A SYSTEM AND METHOD FOR INCREASING TRANSMISSION DISTANCE OF AN OPTICAL SIGNAL
Field of the Invention
The present invention relates to optical communication and more particularly to a system and method for increasing transmission distance and optical signal to noise ratio value using an optical duobinary signal. Background of the Invention
The rapidly worldwide growing data traffic in telecommunication networks results in a sharp increase in the demand for transmission capacity. The main direction of increasing the utilization of an optical network is the use of Dense Wavelength Division Multiplexing (DWDM), i.e., a great number of wavelength channels densely packed with channel spacing of 100 GHz or less and transmission rates of 10 Gbps per channel or more travelling in a single fiber over considerable distance, resulting in an improved utilization of the available bandwidth of the fiber channel.
Other techniques for improving the bandwidth efficiency, in addition to the DWDM technique, are improved modulation formats, channel coding, sophisticated management of chromatic and Polarization Mode Dispersion (PMD), including equalization, wideband optical amplifiers and improved DWDM multiplexers and demultiplexers. It is thus desired to find a technique that can be implemented in a reliable way especially at high frequencies such as 10 Gbps.
A duobinary signal is a partial response signal, also called Quadrature Partial Response (QPR), which overcomes Nyquist's limit for a digital communication link, by using a bit shaping based on the analog sum of the last two bits (the present and the previous transmitted bits). Nyquist limit states that the number of bits you can safely transmit in a channel of bandwidth B Hz that has Inter-symbol interference is 2B per second using a "brick wall" filter having bandwidth of B Hz. The spectrum of duobinary transmission uses approximately 2.8 GHz for a 10 Gbps transmission rate, while the conventional Non-Return to Zero (NRZ) modulation uses approximately 7.5 GHz for a 10 Gbps transmission rate (Nyquist theoretical limit in 5 GHz for 10 Gbps
transmission). Such reduction in the bandwidth permits an increase of the transmission distances, which are limited by dispersion, from 80 km (present distances being achieved by NRZ modulated devices) to over 200 km. The known, prior art principles of generating Optical Duobinary (ODB) signal is described in Figs. 1 through Fig. 8.
Fig. 1 depicts a block diagram for the generation of duobinary modulation. In this figure, two configurations of implementing a duobinary modulated link are illustrated: Fig. IA illustrates the theoretical implementation of a duobinary modulated link where the encoder and the coder are mounted on both sides of the link. The link transfers a three levels signal. Fig. IB illustrates the theoretical implementation of a duobinary modulated link where the encoder and the coder are mounted on one side of the link. The link transfers two levels signal.
A partial response encoded sequence c(k) is related to the binary data sequence d(k) by the following encoding rule:
The encoding rule for a special code is fully described by m coefficients O1n. The duobinary code has n=2 coefficients which means ao = aj = 1, which means that a binary signal is encoded to a three level duobinary signal by adding the current and the bit and the previous bit: c(k) = d(k) + d(k - ϊ) (2)
The output is an analog summation causing c(k) to be a three levels signal 0, 1, and 2 usually denote as -1; 0; 1.
The importance of the three values sequence is that it is a correlated signal, and thus, all possible combinations of the three values cannot occur. For example, the output sequence c(k), cannot contain a 1 followed by a -1, or a -1 followed by a 1. A 1 and a -1 will always have a 0 between them. Similarly, the combinations {1 0 1} or a {-1 0 -1} also never occur at the c(k). Only a {-1 0 1} or a {1 0 -l} can occur.
Two types of decoders are provided in order to decode the duobinary modulated data properly, as described in Fig. 1. Properly decoding/encoding means
minimum Bit Error Rate (BER) after the data has passed the link and recovered at the receiving end.
The first type, which is not being used anymore in optical communication, but works theoretically and may implement in other types of link, is illustrated in Fig. IA. Three main reasons prevent using this communication method in optical communication:
1. The need to change both sides of the link;
2. Transmitting three optical levels, and
3. A catastrophic decoding error.
In the method of Fig IA, a single error at the receiver will propagate forever, causing a catastrophic decoding error (see Fig. 4). Fig. IB illustrates a differential encoder used as a precoder. Such configuration avoids a recursive decoding. To avoid a change on both sides of the link, it is better to move this differential decoding to the transmitter and differentially precode the data. The precoding rule for duobinary coding is: b(k) = d(k)@b(k-ϊ) (3) where d(k) is the transmitted binary data sequence; b(k) is the precoded binary sequence, and θ is the logic XOR.
