 |
Description  |
|
|
FIELD OF THE INVENTION
The present invention relates generally to communication systems and, in
particular, to the transmission of data packets.
BACKGROUND OF THE INVENTION
There exists today a growing need to expand the use of traditionally
office-based computer applications (e.g., word processing programs,
electronic mail, etc.) to remote locations such as the home or car. Data
files and/or messages generated by such applications, residing on hosts
having local-area network/wide-area network (LAN/WAN) connectivity, are
typically transmitted from one location to another using high-speed
protocols such as the so-called Transport Control Protocol/Internet
Protocol (TCP/IP). In order to extend the use of these applications to
remote locations not serviced by a LAN/WAN, it is necessary to establish
connections between LAN/WANs and other non-similar communication networks,
such as wireless communication systems.
A major difficulty in connecting LAN/WANs to wireless communication systems
is the large disparity in their available transmission bandwidths and
hence, their throughput capacities. It is not atypical for a wireless
communication system to have a transmission bandwidth 10 times less that
of a LAN/WAN. This disparity also contributes to the widely differing
protocols used in LAN/WANs and wireless communication systems. The lower
throughput capacities associated with wireless communication systems has
led to the use of circuit-switched techniques, whereas the higher
throughput capacities associated with LAN/WANs has led to the use of
packet-switched techniques.
FIG. 1 illustrates an example of a typical, prior art circuit-switched
communication (100). When initial source data (101) becomes available, a
channel set-up period (107) is required to establish a communication path
between the source and the destination. For example, in the wireless case,
the channel set-up period (107) may be the time required to request and
obtain usage of a particular radio frequency (RF) carrier. Regardless of
the channel type, the channel, once established, remains dedicated for the
exclusive use of the source and destination.
Having established the channel, the initial source data is transmitted
(109) to the destination. As additional source data (103, 105) becomes
available, it is immediately transmitted (113, 117) through the channel.
When necessary, usage of the channel is then discontinued during a channel
tear-down period (119). Advantages of circuit-switched techniques are the
low overhead requirement (i.e., the amount of throughput capacity required
for the transmission of information other than the source data), as well
as the low delay (i.e., the time difference between the availability of
source data and its actual transmission). The periods of channel
inactivity (111, 115) in between periods of data availability, however,
are a disadvantage of circuit-switched techniques. This is a direct result
of the dedicated use of the channel; other sources are unable to utilize
the channel. These advantages and disadvantages make the use of
circuit-switched techniques most efficient for longer communications, such
as file transfers or fax transmissions.
FIG. 2 illustrates an example of a typical, prior art packet-switched
communication (200). A key difference between packet-switched and
circuit-switched methods is that the channel, when used in a
packet-switched manner, is not dedicated and is available for use by
multiple sources and destinations. Once available, the initial source data
(101) is partitioned into data packets (203, 207, 211) for transmission.
The data packets (203, 207, 211) occupy available time-slots that include
capacity for overhead data (201, 205, 209). Due to the commonality of the
channel, the overhead data (201, 205, 209) typically comprises target
destination identification information so that data intended for a
particular destination may be properly routed.
As additional source data (103, 105) becomes available, it is again
formatted into data packets (217, 221, 227) and placed into available
time-slots with their associated overhead data (215, 219, 225). Advantages
of packet-switched methods are that set-up/tear-down periods are not
required. Also, multiple communications may be intermingled on the
channel. Assuming the use of channels having equivalent bandwidths,
packet-switched methods are less efficient relative to circuit-switched
methods due to the additional overhead, typically leading to larger
throughput delays. Delays are further lengthened when time-slot
availability is reduced due to heavy use of the channel. These
disadvantages can be overcome by increasing the packet-switched channel's
transmission bandwidth, if possible, to accommodate the larger overhead
and need for additional time-slots. Typically, packet-switched techniques
are most efficient in the transmission of shorter communications, such as
electronic mail messages or paging services.
