JP2007502054A - Method and system for crosstalk cancellation - Google Patents

Abstract

The signal radio wave of one communication channel may generate crosstalk interference in another communication channel. The crosstalk canceling device can process a signal that causes crosstalk interference and generate a crosstalk canceling signal that corrects the crosstalk when applied to a channel receiving the crosstalk interference. The crosstalk cancellation device includes a model of a crosstalk effect that produces a signal that emulates the actual crosstalk in form and timing. The crosstalk cancellation device includes a controller that monitors the crosstalk corrected communication signal and adjusts the model to enhance crosstalk cancellation performance. The crosstalk canceling device has a self-defining or adjusting mode for transmitting a defined test signal to the crosstalk generating channel and the crosstalk receiving channel.
[Selection] Figure 4

Description

  The present invention relates to the field of communication, and in particular, to improve signal fidelity of communication systems by compensating for crosstalk interference that occurs between two or more communication channels that are common at high data communication rates. About.

(Cross-reference to related applications)
This application claims the priority benefit of US Provisional Patent Application No. 60 / 494,072, filed Aug. 7, 2003, entitled “Method for Crosstalk Cancellation in High Speed Communication Systems”. The contents of US Provisional Patent Application No. 60 / 494,072 are incorporated herein by reference.

  This application is entitled “Multilevel Signal Decoding Method and System” and is a non-provisional US patent application Ser. No. 10 / 108,598 filed May 28, 2002, and “Adaptive Noise Filtering”. And equivalent for optical high-speed multi-level decoding ”and is related to patent application no. 10 / 620,477, filed July 15, 2003, which is not a US provisional application. The contents of US Patent Application No. 10 / 108,598 and US Patent Application No. 10 / 620,477 are hereby incorporated by reference.

  The increasing consumption of communication services is driving the need for increased data environment capacity or bandwidth in communication systems. A phenomenon known as crosstalk often occurs in these communication systems and can undesirably reduce high speed signal transmission and thus communication bandwidth limitations to low levels.

  Crosstalk is a condition that occurs in communication systems where the signal of one communication channel is transmitted through the other channel and destroyed by interference (or bleed-over) from different signals. Interference can occur due to various effects. For example, in electrical systems such as circuit boards, electrical connectors and twisted pair cable bundles, each electrical path is provided as a channel. These copper foils act like antennas while receiving and receiving electromagnetic energy at high communication speeds. Radiated energy from one channel (referred to herein as an “attack channel”) is undesirably combined with another channel (referred to herein as a “sacrificial channel”). This undesirable transfer of received signal energy, known as “crosstalk”, can compromise data integrity in the receive channel. Crosstalk is typically both in that a single channel can emit energy for one or more other channels and can receive energy from one or more other channels. Is tropic.

  Crosstalk has emerged as a significant barrier to increasing the throughput of communication systems. Even when not specifically mentioned, crosstalk often appears as noise. In particular, crosstalk degrades the signal quality by increasing the uncertainty of the value of the received signal and making reliable communication more difficult. That is, the probability that a data error will occur increases. In other words, crosstalk generally becomes more problematic with increased data rates. Crosstalk not only reduces signal integrity, but also the amount of crosstalk often increases the bandwidth of the attacking signal. This makes high data rate communication more difficult. This is the specific case of electrical systems using binary or multi-plane signal transmission. This is because of the following. Copper foil lines through which this type of signal radiates radiate at the high frequencies associated with the level transitions of these signals and receive energy more efficiently. In other words, each binary or multi-plane communication signal is composed of high-frequency signal components that are more susceptible to crosstalk degradation than low-frequency components.

  Crosstalk impairments for increasing data processing rates are a sacrificial signal to terrible attenuation over long signal transmission path lengths (eg, circuit traces several inches long for multi-gigabit / second data rates) Increased by the trend of high-frequency content. That is, the high frequency component of the communication signal not only receives a relatively high level of crosstalk interference, but is also susceptible to interference. Because they are often weak due to transmission loss.

  While these reduced high-frequency components can be amplified through a technique known as channel equalization, this type of channel equalization often increases noise and crosstalk as a byproduct of amplifying high-frequency signals carrying data. To do. The amount of crosstalk present in the communication link often limits the level of equalization that can be used to restore signal integrity. For example, when multi-gigabits / second and data rates are required for next-generation backplane systems, the level of crosstalk energy on the communication channel is the sacrificial signal energy at the high frequencies that underlie this type of high-speed communication. Can exceed the level. In this condition, heterogeneous and wandering signal energy can dominate the energy of the desired data communication signal, making many of these conventional system architectures communicate at these data rates impractical.

  The term “noise” as used herein relates to a completely random phenomenon, unlike crosstalk. In contrast, crosstalk is a deterministic but often unknown parameter. The prior art includes the recognition that the system can theoretically be modified to mitigate crosstalk. In particular, (i) data transmitted via an interfering or attacking channel, and (ii) and crosstalk if there is a signal conversion feature that connects to the attacking channel and occurs in the victim channel Is theoretically determined and canceled. That is, if the data carried by the communication signal that is input to the communication channel is known and the signal conversion imposed on the communication signal by crosstalk is also known, those skilled in the art will know that the crosstalk signal degradation is Understand that it may be canceled. However, achieving this level of signal conversion definition with sufficient accuracy and accuracy to support the de facto implementation of a system that properly cancels crosstalk is difficult with the prior art. Thus, conventional techniques related to crosstalk are generally insufficient for high speed (eg, multi-gigabit / second) communication systems. Thus, in order to increase the data processing rate, techniques for canceling crosstalk are needed to improve the accuracy of the victim signal and to eliminate the barriers often caused by crosstalk.

  It is well understood that physics is raised to crosstalk (eg, electromagnetic coupling of electrical systems or four-wave mixing of optical systems), while this understanding alone is a direct measure for crosstalk transfer functions. Does not provide a simple model. One common reason that traditional modeling is difficult is that the relative geometry of the sacrificial and invading signal paths severely affects the transfer function of the crosstalk effect, and these paths are very engulfed. there is a possibility. In other words, the complexity of the signal path generally verifies the effect on model crosstalk using conventional modeling methods based on analyzing the signal conduit. Furthermore, it is generally undesirable to design a crosstalk canceller for a predetermined specific crosstalk response. This is because (i) the system may have many different responses to different sacrificial-attack pairs (requiring each unique design). And (ii) different systems may require different sets of designs. Thus, (i) a response to the diversity of crosstalk transfer functions that may arise from normal operation of a given system, and (ii) a complex to characterize and apply each sacrificial-attack pair There is a need in the art for a crosstalk cancellation system and method with sufficient flexibility for self-adjustment to avoid manual work.

  While the general concept of crosstalk cancellation is well known in the prior art, conventional crosstalk cancellation is generally not applicable to high speed environments such as channels that support multi-giga band rates. Conventional crosstalk cancellation is typically performed in a fully digital environment where the accessible attack data signal and the received victim signal are digitized and the microprocessor performs a cancellation process. The analog to digital converters and microprocessors required to perform this digital crosstalk cancellation in high speed environments are usually overly complex, resulting in unacceptable power consumption and manufacturing costs.

  What is needed to address these typical deficiencies in the prior art is the ability for crosstalk cancellation that is compatible with high speed environments with low power consumption and reasonable production costs. The capability is also needed to automatically adjust or set the crosstalk cancellation device. Such capability facilitates higher data rates and improves the bandwidth of various communication applications.

  The present invention assists in correcting signal interference, such as crosstalk, that occurs between two or more communication channels. Correcting crosstalk can improve signal quality and enhance the communication bandwidth or information carrying capability.

  A communication signal transmitted on one communication channel can combine unwanted signals to other communication channels, such as crosstalk, and can interfere with communication signals transmitted on that channel. . In addition to occurring between two channels, this crosstalk effect may be more than one channel between combining and between two and three or more multipoint communication channels In, crosstalk may be imposed and coupled to each channel that receives crosstalk from more than one channel. The channel may be a medium such as an electrical conductor or optical fiber that provides a signal path. A single optical fiber or wire may provide a transmission medium for more than one channel, each communicating digital or analog information. Alternatively, each channel may have a dedicated transmission medium. For example, the circuit board may have multiple conductors in the form of circuit traces where each trace provides a dedicated communication channel.