Due to use of a precoder and an encoder in the transmitter, decoding in the receiver is very simple. The data bit d(k) is the absolute value of the encoded symbol c(/c) e {-1,0,1} i.e. d(k) = |c(k)|. Using a Mach Zehnder Modulator (MZM) overcomes the problem of transmitting three optical levels (the MZM square the electrical driving signal while converting it into optical signal (see Fig. 6). Therefore, at the receiver side direct detection photo diode receivers for binary signalling may be used without any modification (the same receiver being used for IMRZ communication).
Fig. 2 depicts graphs showing transmitted bandwidth comparison between NRZ and duobinary modulation. The figure shows the differences between the NRZ
modulation and a duobinary modulation. As can be seen the duobinary modulation bandwidth is about half the bandwidth needed for NRZ modulated signal. The duobinary spectrum is the spectrum of the transmitted signal multiplied with
COS (θ)T 12) , which causes a zero at half of the data rate 1/T, where T is the bit duration. The reduced spectral occupancy by a factor of approximately 2, results in less dispersion sensitivity.
Fig. 3 shows the frequency response of Bessel filters of the 5 and the 9 order. In both filters the 3 dB attenuation (6 dB for voltage response) frequency is at 2.5 GHz. The response is identical up to 6 GHz and the voltage attenuation at this frequency is 28.5 dB. This explains why increasing the filter order will not increase attenuation of the high frequency harmonics of the data. The maximum fundamental frequency component in NRZ data transmission is 5 GHz.
Fig. 4 shows the waveforms along the duobinary generation block and the difference in the behaviour of a duobinary communication system without and with a precoder, as was shown in Fig. IA and Fig. IB respectively.
Figs. 4a through Fig 4c depict the encoding process of a NRZ signal into duobinary signal;
Fig. 4d depicts the decoding result of duobinary signal into NRZ signal using the decoding rules illustrated in Fig. IA;
Fig. 4e depicts the received signal, which includes one error, compare to the signal which was transmitted as indicated in Fig. 4c;
Fig. 4f depicts the sequence of errors being generated when decoding the signal of Fig. 4e. As can be seen the single error caused five consecutive errors;
Figs. 4g through Fig. 41 show the precoding, encoding and decoding process of converting NRZ signal into duobinary signal and back to duobinary, and
Figs. 4m and Fig. 4n depict the consequences when an error occurs while generating the transmitted signal illustrated in Fig. 41. Only one error is generated for one encoded/decoded error signal.
Fig 5 shows a practical block diagram of duobinary transmitter. Fig. 5A illustrates a Duobinary transmitter using a linear/limiting amplifier and Fig. 5B illustrates a duobinary transmitter using a limiting amplifier. The use of a linear/ limiting amplifier or limiting amplifier depends on the Low Pass Filter (LPF) location with respect to the amplifier. If the filter is located at the amplifier's output then an amplifier should be of a limiting amplifier type (as shown in Fig. 5B) and a linear/ limiting amplifier will be used when the filter is located at the amplifier input. The reason for that is the fact that the LPF (with the proper bandwidth, e.g., with a typical 2.8 GHz for duobinary modulation) generates the three levels signal required to drive the MZM. In event that the LPF is located at the amplifier input then only a linear/ limiting amplifier has the capability to maintain the required three levels.
The method of converting the electrical signal into optical signal is shown in Fig. 6. The MZM is biased at Vπδ and the driving voltage is amplified so that its Peak To Peak (PTP) value would be 2 Vπ5 as indicated in the figure. 100% of the transmitted power is achieved when the data is "-1" and "1", while zero power is transmitted for a "0". The information in the phase of the electrical field is lost. The result (in the optical domain) is exactly the square of the electrical field, meaning the transmitted power.
Fig. 7 demonstrates the electrical and optical waveforms in a conventional duobinary transmitter. The LPF forms a three level duobinary signal while narrowing the bandwidth of the transmitted signal. The figure shows the effect of a 5 pole Bessel LPF with a typical 3 dB attenuation frequency at 2.8 GHz. Narrowing the bandwidth (~ half of the equivalent NRZ transmitter) increases the link tolerance to chromatic dispersion significantly. Since the dispersion is relative to the square of the transmitted bandwidth it turns out that the dispersion tolerance increases with duobinary modulation approximately 4 times, meaning the dispersion limited transmitting distance will increase from ~ 65 km to ~ 260 km.