In order to establish connectivity between LAN/WANs and wireless
communication systems, the incompatibilities of their respective
packet-switched and circuit-switched protocols need to be resolved. One
solution to this problem is to directly transmit the packet-switched data,
including the overhead data for each packet, over a circuit-switched
(i.e., wireless) channel. This is inadequate, however, because the
differences in throughput capacities would require an inordinate amount of
packet-switched information to be buffered prior to transmission over the
circuit-switched channel. Even if the circuit-switched channel has
sufficient bandwidth, this solution becomes inefficient due to channel
inactivity during periods of low packet volumes.
Another solution is to establish a circuit-switched transmission for each
data packet or group of data packets. While this solution might be
acceptable for low volumes of packets, it becomes severely inefficient for
increasing packet volumes due to the set-up and tear-down overhead
Therefore, a need exists for a method of transmitting packet-switched data
in an efficient manner over circuit-switched (particularly wireless)
communication channels.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 illustrates a circuit-switched communication in accordance with
prior art.
FIG. 2 illustrates a packet-switched communication in accordance with prior
art.
FIG. 3 illustrates a data network that includes a local/wide-area network
and a wireless communication system in accordance with the present
invention.
FIG. 4 illustrates a logic diagram which may be used by a data network when
transmitting packet-switched information in accordance with the present
invention.
FIG. 5 illustrates a logic diagram which may be used by a target
destination when receiving packet-switched information in accordance with
the present invention.
DESCRIPTION OF A PREFERRED EMBODIMENT
Generally, the present invention provides a method for transmitting
packet-switched information to a target destination via a circuit-switched
channel. This is accomplished by formatting the data to be transmitted
into a plurality of data packets, which data packets each include a first
target destination identifier. Based on the first target destination
identifier, a second target destination identifier, corresponding to the
target destination, is also determined. A circuit-switched channel is
established to the target destination using the second target destination
identifier. The data packets are then transmitted via the circuit-switched
channel to the target destination without either the first or second
target destination identifiers being transmitted therewith. Such a method
allows information normally conveyed in a packet-switched manner, as well
as information normally conveyed in a circuit-switched manner, to be
efficiently transmitted over circuit-switched channels.
The present invention can be more fully described with reference to FIGS.
3-5. FIG. 3 illustrates a data network (300) that includes an information
source (301), a local/wide area network (LAN/WAN) (303), a network
interfacer (305), and a radio frequency (RF) communication system (306) .
The information source (301) may comprise an electronic mail or word
processing application running on a personal computer (PC) or workstation.
The LAN/WAN (303) may comprise a public or private data network such as
ISDN (Integrated Services Digital Network), Frame Relay, or ATM
(Asynchronous Transfer Mode) network. The network interfacer (305) may
comprise a computer-based gateway between the LAN/WAN (303) and the RF
communication system (306), which may comprise a dispatch cellular
service, or a personal communication service (PCS) system.
The RF communication system (306) includes a controller (307), base
stations (313-315), and at least one target destination (323). The target
destination (323) comprises a communication unit (325), such as a Groupe
Speciale Mobile (GSM) digital cellular telephone, or a MIRS.TM. dispatch
radio, both by Motorola Inc., coupled to a data-capable terminal (327),
such as a PC-based facsimile unit. Typically, as the target destination
(323) travels within the RF communication system (306), communications can
be established between the communication unit (325) and the base stations
(313-315) via communication resources (319-320) . In a preferred
embodiment, the communication resources (319-320) comprise pairs of RF
carriers. In addition to allocating the communication resources (319-320)
when necessary, the controller (307) tracks the location of the
communication unit (325). This location information is stored in a
controller database (311) along with the operating characteristics of each
communication unit. The database (311) is typically consulted when a call
is received from or transmitted to the communication unit (325). The
controller (307) and the database (311) may be a Mobile Switching Center
(MSC) and a Home Location Register (HLR) by Northern Telecomm., Inc. When
a communication to/from the information source (301) from/to the target
destination (323) is desired, the network interfacer (305) provides
connectivity between the LAN/WAN (303) and the controller (307).
