 |
Claims  |
|
|
What is claimed is:
1. A reliable, high performance data communication system comprising:
data source means for assembling data into data packets and transmitting
respective said data packets as they become full with a predetermined
amount of data, or if not full, at an occurrence of a reference interval
of time, each of said data packets having data corresponding to data
packet sequencing data and an information field; and
a plurality of receiver means for receiving said transmitted data packets,
each said receiver means having failure detection means for checking for
failures in the reliable receipt of said data packets by monitoring:
said data packet sequencing data; and
for receipt of a respective data packet within each expiration of a period
of time corresponding to said reference interval of time;
whereby both receipt of a data packet having data packet sequencing data
which does not match a next expected data packet sequencing data of a
predetermined sequence, and also a failure to receiver a respective data
packet within an expiration of a period of time corresponding to said
reference interval of time, are indicative of a failure in a reliable
receipt of said data packets.
2. A data communication system as claimed in claim 1, wherein said
reference interval of time corresponds to a lapse of a predetermined
amount of time which is measured from a transmission of a preceding data
packet.
3. A data communication system as claimed in claim 1, wherein each said
receiver means can issue a request for a data packet, said communication
system further comprising:
recovery means for supplying a requested data packet to respective receiver
means which issue a request for said requested data packet in response to
a determination that there was a failure in a reliable receipt of a data
packet.
4. A data communications system as claimed in claim 3, wherein said data
packet sequencing data corresponds to a data packet sequencing number.
5. A data communications system as claimed in claim 4, wherein said
respective receiver means identify a respective requested data packet by
using said data packet sequencing number.
6. A data communications system as claimed in claim 3, wherein said
recovery means comprises:
server memory means for receiving said data packets, and for storing copies
of a predetermined number of most current data packets which have been
transmitted; and
request means, capable of communication with said data source means, for
requesting retransmission of a particular data packet from said data
source meansn in an event said recovery means is unable to satisfy a
respective receiver's request for a particular data packet from said
server memory means, in an event that more than a predetermined number of
respective receivers issue a request for a same particular data packet.
7. A data communication system as claimed in claim 1, wherein each of said
data packets additionally has data corresponding to error checking data,
and wherein said receiver means performs a further check for the reliable
receipt of said data packets using said error checking data.
8. A data communication system as claimed in claim 1, wherein said receiver
means further includes:
buffer means for storing said data packets received from said data source
means.
9. A data communication system as claimed in claim 8, wherein said receiver
means further comprises:
data management means for selectively accessing data in said data packets
stored in said buffer means.
10. A data communication system as claimed in claim 9, wherein said
receiver means further comprises:
data processing means for processing and performing predetermined
operations using said selected data, and means to trigger said processing
and performing upon a detection of data of a predetermined type.
11. A data communication system as claimed in claim 1, wherein each of said
data packets has data corresponding to time stamp data.
12. A data communication system as claimed in claim 1, wherein said
information field of a respective data packet has a plurality of data
items, each of said data items being comprised of data corresponding to a
code field and a corresponding information field.
13. A data communication system as claimed in claim 1, wherein each said
data packet contains a receiver subclass code to allow a plurality of
independent packet streams.
14. A reliable, high performance data communication method comprising the
steps of:
assembling data into data packets, each of said data packets having data
corresponding to data packet sequencing data and an information field;
transmitting respective said data packets as they become full with a
predetermined amount of data, or if not full, at an occurrence of a
reference interval of time; and
receiving said transmitted data packets at a plurality of receiver means,
and at each of said receiver means, checking for failures in a reliable
receipt of said data packets by monitoring:
said data packet sequencing data; and
for receipt of a respective data packet within each expiration of a period
of time corresponding to said reference interval of time;
whereby both receipt of a data packet having data packet sequencing data
which does not match a next expected data packet sequencing data of a
predetermined sequence, and also a failure to receive a respective data
packet within an expiratoin of a period of time corresponding to said
reference interval of time, are indicative of a failure in a reliable
receipt of said data packets.
15. A data communication method as claimed in claim 14, wherein said
reference interval of time corresponds to a lapse of a predetermined
amount of time which is measured from a transmission of a preceding data
packet.
16. A data communication method as claimed in claim 14, wherein each said
receiver means can issue a request for a data packet, said communication
method further comprising the step of:
supplying a copy of a requested data packet to respective receiver means
which issue a request for said requested data packet in response to a
determination that there was a failure in a reliable receipt of a data
packet.
