he
`
`- Appl. No. 09/407,371
`WORLD acerTOationoat ORGANIZATION , ' Doc. Ref. AJI
`.
`\
`PCT
`.
`ot i Pe
`INTERNATIONAL APPLICATION PUBLISHED UNDER HE PATENT COOPERATION TREATY (PCT)
`ee <
`
`
`
`(51) International Patent Classification 5 :
`
`
`
`
`(11) International Publication Number:
`‘WO 93/15572
`
`
`H04J 3/26
`(43) International Publication Date:
`
`5 August 1993.(05.08.93)
`
`
`
`
`(81) Designated States: AU, JP, European patent(AT, BE, CH,
`(21) International Application Number:
`-
`PCT/US93/00641
`
`
`
`
`DE, DK, ES, FR, GB, GR, IE, IT, LU, MC, NL, PT,
`
`(22) International Filing Date:
`25 January 1993 (25.01.93)
`
`
`SE).
`
`
`(30) Priority data:
`
`
`Published
`07/829, 146
`31 January 1992 (31.01.92)
`US
`
`With internationalsearch report.
`
`
`
` (71) Applicant: NELLCOR INCORPORATED (US/US);
`25495 Whitesell Street, Hayward, CA 94545 (US).
`
`
`(72) Inventor: FISCHER, Michael
`; 2910 Hunters Horn, San
`Antonio, TX 78230 (US).
`
`
`(74) Agent: HAUGHEY, Paul, C.: Townsend and Townsend
`Khourie and Crew, One Market Plaza, 20th Floor, Steu-
`art Tower, San Francisco, CA 94105 (US).
`on
`
`$8
`
`0
`
`2
`
` (54) Title: A SERIAL, LAYERED MEDICAL NETWORK
`
` (57) Abstract
`
`
`A network or telemetry system (10) which allows virtual
`~ ®
`_
`2 {nee ' _
`
`
`services at the application or presentation layer to communicate AORRCATOR|_|AAMIAGATOR|—~CRIVIER cs
`with other virtualServices without regard tothe physical inter-
`== Laefe a,
`
`
`
`
`connections. Each message, called a parcel, includes the infor-
`—i !
`mation to be transmitted along with a virtual address header.
`o 8 6 ao9 #973|
`|
`
`
`
`
`The parcel is provided to a gateway (12, 26), which inserts the [rwaemo}LPaA=200)(7oon] |e t a i
`
`E
`parcel without modification into a packet with address informa-
`HAS iy
`!
`é
`
`
`
`tion for the physical through session layers in the packet header. vdsi= 9 20 }
`
`The packet is then transmitted to another network node (62, 64,
`Ys
`A ee
`Vis
`Vas
`¥
`66, 68, 70, 72) which receives and delivers the unmodified parcel
`'
`!
`Fan)
`to the addressed destination virtual service. A numberofparcels
`Ceeel| WwAgfit
`
`\
`from the sameordifferent virtual services can be packed into a
`=
`a
`
`mya * Pim be ag ew ----4
`signal packet (712) for transmission from the gateway in cases
`
`
`amas” Cae}
`wherethese parcelsare all directed to virtual services at the same
`ff
`
`
`
`
`destination node. Oncea sessionis established, such as between orastaton| “2|ow k1
`
`
`
`
`a gateway and a workstation, virtual services at the gateway
`
`
`node and the workstation can communicate with each other
`
`
`without requiring a lot of header overhead for each trans-
`
`mission.
`
`
`
`+
`
`wt.
`
`