  In another aspect of the present invention, the crosstalk cancellation apparatus may input a crosstalk cancellation signal to a channel receiving crosstalk interference in order to cancel or correct the received crosstalk. The crosstalk cancellation signal may be derived or created from a signal that is transmitted to other channels and generating crosstalk. The crosstalk cancellation device may combine between a channel that generates crosstalk and a channel that receives crosstalk. In this configuration, the crosstalk cancellation device may sample or receive some signal that is causing crosstalk, and may be applied to channels that are receiving unwanted crosstalk. A crosstalk cancellation signal may be configured. In other words, the crosstalk cancellation device taps on the channel causing the crosstalk, generates a crosstalk cancellation signal, and receives crosstalk interference for crosstalk cancellation or correction. A crosstalk cancellation signal may be applied to the channel.

  In another aspect of the present invention, the crosstalk cancellation device may generate a crosstalk cancellation signal based on a model of a crosstalk effect. In the form of a signal that estimates, approximates, emulates or resembles a crosstalk signal, the model may generate a crosstalk cancellation signal. The crosstalk cancellation signal may have a waveform or shape that matches the actual crosstalk signal. Settings or adjustments that adjust a model, such as a set that models parameters, may define the characteristics of this waveform.

  The crosstalk cancellation signal may be synchronized with the actual crosstalk signal. That is, the timing of the crosstalk cancellation signal may be adjusted to match the actual timing of the crosstalk signal. A timing delay or other timing parameter may define a relative timing or temporal correspondence between the crosstalk cancellation signal and the actual crosstalk signal.

  In another aspect of the present invention, the crosstalk cancellation apparatus implements modeling and timing adjustments in order that the crosstalk cancellation signal closely matches the actual crosstalk, thereby providing effective crosstalk cancellation. The controller of the crosstalk cancellation device may monitor and analyze the output of the crosstalk cancellation device. That is, the controller may process the crosstalk canceled signal, which is an improved communication signal resulting from applying the crosstalk cancellation signal to a channel having crosstalk interference. The controller may change the modeling parameters and timing delay individually or in tune to minimize any remaining crosstalk remaining after crosstalk cancellation. The state and change of the crosstalk effect may be corrected by adjusting the operation of the crosstalk cancellation device.

  In another aspect of the invention, the crosstalk cancellation device may undergo a correction or preparation procedure initiated internally or externally. The crosstalk cancellation device or other device performing the calibration procedure may begin transmitting a known or predetermined test signal on the communication channel. The test signal may be transmitted to a channel that causes crosstalk or a channel that receives crosstalk interference. Further, different test signals may be transmitted to a channel that receives the generated crosstalk interference, and one test signal may be transmitted to a channel that generates crosstalk. For example, the crosstalk receiving channel may have the same voltage or current signal that symbolizes essentially no data transmission, and the randomized communication signal may be transmitted to the crosstalk generating channel. . The crosstalk cancellation device may utilize these known states to define the timing and shape of the crosstalk cancellation signal that effectively corrects for crosstalk interference. In other words, based on operating the crosstalk cancellation device to send a test signal to the crosstalk generation and crosstalk receiving communication channel, the crosstalk cancellation device defines or refines a model of the crosstalk effect. May be.

  The discussion of modified crosstalk presented in this summary is for illustration purposes only. Various aspects of the invention may be more clearly understood and may be understood from a review following the detailed description of the disclosed embodiments and with reference to the drawings and claims.

  The present invention supports crosstalk cancellation on one or more communication paths of a communication system such as a high-speed digital data communication system. A flexible and adaptable model of the crosstalk effect may output a cancel signal that accurately displays crosstalk interference. Coupling this cancellation signal onto a signal path that has crosstalk can cancel such crosstalk, thereby negating the obstacles that crosstalk can impose on bandwidth.

  Returning now to the discussion of the respective figures shown in FIGS. 1-12B, like numerals indicate like elements throughout the figures, and exemplary embodiments of the invention are described in detail.

  Returning now to FIG. 1, this figure shows a functional block diagram of a communication system 100 having two line cards 101a, 101b communicating through backplane signal paths 120, 130 issuing crosstalk 150, 151. Illustrate. More specifically, FIG. 1 illustrates the appearance of backplane crosstalk 150 and connector crosstalk 151 in the typical case of backplane communication system 100.

  Line cards 101a, 101b are modules, which are circuit boards that typically slide into chassis slots to provide communication capabilities associated with communication channels. The backplane 103 transmits signals between each installed line card 101a, 101b and another communication device such as another line card 101a, 101b or a data processing component in a rack-mounted digital communication system. A set of signal paths (eg, circuit traces) behind the chassis to be

  In FIG. 1, each line card 101a, 101b of the illustrated system 100 transmits and receives a plurality of channels of data, such as the two illustrated channels 120,130. Exemplary channels 130 are: (i) start with transmitter (Tx) 104a on line card 101a, (ii) send line card 101a to backplane 103 via connector 102a, (iii) backplane 103 Across the other connector 102b and line card 101b, and (iv) receive by receiver (Rx) 105b. FIG. 1 shows that such channels are “victim” or “vict.” (Sacrifice transmitter 104a to sacrificial receiver 105b) and “aggressor” or “agg.” (Aggression transmitter 104b to invasion receiver 105a. 2) called two channels.

  When the signal paths 120, 130 are close to each other, signal energy is emitted from the aggression channel 120 and taken into the sacrificial channel 130. That is, in the region of the backplane 103 and the connectors 102a and 102b where the first signal path is located near the second signal path, a part of the signal energy transmitted in the first signal path is the second In the second signal path and may break or damage the signal being transmitted. This crosstalk coupling 150 can occur, for example, on line cards 101a, 101b, on connectors 102a, 102b, on backplane 103, or any combination thereof.

  Although not illustrated in FIG. 1, crosstalk can also occur in the reverse direction. Specifically, the “sacrificial” channel 130 often radiates energy that breaks the “invasion” channel 120. That is, crosstalk often occurs in a bi-directional manner and transfers not only from the first signal path to the second signal path but also from the second signal path to the first signal path. In a system with three or more signal paths coexisting near each other's close (not illustrated), crosstalk may also be transferred between three or more signal paths. That is, one signal can not only impose crosstalk on two or more other signals, but can also receive crosstalk interference from two or more other signals.

  Similar to the multiple physical path case illustrated in FIG. 1 and described above, crosstalk may occur in an invading and sacrificial channel transmitting on a single transmission medium (eg, a single cable or trace). is there. In this scenario, each channel may match a specific signal band (eg, frequency band of frequency division multiplexing system, spectral band described in optical wavelength multiplexing system or time window of time division multiplexing system). ). In other words, two communication channels, one generated crosstalk, and one received crosstalk coexist in a communication medium (eg, optical waveguide or wire) with each communication channel supporting transmission of a dedicated communication signal. there's a possibility that.

  For clarity of explanation, an exemplary embodiment of the present invention based on crosstalk occurring between two channels (each on an independent physical path) is illustrated in FIG. Detailed description is provided herein. In another exemplary embodiment of the present invention, the method and system cancels crosstalk occurring between channels that coexist on a single communication medium. Those skilled in the art will have in this application having two or more channels issuing crosstalk on a single communication medium appended with the detailed description, flowcharts, plots and functional block diagrams included herein. It should be possible to make and use the invention.

  Returning now to FIG. 2, this figure illustrates a functional block diagram 200 of the crosstalk model 210 of the system 100 illustrated in FIG. More specifically, FIG. 2 illustrates a model 210 of the crosstalk effect 151 of the connector 102b based on a single exemplary transfer function 210.

  The invasion transmitter 104 b outputs an invasion communication signal u (t) 215 on the invasion channel 120. The energy from the invading communication signal u (t) 215 is connected to the sacrificial channel 130 by the crosstalk 151 of the connector 102b. The aggression communication signal u (t) 215 is composed of a frequency spread. Since crosstalk 151 is a frequency dependent phenomenon, the frequency of the invasive communication signal u (t) 215 connects to the sacrificial channel with varying efficiency. The frequency model H (f) 210 of the crosstalk effect 151 expresses which extent is coupled to the sacrificial channel 130, each of these frequency components in the form of a signal n (t). This crosstalk signal n (t) 230 is combined with a pure communication signal x (t) extending from the sacrificial transmitter 104a to the sacrificial channel 130. The sacrificial channel 130 transmits the resulting combined signal y (t) to the sacrificial receiver 105b.

  The crosstalk transfer function 210 may be characterized by a frequency response H (f) 210 or its time domain equivalent impulse response h, t. As illustrated in FIG. 2, the response H (f) 210 conveys the transformation that the invasive data signal u (t) 215 experiences at the connector portion of its route from the attack transmitter 104b to the victim receiver 105b. The details of this response 210 usually vary between specific sacrificial-invasion channel pairs. Nevertheless, the general nature of the response is based on geometric constraints and underlying physics. For example, the crosstalk response 151 of the backplane connector may depend on physical system parameters. The backplane crosstalk 150 can also be shaped by a transfer function, and the backplane and connector crosstalk 150, 151 can even be captured by a single (different) transfer function. is there.

