The specifica-
`.tion of this protocol is left open in the GSM standard. The
`SM-SC forwards the short message to a gateway MSC, or GMSC
`(2). In GSM, the locations of all MSs are recorded in a database
`called the home location register (HLR) [20]. The GMSC queries
`the HLR (3 and 4) to identify the destination MS's location (the
`MSC address). The short message is routed to the destination
`MSC, and the associated base stations (BSs) will page the MS
`(5, 6, and 7). The paged MS may confirm the short message
`by sending an acknowledgment back to the originating MS
`using the same message path.in the reverse direction.
`With two-way paging, the billing structure can be signifi-
`cantly changed [21]. A traditional paging carrier charges pager
`wearers on a monthly (flat-rate) base. With paging confirma-
`tion, the message senders' satisfication can be significantly
`improved, and it is possible to implement a calling party pay
`(CPP) billing program.
`In the future, paging will become a typical example of com-
`puter telephony integration. Several software packages are
`already available to bridge the Internet to the paging net-
`works, and Web or e-mail applications in computers can be
`used as input devices to send messages to pagers. If pagers
`are replaced by sophisticated personal digital assistants or lap-
`tops, and paging networks equipped with two-way communi-
`cations capability, the resulting architecture becomes a
`platform for mobile computing. As David Rose, Chairman
`and Chief Executive Officer of Ex Machina, says, "As far as
`paging goes, the best is clearly yet to come" [l].
`Acknowledgments
`The author would like to thank the reviewers for their valuable
`comments that have significantly improved the quality of this
`article. This work was supported in part by the Microelectronics
`
`j
`
`i
`
`
`
`(
`
`I i
`
`I
`
`W Figure 5. The GSMshoi-t message sewice architecture.
`
`. .
`
`_._ __. .. .
`
`~
`
`and Information Systems Research Center, NCTU, and Nation-
`al Science Council, Contract No. NSC 86-2213-E-009-074.
`References
`[ l ] D. Rose, "A Very Brief History of Paging," http://exmachina.com/
`histpageshtml, 1995.
`[2] M. Brown, "Pagin and the Mass Mark