`

`FOR THE PURPOSES OF INFORMATION ONLY
`
`Codes useto identify States party to the PCTon the front pages of pamphlets publishing international
`applications under the PCT.
`
`Viet Nam
`
`France
`Gabon
`United Kingdom
`Guinca
`Greece
`Hungary
`Ireland
`aly
`Japan
`Democratic People’s Republic
`of Korea
`Republic of Korea
`Kazakhstan
`Licchtensicin
`Sei banka
`huacmbourg
`Monzuco
`Mauiigascar
`Mali
`Mongolia
`
`Austria
`Australia
`Barbados
`Belgium
`Burkina Faso
`Bulgaria
`Benin
`Brazil
`Canada
`Ceutral African Republic
`Congo
`Swiverland
`Cole Wivoire
`Cantervan
`Cacchoslovakis
`Czech Republic
`Germany
`Deamark
`Spain
`Finland
`
`Maopritania
`Malawi
`Netherlands
`Norway
`New Zealand
`Poland
`Partugal
`Ruaminia
`Russian Federation
`Sudan
`Sweden
`Slovak Republic
`Senegal
`Soviet Union
`Chad
`‘Togo
`Ukraine
`United States of America
`
`

`

`WO 93/15572
`
`PCT/US93/00641
`
` A SERIAL, LAYERED MEDICAL NETWORK
`
`8.
`
`BACKGROUND OF THE INVENTION
`
`The present invention relates to serial data communication networks,
`and in particular to networks for interconnecting medical instrumentation.
`
`In a typical computer network, computers are connected together over a
`communication medium. Each computer has its own, unique physical address which
`is used for identifying both the source and the destination of any transmission. Data
`and other informationis typically sent in packets, with each packet containing a data
`field and a headersetting forth the source and destination addresses, as well as other
`information. Different protocols exist for the header and for determining when a
`particular source can transmit.
`
`In a numberoffields, such as the medicalfield, it is desirable to be able
`
`to connect remote instruments to a central computer workstation. Typically, the
`instruments will gather data and have minimal processing power. The large bulk of
`data is typically transmitted from the instruments to the computer.
`In addition to the
`data, there may be alarm signals which need to be transmitted and immediately
`received.
`
`It would be desirable to have a system optimized for network
`communication from a number of medical or other instruments to a central computer.
`
`10
`
`15
`
`20
`
`SUMMARY OF THE INVENTION
`
`The present invention provides a network or telemetry system which
`allows virtual services at the application or presentation layer to communicate with
`othervirtual services without regard to the physical interconnections. Each message,
`called a parcel, includes the information to be transmitted along with a virtual address
`header. The parcel is provided to a gateway, which inserts the parcel without
`modification into a packet with address information for the physical through session
`layers in the packet header. The packet is then transmitted to another network node,
`which receives and delivers the unmodified parcel to the addressed destination virtual
`SEIvice.
`
`30
`
`Oy
`
`

`

`WO93/15572
`
`2
`
`PCT/US93/00641
`
`A numberof parcels from the same ordifferent virtual services can be
`packed into a single packet for transmission from the gateway in cases where these
`parcels are all directed to virtual services at the same destination node. Once a
`
`session is established, such as between a gateway and a workstation, virtualservices at
`the gateway node and the workstation can communicate with each other without
`
`Nn
`
`requiring a lot of header overhead for each transmission.
`Instead, the session need
`simply be identified. Each gatewaytypically has one session at a time, but a
`workstation can support up to 64 sessions simultaneously.
`
`For example, a gateway with a pulse oximeter attached mayestablish a
`session with a workstation. The pulse oximeter would provide virtual services for real
`time data streams for oxygen saturation values, ECG values and pulse values. A
`separate virtual service called trend service would periodically store real time data for
`subsequentretrieval. These would communicate with virtual services in the
`workstation over a single established session. The oxygen saturation and ECG may
`communicate with a display control virtual service at the workstation, while the pulse
`value service communicates with an annunciatorservice at the workstation, for
`instance. The workstation can simultaneously carry on other sessions with other pulse
`oximeters or other instruments. A single session maylast the duration of a patient’s
`Stay in a hospital room.
`Unlike the prior art, where separate sessions would typically be needed
`for transmissions between eachpair of virtual services, the present invention supports
`transmissions between multiple virtual services in a single session. This eliminates the
`need for each instrumentorvirtual service to have a large amount of computing
`powerto support its own session. By sharing a session, less overhead in the form of
`computing power to support communication to and from multiple virtual services is
`required; while still permitting the virtual services to be unaware of data handling
`during the communication process.
`The parcels can be of varying size and number. Each parcelincludes
`precedence information in the header whichindicates the relative delivery importance
`of the information contained in the parcel. For example, an alarm indication parcel
`would have a highest precedencelevel, while real-time data would have lower
`precedence (with further distinctions between types of data: general data would be
`higher precedence than detailed data which would be higher precedence than stored
`
`10
`
`15
`
`20
`
`30
`
`

