MXNL06000103A - Optical signal phase regenerator for formats of differential modulation with phase changes - Google Patents

Abstract

The invention relates to a device for regenerating the phase of an optically modulated signal with phase changes and based on two and three replicas, wherein the replicas refer to the number of identical signals that are obtained form the input signal. This regenerator is capable of regenerating the phase and period of any format of modulation of optical communications systems which use differential modulation with phase changes, such as:DPSK, DQPSK, RZ-DQPSK, RZ-D8PSK, D8PSK, RZ-D16PSK, D16PSK. The regenerator design presented involves the regenerator being placed after the multiplexer of a communications system and before the signal modulators and/or decoders. Thus the regenerator receives the signal leaving the multiplexer and this signal is input in an amplitude modulator.

Description

Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes DESCRIPTION OBJECT OF THE INVENTION A regenerative device of the phase of an optical signal modulated with phase changes is presented. This regenerator is capable of regenerating the phase and period of any modulation format of optical communication systems that use differential modulation with phase changes.

BACKGROUND In multichannel optical communication systems, different modulation formats can be used to transmit information over long distances. In digital communication systems the modulation formats use state changes such as changing the power level of the transmitted light, changing the frequency of the light that is transmitted and also changing the phase of the light that is transmitted. Generally, three basic types of digital modulations are distinguished according to the chosen parameter: ASK Modulation (Amplitude-Shift Keying). In this modulation the carrier is passed to represent bit 1, and it is not transmitted to represent bit 0, digitally modulating the amplitude of the carrier.

PSK modulation (Phase-Shift Keying). The carrier is transmitted to represent 1 and inverted carrier in phase to represent 0, resulting in a phase jump of 180 ° in each bit transition from 1 to 0 and from 0 to 1, so it can be considered a modulation digital of the carrier phase.

FSK modulation (Frequency-Shift Keying). A frequency carrier fc is transmitted to represent bit 1 and frequency fcj to represent bit 0, producing a digital frequency modulation.

In figures 1 (a), 1 (b) and 1 (c), the waveforms of the basic types of modulation, Figure 1 (a) represents the modulation with amplitude changes, the figure 1 (b) represents the modulation with phase changes. Figure 1 (c) represents a modulation with frequency changes.

Phase Shift Keying The Phase Shift Keying (PSK) modulation format and variants of it are currently widely used in both military and commercial communication systems. The general analytical expression for PSK is described according to B. Sklar (1988) "Digital Communications: Fundamentals and Applications" (First Edition) New Jersey, Prentice Hall as: where the phase term, f. { ?), will have M discrete values, typically expressed by: For example, for the binary PSK modulation (BPSK) in Figure 1 (b), M is equal to 2. The symbol E represents the energy, T the time duration, with 0 < t = T In the BPSK modulation, the data of the signal to be transmitted are modulated in the phase changes of the wave, 5, (t), between one of these two states 0 (0 °) or (180). As can be seen in Fig. 1 (b), the graph shows a typical BPSK waveform with its abrupt phase changes in the symbol transitions; if the flow of modulated data consisted of an alternating sequence of ones and zeros, there would be abrupt changes in each transition. The waveforms of the signals can be represented as vectors in a graph of polar coordinates; the length of the vector would correspond to the amplitude of the signal and the direction of the vector; for the general case M-ary, corresponds to the phase of the signal relative to other signals M-l. For the specific case of BPSK, the vector graphic would illustrate the two vectors in opposition of phase of 180 °. The signals that can be decomposed by vectors in phase opposition are called antipodal signals. We will show this vector representation later when the DQPSK scheme is presented. The PSK modulation format is usually used to obtain a modulation format that allows more sensitive detection mechanisms within the binary modulation schemes. Next, two different modulation formats are described and explained in which the regenerator proposed in this invention can be applied. This is done with the purpose of being able to explain more clearly, later, the operation of the phase regenerator of this invention and likewise the schemes of the transmitters and receivers that integrate a multichannel communication system when modulation with phase changes is used.

