Multiplex communications – Channel assignment techniques – Carrier sense multiple access
Reexamination Certificate
1997-03-03
2001-01-23
Geckil, Mehmet B. (Department: 2756)
Multiplex communications
Channel assignment techniques
Carrier sense multiple access
C370S446000, C370S447000, C370S448000, C709S225000, C709S233000, C709S234000, C709S235000
Reexamination Certificate
active
06178177
ABSTRACT:
The invention relates to a data-processing network having non-deterministic access, i.e. a network including a plurality of stations communicating with one another by accessing a common medium, using a non-deterministic protocol, i.e. a protocol that does not make it possible to predict the waiting time that will be necessary when a given station requests access to the medium at a given instant. This waiting time is referred to as the “access time” below. For example, a protocol that consists in checking that there is no carrier present on the medium, and in then starting to transmit a message while detecting any collision with another message that happens to start transmitting at the same instant, is a non-deterministic protocol. A station is incapable of predicting transmission of messages by other stations. Access time can be long if traffic on the common medium is heavy. Access time then slows down real-time applications prohibitively.
The most commonly-used common medium is a bus interconnecting all of the stations, or else a “passive hub”, i.e. a star interconnector to which each station is connected via an individual link, so that the links are equivalent to a bus going from station to station. To reduce the risk of collision, a first known solution consists in replacing the passive hub with an active hub, i.e. with routing apparatus, to which each station is connected via an individual link, and which enables the number of collisions to be reduced. It is necessary to use a cascade of hubs if the network includes a large number of stations (e.g. sixteen). That solution is therefore very costly, especially if the network is duplicated to make it more reliable.
Conventionally, each station is connected to the common medium via an interface circuit which manages the access protocol for accessing the medium. To reduce the risk of collision, a second known solution consists in modifying the implementation of the access protocol, in the interface circuit, so as to reduce the number of collisions. That solution thus consists in developing a proprietary circuit instead of using a circuit that complies with a standard and is commercially available. That solution is therefore costly.
SUMMARY OF THE INVENTION
An object of the invention is to provide a network in which access time is made deterministic by eliminating almost all risk of collision, by using means that are cheaper than those of the known solutions.
The invention provides a data-processing network having non-deterministic access but having deterministic access time, the network including a plurality of stations communicating with one another by accessing a common medium using a non-deterministic protocol, said network being characterized in that each station includes:
means for detecting whether there is a risk of collision between a plurality of stations trying to access the common medium; and
means for limiting its mean bit rate over the common medium to a predetermined maximum value at least during the periods in which there is a risk of collision between a plurality of stations trying to access the common medium, said maximum value being chosen such that the sum of the respective maximum values chosen for all of the stations is less than the mean bit rate that can be accommodated by the common medium.
A network characterized in this way encounters almost no collisions because the sum of the respective maximum values chosen for all of the stations is less than the mean bit rate that can be accommodated by the common medium.
In a preferred embodiment, the means for limiting the mean bit rate of a station comprise time-delay means authorizing a station to accesss the common medium after expiration of a delay starting at the end of the most recent access by the station, and having a duration which is proportional to the duration of said most recent access.
A network characterized in this way is less costly to implement because the time-delay means may be constituted by a portion of the software of the processor of the station. A commercially available interface circuit may be used with no modification.
In a preferred embodiment, the means for detecting whether there is a risk of collision between a plurality of stations trying to access the common medium comprise means for determining whether at least one collision has taken place since the beginning of the most recent transmission performed by the station over the common medium, and for concluding that there is a risk of collision if at least one collision has occurred since the end of the most recent transmission performed by said station.
A network characterized in this way is cheap to implement because the means for detecting whether there is a risk of collision may be implemented by using a collision counter that already exists in certain commercially available interface circuits.
The invention can be better understood and other characteristics appear from the following description and from the accompanying figures, in which:
FIG. 1
is a timing diagram showing how an embodiment of a network of the invention operates;
FIG. 2
is a block diagram summarizing this embodiment; and
FIG. 3
is a flow chart of the operations performed in this embodiment.
REFERENCES:
patent: 4271507 (1981-06-01), Gable et al.
patent: 4847835 (1989-07-01), Le Lann et al.
patent: 5047916 (1991-09-01), Kondo
patent: 5056107 (1991-10-01), Johnson et al.
patent: 5369639 (1994-11-01), Kamerman et al.
patent: 5371494 (1994-12-01), Singh et al.
patent: 5381413 (1995-01-01), Tobagi et al.
patent: 5553072 (1996-09-01), Daggett et al.
patent: 5577069 (1996-11-01), Lau et al.
patent: 5592483 (1997-01-01), Hieda et al.
patent: 5600651 (1997-02-01), Molle
patent: 5604869 (1997-02-01), Mincher et al.
patent: 5625825 (1997-04-01), Rostoker et al.
patent: 5642360 (1997-06-01), Trainin
patent: 5661727 (1997-08-01), Kermani et al.
patent: 5706274 (1998-01-01), Angelico et al.
patent: 5734639 (1998-03-01), Bustamante et al.
patent: 5737330 (1998-04-01), Fulthorp et al.
patent: 5774658 (1998-06-01), Kalkunte et al.
patent: 5784559 (1998-07-01), Frazier et al.
patent: 5802041 (1998-09-01), Waclawsky et al.
patent: 5838688 (1998-11-01), Kadambi et al.
patent: 5850525 (1998-12-01), Kalkunte et al.
patent: 5852723 (1998-12-01), Kalkunte et al.
patent: 5854790 (1998-12-01), Scott et al.
patent: 5896561 (1999-04-01), Schrader et al.
patent: 5940399 (1999-08-01), Weizman
patent: 5957854 (1999-09-01), Besson et al.
patent: 5963560 (1999-10-01), Kalkunte
patent: 5968154 (1999-10-01), Cho
patent: 6044396 (2000-03-01), Adams
patent: 6072798 (2000-06-01), Beasley
patent: 0 377 361 A1 (1990-07-01), None
patent: 0 632 619 A3 (1995-01-01), None
patent: 0632 619 (1995-01-01), None
V. Y. Zagurski et al, “Random-Access Method with Deterministic Delay and Control Capability”,Automatic Control and Computer Sciences, vol. 23, No. 3, Jan. 1, 1989, pp. 59-62.
Alcatel
Geckil Mehmet B.
Sughrue Mion Zinn Macpeak & Seas, PLLC
Vaughn, Jr. William C.
LandOfFree
Data-processing network having non-deterministic access, but... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Data-processing network having non-deterministic access, but..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Data-processing network having non-deterministic access, but... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2454453