`

`WO 93/15572
`
`3
`
`PCT/US93/00641
`
`data that could be resent if necessary). The gateway transmits the highest precedence
`level parcels first.
`If buffer space at the gateway is exhausted, parcels having the
`lowest precedence level are discarded.
`If only high precedence parcels are present,
`older parcels are overwritten with the newer parcels from the same source.
`Each virtual service sends parcels to the gateway with precedence
`information in the parcel header. The parcel header also identifies the length of the
`information field. The information field can contain data, a command requesting an
`action, or a reply to a request.
`If the informationfield contains data, a sequence
`numberis included indicating the order in which the parcel was generated.
`Each gatewayhas a table for indicating the location of virtual services
`local to that node and the internal addressing required to deliver parcels to those
`virtual services. This parce] routing is done transparent to the virtual serviceitself.
`The gatewayalso hasa buffer for temporarily storing parcels while they are waiting to
`be multiplexed into a packet. The packet header identifies the number of parcels
`included and the overall length of the packet information field containing the parcels.
`The packet header also contains source and destination handles identifying the
`physical source node and destination node, as well as the particular session (which is
`useful for nodes that support a plurality of sessions). A sequence numberis included
`to identify the packet for detection of packets delivered more than once by the
`physical network hardware.
`
`The present invention detects missing data at the application layer for
`each service. This is done with the assumption that there is a reliable physical/MAC
`layer underneath.
`In this context, "reliable" means that the physicaland MAC layers
`are (when communication is possible) incapable of indicating packet delivery without
`the packet having successfully reached the destination node (at the MAClayer).
`In
`order to do this, the MAClayer, by necessity, is capable of delivering the same packet _
`twice.
`
`For fuller understanding of the nature and advantagesof the invention,
`reference should be made to the ensuing detailed description takenin conjunction
`with the accompanying drawings.
`
`10
`
`15
`
`20
`
`30
`
`th
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`

`

`WO 93/15572
`
`,
`
`4
`
`PCT/US93/00641
`
`Fig. 1 is a block diagram of a system using a protocol according to the
`present invention;
`
`Fig. 2 is a block diagram of a peripheral network adaptor (PNA)ofFig.
`
`1;
`
`Fig. 4;
`
`Fig. 3 is a block diagram of a hub ofFig. 1;
`Fig. 4 is a block diagram of a workstation of Fig. 1;
`Fig. 5 is a block diagram of the intelligent network adapter (INA) of
`
`10
`
`15
`
`20
`
`30
`
`Fig. 6 is a diagram of the ISO layers used by the present invention;
`Fig. 7 is a diagram showing the fields of the parcels and packets of the
`present invention;
`
`Fig. 8 is a diagram of a transmitter state machine; and
`Fig. 9 is a diagram of a receiver state machine.
`
`DESCRIPTION OF THE PREFERRED EMBODIMENT
`Fig. 1 is a block diagram of one embodimentof a system using the
`protocol of the present invention. An Oxinet2 network 10 connects workstations 12
`and 26 to a large number of medical instruments. The network is physically
`configured, for wiring convenience and other reasons, as a numberoflines which are
`physically interconnected through hubs 14 and 16. Communication over network 10 is
`controlled by gateways connected to each node. Each of workstations 12 and 26
`includes an internalintelligent network adapter (INA) which acts as a gateway. Each
`of the other instruments connected to the network either contains its own internal
`gateway oris connected through a gatewaycalled a peripheral network adapter
`(PNA) 18. A numberof instruments are connected to the network through a radio
`frequency link that operates between radio-equipped gateways 19 and radio frequency
`(RF) bridges 20, 22 and 24 that serve as radio hubs and connect the RF network to
`the wired network.
`
`Among the instruments shown in Fig. 1 are pulse oximeters 28 and 30.
`Other pulse oximeters 32, 34 and 36 include a power base with wave form display and
`printer capabilities. Personal monitors 38, 40, 42 and 44 monitor heart rate,
`respiration, oxygen saturation and ECG waveforms. Monitors 38, 40 and 44 include a
`base unit 50, 52, 54, respectively, which provides a bedside waveform display
`
`

