Applied Micro Circuits Corporation

Nortel Networks Corporation

Matthew D Brown

Colin Geoffrey Kelly

Dimitrios Giannakopoulos

Alan Glen Solheim

Ross Caird

Guy Jacque Fortier

Tracy Ma

Joleen Hind

Sheldon James Hood

Chung Y Wu

Matthew Ornes

Stan Harry Blakey

Marika Herod

Yanming Gao

Chris Knapton

Michel Pigeon

Michael Hellmer

Peter Jeremy Anslow

Jay Quirk

Marc A Nadeau

Jim Lew

James Moser

Jean Guy Chauvin

Sean Campeau

Peter T H Kwa

Phil Campbell

Tony Bitzanis

Jean Guy G Chauvin

Steve D Bainton

An Ethernet physical layer (PHY) module is provided with a method for transceiving between a 10GBASE-R client interface and a 100G attachment interface. On each of ten client interface logical lanes a 10GBASE-R signal is accepted. Each 10GBASE-R logical lane is demultiplexed into two 5 gigabit per second (Gbps) pseudo 100GBASE-R logical lanes, creating a total of twenty pseudo 100GBASE-R logical lanes. The pseudo 100GBASE-R logical lanes are arranged into n groups of 20/n pseudo 100GBASE-R logical lanes. Further, the pseudo 100GBASE-R logical lanes from each group are arranged into a 100G attachment logical lane. Finally, a 100G attachment logical lane is transmitted at an attachment interface on each of n physical lanes. In the reverse direction, each of n physical lanes accepts a 100G attachment logical lane at the attachment interface, and a de-aggregation process supplies a 10GBASE-R signal on each of ten client interface logical lanes.

System and method for the aggregation of 10GBASE-R signals into pseudo 100GBASE-R signals

A system and method are provided for framing messages in a data streams encoded with redundant information for transmission and recovering the messages at a receiver. The transmission method accepts an energy waveform representing N words at a first bit rate, encoded with redundant information, where each word includes P number of bits. The N words are transformed, creating N transcoded words, where each transcoded word includes Q number of bits, and where Q&#x3c;P. The N transcoded words are mapped into M lanes in a buffer memory, where M&#x3e;1 and each lane receives a frame of N/M transcoded words. A frame alignment marker is generated and mapped into each frame. Each frame is represented as an energy waveform that is transmitted on a corresponding physical lane at the first bit rate divided by M.

Inverse multiplexing using transcoding and frame alignment markers

A system and method are provided for ordering tap setting modifications in a link partner using a plurality of voltage gain taps, while avoiding minimum and maximum limitations. Provided is a link partner (LP) transmitter having a parallel selectable voltage gain taps. The method sends messages from a network-connected local device (LD) directing the LP to generally change the gain setting of either the pre-tap or the post-tap, as follows. The gain setting of selected tap is changed in the desired direction of modification and then the center tap gain settings is maximally increased until a limit signal is received. If a selected tap increment is desired, the center tap is initially decremented more steps than the desired change in the selected tap.

System and method for transmitter training

A system and method are provided for transporting a serial stream via a lower speed network using multiple parallel paths. At a transmitter, an optical or electromagnetic waveform is accepted representing a serial stream of digital information, and unbundled into n virtual information streams. Each virtual information stream is divided into a sequence of segments. Each segment is encapsulated, creating a sequence of packets by adding a start indicator to the beginning of each segment, and a terminate indicator to the end of each segment. Each packet is disinterleaved across m lanes and reinterleaved into n branches of framed data. Optical or electromagnetic waveforms representing the framed data are transmitted via n network branches. A receiver is also provided, which essentially reverses the above-described transmission method.

Ethernet to multilane optical transport network traffic interface

A system and method are provided for framing messages in a forward error correction (FEC) structure for data streams encoded with redundant signal conditioning information. The method accepts signal conditioning-encoded words at a first bit rate, and eliminates redundant information in the signal conditioning-encoded words, creating N reduced-bit words of k bits. The k-bit words are mapped into a payload field of N*(k/p) p-bit words. Overhead (OH) and FEC parity fields are generated, and a frame is created including the OH field, payload field, and FEC parity field. The bit values in the frame are then pseudorandomly scrambled and the scrambled frame is transmitted at the first bit rate. A system and method are also presented for recovering the signal conditioning-encoded words from an FEC frame.