17. A data communications method as claimed in claim 16, wherein said data
packet sequencing data corresponds to a data packet sequencing number.
18. A data communications method as claimed in claim 17, wherein said
respective receivers identify a respective requested data packet by using
said data packet sequencing number.
19. A data communications system as claimed in claim 16, wherein said
supplying step is accomplished by:
receiving said data packets and storing, in a server memory means, copies
of a predetermined number of most current data packets which have been
transmitted;
if a copy is available in said server memory means, retrieving a copy of a
requested respective data packet from said server memory means; and
in either an event said recovery means is unable to satisfy a respective
receiver's request for a particular data packet from said server memory
means, or in an event that more than a predetermined number of respective
receivers issue a request for a same particular data packet, requesting
retransmission of a particular data packet from said data source means;
and
transmitting said particular data packet to said respective receiver means
which issued a request for said requested data packet.
20. A data communication method as claimed in claim 14, wherein each of
said data packets additionally has data corresponding to error checking
data, and wherein said checking step includes a further check for the
reliable receipt of said data packets using said error checking data.
21. A data communication method as claimed in claim 14, further including
the step of:
storing said data packets received at said respective receiver means in a
buffer means.
22. A data communication method as claimed in claim 21, further comprising
the step of:
selectively accessing, at each said respective receiver means, data in said
data packets stored in said buffer means.
23. A data communication method as claimed in claim 22, further comprising
the step of:
processing and performing predetermined operations, at each said respective
receiver means, using said selectively accessed data.
24. A data communication method as claimed in claim 14, wherein each of
said data packets has data corresponding to time stamp data.
25. A data communication method as claimed in claim 14, wherein said
information field of a respective data packet has a plurality of data
items, each of said data items being comprised of data corresponding to a
code field and a corresponding information field.
26. A relible, high performance data communication system comprising:
data source means for assembling data into data packets and transmitting
respective said data packets as they become full with a predetermined
amount of data, or if not full, at an occurrence of a predetermined
interval of time which is measured from the transmission of the preceding
data packet, each of said data packets having data corresponding to data
packet sequencing data and an information field;
a plurality of receiver means for receiving said transmitted data packets,
each said receiver means having failure detection means for checking for
failures in the reliable receipt of said data packets by monitoring;
said data packet sequencing data; and
for receipt of a respective data packet within each expiration of a period
of time corresponding to said predetermined interval of time;
whereby both receipt of a data packet having data packet sequencing data
which does not match a next expected data packet sequencing data of a
predetermined sequence, and also a failure to receive a respective data
packet within an expiration of a period of time corresponding to said
reference interval of time, are indicative of a failure in a reliable
receipt of said data packets, and each said receiver means can issue a
request for a data packet; and
recovery means for supplying a requested data packet to respective receiver
means which issue a request for said requested data packet in response to
determination that there was a failure in a reliable receipt of a data
packet.
27. A reliable, high performance data communication method comprising the
steps of:
assembling data into data packets, each of said data packets having data
corresponding to data packet sequencing data and an information field;
transmitting respective said data packets as they become full with a
predetermined amount of data, or if not full, at an occurrence of a
predetermined interval of time which is measured from a transmission of a
preceding data packet;
receiving said transmitted data packets at a plurality of receiving means,
and at each respective said receiver means, checking for failures in a
reliable receipt of said data packets by monitoring:
said data packet sequencing data; and
for receipt of a respective data packet within each expiration of a period
of time corresponding to said predetermined interval of time;
whereby both receipt of a data packet having data packet sequencing data
which does not match a next expected data packet sequencing data of a
predetermined sequence, and also a failure to receiver a respective data
packet within an expiration of a period of time corresponding to said
reference interval of time, both being indicative of a failure in a
reliable receipt of said data packets, and each said receiver means can
issue a request for a data packet; and
supplying a copy of a requested data packet to respective receiver means
which issue a request for said requested data packet in response to a
determination that there was a failure in a reliable receipt of a data
packet.