`

`WO 93/15572
`
`5
`
`PCT/US93/00641
`
`capability. A display 56 is shown separately hooked up to hub 14. Also shown are
`blood pressure monitors 57, 58 and 60. The PNA gateway devices 18 and RF PNA
`gateway devices 19 allow older, existing instruments out in the field to be adapted for
`communication over the network of the present invention.
`Each workstation has other elements coupled to it. Work station 12 has
`a pair of annnunciator displays 62 and 64, a pair of interactive displays 66 and 68, a
`printer 70, and is coupled to a hospital’s patient data management system 72.
`Workstation 26 is connected to a single interactive display 74 and a pair of
`annnunciator displays 76 and 78. Annunciator displays only present alert conditions.
`The interactive displays permit a user to request the displayofalerts, status, data,
`trends and waveforms.
`
`10
`
`15
`
`20
`
`Fig. 2 is a block diagram of a peripheral network adapter or gateway 18
`or 19 of Fig. 1. The connection to its associated instrumentis throughaserial port
`210. The connection to the network is through a network interface 212 for wired
`gateways 18 or a radio interface 214 for RF gateways 19.
`The PNA has a microcontroller 218 which is connected to push buttons
`220, status LEDs 222, and EEPROM memory 224 that is used to store configuration
`parameters. A separate flash EPROM memory 226is used to hold operating
`firmware. An SRAM memory 228 acts as the buffer for the PNA, storing parcels
`which are assembied into packets and holding trend data for up to 24hours for
`readout by the workstation upon request. The PNAalso includes a clock/calendar
`circuit 230, and a power supply 232 with a connection to an external AC adapter 236.
`A battery 234 is used to maintain the contents of the clock/calender and SRAM
`circuits for at least 24 hours without AC power.
`For a wired connection to the network, a separate ARCNET controller
`circuit 238 handles the network transmissions, implementing the ARCNET MAC
`layer. "ARCNET"is a registered trademark of Datapoint Corp. ARCNETis a
`widely ‘used LAN described in the ARCNET Designers Handbook, Datapoint
`Corporation, 2nd Ed., 1988, order no. 61610. ARCNETis also covered by a
`proposed ANSIStandard, draft rev. 1.6, 1-5-91, available from the ARCNET Trade
`Association, 3365 North Arlington Heights Road,Suite J, Arlington Heights, Il 60004.
`
`30
`
`