There is a wide variety of modulation formats that use phase changes to transmit information. In the following sections we summarize the most modern modulation formats and in which a large number of scientific articles have recently been published. These formats work through phase changes such as Differential Phase Shift eying (DPSK) and Differential Quadrature Phase Shift eying (DQPS).

PSK Differential Detection (Differential Phase Shift Keying (DPSK)) The essence of PSK differential detection is that the identity of the data is referred to the phase changes between symbol and symbol. The data is differentially detected by examining the signal, where the transmitted signal is first differentially encoded. For the case of the DPSK modulation, the coded bit sequence, c (k), can be, generally, obtained from the following two logical equations: c (k) = c (k - \) ® m (k) or c (k) = c (k -]) ® m (k) where the symbol T represents a sum in module 2 and the over-bar denotes the logical complement. In these expressions, m (k) represents the original sequence of data to be transmitted bit by bit, c (k) represents the coded bit obtained based on the logical operations indicated by the previous equations and c (k - \) refers to the bit encoded previously obtained to the c (k) bit. Subsequently, the information of the coded signal c (k) is translated into a sequence of phase changes, where bit '1' is characterized by a phase change of 180 ° and bit O 'is characterized by a phase of 0 °. It should be noted how the process of differential coding of a baseband bit sequence prior to modulation constitutes one of the simplest forms of coding as protection against errors. The bit streams that are transmitted through Many of the communication systems can intentionally invest their value within the channel. Many signal processing circuits can not discern whether any of the transmitted bits have inverted their value or not. This characteristic is known as phase ambiguity. Differential coding is used as protection against this possibility. Next, the differential encoding process of the information bits previously to be transmitted with DPSK format is detailed by means of a numerical example. As already mentioned, a differential coding system consists of an addition operation in module two as illustrated in Figure 2; where c (k) = c (k - \) ® m (k), where: m (k), input data sequence; c (k), bit by bit encoded bit sequence; c (k - 1), coded bit obtained prior to c (k) The way in which a differential encoder operates is described below. Consider the sequence of bits that is represented in Figure 3. The described coding circuit has a reference bit that can be? ' or T indistinctly. The incoming bit to the encoder system is added to the reference bit, forming the second bit of the encoded data sequence. This bit obtained is added to the next bit of information by continuing the process described in Figure 3, such as the Charan Langton reference, "Tutorial 2 - What is Differential Phase Shift Keying?".

The decoding process performed in the receiver is the reverse process to that described above. The incoming bit sequence is added together with the purpose of recreating the original sequence of data as can be seen in Figure 4. As each bit can be displayed, it is added to the adjacent bit that has a 1 bit retrace. On the other hand, there are also two possibilities, that the bits have been transmitted correctly without errors, and in the opposite case, that the received data sequence has errors (containing bits that have inverted their value along the channel of transmission) as can be seen in figure 5.

The decoder circuit of the receiver operates in the following way, according to Charan Langton, "Tutorial 2 - What is Differential Phase Shift Keying?". for each the two possibilities shown in figures 4 and 5. In figure 4 there is a sequence of bits received without error and in figure 5 there is a sequence of bits received with error. In both cases, the benefit of using differential coding allows the original signal transmitted to be recovered.

The application of differential coding as a coding for phase changes leads to the obtaining of differential modulation formats (DPSK, DQPSK ...). The schema of a DPSK detector is illustrated in Figure 6. by the corresponding block diagram referred to by Sklar in 1988 as: "Digital Communications: Fundamentals and Applications" First Edition, New Jersey: Prentice Hall.