  An exemplary (non-limiting embodiment of the present invention) of a crosstalk canceling device that corrects crosstalk occurring in a line card-backplane connection will be described later with reference to FIGS. 3 to 12B. In order for this disclosure to fully and completely convey the scope of the invention to those skilled in the art, the embodiments disclosed herein are provided. Those skilled in the art may apply the present invention to relate to crosstalk occurring on the backplane or elsewhere in the communication system, and the present invention may correct various types of crosstalk. Would admit.

  Returning now to FIG. 3, this figure illustrates a plot 300 of the crosstalk response 210 for the backplane-line card connector 102b in accordance with an exemplary embodiment of the present invention. This plot 300 illustrates a laboratory measurement of the power of the crosstalk signal 151, more specifically the power transferred from the invader channel 120 to the victim channel 130 at the connector 102b. The horizontal axis is the frequency measured in gigahertz (GHz). The vertical axis describes the signal power in decibels ("dB"), more specifically, the bottom of the squared crosstalk frequency response 210 is described as 10 times the logarithm of 10. Thus, this plot 300 illustrates the level of crosstalk power transferred from one channel 120 to another channel 130 for each frequency component of the invasion signal u (t) 215.

  In connectors 102a, 102b, the dominant mechanism for crosstalk 151 is typically capacitive coupling between connector pins. This mechanism is clearly evident in FIG. 3 as the general high pass nature of the response of plot 300. In other words, the plot 300 shows that higher signal frequencies above about 1 GHz tend to transfer energy through the crosstalk mechanism 151 more immediately than frequencies below about 1 GHz. The left side of plot 300, which is approximately below 1 GHz, provides a reduced power supply crosstalk signal of less than approximately −25 dB. Thus, the plot 300 shows that frequency components below approximately 1 GHz of the communication signal u (t) 215 transfer a relatively small amount of power delivered to the sacrificial channel 130 via the connector crosstalk 151. Show. The absolute value of the crosstalk 151 increases between approximately 0.25 GHz and 1 GHz. Thus, based on this plot 300, the components of the victim communication signal x (t) 214 having a frequency between approximately 1 GHz and 4.25 GHz are derived from the invading communication signal u (t) 215 having a similar signal frequency. It is particularly susceptible to the crosstalk effect 151.

  Further, the variation in the frequency response plot 300 for frequencies above 2 GHz illustrates that the crosstalk effect 151 is severely affected by other effects than simple capacitive coupling between a pair of pins. In other words, above 2 GHz, the plot 300 departs from the classical capacitive coupling response and generally increases with increasing frequency asymptotically (and monotonically). In contrast, the illustrated plot 300 produces a peak and trough pattern, such as a minimum at a higher frequency, for example, approximately 4.6 GHz.

  As noted above, sufficient crosstalk cancellation relies heavily on accurately modeling the system's crosstalk response. The crosstalk cancellation performance depends on the model accuracy, especially for frequencies where the crosstalk effect is strong, i.e. frequencies above approximately 1 GHz.

  In plot 300, the higher order effects of the peaks and valleys described above are generally characteristic of the relative geometric relationships that are not immediately known, between sacrificial signal path 130 and invasion signal path 120. Very dependent on. In other words, deriving an accurate and sufficient crosstalk model based on the geometric or physical analysis of the communication path is a problem without empirical data or test measurements regarding the actual crosstalk effects on the signal. May be included.

  Stated otherwise, the plot 300 of FIG. 3 shows that the higher frequency components of the communication signals 214, 215 are particularly prone to crosstalk 151 and the crosstalk response 210 for these higher frequency components. Modeling illustrates that it involves relating to the inherently irregular nature of this high frequency response. This type of model accurately represents these high orders (irregular response characteristics) because an accurate model of the system's crosstalk response 210 may provide a basis for sufficient crosstalk cancellation. You need to do. Even though passive circuit analysis does not easily derive a model with the required accuracy, the actual signal response may serve as a basis for creating a suitable model.

  In one exemplary embodiment of the present invention, the crosstalk model of the crosstalk cancellation device is defined based on crosstalk measurement data such as the measurement data shown in the exemplary plot 300 in FIG. Can do. As an alternative to obtaining this type of measurement data in the laboratory, as discussed below with respect to FIGS. 9 and 11, such as in an embodiment where the crosstalk cancellation device is switched to correction mode, the data is stored in the field. It can be obtained during operation.

  Returning now to FIG. 4, this figure illustrates a functional block diagram of a crosstalk cancellation system 400 according to an exemplary embodiment of the present invention. As described above, the present invention can provide crosstalk cancellation for a high speed digital communication system, such as the communication system 100 illustrated in FIGS. 1 and 2 and discussed above. More specifically, as described above with respect to FIGS. 1, 2, 3 and 4, FIG. 4 illustrates the crosstalk cancellation located in the canceling crosstalk 151 occurring in the backplane-line card connector 101b. An apparatus or crosstalk canceller ("XTC") is illustrated.

  In the sacrificial channel 130 for reception by the sacrificial receiver 105b, the digital data x (t) propagates. The sacrificial channel 130 is also derived from the digital data u (t) 215 output by the aggressor transmitter 104b and carries an unnecessary crosstalk signal n (t) 230 that is not intended for reception at the sacrificial receiver 105b. The intended data stream signal x (t) 214 and crosstalk signal n (t) 230 additionally form a composite signal y (t) 260. The crosstalk canceller 401 receives the composite signal y (t) 260, corrects the crosstalk interference n (t) 230 from this signal 260 via cancellation, and outputs a corrected signal z (t) 420, The victim receiver 105b receives it. That is, the crosstalk canceller 401 actually leaves the desired data signal 214 in its entirety and effectively propagates the signal 260 propagating in the victim channel 130 to effectively cancel the crosstalk signal element 230. The estimated value of the crosstalk 230 is applied.

The steps performed by the crosstalk canceller 401 consist of:
(I) Representative of independent inputs y (t) 260 (sacrificial signal destroyed by crosstalk 151) and u (t) 215 (invasion signal spreading in the invasion channel causing crosstalk signal 230) Accept as a part,
(Ii) transform the transmitted aggression signal u (t) into a crosstalk estimate that emulates the signal transformation 210 actually generated in the system 200 via the crosstalk effect 151;
(Iii) subtract the modeled crosstalk from the sacrifice y (t) 260 to cancel the crosstalk signal n (t) 230 component; and
(Iv) The corrected signal z (t) is output to the sacrificial receiver 105b which is a conventional receiver without a specific technique for crosstalk correction.

  Returning now to FIG. 5, this figure illustrates a functional block diagram of a crosstalk cancellation system 500 in accordance with an exemplary embodiment of the present invention. More specifically, FIG. 5 shows an example of a crosstalk canceller 401 comprising a crosstalk model 501, a weighted node 502, an electrical control “function” of the controller 503, or three functional elements 501, 502, 503 of a control module. A design overview is illustrated. Model 501 generates a crosstalk estimate signal w (t) and weighting node 502 applies this crosstalk estimate 520 to the victim channel 130. The controller 503 adjusts the parameters of the model 501 based on the output z (t) of the weighting node 502.

  Model 501 emulates an invasive transformation function H (f) 210 in the form of an adjustable frequency response function G (f) 501. That is, the model 501 generates an artificial crosstalk signal w (t) 520 that is generated by electromagnetic coupling of the connector 102b between the invading channel 120 and the sacrificial channel 130. Model, simulation or emulation of the crosstalk signal n (t) 230 interfering with The model frequency response G (f) 501 applies a frequency dependent response similar to the plot 300 illustrated in FIG. 3 and effectively applies the invasion data signal u (t) 215 in the manner discussed above. Apply a filter.

  Since the same invasive data stream u (t) 215 drives the actual crosstalk response H (f) 210 and the model 501 of the crosstalk canceller, the output w (t) 520 of the model 510 is an ideal case, It is equal to the invasion signal component n (t) 230. That is, G (f) 501 is equal to H (f) 210 in a theoretical or ideal case where all system parameters are known and fully modeled in a noise-free environment. Furthermore, in this ideal scenario, the output signals n (t) 230 and w (t) 520 from H (f) 210 and G (f) 501 respectively are also equal to each other. In a real situation with a large number of unknown effects and uncertainties, G (f) 501 is H () with sufficient accuracy and accuracy to support high data rate communication with essentially no error. f) Approximate to 210.