28. A reliable, high performance data communication system comprising:
a communications network interconnecting a data source means with a file
server means and a plurality of receiver means;
said data source means periodically broadcasting packets of information to
said file server means and said plurality of receiver means;
each of said packets including a packet sequence number, a time stamp, and
a plurality of information fields, each information field including a
unique code field and a corresponding data field;
a ring buffer in each respective one of receiver means in said plurality of
receiver means for storing a plurality of packets received over said
communications network;
data management means for performing a code selection function in each said
receiver means for selecting from said plurality of information fields in
each packet stored in said ring buffer, data fields corresponding to
selected ones of said unique codes;
an ordered data table means in each said receiver means for storing in an
ordered sequence said selected data resulting from performance of said
code selection function;
data processing means in each receiver means for utilizing said selected
data in said ordered data table means;
said file server means storing all of said packets received from said
communications network; and
negative acknowledgement means in each said receiver means for anticipating
a receipt of said periodic packets from said data source means and
detecting an omission of an anticipated receipt of a data packet, and in
response thereto, generating a request for retransmission to said file
server means, which performs retransmission through either a broadcast or
via a session type connection;
whereby a high performance, reliable data communication from said data
source means to said receiver means is established.
29. The data communication system of claim 28, wherein said file server
means performs a negative acknowledgement assessment function for
determining whether negative acknowledgements from a subplurality of said
plurality of receiver means is likely to indicate a failure of
broadcasting of a packet from said data source means to said receiver
means; and
wherein said file server means further comprises request means for
establishing a communications session between said file server means and
said data source means in response to an indication from said negative
acknowledgement assessment function, for accessing a packet of information
from said data source means corresponding to said failed broadcast.
30. The data communication system of claim 28, wherein said data management
means performs a change assessment function for accessing data stored in
said ordered data table means, and determining whether changes in a value
of said data, corresponding to a selected code, are greater than a
predetermined threshold value;
said ordered data table means being responsive to said change assessment
function for outputting new values of data identified as surpassing said
threshold; and
said data processing means being responsive to an output from said ordered
data table means to execute tasks corresponding to said changed data which
has surpassed said threshold value.
31. The data communication system of claim 28 wherein:
said periodic broadcast of said data packets by said data source means is
based upon a reference interval of time wherein a next data packet is
required to be broadcast at the end of said reference interval of time as
measured from a broadcast of a most recent data packet;
whereby said data source does not fail to broadcast a data packet less
frequently than once per said reference interval of time.
32. The data communication system of claim 28, wherein said communication
network comprises a token ring local area network.
33. The data communication system of claim 28, wherein said negative
acknowledgement means of respective receiver means issue said request for
retransmission in response to a detection of an expiration of a reference
interval of time as measured from receipt of a most recent packet.
34. The data communication system of claim 28, wherein said negative
acknowledgement means of respective said receiver means issue said request
for retransmission in response to a determination of a missing packet
sequence number.
35. The data communication system of claim 28, wherein said negative
acknowledgement means and said respective receiver means issue said
request for retransmission in response to an absence of an anticipated
periodic packet which has not been received from said data source means. |
|
 |
|
 |
Claims |
|
|
|
Description |
|
|
TECHNICAL FIELD
The present invention relates to a Multicast Data Distribution System
(MDDS) and method and, more particularly, to a data distribution system
and method for the timely, efficient, and reliable distribution of data to
an unlimited number of remote receivers.
BACKGROUND ART
Numerous data communication systems have been proposed for a wide variety
of different applications, for example:
U.S. Pat. No. 4,317,957, issued to Sendrow on Mar. 2, 1982, discloses a
system for protecting transactions and providing authentication of users
in an on-line system for automatic tellers. A Personal Identification
Number (PIN) which is concatenated with an anti-counterfeiting code, a
Personal Identification Number (PIN) and a time stamp are used to validate
automatic teller transactions.
U.S. Pat. Nos. 4,454,508 and 4,459,588, issued to Grow on June 12, 1984 and
July 10, 1984, respectively, disclose timed token ring protocols. Any
station can transmit information upon the receipt of the token if the time
since the previous receipt of a write token is less than the unused amount
of capacity left on the ring.
U.S. Pat. No. 4,569,042, issued to Larson on Feb. 4, 1986, discloses a data
packet switching network in which continuity packets bearing time stamps
are employed to determine call path transit time delay.
U.S. Pat. No. 3,587,044, issued to Jenkis on June 22, 1971, discloses a
system for the transfer of digital communication data through a
communication medium having a significant delay. The disclosed system has
means for interlocking the data transfers to assure accurate data
transmission and receipt.
U.S. Pat. No. 4,538,147, issued to Grow on Aug. 27, 1985, discloses a
method of allocating available bandwidth between a plurality of stations
which are configured in a write-controlled loop communication network.
U.S. Pat. No. 4,587,650, issued to Bell on May 6, 1987, discloses a method
of simultaneously transmitting isochronous and non-isochronous data on a
local area network through the use of a unique signal pattern.