There are significant differences of the DPSK detector shown in Figure 6 versus a coherent PSK detector. In a coherent PSK detector, it is attempted to correlate the phases of the signal sent with a reference signal or local oscillator. Correlating the phases of two optical signals is an extremely difficult process. In fact, that synchronization of the phases of two optical signals is the main reason why coherent detection systems could not be developed to commercial equipment phases. In the case of a DPSK detector the reference signal is simply a delayed version of the previously received signal. In other words, during each symbol time, each received symbol is compared with the previous symbol and the correlation or anticorrelation between them is observed (1 80 ° out of phase). The DPSK modulation format in contrast to PSK is much less demanding than PSK since the information is encoded as a change (or no change) in the optical phase of the signal. DPSK is directly related to the systems of high transmission rates since the phase fluctuations between the bits of two signals are reduced. Although the non-synchronized demodulation of a PSK signal is not strictly possible because the information resides in the phase of the carrier signal, the detection by Phase comparison associated with DPSK reduces the synchronization problems associated with coherent PS systems.

Format with Quadrature Phase Shift Keying (QPSK) The reliable behavior of a system, represented by a low probability of error, is one of the important points to take into account in the design of a digital communications system. Another important characteristic to consider is the efficiency in the utilization of the bandwidth or spectral efficiency defined as the ratio of bit transmission between the separation between channels (or carriers) in a multichannel system. In the Quadrature Phase-Shift Keying (QPSK) format, as in the binary PSK format, the information to be transmitted is contained in the phase of the signal being transmitted. In particular, the phase of the carrier signal acquires one of the following phase values, which are equispaced, p / 4, 3p / 4, 5p I 4 and? P? radians. For these values, the transmitted signal can be defined according to Simón Haykin, "Communication systems", 4th edition, Ed. John Wiley & Sons, pp. 31 1 as: where E represents the energy per symbol of the transmitted signal and T the symbol duration. The frequency of the carrier signal fc is equal to «c / r for a fixed integer nc. Each phase value corresponds to a single pair of bits.

Spatial Diagram of the QPSK Signal By using trigonometric identities and starting from the previous equation, the energy of the transmitted signal sj can be redefined (for the interval 0 = t = T by the expression defined by Simón Haykin, "Communication systems", 4th edition, Ed. John Wiley & Sons, pp. 31 1 Following this representation, two fundamental observations can be made: There are two basic functions orthogonal to each other, y ^ 2 (0> contained in the expression of 5, (t).) Specifically, ^, (t) and f2 (?) Are defined by a pair of carriers in quadrature, as reference Simon Haykin, "Communication systems", 4th edition, Ed. John Wiley &Sons, pp. 31 1 There are four information points, which are associated with the signal vectors defined according to Simon Haykin, "Communication Systems", 4th edition, Ed. John Wiley & Sons, pp. 31 1, as: The QPSK format has two-dimensional constellations (N = 2) and four information points (M = 4). Whose phase angles increase in the direction shown in Figure 7 according to Simon Haykin, "Communication Systems", 4th edition, Ed. John Wiley & Sons, pp. 3 1 1.

As with the PS modulation format, the QPSK signals have a minimum average power.

Differential Format with Phase Quadrature known as DQPSK for its acronym in English Differential Quadrature Phase Shift Keying Since this modulation format forms the basis of the investigation, analysis, and comparison in the present invention, the details of the RZ-DQPSK modulation, described by O. Vassilieva, et al in "Non-Linear Tolerant and Spectrally Efficient 86Gbit / s RZ-DQPSK Format for a System Upgrade "(OFC 2003) and by RA Griffin, et al" Optical differential quadrature phase-shift key (oDQPSK) for high capacity optical transmission ", in Proceedings OFC '2002, pp . 367-368), where the architecture of the transmitter and receiver schemes is exhaustively described.

In the DQPSK modulation format, the information is encoded in the phase of the optical signal so that the phase can take one of these four possible values: 0, 7z72, p and 3p72 radians. Each value of the phase corresponds to a pair of bits, the symbol rate being exactly half the bit rate. This feature causes any type of DQPSK format to be particularly interesting since the effective "bit rate" of the transmission (B) requires only the use of B 12 of the electronic symbol ratio. For example, it is possible to transmit at a bit rate of 40 G bit per second with electronics that work at 20 G Hertz because in each symbol (identified by a phase change) transmitted two information bits are sent.