Transparent in-band forward error correction for signal conditioning-encoded signals

An electronic signal level detection system and method are provided. The method receives an analog input signal having a variable voltage and compares the input signal voltage to a threshold. A detection signal is generated for input signal voltages exceeding the threshold in a periodic first time frame. In a second periodic time frame (following the first time frame), a count is updated in response to the generated detection signals. The count is used to create a metric representative of the difference between the input signal voltage and the threshold. The count is incremented in response to the generating a detection signal ('1') in the first time frame, and decremented in response to not generating a detection signal ('0') in the first time frame.

Signal level detection method

Transport of multilane traffic over a multilane physical interface

A system and method are provided for negotiating a link data rate in a communication system using a plurality of data rates. In a system including a first device network-connected to a second device, auto-negotiation (AN) messages are mutually transmitted. The AN messages indicate rate information such as preferred data rate capabilities, if the device has a dual-rate capability, or a single data rate capabilities. If the AN messages are mutually transmitted, a negotiated link data rate is established. However, if one of the devices cannot send AN messages, the other device times-out, and a link data rate is established at the data rate transmitted by the device that is not AN-capable.

System and method for link rate negotiation

A system and method are provided for deinterleaving differential inverse multiplexed (DIM) virtual channels in a 40G Ethernet receiver. The method accepts a 10.3125 gigabits per second (Gbps) (10G) Ethernet virtual channel with 64B/86B blocks, including periodic Lane Alignment Marker (LAM) blocks. The 10G virtual channel is deinterleaved into two 5.15625 Gbps (5G) virtual channels by: 1) deinterleaving consecutive blocks from the 10G virtual channel into the 5G virtual channels in an alternating order, and 2) reversing the order of deinterleaving in response to each detected LAM block. Then, the method supplies the 5G virtual channels (i.e. to a MAC module).

Differential inverse multiplexing virtual channels in 40G ethernet applications

System and method for inverse multiplexing using transcoding and frame alignment markers

A system and method are provided for negotiating a link data rate in a communication system using a plurality of data rates. In a system including a first device network-connected to a second device, auto-negotiation (AN) messages are mutually transmitted. The AN messages indicate rate information such as preferred data rate capabilities, if the device has a dual-rate capability, single data rate capabilities, or is capable of communicating over a plurality of physical medium lanes. If the AN messages are mutually transmitted, a negotiated link data rate is established. However, if one of the devices cannot send AN messages, the other device times-out, and a link data rate is established at the data rate transmitted by the device that is not AN-capable.

System and method for multilane link rate negotiation

System and method for signal level detection

A protocol independent multiplexer is described that allows for multiple different protocols that operate at different bit rates to be combined and output in a format that may have yet another bit rate. The multiplexer includes a series of inputting devices that are each coupled to a respective buffering device, a mapping device coupled to each of the buffering devices, and an outputting device coupled to the mapping device. Each of the inputting devices receive an input optical signal and forwards recovered data information to the corresponding buffering device. The buffering devices store the data information and output to the mapping device, the outputting being controlled by the mapping device to ensure that the buffering devices remain approximately half full. The mapping device formats the data information into individual data units and outputs the data units to the outputting device which subsequently multiplexes the data units. One key advantage of this protocol independent multiplexer is that only one piece of hardware is required for the operation of numerous different protocols.

Protocol independent sub-rate device

A client signal received at an ingress interface is adapted to a higher-rate transport signal. Clock frequency acceleration is achieved by an M/N-multiplying PLL where the magnitude of M and N can be restricted without causing the rate of the resulting transport signal rate to deviate unacceptably from a nominal transport signal rate. Each frame of the transport signal has a payload section with a fixed number of transport payload bytes, each of which is either a dummy byte or a client byte. The number of client bytes per transport frame is within one byte of the number of client bytes actually received at the ingress interface during the duration of the frame. The designation of each frame as a low-fill frame or a high-fill frame is automatically regulated by checking the fill level of a memory element and is redundantly encoded by the ingress interface and transmitted to an egress interface as part of the frame.

Processing of arbitrarily formatted client signals to achieve compatibility with existing transport networks

The present invention involves the passing of clock information from a point where a tributary payload is mapped into a carrier to a point where a tributary payload is de-mapped from the carrier. A synchronization signal is generated at the mapping end. Rather than sending clock phase information implicitly by means of the data, as is normally the case, a value of the synchronization signal at a given time, is sent explicitly, independent from the payload data.

Growing community of inventors

Kanata, Canada

Matthew D Brown