BIDIRECTIONAL DATA COMMUNICATION SYSTEM
BACKGROUND OF THE INVENTION '
Field of the Invention
The present invention relates to data communication networks and, more particularly, to interfaces conformed for connecting a data processing system to a bidirectional network in which the communication directions are separated by channel. Reference to Related Applications
This application is a continuation of U.S. Patent Application Serial No. 08/011,685 filed on February 1, 1993 entitled "Bidirectional Data Communication System" . Description of the Prior Art
Communication between data processing terminals, central data bank storage facilities, and central processing stations is a matter of increasing concern as various personal computers and work stations are multiplied. As the density of these competing stations increases, the burden imposed on the communication network increases exponentially. Thus, the bandwidth of the communication network, the interfaces with the devices tied to it, and the communications protocols of each communicating device all need fundamental attention in order to avoid logical mazes and logical indeterminants.
Characteristically, each data processing device either transmits data onto a network, or receives data from it. Between these two communication states an active data processing device may effect internal operations of only a few or a large number of switching cycles and thereafter may require, again, the further use of a network. The third alternative, where the data processing device is idle or passive, is insignificant to the communication burden imposed.
Since the network itself is normally incapable of any data processing, a complete network architecture requires that for each transmission there be at least one receiver, and that for each receiver there be at least one transmitter. In consequence, each of the devices actively on the network must resolve the status of the network data that is to be received, and also the status of the network for transmission. One further condition, a condition not logically determined at the processing device, is the geometric arrangement between any two devices communicating on the network.
As the number of those using data processing devices and terminals increased the communication burden on any common network also increased. Furthermore, the proliferation of personal computers was then followed by an increase in the facility of their use. Thus, it is not just the number of stations on a network but also the task complexity and the consequent data transfer demands from each terminal that have increased.
In response to these communication demands various network arrangements have been devised which attend to the channel width of communication, or bandwidth. One example of the bandwidth expansion is the current use of fiber optics. While the much higher range of communicating channels has greatly simplified the physical network complexity, each network nonetheless must resolve the competing demands of the devices on it. This competition for the network, however, is at the processing rate of the slowest device and it is this competition process that now provides the most significant burden on shared communication.
In the past, techniques have been devised which resolve competition between devices for the same communication line. One example of such a technique is set out in U.S. Patent No. 4,063,220 issued to Metcalfe et al. While suitable for the purposes intended, the foregoing teachings attend to the inherent channel competition by devices operating on similar cycle rates. This is resolved by random restart intervals in each competing device which, of necessity, involves multiples of the processing rate.
Accordingly, the network competition entails logical processing intervals, a burden that substantially affects communication difficulties.
Inherent in any communication between processing stations is the transmit-receive pairing summarized above, as is also the direction of the invoked communication path. It has been found that a network arrangement that satisfies both the pairing and the direction constraints uniquely also resolves communication efficiency, and it is this arrangement and the logical interconnection thereof that are described herein.
SUMMARY OF THE INVENTION
Accordingly, it is the general purpose and object of the present invention to provide a paired network arrangement which includes logical interconnections at each communicating device.
Other objects of the invention are to provide a paired interconnection at the input/output port of a processing device for selection and connection to a paired communication network.
1 Yet further objects of the invention are to provide a 2 paired communication network in which the receive and transmit 3 functions are logically selected. 4 Briefly, these and other objects are accomplished δ within the present invention by providing a communication network 6 comprising a pair of separated channels, one channel being 7 connected for transmitting data in a first direction while the 8 second is aligned for oppositely directed transmission. The 9 foregoing channels may be variously implemented as a hard wire or
10 fiber optic filament, or by way of frequency or pulse code 11 separation of a carrier medium. 12 The paired channel network thus implemented is then 13 interconnected at each station communicating therewith in a 14 polarized connection; the first channel being connected across a lδ receiver-transmitter aligned in a first direction, while the 16 second is similarly connected but in a direction opposite to the 17 first. Thus, at each station interface there are at least two 18 receivers and two transmitters, connected in a receiver- 19 transmitter connection when the device is passive, and logically 20 switched to combine the receivers and the transmitters together 21 when the station acquires the network. 22 It is to be noted that while the station is inactive, 23 the receiver-transmitter pairs at the station interface simply 24 relay the communicated data on the network channels, in 2δ accordance with their connection polarity. Once these 26 connections are switched, however, the network is effectively 27 disrupted for all other communication excepting that issued by 8 the station effecting the switching. Thus, the switching logic 9 at each network acquisition functions in an unambiguous logical 0 arrangement. This same logical arrangement can then also include 1 network data collision functions like those earlier devised. 2
The switched interconnection, however, occurs locally and therefore reduces the incidence of collisions caused by extended propagation delays on a large network. The combination of the paired network and the local relays at each station, moreover, fix the polarity of communication both at the interface and in the network itself.