U.S. Pat. No. 4,661,952, issued to von Sichart et al on Apr. 28, 1987,
discloses a method of transmitting data on a telecommunications exchange
via a ring line system which allows information items to be transmitted in
periodically recurring pulse frames.
U.S. Pat. No. 4,663,748, issued to Karbowiak et al on May 5, 1987,
discloses a communication system suitable for use as a local are network
(LAN). The Karbowiak et al system has a network comprised of a plurality
of nodes such that the system is capable of reconfiguring itself after
node failure to minimize system disruption.
U.S. Pat. No. 3,387,268, issued to Epstein on June 4, 1968, discloses
remotely locatable apparatus which receives a continuous stream of coded
alpha numeric data from a central station for the monitoring of stock
trading transactions.
U.S. Pat. No. 4,644,542, issued to Aghili et al on Feb. 17, 1987, discloses
a method for the reliable broadcasting of information in a distributed
network, such that, despite the presence of faulty processors in the
system, the fault-free processors of the system can still obtain
consistent views of the information which is available on the system.
U.S. Pat. No. 4,569,015, issued to Dolev et al on Feb. 4, 1986, discloses a
method for reliably achieving agreement between multiple processors, by
having each processor add a distinctive unforgeable signature to a message
contents as the message is passed on.
U.S. Pat. No. 4,418,384, issued to Holtey et al on Nov. 29, 1983, discloses
a communication subsystem which automatically aborts a sequence of bits
when the subsystem senses that it is not receiving data from a
microprocessor fast enough to maintain synchronous transmission over the
communication line.
U.S. Pat. No. 4,325,120, issued to Colley et al on Apr. 13, 1982, discloses
an elaborate data processing system which has processors which recognize
two basic types of objects (i.e., an object being defined as a
representation of related information maintained in contiguously-addressed
set of memory locations). The first type of object contains ordinary data,
while the second type of object contains information for locating and
defining the extent of access to object associated with that access
descriptor.
U.S. Pat. No. 4,354,225, issued to Friedler et al on Oct. 12, 1982
discloses a data processing system having a plurality of intercoupled
processors working in conjunction with an intelligent main memory system.
U.S. Pat. No. 4,276,594, issued to Morley on June 30, 1981, discloses an
elaborate digital computer and associated method steps.
In the world of data communications, there exists a need for generic data
communication system and method having the ability to simultaneously
satisfy a number of important requirements which are critical in many data
communication applications. More particularly, there often exists a need
for a data communication system which can provide for the regular and very
high performance delivery of updated data to an unlimited number of
receiver installations. In many of these environments, respective
receivers are not interested in all the data which is available but,
instead, are interested only in selected data needed to perform the
receiver's intended function. Furthermore, it is often highly critical
that the data delivery be reliable, and that the data communication system
have a facility which can provide a data guarantee. A number of different
approaches, which attempt to provide the above requirements, are discussed
below.
More particularly, there are several data communication approaches which
have been used to provide selected data to selected receivers.
In a first approach, the delivery of selected data to selected receivers is
performed by the data transmission device or by an intermediate device at
some intermediate point along the communication network. More
particularly, in implementing such an approach, a data source device or
communication node device would apply data management routines to the raw
body of data, and then selectively channel data to respective receivers
along the communications network. The selective channeling of the selected
data can be performed using either time-multiplexing by dividing the
available broadcast time of the communication network, or
space-multiplexing by dividing the communication network into respective
communication links.
Such an approach was found to be disadvantageous in terms of data delay,
because respective portions of the data delivery device (i.e., data source
or communication node device) operating time must be dedicated solely to
satisfying the need to select data for the respective receivers. Rather
than obtaining swift receipt of the data from the data collection point,
each respective receiver has to wait in turn while the delivery device
performs a number of data management and selective channeling routines. In
effect, it can be seen that each respective receiver suffers a cumulative
data delay penalty due to other receivers on the system. As the number of
respective receivers which a delivery device has to service increases,
data delay increases proportionally.
In a second approach, the selective delivery of data is provided by an
intermediate file server which serves as a data library by receiving data
transmissions from a data source and maintaining updated tables of data.
The respective receivers must make individual queries to the file server
as to the data of interest. The file server handles queries in the order
of receipt by checking the status of the selected data in question, and
forwarding the status to the requesting receiver.
This second approach has also been found to induce data delay for reasons
which are similar to the above. A respective portion of file server
operating time must be dedicated solely to the answering of each
individual request. Rather than obtaining swift receipt of the select
data, each respective receiver must wait in turn for the answer to its
request. As the number of queries handled by a file server increases, data
delay increases proportionally.