`

`WO 93/15572
`
`6
`
`PCT/US93/00641
`
`Microcontroller 218 controls the assembling of parcels into packets for
`
`transmission.
`
`If a radio interface is used, RF network interface 240, radio modem
`
`242, and RF antenna 244 transmit to a RF bridge as shown in Fig. 1. The RF bridge
`contains a microcontroller similar to microcontroller 218 and an ARCNETcontroller
`similar to controller 238. Also shown in the radio interface 214 is an infrared location
`
`interface 246. This is used to receive signals from an infrared transmitter in a
`
`*
`
`hospital room to indicate the location of the instrument.
`
`PNA microcontroller 218 acquires data from an instrument over EJA-
`
`232 link 210. The link is configured to be in a mode compatible with the link of the
`
`10
`
`instrument. A peripheral transaction server (PTS) in microcontroller 218 is
`
`programmedto receive data from the instrument. The data acquired from the
`
`instrumentis stored in a circular buffer in the microcontroller where the raw
`
`instrument data is assembled and routed to a buffer position in SRAM 228.
`EEPROM224is a configuration memory which holds session andinitialization
`
`-15
`
`parameters. The PNA microcontroller executes firmware stored in flash EPROM 226
`
`that includes functions that translate instrument-specific data formats into the uniform
`parce] formats used by the workstations. This firmware also maintains a record of
`
`actual sensor data each 10 seconds in a special area of SRAM 228 allocated for such
`
`trend storage, and permits these trends, extending back up to 24 hours, to be read out
`
`20
`
`by the workstation.
`
`Fig. 3 is a block diagram of hub 14 of Fig. 1. The network connections
`
`are madethrough the transceivers for ports 1-16 as shown on theright side of the
`figure. These ports connect to the workstations and the various instruments and
`
`PNAs as shown in Fig. 1. A logic block 310 interconnects the lines depending upon
`whois transmitting to whom. Thelogic is controlled by an ARCNET controller 312
`and a separate microcontroller 314. There is also provided a signal retiming (delay
`line) logic 316 and network address detection logic 318. Status LEDs 320 arealso
`
`30
`
`provided, along with a power supply 322. The hub listens for transmissions from the
`different ports. When a transmission is detected, the hub will receive on that port
`and re-transmit on the remaining ports.
`Fig.4 is a block diagram of a workstation 12 of Fig. 1. The connection
`to the ARCNET network 10 is through a intelligent network adapter (INA) 410. The
`workstation is operated under the control of a main, root processor 412 with main
`
`

`

`WO 93/15572
`
`7
`
`PCT/US93/00641
`
`memory 414. An internal bus 416 interconnects the elements under the controlof a
`system bus controller 418. A graphics adapter 420 provides data to a video display or
`a touch services board 422. An SCSI controller 424 provides an interface to disk
`memory 426. Different virtual services in workstation 12 communicate with virtual.
`counterparts in the instruments attached to the network using the protocolofthis
`invention. An example ofa virtual service is a program running on the workstation
`that performsa particular function.
`
`Fig. 5 is a block diagram of the INA 410 of Fig. 4. The connection to
`the network is through an ARCNETcontroller 510 and transceiver 512. The
`connection to the internal bus 416 of the workstation is through a memory space
`interface and relocation register 514 and an J/O space interface 516. The INA
`operates under the control of its own processor 518, and includes EPROM memory
`520, SRAM memory 522, and EEPROM memory 524. Also included is a watchdog
`timer 526, switches and LEDs 528 and an optional serial interface 530 for debug
`purposes.
`INA 410 acts as the gateway for workstation 12, assembingparcels into
`packets for transmission, and deassembling received packets.
`
`Transmissions primarily take place from the instruments to the
`workstation, and from the workstation to the instruments. An instrument sending
`data or other information transmits a parcel to its associated gateway. The gateway
`receives several parcels from various instruments or the same instrument and packs
`them into a packet in the orderof arrival.
`If more parcels arrive between packet
`transmission opportunities than will fit into a packet, precedence is used to cause the
`most important parcels to be packed into packets first. The packetis then
`transmitted to the remote workstation as discussed in detail below. The instrument
`need only provide the virtual service address of the application layer program in the
`workstation, with the intermediate levels through physical layer transmission being
`taken ‘care of by the gateway and the network.
`The gateway will examine the precedence information in the parcels it
`receives and send alarmsfrom its associated instruments first, in accordance with the
`precedence rules discussed below. The parcels awaiting transmission in the gateway’s
`buffer are overwritten in accordance with the precedencerules in the event of buffer
`
`10
`
`15
`
`20
`
`30
`
`