Generation and Detection of DQPSK Signals In DQPSK modulation format like DPSK, it is necessary to pre-encode the data in the transmitter in order to use a simple direct detection in the receiver. In the case of DQPSK, the necessary precoding function entails the implementation of a considerably more complex digital logic circuit associated with DPS. Given that it is a multiphase modulation, with four different phase levels, the precoding function will have two binary data inputs, facilitating two outputs with the data already encoded as described by RA Griffin, et al, in "Optical differential quadrature phase -shift key (ODQPSK) for high capacity optical transmission ", in Proceedings OFC '2002, pp. 367-368).

Spectrum Power Signals RZ-DQPSK The spectrum of a RZ-DQPSK signal at the output of a transmitting system, as well as its corresponding electrical signal in the receiver, can be observed in Figures 8 (a) and Figure 8 (b) respectively.

In Figure 8 (a) it can be seen that the bandwidth occupied by the modulated signal is extremely wide. Most of the high-capacity optical systems are based on wavelength multiplexing in order to obtain higher transmission rates, wavelength division multiplexing (multiplexing by division in wavelength). In this way, due to the need to multiplex several channels in the same optical link, it is required that each of them be limited in band. Therefore it is necessary to make an optical filtering. This optical filtering is carried out in the multiplexers and demultiplexers of an optical system so that the interference of one channel over the others is minimized until the inter-channel interference requirements are met.

This optical filtering causes transient responses to phase changes in the modulated signal as can be seen in Fig. 8 (b). These transients, both in phase and in power, deteriorate the reconstruction of the information at the output of the system, thus limiting the maximum transmission capacity for a given link distance. The main objective of the signal regenerator for differential phase signals, presented in this invention, is to mitigate the effects of these transients, allowing a greater transmission capacity for the same distance or a greater transmission distance for the same or greater capacity. It is important to mention that patent US 6,323,979 describes a regenerator that uses optical phase modulation, using solitotes, in a fiber optic transmission system, where the signal is modulated by a clock. There are many differences with respect to this invention in fact they are completely different. Note that the modulation format in US Pat. No. 6,323,979 is by phase distribution, solids sent is used, the phase difference between the information contained in the soliton and the signal clock is used to synchronize the clock in the receitor. Those details show that the patent is very different from ours.

BRIEF DESCRIPTION OF THE FIGURES Figure 1 (a). Basic Forms of Digital Transmission AS.

Figure 1 (b). Basic form of Digital PS transmission.

Figure 1 (c). Basic form of Digital Transmission FS.

Figure 2. Coding Scheme for DPSK.

Figure 3. Numeric Coding Example for DPSK.

Figure 4. Numerical Decoding Example for DPSK (Sequence Received without Mistakes).

Figure 5. Numerical Decoding Example for DPSK (Sequence Received with Mistakes).

Figure 6. Diagram of Blocks of a Differential Detector.

Figure 7. Phase diagram for a QPSK System.

Figure 8 (a). Signal Spectrum for DQPSK.

Figure 8 (b). Electrical signal received from DQPSK after being optically filtered.