In analyzing the above two approaches, it can be seen that each of the
receivers suffers a data delay penalty due to the presence of other
devices on the data communication system. More particularly, it can be
seen that a tremendous amount of delay time suffered by each receiver is
caused by time used to selectively manage, channel, or answer the data
requests of the other receivers.
Turning now to providing a data guarantee requirement in a data
communication system, there are several approaches which can be used to
check whether all data has been received by a receiver.
One approach is to have each receiver, upon receipt of data from the data
source, send an acknowledgement back to the data source. The data source
monitors the acknowledgements from the respective receivers, and in the
event of a failure, retransmits the data to the respective receiver or
otherwise informs the receiver of the data failure. This approach is
disadvantageous in several respects.
First, important communication medium resources (i.e., transmission
bandwidth or transmission time) are being absorbed through the use of the
communication medium to send acknowledgements, and/or to transmit a
duplicate copy of the failed data or a data failure warning. Rather than
being able to devote a maximum amount of the communication medium
resources to support a continuous transmission of updated data, a
percentage must be dedicated to the acknowledgement mechanism. As the
number of respective receivers sending acknowledgements increases, more
and more of the communication medium is used, resulting in a proportional
increase in data delay.
A second and more important disadvantage using the acknowledgement scheme
is that each receiver performs only an echo function (i.e., echoing an
acknowledgement of receipt to the data source), and does not participate
in the determination of a failure in the data receipt. In the event of an
interruption which disturbs data delivery, a receive will continue to
operate on the assumption that it has received the most recent data, until
such a time when the data source informs it otherwise. If a failure in the
communication link has isolated the receiver from the data source, the
notification of a failure from the data source will never arrive.
Another approach which can be used to provide data guarantee is to perform
some sort of data comparison which can be performed at at least two
different locations, i.e., at the the point of data transmission or the
point of data receipt.
If the comparison is to be performed at the point of data transmission,
each receiver must echo the data back to the data source in a manner
similar to the acknowledgement routine. Although this method provides a
higher degree of data guarantee than echoing just an acknowledgement
(i.e., via a one-to-one data comparison), this approach is still deficient
in the same manner as the acknowledgement routine.
The arrangement of having a data comparison performed at the point of
receipt provides advantages over the above approach.
In utilizing this type of arrangement, the communications system must
incorporate some sort of routine whereby each receiver is able to obtain
two independent copies of the data which can be compared. This can be
accomplished by the simultaneous transmission of the data along parallel
communication paths or repeated transmissions of the data along a single
communication path. Although each of the receivers is now able to make a
determination of a data failure at the time it occurs thus participating
in the data guarantee routine, this approach also has disadvantages.
More specifically, the use of parallel paths to provide dual copies of the
data is disadvantageous, because a dual expenditure is incurred in the
development and maintenance of two communication paths. The use of
repeated transmissions to provide dual copies is disadvantageous because
transmission medium resources (i.e., transmission bandwidth or
transmission time) are absorbed by the retransmission process. As
described above, this prevents maximum use of transmission resources to
support a continuous transmission of updated data, thus resulting in data
delay. In addition to the above disadvantages, both arrangements have a
more important disadvantage, in that, if there is a long-term failure in
the single transmission path used with a retransmission approach, or if
there are simultaneous failures in the parallel transmission paths used in
a parallel transmission approach, the isolated receiver will not receive
either copy of the data and, hence, will not know that data has been
missed.
In many data communication systems, the data source means is heavily
involved in the data recovery process. In these types of systems, an
individual receiver with missing data initiates a direct communication
with the data source along the data communication network in order to
request a copy of any data which has been missed. In order for the data
source to provide the data guarantee, the data source then must retransmit
the data along the data communication network. Such an approach has a
major disadvantage.
More specifically, important communication medium resources (i.e.,
transmission bandwidth or transmission time) are being absorbed by the use
of the communications medium to provide the data guarantee. Rather than
being able to devote a maximum amount of the medium resources to support a
continuous transmission of updated data, a percentage of the resources
must be dedicated to the data guarantee. As the number of occurrences of
data recovery increases, more and more of the communications medium is
used, resulting in a proportional increase in data delay.
Because of the above disadvantages and shortcomings, there still exists a
need for a system and method which provide for the widespread, timely and
reliable distribution of data to an unlimited number of remote receiver
installations, while at the same time, providing for the delivery of data
with minimum delay, and for an arra | | |