  The difference node 502 subtracts the emulated invasion signal w (t) 520 or the emulation signal 520 from the composite signal y (t) 260 and further eliminates crosstalk interference from the received victim signal y (t) 260. Or reduce. In a physical implementation that functions in an actual operating environment, the model G (f) 501 does not necessarily match the real response H (f) 210. The controller 503 adjusts the model 501 to minimize this error regarding inaccuracy between the actual crosstalk effect H (f) 210 and the emulated or modeled crosstalk effect G (f) 501. To do.

  The implementation of the weight node 501 is usually simple to those skilled in the art. However, special care must be taken to maintain high sensitivity to the two inputs. In addition, it is not uncommon for the crosstalk signals 230 and 520 thus modeled to have a small amplitude, particularly at high frequencies. At first glance, these high frequencies are often amplified by an equalizer (not shown) for a very short time on the surface. Thus, although neglected high frequency crosstalk may be small before equalization, it can be very significant after equalization. A weighting node must be implemented to accommodate this type of high frequency response.

  Some corrected signal z (t) 420 (ie, the output of difference node 502) is tapped and fed to the controller, where essentially the same signal 420 that the victim receiver 105b receives is sent to the controller 503. Supply. The controller adjusts the parameters of the modeling filter 501 characterized by the response G (f) 501 in order to maximize the fitness to the actual response H (f) 210. In particular, the controller 503 receives the crosstalk corrected signal z (t) 420 as input and processes, monitors or analyzes the signal 420 to determine signal fidelity. In other words, the controller 503 evaluates the performance of the model by analyzing the extent in which the model output 520 cancels the crosstalk signal 230. The controller 503 also adjusts the model 501 to enhance crosstalk cancellation and provide dynamic performance for condition changes.

  Since the output of the controller 503 includes the parameters of the modeling filter 501, the controller can adjust the modeled response G (f) 420. Thus, the controller 503 maximizes the fidelity of the corrected signal 420, i.e., minimizes crosstalk on z (t) 420, thereby reducing G (f) 420 and H (f) 210. And the modeling filter 501 can be processed. Stated otherwise, the controller 503 monitors the corrected crosstalk canceled signal z (t) and enhances the crosstalk model G (f) 420 to enhance crosstalk cancellation and signal quality. Adjust to. Thus, in one exemplary embodiment of the present invention, to correct for modeling errors, fluctuating dynamic conditions and other effects, adjust crosstalk cancellation, correct by yourself, You may include feedback that can be set with.

  The system illustrated in FIG. 5 may be implemented using analog integrated circuit engineering primarily to provide a relatively low degree of complexity, power consumption and cost. In an embodiment, model 501 and difference node 502 are completely analog. In another embodiment, the particular situation of model 501 is implemented digitally to take advantage of the digital nature of the invasion data source 104b.

  Controller 503 typically includes analog and digital circuits. Due to the specific situation of the analog preprocessing of the controller 503, this digital circuit can operate at a low speed associated with the communication data rate, thus facilitating practical implementation. In particular, digital circuits can operate at speeds on the order of less than the channel baud rate. In one exemplary embodiment of the present invention, the digital circuitry of controller 503 operates in at least one order below the channel baud rate. In one exemplary embodiment of the present invention, the digital circuitry of controller 503 operates in at least two orders below the channel baud rate. In one exemplary embodiment of the present invention, the digital circuitry of controller 503 operates in at least three orders below the channel baud rate. Further details of controller 503 and model 501 that together produce a crosstalk cancellation solution that achieves low power and low cost Exemplary embodiments are discussed in more detail below.

  Returning now to FIG. 6, which is a functional block diagram of a tapped delay line filter 600 according to an exemplary embodiment of the present invention. The tapped delay line filter 600 delays the input signal 215 by a series of delay stages 601a, 601b, and 601c, and generally outputs from the delay stages 601a, 601n, and 601c by amplifiers 602a, 602b, 602c, and 602d, respectively. And generating the output signal 620 from the input signal 215 by adding or combining these scaled signals. The tapped delay line filter 600 may be an analog component of the model 501 that generates a signal v (t) 620 having a shape or waveform that approximates the negative crosstalk signal n (t) 230. . That is, the tapped delay line filter 600 may be a typical waveform shaper implemented via an analog component.

  As described above, accurately modeling the actual crosstalk response 210 facilitates sufficient removal of the crosstalk interference 230 due to crosstalk cancellation. Rather than improving the crosstalk cancellation device (not illustrated), this type of device may be based on an inaccurate crosstalk model (not illustrated) that downgrades signal quality. For example, a “correction” signal intended to cancel the crosstalk interferes with the received victim signal while leaving the crosstalk signal targeted for essentially complete cancellation as a result of the wrong model Might add. Thus, the crosstalk model for an embodiment based on a filtering mechanism should be flexible enough to assist in modeling the various crosstalk transfer functions that may be encountered in an application. It is. That is, for example, a flexible crosstalk model is desirable over a rigid model that cannot be readily adapted to various applications, operating conditions and environments.

  In one exemplary embodiment of the present invention, as illustrated in FIG. 6, an analog tapped delay line filter 600 (also known as a transversal filter) includes electrically controllable gain coefficients 602a, 602b, 602c, 602d and model the invasive crosstalk transfer function 210. The filter 600 can provide a desirable level of flexibility and adaptability that supports a wide range of operating conditions and situations. More specifically, the tapped delay line filter 600 can generate a waveform that approximates the waveform of the crosstalk signal 230 imposed on the sacrificial channel 130.

The illustrated filter 600 includes N delay elements 601a, 601n, 601c (providing each time delay δ (delta)) and α _n (alpha) expressed by coefficients up to n = 0,. _n ) is a typical tapped delay line filter. The output v (t) 620 of the tapped delay filter 600 is

It is described.

Changing the values of the gain coefficients α _o , α _l , α ₂ ... (Alpha _o , alpha _l , alpha ₂ ... Alpha _n ) can cause a corresponding filter 600 response change. The tapped delay line filter 600 can model the impulsive impulse response up to Nδ (N times the delta), ie up to the time range of the filter 600. Furthermore, the frequency content of the invasion response 210 (illustrated in FIG. 3 and discussed above) is modeled up to a frequency of f = 1 / 2δ (the frequency is equal to the reciprocal of twice the delta). Can do. Thus, δ (delta) is less than f = 1 / (2δ) (the frequency is equal to the reciprocal of twice delta) at the highest frequency of interest in the victim signal x (t) 214. Things are chosen. Furthermore, N should be chosen such that the majority of the invasive impulse response is contained within a time range of Nδ (N times delta). Equivalently, the aggression frequency response 210 should not produce a large phase variation under a frequency of f = 1 / (Nδ) (the frequency is equal to N times the delta). These states for choosing N and δ (delta) contrast with the state of the invasion signal. A well-designed receiver can easily suppress these higher frequencies without degrading the sacrificial signal quality, so it is not critical if invasive noise remains above the specified frequency.

  Although the tapped delay line filter 600 can emulate, estimate or mimic the resulting pulse shaping by the invasion response 210, this filter 600 typically has an unrealistic number of taps or delay stages. No highly variable time delay can be adequately mentioned. As illustrated in FIG. 5 and discussed above, the time delay includes (i) a circuit tap directed to a portion of the invading data signal u (t) 215 to the crosstalk canceller 401, and (ii) a crosstalk canceller. This is directly related to the length of the signal path that gradually advances between the weighted nodes 502 of 401. More specifically, the modeled time delay is such that the modeled actual signals 230, 520 are properly synchronized or adjusted with respect to each other for effective mutual cancellation. Should be close to the time delay of the crosstalk signal n (t) 230 of Although the output 620 of the tapped delay line filter 600 can be used directly as the output w (t) 520 of the model 601, the delay line filter tapped to the crosstalk signal 120 on the victim channel 130. Synchronizing to the output 620 can enhance crosstalk cancellation and provide increased signal fidelity to the sacrificial receiver 105b, improving overall modeling flexibility.

  Since the location of the actual crosstalk signal 230 and its modeled object 520 binding point can be very variable within the sacrificial-invasion set, their respective delays are unclear. There may be a possibility or it is likely to be uncertain. Even in the relatively simple case of strong coupling through the backplane-line card connector 102b, the length of the signal path on the line card 101b is often a variable. Thus, the time delay can be difficult to predict without the specific knowledge and analysis of the line card layout. To address this uncertainty in time delay, an adjustable delay 701 can be incorporated into the crosstalk modeling filter 501 as illustrated in FIG.