Figure 9. Block diagram of the optical signal regenerator for two replicas. Figure 10 (a). Signal modulated in phase before being filtered optically. Figure 10 (b). Signal modulated in phase after being filtered optically. Figure 1 1 (a). Input to the amplitude modulator, optical signal. Figure 1 1 (b). Input to the amplitude modulator clock signal. Figure 1 1 (c). Output signal of the amplitude modulator. Figure 12. Definition of the ports of an optical coupler. Figure 13 (a). Output signals of the amplitude modulator: signal with phase shift of 180 degrees. Figure 13 (b). Output signals of the amplitude modulator, delayed replica T / 2. Figural 3 (c). Output signals of the amplitude modulator, sum of both signals. Figure 14 (a). Control system response: attenuation. Figure 14 (b). Control system response: gain. Figure 14 (c). Output voltage (Vout) vs Input voltage (Vin). Figure 1 5 (a). Signals at the exit of the photodetector. Figure 1 5 (b). Signals to the output of the control system .. Figure 16. Output signal of the regenerative system. Figure 17. Block diagram of the optical signal regenerator for three replicas. Figure 1 8. Clock signals and their displaced replica. Figure 19. Diagram of the fiber optic channel used. Figure 20. Scatter map of the optical communication system. Figure 21 Accumulated dispersion for different channels. Figure 22 (a). Eye of the DQPSK signal without regeneration.

Figure 22 (b). Eye of the DQPS signal with regeneration.

Figure 23 (a) Results of DQPSK, Q factor versus OSNR.

Figure 23 (b). EOP results against residual dispersion.

Figure 23 (c) EOP results against differential group delay.

Figure 23 (d). EOP results against transmission power.

Figure 24 (a) Eye of the signal D8PSK after the transmitter, that is to say 0 km.

Figure 24 (b). Eye of the signal D8PSK with regeneration at 400 km.

Figure 24 (c) Eye of signal D8PSK with regeneration 600 km.

Figure 25 (a) Results of D8PSK, Q factor versus OSNR.

Figure 25 (b). EOP results against residual dispersion.

Figure 25 (c) EOP results against differential group delay.

Figure 25 (d). EOP results against transmission power.

DETAILED DESCRIPTION OF THE INVENTION The block diagram of the optical signal phase regenerator characteristic of the present invention is shown in Figures 9 and 17, respectively. Figure 9 presents the regenerator based on two replicas. Replicas refers to the number of identical signals that are obtained from the input signal. Figure 17 shows the regenerator of signals operating for three replicas. Subsequently, the regenerator operation for three replicates is described in detail, which uses operation principles similar to the regenerator of two replicas.

In Figure 9 the regenerator has two inputs the first of which is an optical signal (A l), the second is an electrical control signal (B l). The electrical control signal (131), is generates when a clock signal (5) is sent to the time retarder (6) which is controlled by an Electronic Control Circuit (7) so that the high clock level coincides with the maximum value of the envelope and with the most stable value of the phase for each symbol. The Electronic Control Circuit (7) measures the quality of the signal received at the receiver (1 8) and sends a signal to the temporary retarder (6). The signals A and B l, enter the Amplitude Modulator (8) and the output signal (801) is sent to an optical coupler (9) that is connected to a 180 ° signal shifter (10) and a temporary retarder of half period (1 1) the signals of each one, later pass through an optical coupler (12) that generates two output signals, one of them is directed to an optical amplifier (1 3), generating the output signal (1301) ) that can follow two routes, the first of which is to pass through optical fiber (14) and reach the amplitude modulator (15); and the other route is to go through a PIN detector photo (16) whose function is to convert the optical signal (1301) to an electrical signal (1601), which passes through a level inverter (17) and the amplitude modulator ( fifteen). Both routes result in an optical signal with differential modulation (1 01), which enters the Receiver (1 8).

Continuing with the description of the regenerator of Figure 9 below we will show step by step the regeneration process for two replicas. Later, the behavior of the regenerator is evaluated by simulation using a very realistic optical communication system. In the results presented below it is verified that the system is capable of improving reception not only for DQPSK but also for another type of phase modulation known as D8PSK.