The foregoing functions may be conveniently implemented in a logical arrangement at each communicating interface by collecting in an OR connection the outputs of both of the network channel receivers, and at an AND-OR logical connection the inputs to the transmitter pair. More precisely, the receivers of each station are connected to an OR gate which then feeds the input terminals of the input/output port of the data processing system tied to the network.
The network transmitters are, in turn, connected to a corresponding output of another set of OR gates which collect at their inputs the outputs of two AND gates on each network channel. One of the AND gates on each channel then receives the data output of the local communicating device, in combination with an enable signal enabling transmission into both of the network channels. Concurrently, the enable signal is inverted to the other AND gate, thus suppressing direct network transmission across the interface. In this manner the interface either ties the local station to the network in both directions, to the exclusion of others, or effects a receive-transmit connection directly thereacross. This unambiguous switching then provides a very predictable interface function throughout the whole communicating system.
Of course, other equivalent logical arrangements can be utilized to achieve the same result. The foregoing summary therefore sets out an example only of the general function herein.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a diagrammatic illustration of a communication network arrangement according to the present invention; and
FIG. 2 is a logic diagram of a typical network interface according to the present invention.
DESCRIPTION OF THE PREFERRED EMBODIMENT
In order to clearly set out the substance of the present invention, the following understanding of terms is adopted, wherever applicable: by the term "network" we refer to a communication system that is logically passive, and that is conformed to convey digital data including both the address of the terminal to which it is directed and the message that follows the address; by the term "medium" we refer to any electromagnetic, optical, sonic or ultrasonic physical mechanism for conveying information and, although set out herein in the form of a fiberoptic mechanism, it is intended that other well known techniques be included in the scope of the teachings herein;
by the term "station" we refer to a logically active device including a data processing system, a data storage system, or any other combination of processing and storage that is tied to receive or transmit data onto the network; by the term "data" we refer to any logical item of information, including information needed to control the operation of any station or device; by the term "interface" we refer to a set of logical interconnections between the station and the channels of the network; by the term "relay" we refer to a set of electronic devices at each interface channel that receive data on one side and retransmit that same data into the network channel on the other side.
System Description:
As shown in FIG. 1, the inventive communication system, generally designated by the numeral 10, comprises a first network segment 11 including one network channel 12, polarized in a first direction, and a second channel 14 of opposite polarity. Other network segments, shown by example as network segments 21-1 and 21-2 aligned in crossing orientation with the first direction, each comprises a first and second channel 22 and 24, polarized in opposite directions. Channels 12, 14, 22, and 24 are each isolated from the others, and may take the form of a fiber optic filament in the example herein. Thus, a set of interconnections is required between channels 12 and 22, and channels 14 and 24. These interconnections may take the form of a communication hub, shown generally at 30.
More precisely, hub 30 effects the logical function of an OR gate 32 collecting at its input the signal all the channels 22, and other like channels, and driving by its output channel 14, and an OR gate 34 collecting at its input channels 12 and 22 and driving channel 24. In each instance OR gates 32 and 34 collect at its input those channels of corresponding polarity excepting the channel of its own segment.
It should be noted that the above crossed network arrangement is intended to serve as an example of a whole range of network combinations, the teachings herein further disclosing serial connection to illustrate all the conjunctive, AND, and disjunctive, OR, Boolean states for complete logical implementation.