`

`WO 93/15572
`
`8
`
`PCT/US93/00641
`
`exhaustion. This prevents a real time data waveform from preventing an alarm from
`getting through, for mstance.
`1.
`Definitions
`
`The meanings of some terms used are defined below. Termsbeing
`
`defined are presented in bold, while the first references to terms whose definitions
`
`appear subsequentto their initial usage are presentedin italics.
`
`*
`
` Oxinet2 is the name of a network using the protocol of the present
`
`invention.
`
`The Oxinet2
`
`10
`
`*
`
`protocol is a definition of communication over Oxinet2 networks.
`
`An Oxinet2
`
`*
`
`network is a set of entities communicating among each other, via a
`
`serial interconnection medium, using packets conforming to the Oxinet2
`
`network protocol.
`
`15
`
`The elements that usethe Oxinet2 network communication facilities are
`
`*
`
`services, which are application layer entities that exchange parcels with
`
`other services.
`
`Each Oxinet2 network is comprised of one or more
`
`*
`
`segments, which are subsets of the network that communicate via a
`
`20
`
`single instance of a particular link layer facility. Examples of link layer
`
`facilities currently used for Oxinet2 segments are
`ARCNET (a registered trademark of Datapoint Corporation), used to
`provide 2.5Mbpsof raw transfer bandwidth over coaxial cable to
`locations where wired interconnection is available, and the
`Oxinet2 RF network, used to provide a spread-spectrum radio link to
`
`mobile devices and/or locations where wired interconnection is not
`
`available.
`
`*
`
`Each segment using the Oxinet2 RF network operates on a
`
`particular channel, which is a subset of the available portion of
`
`30
`
`the RF spectrum, using a particular frequency range and
`
`spreading sequence, to offer 200Kbps of raw data transfer
`
`bandwidth.
`
`Communication between Oxinet2 network segments takes place through
`
`

`

`WO 93/15572
`
`9
`
`PCT/US93/00641
`
`bridges, which are clusters that forward packets between network
`segments, performing any necessary buffering and conversion of frame
`formats. Bridges direct packets according to
`*
`routes, that associate network-specific addresses on each network
`segmentattached to the bridge with the unique station ID in the
`packet header.
`Onetype of bridge used for Oxinet2 networks is an RF bridge,
`which forwards packets between an ARCNET segment and an
`RFchannel that constitutes an Oxinet2 RF network segment.
`The session administrator (SA) is a service facility, typically
`implemented in software on an Oxiview workstation, that
`receives requests to initiate communication activities:
`determines whetherto create sessions in response to these
`requests,
`determines whether each new session is a reconnection of a pre-
`existing session, and
`provides the necessary information to permit bridges to
`determine the appropriate route for the communication activity.
`Communication between Oxinet2 networks takes place through
`gateways, whichare stations that forward parcels between the network
`and virtual services, performing any necessary multiplexing and
`demultiplexing functions.
`The basic addressable units on a Oxinet2 network are
`*
`Stations, each of which are uniquely-identified physical entities from
`among all manufactured equipment which maybe attached to, and
`exchange information over, an Oxinet2 network.
`One or more physically connected stations constitute a
`**
`cluster, which is a component, or a set of interconnected components,
`attached to an Oxinet2 network. Examples ofclusters include
`*
`Oxiview workstations (including their attached peripheral
`devices), each of which is a single station;
`PNA-200 peripheral network adapters (including their attached
`peripheral devices).
`
`* *
`
`x
`
`10
`
`45
`
`20
`
`30
`
`%
`
`