  Returning now to FIG. 7, this figure is a functional block diagram of the crosstalk modeling filter (“XTMF”) of the crosstalk cancellation device 401 with adjustable delay 701 according to an exemplary embodiment of the present invention. is there. Adjustable delay 701 may precede or follow tapped delay line filter 600 (as shown in FIG. 7). In one exemplary embodiment of the present invention, placing an adjustable delay 701 on the input side of an analog tapped delay line filter 600 (as shown) can simplify implementation. This simplification may result from the individual characteristics of the digital signal u (t) 215, where it should be easily maintained by quantizing or hard limiting the output of the delay device 701. is there. Alternatively, if adjustable delay 701 follows tapped delay line filter 600, signal v (t) 620 is analog at the input to adjustable delay 701, according to the illustrated configuration. Inputting an analog signal to adjustable delay 701 can impose the need for a linear response over a wide range of signal values and frequencies, which is difficult to achieve due to large delay values. there is a possibility.

  A correction signal w (t) 520 close to the crosstalk signal n (t) 230, which is undesirably spread with the sacrificial signal 130 along with the intended data signal x (t) 214, and an adjustable delay 701 is provided. The waveform of the synchronization signal 520 is synchronized with the waveform of the unwanted crosstalk signal 230. In other words, the adjustable delay 701 times or adjusts the correction signal 520 so that it is synchronized in time with the actual crosstalk interference 230.

  Based on the function of the tapped delay line filter 600 and adjustable delay 701, the crosstalk modeling filter 501 has a cancel signal having a form and timing that exactly matches the actual crosstalk signal n (t) 230. w (t) 520 is output. When inserted or applied to the sacrificial channel 130 via the subtraction node 502 as illustrated in FIG. 5 or discussed above, the cancellation signal w (t) 520 is the actual crosstalk signal. 230, thereby enhancing the quality of the communication signal z (t) 420 delivered to the victim receiver 105b.

  As discussed above with reference to FIG. 5 and further below with reference to FIG. 8, the controller 503 includes a tapped delay line filter 600 and an adjustable delay 701 with their respective performance. Is finely adjusted to enhance the fidelity of the correction signal 420 delivered to the victim receiver 105b.

  Returning now to FIG. 8, this figure is a functional block diagram of a crosstalk modeling filter 501 'of a crosstalk cancellation apparatus 800 comprising a high pass filter 801 according to an exemplary embodiment of the present invention. The high-pass filter 801 is generally a fixed or non-adjustable filter. In FIG. 8, in an exemplary embodiment configuration of the illustration, an adjustable delay 701 provides a tapped delay line filter 600. It thus provides a reliable application, as was certainly discussed above with respect to FIG.

  As illustrated in FIG. 7, including an optional high pass filter 801 in the exemplary crosstalk modeling filter 501 'may enhance the performance of some applications or operating environments. The high-pass filter 801 is a device that receives a signal having a frequency component range, reduces the frequency component below the frequency threshold, and transmits the frequency component above the frequency threshold.

The tapped delay line filter 600 has a frequency range

Although having the above flexible modeling response, they are often less flexible at low frequencies, such as f <1 / (Nδ) (the frequency is less than twice the inverse of δ). Thus, a low frequency characteristic that accurately models the crosstalk response 210 may require a large number N of filter taps that increase filter complexity and a longer delay increment that causes high frequency flexibility. May require δ (delta). In many applications, it is preferable to avoid such tradeoffs. As discussed above with respect to FIG. 3, for electrical systems, the low frequency crosstalk characteristics are usually dominated by capacitive coupling effects, and thus, such as simple primary resistor-capacitor (“RC”) high pass filters, etc. Can be accurately modeled by a high-pass filter. That is, inserting the high pass filter 801 into the crosstalk modeling filter 801 can provide a high level of performance without requiring a cumbersome or uneconomic number of tap filters in the tapped delay line filter 600.

  Similar to the exemplary embodiment of the crosstalk modeling filter 501 represented in FIG. 7, the ranking of the tapped delay line filter 600, adjustable delay 701 and high-pass filter 801 supports various preparations. The order may be changed for this purpose. That is, the present invention places physical components corresponding to each functional block 701, 600, 801 illustrated in FIG. 8 in any parallel or series configuration that provides acceptable performance for the intended application. To help. Nevertheless, certain configurations or rankings may provide certain advantages or trade-offs for selected application situations compared to other configurations.

  In FIG. 8, the exemplary exemplary inline configuration places adjustable delay 701 on the input side of tapped delay line filter 600 and high pass filter 801 on the output side of tapped delay line filter 600. . With this ranking, the implementation of adjustable delay 701 can be simplified by taking advantage of the discrete-amplitude characteristics of its input and output signals. Through a digital delay element, the tapped delay line filter 600 can also take advantage of the discrete-amplitude input provided by the adjustable delay 701. In that RC implementation, the high pass filter 801 is an analog device that does not receive the benefits from providing it with a discrete amplitude input. As described above, it is generally not disadvantageous to arrange the high-pass filter 801 at the output side of the crosstalk modeling filter 501 'or at another position.

  As discussed further above with respect to FIG. 5, the control module 503 takes the crosstalk compensation digital z (t) as an input and outputs control signals 820, 803 to adjust the crosstalk response model 501. . The output 820, 830 of the control module to the crosstalk modeling filter 501 is (i) a “delay control” signal 830 for controlling the time delay implemented by the adjustable delay component 701, and (ii) tapped. A set of “filter control” signals 820 that control the gain on variable coefficient amplifiers 602a-d of delay line filter 600. That is, the controller 503 outputs a modeling parameter to the tapped delay line filter 600 and a timing parameter to a further adjustable delay 701.

  These output control values are determined based on observation, processing and / or analysis of the corrected signal z (t) 420. A non-provisional US patent application number 10 / 108,598 filed March 28, 2002 entitled "Method and System for Outputting and Encoding Multilevel Signals" is for evaluating signal fidelity. A realistic exemplary system and method is disclosed. No. 10 / 627,477, filed July 15, 2003, entitled “Coding Adjustable Noise Filtering and Equalization with Optical High-Speed Multilevel Signals”, which is not a commonly owned US provisional application. Discloses a practical exemplary system and method for controlling device parameters of the crosstalk modeling filter 501. The disclosures of US Patent Application No. 10 / 108,598 and US Patent Application No. 10 / 627,477 are hereby fully incorporated by reference. One or more of crosstalk model 501, tapped delay line filter 600 and adjustable delay 701 are disclosed in US patent application Ser. No. 10 / 108,598 or US patent application Ser. No. 10 / 627,477, respectively. May be controlled and / or adjusted using existing methods and / or systems. The time delay adjustment of the adjustable delay 701 may be determined by considering the delay control as a variable that is removed in its entire range of potential values, for example, following the disclosure of these patent applications.

  Returning now to FIG. 9, which illustrates the exemplary crosstalk modeling filter 501 ′ illustrated in FIG. 8 or the exemplary crosstalk modeling filter 501 illustrated in FIG. 7 and their associated adjustments. An exemplary system 900 for controlling possible delays 701 is illustrated. More specifically, FIG. 9 is a functional block diagram of the control module 900 of the crosstalk cancellation apparatus 401 according to an exemplary embodiment of the present invention. In FIG. 9, the exemplary controller 900 illustrated facilitates relatively simple theoretical analysis and implementation, such as in US patent application Ser. No. 10 / 627,477, discussed above, or Benefits may be provided for certain applications on the control method and system disclosed in US patent application Ser. No. 10 / 108,598.

  The controller 900 of FIG. 9 includes a filter 901 having a frequency transfer response P (f) that receives the signal z (t) with the crosstalk scheduled for reception by the victim receiver 105b canceled. The filter 901 may be a spectrum weighting filter based on this frequency transfer response. The output of this filter 901 is coupled to a power supply detection or signal squaring device 902 that provides an output to a low pass filter 903. The low-pass filter 903 is a device that receives a signal having a range of frequency components, reduces frequency components above the frequency threshold, and transmits frequency components below the frequency threshold.

  An analog-to-digital converter (“ADC”) receives the output of the low pass filter and generates a corresponding digital signal that is supplied to the digital controller 905. Digital controller 905 in turn generates a digital control signal for each adjustable delay 701 and tapped delay line filter 600. Each digital / analog converter (“DAC”) 906a, 906b converts these signals to the analog domain for respective transmission on the delay control line 830 and the filter control line 820. While adjusting the delay line filter 600 tapped with the analog filter control signal, the analog delay control signal adjusts an adjustable delay 701.

  It is useful to discuss a simple operable embodiment imposed on crosstalk on a channel in a transient state that does not convey data. More specifically, the attack transmitter 104b transmits data in a wide spectrum content or signal frequency range, such as pseudo-random or encoded pseudo-random data, for example, while the victim transmitter 104a does not transmit any data. Consider. That is, referring back to FIG. 5, temporarily referring to signal x (t) 214 while u (t) 215 is a digital data signal having a broad analog spectrum content that randomly arises from various digital data patterns. Is zero. In this case, signal y (t) 260 is simply the invaded n (5) 230 received, and signal w (t) 520 is the modeled invasion. Thus, the signal z (t) 420 is an error that is actually modeling the canceling device. In the ideal situation in the theory of complete crosstalk cancellation, z (t) 420 is at zero.