The network interfacer (305) provides a link between LAN/WAN (303) and the
RF communication system (306) thereby allowing communications to occur
between the information source (301) and the target destination (323). The
network interfacer may be a group of modems providing multiple access
paths to the public switched telephone network (PSTN), or it may be a
packet gateway/router that incorporates many different protocols to
communicate with public or private data networks.
FIG. 4 illustrates a logic diagram which may be used when transmitting
information between the information source (301) and the target
destination (323) in accordance with the present invention. The
information transfer illustrated in FIG. 4 is hereinafter described in the
context of an application using the TCP/IP protocol in a simplex manner
(i.e., in one direction only). However, it is important to note that the
process illustrated in FIG. 4 may also work for other protocols, such as
the AppleTalk Transaction protocol, operating in a duplex manner, i.e.,
from the information source (301) to the target destination (323), and
vice versa.
At step 401, it is assumed that information, intended for transmission to
the target destination (323), is provided by an application operating at
the information source (301). This information is in a digital form and
may comprise a short electronic mail message or a long document file. The
information source (301) formats the information into data packets (402).
Each data packet includes a header that comprises a first target
destination identifier, such as the IP address, for uniquely identifying
the data-capable terminal (327).
Having produced the data packets, a session is then initiated from the
information source (301) to the data-capable terminal (327) by further
formatting the data packets into a session (403). The information source
(301) opens a conventional TCP session having a session header, which
session header includes the relevant TCP and IP information. The session
header provides information on the type, characteristics and length of the
session. Also, a session conclusion indicator may be included in the
session. In sessions where a session conclusion indicator is not
available, the type and/or characteristics of the session (i.e., the TCP
port number) can be used to predict the session conclusion. It is
important to note that the boundaries of the session (i.e., the session
header and conclusion indicator) are used to control the set-up and
tear-down of the communication path. After session formatting, the packets
are forwarded to the LAN/WAN (303) which in turn utilizes the IP
information to route the packets to the network interfacer (305).
Upon reception of the data packets and session header, the network
interfacer (309) requests the controller (307) to establish a
communication path to the communication unit (325) associated with the
first target destination identifier and hence, the data-capable terminal
(327) . To this end, the controller (307) maps the first target
destination identifier into a second target destination identifier (404),
such as a telephone number of the communication unit (325), by consulting
the controller's database (311).
Prior to establishing communications with the communication unit (325), the
interfacer (309) determines the session format and type based on the
message type. For example, if the TCP port number indicates that a message
is of an electronic mail type, the interfacer (305) may determine that a
particular type of session format has been used. Information about the
session format is later used to detect conclusion of the session.
Using the second target destination identifier, i.e., a phone number, the
controller (307) establishes a circuit-switched channel (406) to the
communication unit (325). The circuit-switched channel is established by
the controller (307) using well-known techniques of assigning a
communication resource (319-320) to a base station (313-315) . Prior to
transmitting any information to the communication unit (325), the
controller (307) checks availability of the communication unit (325) and
presents this information to the network interfacer (305) . In a preferred
embodiment, the network interfacer (307) requests the WAN/LAN (303) to
buffer the packets when the communication unit (325) is unavailable.
Alternately, the network interfacer (305) may respond to the information
source (301) with a TCP "open failure" message when the communication unit
(325) is unavailable.
Assuming that the communication unit (325) is available, the network
interfacer (305) forwards any data packets having the first target
destination identification over the circuit-switched channel (407). Since
the target destination is now uniquely determined by the circuit-switched
channel, the first and second target destination identifiers are not
transmitted with the data packets, thereby improving throughput capacity
of the circuit-switched channel.
As is well known in the art, the RF communication system (306) continuously
evaluates the condition of the circuit-switched channel during
transmission of the data packets. As long as the circuit-switched channel
is in good condition (i.e., reliable), the data packets are transmitted
over the channel. When it is determined that the channel has degraded
(408), due to fading or interference for instance, the communication unit
(325) can use well-known protocols to re-establish the circuit-switched
channel under a site having good signal conditions.