Within each network segment the networks may be tied to local stations, in each instance across an interface shown as an interface 50. These interfaces are repeated at each instance, and like numbered parts therefore are intended to function in similar manner. By reference to the network segment 11, and by equal example elsewhere, interface 50 includes a first receiver 52 and a transmitter 62. Similarly, channel 14 is tied to a receiver 54 and a transmitter 64, which with the other terminating devices form a relay arrangement in the segment 11. While termination devices may be variously implemented, in the network described herein devices like those sold under the model designation R-2521 and sold by Hewlett-Packard have been found suitable to effect the optical conversion of receivers 52 and 54 and devices sold under the model designation T-1521 and sold by Hewlett-Packard have been found suitable to effect the transmitter function of transmitters 62 and 64. Thus, each interface 50 includes a paired set of receivers and transmitters,
which may be geometrically arranged at the channel ends to polarize the direction of signal propagation within the respective channel.
Accordingly, channels 12 and 22 are polarized in a direction opposite to channels 14 and 24. In this manner, various open-ended network circuit combinations may be effected, each of the interfaces thereon determining the signal direction and, by the description following, signal termination.
To effect the latter, the outputs of receivers 52 and 54, converted to the signals of the logic elements in the interface, are collected together within a logic stage 110. The transmitters 62 and 64 are similarly fed from the logic stage 110. Generally the logical connections within the logic stage effect a direct relay from receiver 52 to transmitter 62, and from receiver 54 to transmitter 62, when the using station 150 is idle. Once the using station turns active a set of logic conditions ties transmitters 62 and 64 in a logical connection and also receivers 52 and 54.
Interface Description:
As shown in more detail in FIG. 2, the foregoing logical switching may be effected by collecting the outputs of receivers 52 and 54 at the input of an OR gate 111. Concurrently, receiver 52 also drives one input of an AND gate 112 and receiver 54 drives an AND gate 114. Gate 112 is then paired with another AND gate 122 at the input to an OR gate 132 connected to the transmitter 62. Thus, OR gate 132 will include in its signal output the signal from receiver 52, and passed by gate 112. The signal output of OR gate 132 is then connected to the transmitter 62, completing the relay function earlier summarized.
The other gating input to AND gate 112 is an inverted signal from the port 151 of the using station 150, indicating the communication state thereof. This signal, in accordance with conventional practice, indicates that transmission is enabled from the using device, shown as signal TXEN. Accordingly, when inverted at gate 112 the relay of the signal path across the receiver 52 to transmitter 62 is disabled. Consequently, once the station 150 switches to a transmitting state, as indicated by the state of the port signal TXEN, the network continuity is disrupted at the local interface. Signal TXEN is concurrently connected to the other AND gate 122 which also receives the transmitted data signal TX from the port 151, carrying the locally generated data transmission. Gates 112, 122, and 132, therefore, switch the network channel 12 from a passive relay state to a state wherein the local station transmits its own data transmitter 62.
A similar logical switching is effected on channel 14, between receiver 54 and transmitter 64. More precisely, AND gate 114 collects at its input the inverted signal TX and TXEN, with the outputs of gates 114 and 124 collected at an OR gate 134 and then fed to the transmitter 64. Thus, like the relay function on channel 12, gates 114, 124, and 134 switch the network from a passive relay function to a state where only the locally generated data is impressed. It should be noted that the foregoing switchover occurs solely as result of the local state of signals TX and TXEN.
Of course, while the local device is passive it is necessary to detect any data on the network that may be directed to it. This is effected by collecting the signals on both channels 12 and 14, locally sensed at receivers 52 and 54, and collected at the OR gate 111. This combined received data is
It should be noted that the same logical architecture may be carried out in various electrical conventions, resulting ir a simple and reliable interface by which variously implemented data processing systems are tied to a common network.
Obviously many modifications and changes may be made to the foregoing description without departing- from the spirit of the invention. It is therefore intended that the scope of the invention be determined solely of the claims appended hereto.