  In other words, transmitting a signal having a wide range of frequencies on the invading channel 120 while transmitting essentially the same voltage on the sacrificial channel 130 provides pure crosstalk on the sacrificial channel 130, n (t) 230 is equal to y (t) 260. If the crosstalk canceller 401 outputs a cancel signal w (t) 520, making the pure crosstalk signals n (t) 230, z (t) 420 equal also has essentially no signal energy. Thus, in this state, the signal energy of z (t) 420 represents modeling or delay inaccuracy of the crosstalk modeling filter 501.

  The control module 900 may implement a state of sending a definition signal on the invasion channel 120 to send a constant voltage or essentially other than a data signal on the victim channel 130. In order to minimize the signal z (t) 420 received by the sacrificial receiver 105b, the control module 900 may then adjust the tunable parameters of the crosstalk modeling filter 501 ′, and in this way, the actual cross A crosstalk cancellation signal w (t) 520 that matches the talk signal n (t) 230 is provided, and a modeled crosstalk response G (f) 501 that more effectively matches the actual crosstalk response H210. provide. More generally, in order to characterize the crosstalk effect 151 and to control, optimize or adjust other forms of crosstalk cancellation or other crosstalk correction, the control module 900 can control the invasion channel 120, the sacrificial channel 130. Or it causes the transmission of a defined or well-known signal pattern in both the invasion channel 120 and the sacrificial channel 130. Furthermore, the control module 900 may have a learning or adaptive mode in the form of a setup mode or a self-configuring procedure and may implement automatic or self-correction.

  In connection with FIG. 9 and over the example of imposing crosstalk on the channel void of data, a general discussion will be made to highlight any higher importance of a particular frequency over others. The error signal z (t) 420 may be weighted with a spectrum, with an arbitrary filter 901 whose response is shown as P (f). For example, an error may be required for the high pass filter error signal z (t) 420 to emulate the effect of equalization of the victim receiver 105b. The error signal z (t) 420 (potentially spectrally loaded) is then squared or power detected, i.e. the output of the squaring device 902 is signal power. The power signal is then passed through a low pass filter 903 (or integrator) with a relatively low cutoff frequency to obtain an integrated power supply (ie energy) for the error signal z (t) 420. Thus, the signal at this point matches the analog estimate of the statistical variance of the error signal z (t) 420 (ie, the square of the standard deviation).

  As familiar to those skilled in the art, error variance is a useful metric for measuring fidelity. Since the cut-off frequency of the low-pass filter 903 is very long (typical order under the symbol transmission rate), the transient effect of any modeling filter is almost constant after changing the attenuation. In this way, the analog dispersion signal may be sampled by a simple slow high resolution analog to digital converter 904. The analog-to-digital converter 904 digitizes the signal output into a simple microprocessor, state machine, finite state machine, digital controller or similar device (referred to herein as a digital controller) for error variance information. I will provide a. After recording the error variance for the current set of response modeling parameters, the digital controller 905 then digitally passes the new parameters to a set of DACs 906 that provide the corresponding analog signal to the invasion emulation module 501. A new filter configuration may be specified by outputting.

  The digital controller 905 can (i) set the parameters of the crosstalk modeling filter 501 and (ii) directly observe the effect of the current parameters on the modeling error variance, so that the digital controller 905 returns to the actual response 210. A parameter set that maximizes the fit of the invasion response model 501 can be found. All combinations of model parameters may be tested in many instances because trial and error processing is not overly complicated. However, other empirical search / optimization methodologies known to those skilled in the art may be used instead. In one exemplary embodiment of the present invention, the coordinate descent method described in US patent application Ser. No. 10 / 620,477 discussed above can be searched and optimized to identify acceptable model parameters. I will provide a.

  As described above, the control module 900 may include a combination of analog and digital circuits to provide a virtual control implementation. The filter 901 and the power supply detection device 902 jointly output a high-speed analog signal. The low-pass filter 903 receives a high-speed analog signal as an input and outputs a low-speed analog signal. Filter 901, power supply detector 902, and low pass filter 903 collectively receive a projection of the high speed signal onto the low speed signal by extracting the relevant statistical information from the high speed signal and presenting it in a more concise form. The ADC 904 receives this low speed analog signal as input and outputs a corresponding digitized approximation. As a result, the controller 905 receives and processes this low speed digital signal. Because digital signals are slow, the associated processing circuitry is less complex than would be required if the signal was fast. The digital control device 905 outputs low-speed digital control signals to the digital / analog converters 906a and 906b that sequentially output low-speed analog signals. As a result of simple high-speed analog preprocessing and low-speed digital processing in series, the control module 900 provides a signal analysis based on powerful statistical property tests that can facilitate crosstalk cancellation on high-speed communication systems. It implements a robust control methodology with relatively little circuit complexity.

  To generate error variance, FIG. 9 uses a power supply detection (or signal square) device, while a full wave rectifier (which receives the absolute value of the signal) may be used instead. Due to the full wave rectifier based implementation, the output of the low pass filter 903 now represents a valid fidelity criterion even though it no longer matches the error variance. In particular, it is one standard for the error signal 420, and thus the fidelity metric still has appropriate mathematical properties. Those skilled in the art will appreciate that the “one reference” of the signal generally determines that it comprises integrating the absolute value of the control signal. (I) One reference signal may have a decreasing dynamic range (and loosen the resolution constraints of the analog to digital converter 904), and (ii) a full wave rectifier may be easier to implement than a power supply detector. For reasons such as not present, this substitution may be advantageous for certain applications. Such modifications are considered to be within the scope of the present invention.

  Similarly, the power detector 902 may also be replaced with a half-wave rectifier or any similar device used to evaluate the signal absolute value. The division into the crosstalk canceller 401 into the functional blocks, modules and respective sub-modules shown in FIGS. 5 to 9 is conceptual and ensures the functionality or physical grouping of the components. It will be appreciated by those skilled in the art that this does not necessarily indicate a critical boundary. Rather, the representation of the exemplary embodiments as diagrams based on functional block diagrams facilitates the description of the exemplary embodiments of the present invention. In practice, these modules may be combined or separated, while being redistributed to other modules that do not depart from the scope of the present invention.

  In one exemplary embodiment of the present invention, the crosstalk cancellation system is a single integrated circuit (“IC”), such as a monolithic IC. Each crosstalk cancellation device, control module, and crosstalk modeling filter may also be a single IC. Such an IC may be a complementary MOS ("CMOS") IC and may be manufactured, for example, in a 0.18 micron process.

  Now, the process for canceling crosstalk and the process for adjusting the crosstalk canceller are described with reference to FIGS. 10 and 11 respectively. As described herein, certain steps in the described process must naturally precede other functions for the present invention. However, if such an order or sequence does not change the functionality of the invention, the invention is not limited to the order of the steps described. That is, it is recognized that some steps may be performed before or after other steps or in parallel with other steps without departing from the scope and spirit of the invention.

  Returning now to FIG. 10, this figure is a flowchart illustrating a process 1000 for canceling crosstalk 151, named crosstalk cancellation, according to an exemplary embodiment of the present invention. In step 1010, which is the first step of process 1000, the invasion transmitter 104b transmits an invasion communication signal u (t) on the invasion channel 120. This communication signal 215 may be an analog or digital signal carrying data.

  In step 1015, the crosstalk effect 151 couples energy from the aggression communication signal u (t) 215 to the sacrificial channel 130 as crosstalk n (t) 230. The coupling mechanism may be the electromagnetic coupling described in the exemplary case of an electrical data signal extending to the backplane 103 or other optical or electrical crosstalk mechanism. The energy transfer of the crosstalk effect 151 generates a crosstalk signal n (t) 215 in the sacrificial channel 130 in a manner that results in signal propagation towards the sacrificial receiver 105b.

  In step 1020, the sacrificial transmitter 104 a transmits the sacrificial communication signal x (t) 214 on the sacrificial channel 130. The sacrificial communication signal 214 may be an analog or digital signal. In step 1025, the crosstalk signal n (t) 230 coexists or mixes with the sacrificial communication signal x (t) 214 in the sacrificial channel 130. Composite signal y (t) 260 is the result from the combination of these signals 214, 230.

  In step 1030, the crosstalk model 501 obtains a sample of the invasive communication signal u (t). In other words, the tap or other node directs a representative portion of the invasive communication signal 215 to the crosstalk canceller 401 for reception and processing by the crosstalk model 501.