`

`WO 93/15572
`
`The term
`-
`
`10
`
`PCT/US93/00641
`
`local refers to facilities or services which are internal to the entity of
`
`reference (typically a cluster), whereas
`remoterefers to facilities, services, or entities which are external to the
`
`entity of reference.
`Satellite refers to PNA-200s in contrast to Oxiview workstations and RF
`
`Bridges. For example a
`
`*
`
`satellite RF station, is a station with a transceiver for the Oxinet2
`
`RF network. Thesatellite RF station is remote from the
`
`reference points of Oxiview workstations and RF bridges.
`The implementation of network facilities is described in terms of a sequence of seven,
`hierarchical layers. From top to bottom these are the:
`*
`
`application layer, which implements the user-accessible functionality of
`
`the workstation, instrument, or peripheral device;
`
`presentation layer, which provides an application programming .-
`
`interface (API) between application layer software and the lowerlayers
`implementing the Oxinet2 protocol;
`.
`session layer, which performs parcel multiplexing and demultiplexing
`functions on network stations;
`transport layer, which consists of transmitter and receiver control state
`
`machines for each half-session on network stations;
`network layer, which provides routing andsession establishment services
`
`on networkstations;
`
`link layer, which is generally subdivided into the
`
`*
`
`logical link contro] (LLC) layer (sometimes termed a “sub-layer"),
`which is functionally nul] on Oxinet2 networks, and
`
`media access control (MAC)layer (sometimes termed a "sub-
`layer"), which concerns arbitration, framing, flow control, and
`acknowledgementof transfers over a particular medium; and
`physical layer, which concerns details of the particular media
`
`.
`
`*
`
`(ARCNET, Oxinet2 RF network).
`Each instance of communication on an Oxinet2 network involves sending a
`
`packet, which consists of an
`
`10
`
`15
`
`30
`
`

`

`WO 93/15572
`
`11
`
`PCT/US93/00641
`
`Oxinet2 header, that holds network, transport, and session layer
`control information for the Oxinet2 network protocol; and an
`Oxinet2 information field, that carries the data stream being
`communicated between presentation layer entities by the Oxinet2
`network protocol.
`*
`The Oxinet2 information field of each packet may contain
`one or moreparcels.
`.
`A transmission opportunity is a point in time whena stationis
`permitted to transmit a packet over a network segment according to the
`MAClayer protocol for that network segment. For ARCNET,
`transmission opportunities occur at least once every 50ms, whereas for
`the Oxinet2 RF network, transmission opportunities can be separated by
`up to 1 second.
`Each instance of communication between virtual services on the Oxinet2 network
`involves sending a
`*
`parcel, which is the lowest level data structure exchanged between
`application layer services. Parcels consist of a
`*
`parcel header, that holds presentation layer control information
`pertaining to the parcel; and a
`parcel information field, that carries the data being transported
`between application layer entities by this parcel.
`*
`the precedence of a parcel refers to one of four levels of
`relative priority for use in
`*
`ordering the transmission of parcels overthe
`network, and
`
`the selective discarding of parcels awaiting
`multiplexing for network transmission (by gateways)
`if buffer space is unavailable.
`Subject to availability of space in the Oxinet2 packet
`information field, multiple parcels may be transferred in a
`single packet; however, single parcels are never split across
`packet boundaries.
`There are three logical types ofparcels:
`
`*
`
`10
`
`15
`
`20
`
`30
`
`

`

`WO 93/15572
`
`12
`
`PCT/US93/00641
`
`request parcels, that are specific demands for which a subsequent reply
`
`parcel is expected;
`
`reply parcels that are sent strictly in response to a predecessor request
`
`parcel; and
`
`data parcels, that transfer data and/or state information at periodic time
`
`intervals or upon the occurrence of particular, aperiodic events.
`For example, a request parcel is sent from a PNA-200 to the session
`administrator service on an Oxiview workstation in order to establish a
`
`session. When the session is established, a reply parcel is sent from the
`
`10
`
`session administrator to the satellite PNA-200.
`
`Within each station, parcels are transferred through the
`*
`
`parcel hole, which is a presentation layer abstraction that
`
`*
`*
`
`accepts parcels for transmission from the application layer, and
`provides received parcels and networkstatus to the application
`
`layer.
`For actual transmission over any particular MAClayer, each packet or parcelis
`a»
`
`encapsulated in a
`
`frame, that surrounds the packet or parcel with
`
`framing information in the form of the physical and link layer header
`
`and trailer fields required by the particular network being used.
`
`* *
`
`Communication of Oxinet2 packets takes place between
`+
`(session) partners, that are stations capable of serving as one end of a
`full-duplex communication activity known as. a
`
`session. Each session is a point-to-point, full-duplex virtual circuit
`
`established between a pair of Oxinet2 partners. Except in the case of an
`Oxiview workstation, there is never more than one session active at any
`
`cluster, independent of the numberofstations that constitute the cluster
`
`or are attached to the network.
`
`The action that establishes a session is a connect. A connect request
`
`(CONN_REQ)is sent from the active gateway of a PNA-200 to an
`Oxiview workstation, where session initialization activities take place
`before returning a connect reply (CONN_REP) to the requester.
`
`The term
`
`15
`
`20
`
`30
`
`