Finally, Fig. 8 presents a prior art NRZ and duobinary optical eye pattern, illustrating the differences between the NRZ and duobinary optical power eye
diagram. A transmitted "0" in a conventional Duobinary signal has more optical power compared to the transmitted "0" in the NRZ format. Disclosure of the Invention
This invention is directed to duobinary modulation techniques, which increases the bandwidth efficient modulation format at channel rates of 10 Gbps and 40 Gbps.
Although the duobinary modulation is more tolerant to dispersion effect in the fiber, it is more sensitive (compared to NRZ modulation) to optical noise results in lower values of Optical signal to noise ratio (OSNR).
It is therefore a broad object of the present invention to ameliorate the drawbacks of the prior art optical duobinary transmission systems and to provide an optical duobinary system and method for increasing the transmission distance limited by low OSNR value.
It is a further object of the present invention to provide an optical duobinary transmission system and a method for increasing the transmission distance limited by low OSNR value by utilizing a modified waveform of an ODB signal.
It is still a further object of the present invention to provide an optical duobinary transmission system and a method for increasing the transmission distance limited by low OSNR value by reducing the optical power level of the transmitted logical "zero" while maintaining the optical power level of logic "one" substantially the same.
In accordance with the present invention there is therefore provided a system for increasing transmission distance limited by Optical Signal to Noise Ratio (OSNR), comprising an optical duobinary transmitter including a duobinary precoder, a Low Pass Filter (LPF) producing a duobinary signal and a modulator; at least one amplifier and a voltage level suppressing logic "zero" voltage unit connected between said LPF and modulator for modifying the waveform of the signal fed into said modulator.
The invention further provides a method for increasing transmission distance limited by low OSNR value, comprising providing a duobinary transmitter having logic "zero" and logic "one" levels, an optical duobinary transmitter including a
duobinary precoder, a Low Pass Filter (LPF), a modulator, at least one amplifier and a voltage level suppressing logic "zero" voltage unit, and reducing the optical power level of the transmitted logic "zero" while maintaining the optical power level of logic "one" substantially the same. Brief Description of the Drawings
The invention will now be described in connection with certain preferred embodiments with reference to the following illustrative figures, so that it may be more fully understood.
With specific reference now to the figures in detail, it is stressed that the particulars shown are by way of example and for purposes of illustrative discussion of the preferred embodiments of the present invention only, and are presented in the cause of providing what is believed to be the most useful and readily understood description of the principles and conceptual aspects of the invention. In this regard, no attempt is made to show structural details of the invention in more detail than is necessary for a fundamental understanding of the invention, the description taken with the drawings making apparent to those skilled in the art how the several forms of the invention may be embodied in practice.
In the drawings:
Figs. 1 to 8, are respectively, block and waveform diagrams showing prior art ODB transmitters and transmission signals;
Fig. 9 is a block diagram of an embodiment of an ODB transmitter and waveforms according to the present invention;
Fig. 10 is a circuit diagram of the transmitter of Fig. 9;
Figs. 11 and 12 are block and circuit diagrams illustrating another embodiment of an ODB transmitter and waveforms according to the present invention;
Figs. 13 and 14 are block and circuit diagrams of a further modification of the transmitter of Figs. 9 and 10;
Figs. 15 and 16 are block and circuit diagrams of still a further embodiment of an ODB transmitter and waveforms, according to the present invention;
Fig. 17 shows an NRZ and optical eye pattern resulting from use of the system and method according to the present invention, and
Fig. 18 is a graph showing the spectrum of a duobinary signal according to the present invention, as compared with the spectrum of a prior art duobinary signal. Detailed Description of the Preferred Embodiments
As mentioned above, the drawback in implementing the prior art duobinary transmitter, is the fact that a logic "0" is not transmitted at zero optical power, or at least at the equivalent optical power when transmitting a "0" in an NRZ modulation. This means a reduction in the OSNR and the Extinction Ratio (ER), which reduction in these parameters reduces the transmitting distance. The present invention is a way to increase the OSNR and ER by reducing the optical power level of the transmitted logic "zero" with a minimum effect on the optical power level of the transmitted logic
"one".
The invention has its affect in the electrical domain of the transmitter due to the fact that the optical bandwidth of the duobinary modulation has been optimized as an MZM-based duobinary modulation, does not induce optical frequency deviation (chirp).