In step 1035, the crosstalk model 501 processes the sampled portion of the invasive communication signal u (t) 215 through the tapped delay line filter 600. Modeling parameters such as the gain or scaling constant of the tapped delay line filter provide a basis for generating the waveform estimate v (t) 620 of the crosstalk signal n (t) 215. More specifically, the coefficients α _o , α _l , α ₂ ... (alpha _o , alpha _l , alpha ₂ ... alpha _n ) of the variable coefficient amplifiers 602a, 602b, 602c of the tapped delay line filter are: A waveform v (t) 620 that approximates the crosstalk signal 215 is defined.

  In step 1040, the adjustable delay 701 of the crosstalk model 501 causes the time delay to be synchronized with this waveform 620 to the widening interfering crosstalk signal n (t) 230 in the victim channel 130. Applies to (t). At step 1045, the weighting node 502 of the crosstalk canceller 401 uses the resulting crosstalk cancellation signal w (t) 520 as the victim channel 130 and the combined crosstalk and the communication signal y ( t) Apply to 260. The crosstalk cancel signal w (t) 520 cancels the least part of the spread crosstalk signal component 2 (t) 520 spreading in the sacrifice channel 130. Reducing this crosstalk interference 520 improves the signal fidelity of the communication signal z (t) 420, which is the output by the crosstalk canceller 410 for delivery to the victim receiver 105b.

  At step 1050, the controller 503 processes or analyzes the crosstalk corrected signal z (t) 420 to determine the effect of the crosstalk cancellation. In other words, the controller 503 determines whether the crosstalk canceller is applying a crosstalk cancellation signal w (t) 520 that exactly matches the actual crosstalk n (t) 230 in both waveform and timing. In order to evaluate signal fidelity.

  In step 1055, the controller 503 optimizes the modeling parameters, particularly the tapped delay line filter 600, to optimize the waveform match between the crosstalk cancellation signal w (t) 520 and the actual crosstalk signal n (t) 230. The coefficients of the variable coefficient amplifiers 602a, 602b, 602c, and 602d are adjusted. The controller 503 further adjusts the variable or adjustable time delay of the adjustable delay 701 in order to synchronize the crosstalk cancellation signal w (t) 520 with the actual crosstalk signal n (t). That is, the controller 503 performs parameter adjustment on the crosstalk modeling filter 501 to enhance the fidelity of the net communication signal z (t) 420 delivered to the victim receiver 105b, thereby operating the crosstalk canceller 401. Adjust.

  Following step 1055, process 1000 repeats steps 1010 through 1055. The crosstalk canceller 401 continues to implement crosstalk 230 cancellation and response implementations that adapt to dynamic conditions, thereby providing a high level of communication signal fidelity in progress.

  Returning now to FIG. 11, this figure is a flow chart illustrating a process 1100 for correcting the crosstalk cancellation device 401, named crosstalk cancellation correction, according to an exemplary embodiment of the present invention. In step 1110, which is the first step of the process 1100, the controller 503 begins a correction sequence. The controller 900 instructs the invasion transmitter 104b to output a signal on the invasion channel 120 that has a known or defined test pattern, such as a bit pattern of random or pseudo-random data. This test or correction signal may have the format of the invasion communication signal (t) 215 or may be a proprietary format for characterizing the crosstalk response H (f) 210. That is, the controller 900 can control transmission of a signal having a predetermined voltage pattern on the invasion channel 120.

  At step 1115, the controller 900 instructs the sacrificial transmitter 104b to output a known sacrificial test or reference signal on the sacrificial channel. The test signal may be a predetermined communication signal or simply a constant voltage, data null.

  Sending a known test signal on the sacrificial channel 130 facilitates separating the crosstalk response H (f) 210 from other effects that may generate code distortion on the sacrificial channel 130. That is, the controller 900 can control transmission of a signal having a predetermined voltage pattern on the sacrifice channel 130.

  At step 1120, crosstalk n (t) 230 from the aggression signal u (t) 215 is coupled to the sacrificial channel 130. With the sacrificial channel 130 carrying a constant voltage as the sacrificial signal x (t) 214, the composite communication and crosstalk signal y (t) 260 on the sacrificial channel 130 is essentially a crosstalk signal n (t 230.

  In step 1125, the crosstalk canceller 401 generates an estimated value w (t) 520 of the crosstalk signal n (t) 230 for crosstalk cancellation.

The crosstalk canceller 401 generates an estimated value 520 using modeling and delay parameters that have a waveform and timing at which the crosstalk signal n (t) and the crosstalk cancellation signal w (t) match. The crosstalk compensator 401 applies the crosstalk estimate 520 to the victim channel 130 and cancels at least a portion of the crosstalk 230 that is spread over it. The resulting crosstalk canceled signal z (t) 420 is transmitted to the sacrificial receiver 105b.

  In step 1130, the controller 503 processes and analyzes the crosstalk canceled signal z (t) 420 output by the crosstalk canceller 401. Based on the analysis, the controller 503 adjusts modeling and delay parameters to minimize the energy of the crosstalk canceled signal z (t) 420. That is, the controller 503 changes the operation parameter of the crosstalk canceller 401 so as to reduce the residual crosstalk. This control operation matches the crosstalk compensation signal w (t) 520 to the actual crosstalk n (t) 230 passed in the sacrificial channel 130.

  At step 1140, the controller 503 completes the calibration cycle and provides notification to the attack and victim transmitters 104a and 104b that the crosstalk canceller 401 sets to process live data. In response to this notification, at step 1145, the victim transmitter 104a and the aggressor transmitter 104b transmit live data on their respective channels 130, 120, respectively.

  At step 1150, crosstalk from live data 215 being transmitted on the invader channel 120 is coupled to the victim channel 130. At step 1155, the crosstalk canceller 401 processes the sample of live data 215 being transmitted on the invasion channel 120 and uses the modeling and delay parameters defined or updated during correction to emulate the crosstalk 230 or An estimated value 520 is created.

  In step 1160, the crosstalk canceller 401 applies the crosstalk estimate 520 to the victim channel 130 for crosstalk cancellation to indicate the victim receiver 105 having a hi-fi signal. Following the step 1160, the process 1100 ends. The controller 503 may repeat the calibration procedure at a defined or regular time interval, or when the controller's monitoring function determines that signal fidelity is compromised or falls below a threshold.

  Returning now to FIGS. 12A and 12B, these figures illustrate test data for a communication system, respectively, before and after implementing crosstalk cancellation in accordance with an exemplary embodiment of the present invention. These figures show an eye diagram 1200, 1250 of measured data captured under laboratory conditions. As known to those skilled in the art, eye diagrams 1200, 1250 provide a visual indication of signal quality. The level of openness of the “eye” 1225, 1275 of the eye diagrams 1200, 1250 correlates with the level of signal quality. That is, eye diagrams that are noisy, deformed, or closed eyes generally indicate a signal failure.

  FIG. 12A is an eye diagram 1200 from a 5 Gbit / s binary communication system operating under laboratory conditions that are considered representative of field conditions. The aggression signal 120 has an amplitude of 1200 millivolts, and the sacrificial signal 130 has an amplitude of 800 millivolts. FIG. 12A illustrates an eye diagram 1200 of the received signal 260 after equalization and cross amplification without crosstalk compensation. FIG. 12B illustrates an eye diagram 1250 of the received signal 420 after application of crosstalk cancellation in accordance with an exemplary embodiment of the present invention with equalization and limited amplification. Similar to the eye diagram of FIG. 12A, the invasion signal 120 has an amplitude of 1200 millivolts, and the sacrificial signal 130 has an amplitude of 800 millivolts.

  The signal path includes limiting amplifiers in both the crosstalk-corrected eye diagram 1250 and the non-crosstalk-compensated eye diagram 1200, so that the top and bottom horizontal “eyelid” thickness of each eye diagram 1200, 1250. Does not provide a useful gauge of signal quality. Rather, the signal enhancement provided by crosstalk cancellation is evident from the wide open eye 1275 of the crosstalk corrected eye diagram 1250 in conjunction with the narrow noisy eye 1225 of the eye diagram 1225 without crosstalk correction. It is.

  Furthermore, in order to characterize the communication performance improvement achieved by crosstalk cancellation in accordance with an exemplary embodiment of the present invention, before and after crosstalk cancellation, bit error rate measurements are taken from this test system under the same test conditions. Obtained. Without crosstalk cancellation, the communication system produced an average 1-bit error for every 100,000 bits transmitted. With crosstalk cancellation, the communication system produced an average 1-bit error for every 100,000, 000,000,000 bits transmitted.