In Figure 10 (a), an ideal signal modulated in phase with an excessively high wavelength in relation to the symbol time is shown. This figure helps us to understand what happens with the phase and the envelope (Power) of the carrier signal. On the other hand, in Figure 10 (b) the signal of Figure 10 (a) is shown but now under the effect of optical filtering. As you can see the effect of the optical filtering severely damages the signal because it produces changes in the envelope as well as smoothing in the phase changes. Optical filtering is essential in multichannel optical communication systems because multiplexers and demultiplexers filter signals as published by G. A. Castañon, et. al, in "Requirement of filter characteristics for 40 Gbit / s-based DWDM systems", in Proceedings ECOC'2001, Vol. 1, pp.60-61. and in "Impact of Filter Dispersion Slope in NRZ, CS-RZ, IMDPS and RZ formats on Ultra High Bit-Rate Systems", in Proceedings of the European Conference on Optical Communications ECOC 2002, Copenhagen Denmark, September 8-12, 2002).

It is important to mention that the regenerator shown in Figure 9 must be placed after the multiplexer of a communication system and before the optical decoders of the signal. The signal (A l) that is indicated in Figure 9 is input to the regenerator and this signal is obtained from the multiplexer of a multichannel optical system. This signal (A l) is input to an amplitude modulator (8) controlled by a clock signal (5) with duration of T / 2. It is worth mentioning that T is the period of a symbol. To obtain the desired result, you have to synchronize the input clock to the amplitude modulator (8) with the optical signal (A l) so that the high level interval of the clock matches the center of the symbol as shown in Figure 1 1 (a) where the input signal to the regenerator, and you can see the falls of the level of the envelope, and smoothing in the phase of the signal. Figure 1 1 (b) shows the clock signal in which it can be seen that the high level coincides with the least damaged sections of Figure 11 (a). If we observe the envelope representation / phase of the signal, it is observed that the high clock level coincides with the maximum value of the envelope and with the most stable value of the phase for each symbol.

Figure 11 (c) shows the output signal of the amplitude modulator. It can be seen that even the envelope shows abrupt changes and it can also be observed that we have quite abrupt changes in the phase of the signal, clearly distinguishing the phase level of the signal. In this way, the transients caused by the filtering have been eliminated, leaving only those intervals where the optical signal shows more stable behavior. This is the type of output that is obtained after the amplitude modulator indicated by the index (801) in Figure 9.

Once the interval of the optical signal has been obtained where the information is less damaged, figure 11 (c). The signal that leaves the amplitude modulator (801) is replicated by means of an optical coupler (9) with two outputs that can be seen in Figure 12. If we cancel the entry in the input port called reflection, in each of the outputs there is a copy of the input signal. In the direct port, defined in Figure 12, there is an identical phase signal but with power 3 dBs lower with respect to the input power. In the coupling port, in addition to having a loss of 3 dBs, there is an additional offset of 90 ° with respect to the input signal.

Each of the replicas is processed separately in Figure 9. The signal obtained in the direct port must be displaced for a time equivalent to half a symbol period, that is, T / 2 (ll) is delayed, so that it completes the original signal at the moment of joining the two replicas in the optical coupler (12). The other signal that comes out of the coupling port must be offset 1 80 ° (10) to compensate for the phase shifts introduced by both the first coupler and the second. Each of them introduces a 90 ° phase shift, so the total phase shift must be 180 °. Finally both processed signals are input to the inputs of the coupler (12) to obtain the full period in the direct port of the second coupler. The second coupler (12) what it does is add the two input signals together. The two signals produced by the first coupler (9) are shown in Figure 13 (a) and Figure 13 (b). Note that the signal in Figure 13 (b) shows a delay of T / 2, Figure 13 (a) shows the signal with 180 ° phase shift and Figure 13 (c) shows the sum of the two signals and this sum is produced by the second coupler (12). Note that the phase of the signal can be regenerated perfectly but even the envelope shows transients. This is the signal that is obtained at the output of the second optical coupler.

To eliminate the effect of the transients of the envelope (power transients), a power control is carried out that maintains the level of the envelope constant at the output. Considering the difficulty of implementing optical gain amplifiers variable, a system based on variable attenuation is used and signal sections with excessively high power are attenuated more than those with lower power levels.