When it is determined that the channel has degraded (408), due to fading or
interference for instance, the controller (307) buffers (409) any data
packets received from the LAN/WAN (303) intended for the target
destination (323). While the data packets are buffered (409), the
communication unit (325) uses well-known techniques to re-establish (410)
the circuit-switched channel under a site having better signal conditions.
After the communication unit (325) re-establishes the circuit-switched
channel (410), the controller (307) forwards any buffered data packets for
transmission and returns to transmitting any incoming data packets (407).
When it is determined that the channel has not degraded (408), the network
interfacer (309) determines if the session conclusion indicator has been
detected (411). The session conclusion indicator may be included as part
of the session or it may be inferred from the type and format of the
session. For example, in a file transfer under a TCP/IP session, the first
data packet of the session includes the maximum number of data packets to
be transferred during that session. The network interfacer (309) counts
the number of data packets received and, when the maximum number is
reached, indicates that the entire file has been transmitted and, in this
case, that the session conclusion indicator has been detected.
In another example, a UDP (User Datagram Protocol) trivial file transfer
simply transmits blocks of data packets with session packets indicating
the start and end of the block. The session format usually transmits
blocks of a fixed length (e.g., 600 bytes) . In this case, the network
interfacer (305) checks the length of every block received. If the length
is 600 bytes, then there is a higher probability that another block will
be coming in the near future. Conversely, if a block is less than 600
bytes, it is highly probable that this block is the last one in the
session and a session conclusion indicator is therefore detected by the
network interfacer (309).
Regardless of the method used, when a session conclusion indicator is not
detected (411), processing continues at step 407 as described previously.
When a session conclusion indicator is detected (411), the network
interfacer (305) requests the controller (307) to discontinue the
circuit-switched channel. In this manner, the channel does not remain in
use during periods of channel inactivity, and thus may be reassigned for
other purposes. This is an advantage over prior art solutions in which no
other usage of the channel was allowed due to the exclusive nature of the
channel assignment. As previously discussed, it is understood that the
procedure described above with reference to FIG. 4 may also be utilized to
transmit information in the opposite direction, i.e., from the
data-capable terminal (327) to a target destination affiliated with the
LAN/WAN (303).
FIG. 5 illustrates a logic diagram which may be used by the communication
unit (323) when receiving information in accordance with the present
invent ion. At step 501, the communication unit (323) receives
notification of an incoming circuit-switched call and it establishes the
RF channel. The communication unit (323) then receives the data packets
(502) and recombines the first target destination identifier, as well as
the TCP information, with the received data packets (503). The first
target destination identifier, although not transmitted, can be recombined
with the data packets since it is previously known by the communication
unit (325). The TCP information, on the other hand, is available because
it has been included with the initial session information. In this manner,
the circuit-switched transmission process is transparent to the
data-capable terminal (325) because data packets are presented to the
data-capable terminal (327) with the first target destination identifier
as they would be if received directly from the LAN/WAN (303) . The
communication unit (323) will continue delivering the data packets until
it detects that the circuit-switched channel has been discontinued (504)
by the network interfacer (305) and the controller (307).
The present invention provides a method for transmitting packet-switched
information to a target destination via a circuit-switched channel. With
such a method, the efficient transport of packet-switched data within a
circuit-switched RF communication system is realized because the need to
re-establish a circuit-switched channel for each data packet is
eliminated. Additionally, the inefficiencies inherent to directly
transmitting packet-switched data over a circuit-switched channel, such as
undue buffering and reduced throughput capability, are overcome because
the additional overhead associated with packet-switched data (as in the
aforementioned first and second target destination identifiers) is not
transmitted. The nature of a circuit-switched channel ensures that the
data will be reliably routed to the correct destination. By observing the
data session type and format, conclusion of a session may be reliably
detected allowing communication channels to be discontinued, thereby
avoiding channel inactivity. Also, by recombining the data packets with
the appropriate packet-switched overhead (i.e., TCP/IP information), usage
of the circuit-switched channel is transparent to the target destination.
* * * * *
|
|
 |
|
 |
Description |
|