  Although the system of the present invention may consist of circuitry that cancels, corrects or corrects crosstalk imposed on one communication signal by other signals, those skilled in the art will be able to It is understood that the present invention is not limited to this application, and that the embodiments described herein are illustrative and not limiting. In addition, it should be understood that various other alternatives to the embodiments of the invention described herein may be used in practicing the invention. The scope of the present invention is intended to be limited below by the claims only.

FIG. 1 illustrates a functional block diagram of a communication system having two line cards communicating through a backplane and inviting crosstalk. FIG. 2 illustrates a functional block diagram of the crosstalk model of the system illustrated in FIG. FIG. 3 illustrates a crosstalk response plot for a backplane-line card connector according to an exemplary embodiment of the present invention. FIG. 4 illustrates a functional block diagram of a crosstalk cancellation system according to an exemplary embodiment of the present invention. FIG. 5 illustrates a functional block diagram of a crosstalk cancellation system including functional blocks of a crosstalk cancellation apparatus according to an exemplary embodiment of the present invention. FIG. 6 is a functional block diagram of a tapped delay line filter according to an exemplary embodiment of the present invention. FIG. 7 is a functional block diagram of a crosstalk modeling filter of a crosstalk cancellation apparatus with adjustable delay according to an exemplary embodiment of the present invention. FIG. 8 is a functional block diagram of a crosstalk modeling filter of a crosstalk cancellation apparatus with a high-pass filter according to an exemplary embodiment of the present invention. FIG. 9 is a functional block diagram of a control module of the crosstalk cancellation apparatus according to an exemplary embodiment of the present invention. FIG. 10 is a flowchart illustrating a process for canceling crosstalk in accordance with an exemplary embodiment of the present invention. FIG. 11 is a flowchart illustrating a process for adjusting a crosstalk cancellation apparatus according to an exemplary embodiment of the present invention. 12A and 12B illustrate communication system test data before and after performing crosstalk cancellation in accordance with an exemplary embodiment of the present invention.

Claims (43)

A signal processing system that applies a crosstalk estimate to the first communication channel to correct for crosstalk coupled from the second communication channel to the first communication channel,
A waveform shaper including a plurality of delay stages coupled to a plurality of respective amplifiers;
An adjustable delay coupled to the waveform shaper, coupled to the second communication channel and operative to produce the crosstalk estimate;
A controller operable to process the crosstalk corrected communication signal to adjust the waveform shaper and the adjustable delay;   The signal processing system according to claim 1, wherein the waveform shaper includes a tapped delay line filter.   The signal processing system of claim 1, wherein the adjustable delay receives input from the waveform shaper.   The signal processing system of claim 1, wherein the adjustable delay provides an input to the waveform shaper.   The model further comprises a fixed filter coupled to one or more of the waveform shaper and the adjustable delay, transmitting a frequency above a limit frequency and reducing a frequency below the limit frequency. Item 1. The signal processing system according to Item 1.   The signal processing system of claim 1, wherein the controller comprises a digital controller.   The signal processing system of claim 1, wherein the controller is further operable to control transmission of a signal having a predetermined voltage pattern on the first and second communication channels.     The signal processing system of claim 1, wherein the controller operates to adjust the model. A method for processing a first communication signal having crosstalk from a second communication signal is:
Taking a sample of the second communication signal;
Processing a sample of the second communication signal to generate a crosstalk emulation signal;
Applying the crosstalk emulation signal to the first communication signal to cancel at least some crosstalk;
Generating a third communication signal based on application of the crosstalk emulation signal;
Analyzing the third communication signal;
And adjusting the crosstalk emulation signal based on the analysis. The method of claim 9, wherein
The step of processing comprises processing the sample with a model comprising at least one modeling parameter;
The adjusting step includes a step of adjusting the at least one modeling parameter. The method of claim 9, wherein
The step of adjusting comprises timing the crosstalk emulation signal to match the crosstalk. The method of claim 9, wherein
The step of processing comprises synchronizing the crosstalk and the crosstalk emulation signal to a level of synchronization;
The adjusting step includes the step of improving the level of synchronization. The method of claim 9, wherein
The step of generating the crosstalk emulation signal comprises a step of generating a waveform different from the crosstalk by a difference,
The adjusting step includes a step of reducing the difference. The method of claim 9, wherein
Analyzing the third communication signal comprises monitoring for residual crosstalk of the third communication signal;
The step of adjusting the crosstalk emulation signal includes a step of reducing the residual crosstalk.   10. The method of claim 9, wherein applying the crosstalk emulation signal to the first communication signal comprises subtracting the crosstalk emulation signal from the first communication signal. The signal processing method is
Sending a test signal over the first communication channel;
Coupling a part of the test signal from the first communication channel to the second communication channel due to the crosstalk effect;
Defining a model of the crosstalk effect based on processing a portion of the test signal coupled to the second communication channel due to the crosstalk effect;
Transmitting a first communication signal on the first communication channel;
A portion of the first communication signal is coupled to the second communication channel due to a crosstalk effect;
Transmitting a second communication signal on the second communication channel;
Processing a portion of the first communication signal with the model and outputting an estimate of a portion of the first communication signal coupled to the second communication signal due to the crosstalk effect;
And it comprises the step of correcting the crosstalk effect by applying the estimated value to the second communication channel. The method of claim 16, wherein
Transmitting the test signal on the first communication channel further includes transmitting data to the first communication channel and supplying essentially the same voltage to the second communication channel. With. The method of claim 16, wherein
Defining the model of the crosstalk effect includes adjusting modeling parameters of the model.   17. The method of claim 16, wherein defining a model of the crosstalk effect includes adjusting a signal delay of the model.   17. The method of claim 16, wherein defining a model of the crosstalk effect includes adjusting a gain of the model.   17. The method of claim 16, wherein applying the estimate to the second communication channel comprises combining the second communication signal and the second communication signal due to the crosstalk effect. Subtracting the estimated value from a portion of. A circuit for correcting crosstalk coupled from the first communication channel to the second communication channel is:
Adjustable analog filters that can be controlled by modeling inputs;
An adjustable delay coupled to the adjustable analog filter;
A modeling filter coupled to the first communication channel and operative to generate an estimate of the crosstalk;
An analysis circuit operable to analyze a crosstalk correction signal and adjust the adjustable analog filter and the adjustable delay;   23. The signal processing system according to claim 22, wherein the analog filter includes a tapped delay line filter.   23. The signal processing system of claim 22, wherein the adjustable delay receives input from the analog filter.   23. The signal processing system of claim 22, wherein the adjustable delay provides an input to the analog filter.   23. The signal processing system of claim 22, wherein the modeling filter is further connected to one or more of the analog filter and the adjustable delay, and transmits a frequency above a limit frequency, A fixed filter that reduces the frequency is provided.   23. The signal processing system according to claim 22, wherein the analysis circuit includes a digital controller.   23. The signal processing system of claim 22, wherein the analysis circuit is further operative to control transmission of signals having predetermined voltage patterns on the first and second communication channels.   23. The signal processing system of claim 22, wherein the analysis circuit is further operative to adjust the modeling filter. In a data communication system, a device for correcting crosstalk is:
A crosstalk modeling filter that models the crosstalk response;
A difference node that subtracts the modeled crosstalk response from the communication signal comprising the crosstalk to provide a crosstalk corrected signal;
A control circuit that adjusts the crosstalk modeling filter to process the crosstalk corrected signal and reduce residual crosstalk of the crosstalk corrected signal;   30. The invention of claim 30 wherein the crosstalk modeling filter comprises an adjustable delay filter and an analog tapped delay line filter.   In the invention of claim 31, the adjustable delay filter circuit includes a digital circuit that receives the data signal causing the crosstalk and limits amplification of the data signal.   In the invention of Claim 31, the crosstalk modeling filter further comprises a first order high pass filter.   In the invention of claim 30, the control circuit comprises an analog low-pass filter, an analog-digital converter, and a digital controller.   35. The invention of claim 34, wherein the data communication system communicates that the baud rate and data of the analog-to-digital converter are moving at a lower speed than the baud rate.   The data communication system transmits data at a baud rate, and the digital controller operates at a speed lower than the baud rate.   In the invention of claim 34, the apparatus for measuring the absolute value of the signal based on the signal amplitude supplies an input to the analog low-pass filter.   In the invention of claim 37, the device comprises a power supply detector.   The invention of claim 37, wherein the device comprises a full wave rectifier.   The invention of claim 37 wherein the device comprises a half wave rectifier.   37. The invention of claim 37 wherein a spectral weight filter provides input to the device.   35. The invention of claim 34, wherein the control circuit comprises a finite state machine.   35. The invention of claim 34, wherein the control circuit includes a microprocessor.

Images