Next, the control signal for the variable attenuator is described, following with Figure 9 To do the above, we must first amplify the signal to ensure that the level of the envelope is greater than the level required at the output including the minimum of the envelope. To increase the power level, an optical amplifier (13) doped with erbium is used. After amplifying the optical signal it is divided into two and one of the outputs of the optical splitter is connected to a photo-detector (16). The photo-detector (16) will give an electrical signal proportional to the envelope of the optical signal at the input, as shown in Figure 1 5 (a). This electrical signal is processed to obtain the signal to be input to the corresponding amplitude modulator. The control system (17) is composed of a scale element and a level inverter (l / x) limited so that low powers do not cause excessively high peaks. The curves of Figure 14 (a), Figure 14 (b) and Figure 14 (c); define the behavior of the control system as a function of the output voltage of the photodetector. Figure 14 (a) shows the attenuation, Figure 14 (b) shows the gain curve and Figure 14 (c) shows the output voltage (Vout) against the input voltage (Vin). This level inverter is responsible for the control system (17) present the curves described in Figure 15 (b). Bearing in mind that the amplitude modulator (15) of Figure (9) is a passive element, the control signal can not exceed the unit value, so we must limit the inverter to ensure that it never gives an excessive value. To compensate for the delay produced in the control system (17), an optical fiber section (14) is introduced whose function is to delay the signal the electronic processing time so that the control and optical signal are synchronized in the second amplitude modulator (15).

As can be seen in Figure 15 (a), the signal at the output of the photodetector (16) corresponds to the envelope of the optical signal in which the variation of the signal. The control signal obtained from it, Figure 15 (b), shows that for the lower power levels, the control signal has a maximum value which means a minimum attenuation while for high power values, the signal of control is minimal, which means greater attenuation.

The resulting control signal (1701) will be injected at the electrical input of the respective amplitude modulator (15) of Figure 9 so that the average output power of the optical signal is uniform. Obtaining as a result the signal shown in Figure 16.

The output signal of the regenerative system, Figure 16, shows a signal practically identical to the input signal to the optical communication system and shown in Figure 10 (a). As can be seen in Figure 16 the output signal of the regenerator is very good, however, we must take into account problems not considered. Certain non-idealities such as the non-ideal overlapping of the replicas or certain power losses, can make the behavior of the regenerator not perfect and the reconstruction is not very good. However, the regenerator in general terms improves the input signal. To show the limits of the operation, results obtained by means of a simulator specialized in optical communications are presented later.

The following describes the architecture of a regenerator with three replicas: In the previous section an ideal phase regenerator based on two replicas was presented. However, one can think of generalizing the system for a larger number of replicas such as the regenerator shown in Figure 17.

The working principle of the regenerator based on three replicas is slightly different from that of two replicas explained above.

This regenerator with three replicas, (like that of two replicas) has two entries; the first is an optical signal, called (Al) and the second is an electric control signal called (B l). The electrical control signal (B l) is generated when a clock signal (1 9) is sent to the time delay (20) which is controlled by an Electronic Control Circuit (21) so that the high clock level coincides with the maximum value of the envelope and with the most stable value of the phase for each symbol. The Electronic Control Circuit (21) measures the quality of the signal received in the receiver (36) and sends a signal to the temporary retarder (20). The signals A l and B l, enter the Amplitude Modulator (22) and is modulated by a clock signal (19) whose duration is T / 3. Period T is the period of a symbol. What the amplitude modulator does is take a third of the signal symbol. As in the case of the regenerator of two replicas, what is taken from the input signal is the central part of a symbol where the phase information of the symbol is less damaged.

The replica or copy of the signal (2201) generated by the Amplitude Modulator is sent to the first optical coupler (23) which produces two replicas. The first replica is simultaneously sent to a second coupler (24) which generates a third replica in addition to passing the second input replica. Then the reconstruction of the signal is done by means of two couplers (28 and 29) in cascade as can be seen in figure 17.