`

`10
`
`15
`
`20
`
`WO 93/15572
`
`B
`
`PCT/US93/00641
`
`half-session is sometimes used to refer to one of the two, simplex
`communication paths that comprise a session. Each half-session is
`controlled by a pair of
`state machines, one controlling the transmitter and one controlling the
`receiver.
`
`Communication within an Oxinet2 session takes the form of a
`*
`
`data stream, which is a sequence of parcels accepted from the
`presentation layer at the transmitting end and delivered to the
`presentation layer at the receiving end,
`*
`unmodified, other than possible discarding of entire parcels in
`cases of inadequate buffer space; and
`in the order received, other than possible delivery of higher-
`precedence parcels ahead of lower-precedence parcels.
`Packing of the parcels into packets for transmission, and the subsequent.
`unpacking upon reception, is transparent to the usersof the data .
`stream. The orderof parcels through the data stream is Strictly
`maintained in the order that they are supplied to the session layer by
`the presentation layer. There are no reserved codes nor reserved data
`sizes, other than an upper bound on the length of each parcel. The
`contents of Oxinet2 data streamsare totally arbitrary as viewed from the
`presentation layer.
`A goal of Oxinet2 network communication is to minimize the
`*
`observable delay, which is the time between a user’s perception of the
`occurrence of an event (such as an alarm condition) at a satellite
`monitor and the reporting of the occurrence of that same event to a
`care provider at the connected workstation.
`Each Oxinet2 service is identified by a
`_*
`service name, that is assigned globally.
`Communication between services is directed using a three-component address that
`consists of a
`
`destination tag (DTAG),
`
`recipient service ID (RID), and
`
`action code (ACT).
`
`* * *
`
`

`

`WO 93/15572
`
`641
`14
`The 8-bit tag identifies an instrument, workstation, or bridge. Both a
`*
`source tag (STAG) and a
`
`PCT/US93/00
`
`*
`
`*
`
`destination tag (DTAG)are included in each parcel.
`
`Tag values are assigned globally.
`
`The 8-bit service ID is a virtual label that uniquely identifies a particular
`service within the context of the module identified by the associated tag.
`Both an
`
`*
`
`*
`
`*
`
`originator service ID (OID) and a
`
`recipient service ID (RID) are included in each parcel.
`
`with the exception of a few globally-assigned service IDs, the
`
`service ID values are assigned dynamically by the session
`
`manager whenevera satellite cluster is initialized.
`
`The 16-bit action code (ACT) uniquely identifies the action to be
`performed by the recipient of the parcel, and is used, within the
`recipient module to direct the parcel to the task appropriate for the
`designated action.
`
`A partneris identified by a
`*
`
`32-bit handle, which is a globally unique identifier value.
`The handle is subdivided into a
`*
`
`24-bit station ID (STID), which uniquely identifies the station at which
`the partner exists; and an
`8-bit session number (SSN), assigned uniquely by the Oxinet2 network
`layer at the designated station.
`In the case of multi-station clusters, the station ID in the handles used
`
`10
`
`15
`
`20
`
`for session communication to and from that cluster identifies the station
`
`serving as the session managerfor the cluster.
`In termsof facilities provided by the Oxinet2 protocol, a workstation is
`a single station as far as any satellite station is concerned.
`Thestation ID consists of an
`
`8-bit station type (STP), which identifies the kind of station and a
`16-bit serial number (SRL), unique within each station type value.
`More than onestation type value may be assig