Referring now to Figs. 9 and 10 there is shown a duobinary transmitter 2, according to the present invention, which includes a precoder 4 having an NRZ data input 6 and data output 8 leading to a limiting amplifier 10 having a driver output impedance of 50 Ω. The output from the limiting amplifier 10 is fed via a DC blocking capacitor 12 to a LPF 14, the output of which is directed to a controlled voltage level suppressing unit 16, e.g., two serially connected clamping diodes (Figs. 9 to 14), or an analog switch (Figs. 15 and 16). The three-level signal exiting the unit 16 drives a MZM 18, which modulates the CW laser source 20. As can be seen in Fig. 9, the values of the optical power for "0" have been reduced and the voltage drop for the transmitted "1" and "-1" can be compensated by increasing the drive voltage.
Figs. 11 and 12 illustrate another embodiment of the duobinary transmitter of Figs. 9 and 10. Here there are provided two amplifiers, the first limiting amplifier 10
is interposed in circuit between the precoder 4 and the LPF 14 and a second linear/ limiting amplifier 22 is connected at the output of the unit 16. The gain of the first stage should be at a value so that voltage excursion at zero level will not exist. This implementation consists of a limiting amplifier as the first amplification stage, while the second amplification stage is a linear/limiting amplifier. This arrangement erases the voltage excursions at the zero level, as illustrated by the voltage waveforms.
A further modification is illustrated in Figs. 13 and 14 in which the linear/limiting amplifier 10 is interposed between the unit 16 and the MZM 18.
An embodiment showing a controlled "zero" voltage level suppressing unit 16 utilizing an analog switch is illustrated in Figs. 15 and 16. The analog switch is controlled by a control circuit 24 receiving input data from the NRZ signal while being synchronized by the same input clock of the precoder 4, thus obtaining the same zero timing at the LPF 14 output. The analog switch can be implemented by using a high speed transistor, PIN diode or any other equivalent electronic circuit.
As mentioned above, the output of the LPF is a three levels signal. The mid level (designated as "0" at the filter output waveform in Figs. 9, 11 and 13) including ripples caused by residual of the suppressed high frequency harmonics existing in the transmitted data. The high frequency harmonics suppression is advantageously performed by a Bessel LPF. Increasing the number of the filter's pole should suppress the residual value but actually is not effective due to the fact that a Bessel or a linear phase filter is used. Such filters are required due to the fact that they keep the Group Delay (GD) value of the random data to a minimum required value. The drawback of using such filters is their low attenuation rate with respect to frequency in the stop band, as shown in Fig. 3, which illustrates that no difference in attenuation exists between the 5th to the 9th order of a Bessel filter at frequencies beyond the 3 dB attenuation frequency and 6 GHz. Attenuation beyond 6 GHz has negligible suppression effect on the transmitted data. Thus, the present invention utilizes a voltage level suppressing logic "zero" voltage unit, which will prevent ripples having amplitude values below a threshold voltage of the diodes (when clamping diodes are
used) from propagating into the MZM. Likewise, the analog switch will erase any value of these ripples by shorting them to analog ground. The ripples caused by the filter will not appear in the optical domain.
The voltage drop across the clamping diodes will affect the transmitted power of the logical "1". In the implementation described in Fig 9 and 10, compensating the driving voltage for logical "1" will affect the driving voltage of the logical "0" significantly less, which means that the effect of suppressing of the voltage excursion of the driving voltage at the "0" level is maintained, in spite of changing the driving voltage for compensating the "1". In the implementation in Figs. 11, 12, 13 and 14 the voltage drop across the "1", is compensated by the second linear/limiting amplifier without any effect on the voltage corresponding to logical "0".
The above results can be seen in Fig. 17 when comparing it to the eye pattern of Fig. 8. The optical power for transmitting a logic "0" is zero.
Fig. 18 shows that the spectrum of a duobinary output signal according to the present invention is not affected as compared to a combined output duobinary signal. This means that the increased dispersion tolerance which was gained by using a conventional duobinary modulation is not affected by using a duobinary modulation, having a voltage level control unit according to the present invention.
The present invention is applicable to any communication bit rate, e.g., bit rates between 10 Gbps to 40 Gbps provided the 3 dB attenuation of the LPF is adjusted to approximately 25% of the transmitted bit rate.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrated embodiments and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.