Electrical computers and digital processing systems: multicomput – Network-to-computer interfacing
Reexamination Certificate
1998-09-24
2001-08-07
Geckil, Mehmet B. (Department: 2152)
Electrical computers and digital processing systems: multicomput
Network-to-computer interfacing
C709S236000
Reexamination Certificate
active
06272552
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates generally to media for interconnecting computer network devices, such as computers, hubs, and servers. More particularly, the apparatus herein utilizes two otherwise unused pairs of wires in an existing network interconnection cable, to provide a second communication channel, or link. Use of the second channel doubles the effective bandwidth of the communication link between the network components, thereby eliminating “bottlenecks” and delays in data transfer.
2. Description of Prior Art
Networks are typically used where many personal computers, printers, plotters, FAX machines, scanners, and file servers must function conjunctively in a work environment. The network allows the devices or components of the system to be interconnected, so that data can be accessed and transferred throughout the system. The network also permits users to share network devices, such as printers, FAX machines, and the like.
A computer network which encompasses a relatively small area is known as a local-area network, or LAN. Most LANS are restricted to a single building, or a group of adjacent buildings. However, any number of LANS can be interconnected to each other over any distance by using telephone lines or radio frequency transmissions. A group of LANS connected in the manner is called a wide-area network, or WAN.
Different types of LANS have been developed over the years, differing principally in topology, protocols, and media.
Topology refers to the geometric arrangement of devices in the network. For example, network components may be arranged in a bus or straight line, in a star pattern, or in a ring-like configuration.
Protocols pertain to the rules and encoding specifications for sending and receiving data throughout the network. Protocols also determine whether the network employs a peer-to-peer architecture, where equal responsibilities are assumed by each node, or a client/server architecture, where clients rely upon servers for network resources.
The media of a network is the method by which the devices are interconnected. Commonly used media include twisted-pair wire, coaxial cables, fiber optic cables, or radio frequency transmissions.
The Ethernet protocol, originally developed in 1976, remains one of the most common protocols still used today. An Ethernet network uses either a bus or star topology, and supports data transfer rates ranging from 10 to 1000 Mbps (Megabits/second). The Ethernet specification served as the basis for the IEEE (Institute of Electrical and Electronic Engineers) 802.3 standard. This standard specifies the physical and lower software layers required for data transfer in an Ethernet network. The Ethernet also uses the CSMA/CD (Carrier Sense Multiple Access/Collision Detection) access method, to handle overlapping and simultaneous device access to the network.
The 10Base T (10 Mbps) and 100Base-T (100 Mbps) sections of the IEEE 802.3 CSMA/CD standard describe the mechanical and electrical characteristics of interconnections for the network devices. These characteristics include a point-to-point connection between devices, using a cable having a predetermined maximum length. The cable consists of four wire pairs (“twisted pairs”), contained in a common sheathing. A standardized electrical connector (RJ-45) and a wiring assignment for that connector are also specified. For 10Base T and 100Base-T, the IEEE 802.3 standard specifies the use of only two of the four twisted pairs of wires. While the interconnection of the unused twisted pairs to the standardized connector is specified, their use is not.
SUMMARY OF THE INVENTION
The apparatus of the present invention employs the two unused twisted pairs in an existing Ethernet specified cable to provide a second channel, or link, for data communication in a 100Base-T circuit. This adaptation avoids the need to run additional cabling between devices to provide a second channel. By utilizing the apparatus of the present invention, the effective bandwidth of the communications link is doubled, with relatively few changes to the existing 100Base-T system.
A conventional Media Access Controller (MAC) is modified to include a second communication channel. The modified MAC transfers data packets into an internal transmit FIFO (First In, First Out) storage, having primary and secondary data streams. As words enter the FIFO, the MAC transfers words simultaneously from the two streams to respective primary and secondary MII (Media Independent Interface) circuits. The MII circuits are interconnected, in turn, to primary and secondary PHY (Physical Layer). In this manner, nibbles of data in the primary stream of the FIFO are transferred to the primary PHY, and nibbles in the secondary stream of the FIFO are transferred to the secondary PHY.
The first two pairs of wires in the Ethernet specified cable are connected to the Rx/Tx (Receive/Transmit) connections on the primary PHY. The second two pairs of wires in the cable are connected to the Rx/Tx connection on the secondary PHY.
These components and their respective interconnections are mirrored at the other network device to which the cable is connected.
The receiving MAC at the other network device is responsible for re-assembling the data packets received from the primary and secondary PHY links. FIFO channel skew compensation is provided within the receiving MAC, in the event the lengths in the total wire runs for the two channels are different. In addition, the receiving MAC waits for a predetermined amount of time after receipt of the last nibble to ensure that all data from both the primary and secondary nibbles have been fully received.
In this manner, the two channels act as one higher bandwidth channel. An error detection system in the MAC monitors both channels. If an error is detected, error packet processing is performed in the same manner as with a single channel design. The dual-channel link of the present invention is preferably run in a full-duplex mode, sending and receiving simultaneously. However, the dual channel operation can be applied to a half-duplex channel as well. It is also contemplated, although not required, that auto-negotiation, as defined in the IEEE 802.3u standard, would be performed on both the primary and secondary channels in the course of data transfer.
REFERENCES:
patent: 6061362 (2000-05-01), Muller et al.
patent: 6067585 (2000-05-01), Hoang
patent: 6085241 (2000-07-01), Otis
patent: 6094436 (2000-07-01), Runaldye et al.
patent: 6094439 (2000-07-01), Krishna et al.
Melvin Bruce W.
Singh Bharat K.
Geckil Mehmet B.
Hewlett--Packard Company
LandOfFree
Dual channel 100Base-T link for computer networks does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Dual channel 100Base-T link for computer networks, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Dual channel 100Base-T link for computer networks will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2440964