To reconstruct the output signal correctly, the three replicas are added where: the signal (2301) of the first replica is not delayed; the second replica (2401) is delayed one third (25) of symbol period; the third replica (2402) is delayed two thirds (26) of period. In addition, the first replica (2301) is not offset because it crosses two couplers (23 and 29) and in none of them there is gap, because in both enters the port of entry and exits through the direct. The second replica (2401), crosses the four couplers (23, 24, 28 and 29) so it has a loss of 12 dBs (3 dBs for each coupler) and the configuration used only in two of them is offset 90 ° (23 and 29), that is, a total phase shift of 1 80 ° (27) is required to compensate for the phase shift introduced by the couplers (23 and 29). The third of the signals (2402) also crosses the four couplers (23, 24, 28 and 29) and suffers 12 dB-s of losses. However, in this case, the four couplers introduce a phase shift of 90 ° each, so that the total phase shift is 360 ° and, therefore, the signal will not have to be compensated with an extra phase shift. Both the second and the third replicas go through four couplers while the first crosses only two. This implies that the first of the replicas suffers 6 dBs less losses than the other two. Therefore, 6 dBs should be attenuated as indicated in element (30) so that the three signals have the same power level at the output of the couplers.

For both replicates as well as for three, the process of combining the replicas can be problematic because it can not be guaranteed that an entire number of periods of the optical signal will be replicated. However, this will have no effect on the output since as it is using differential modulation the resulting signal is obtained from the phase difference between two symbols. The discontinuity in the phase, produced by the combination of the replicas in the couplers, of a symbol is compensated by the discontinuity of the subsequent symbol.

Clock signal Although the clock signal (19) of figure 17, it can be understood as a train of rectangular pulses, for sufficiently high transmission rates that they have to consider the effects due to the finite rise slope of the amplitude modulators. To compensate for the effect of the finite slope of rising and falling, replicas have to overlap. The following table 1 shows the rise time of the amplitude modulators of the presented regenerators. These rise times can be relatively easily reached by commercial amplitude modulators. The table shows the ideal case and also the rise time required for the presented regenerators to work perfectly. The required rise and fall time is 3 picoseconds, in the cases presented of two and three replicas.

Table 1. Parameters of the clock signals to be used.

Synchronization of the clock signal for 2 and 3 replicas.

The synchronization of the clock signal with that of symbol period information in the optical carrier is implemented by means of an Electronic Control Circuit (2 1) in Figure 19 and (7) in Figure 9, which as a function of the signal Optical (A l) control the start instants of the pulses of the clock signal. This is not a synchronization with the phase of the optical signal as would be done in a coherent detection system, but the clock signal is synchronized with the symbol period T which contains the phase information. What is desired is to take T / 2 (in the case of two replicas) of information from the central part of the symbol. This central part of the symbol is the part where the phase information of the symbol is less damaged. It is important to consider that the regenerator is in the terminal part of an optical communication system, that is, before the demodulators and detectors. From the signal received in the detectors, the clock of the signal can be extracted electronically and that clock can be sent back to the first amplitude modulator of the regenerator so that it takes the T / 2 samples of each symbol. The same can be done with the second option of regenerator, of this invention, which is that of three replicas. As seen in Figure 9 and 17 the time delay is controlled by an Electronic Control Circuit (7) and (21) respectively that measures the quality of the signal received in the receiver (18) and (36) respectively and sends a signal to the retarder of the clock signal so that the high clock level coincides with the maximum value of the envelope and with the most stable value of the phase for each symbol.

Phase Regenerator results for differential modulation formats with phase changes.

The optical transmission system used in the simulations is presented in Figure 19 in which the different transmission sections divided by the optical amplifiers are shown. Twelve sections are used to check the operation of the system, which implies a total transmission distance of 600 km.

As is known, one of the main effects that limit an optical transmission system is the chromatic dispersion. That is, we must minimize the residual dispersion in order to achieve higher transmission rates. To reduce the residual dispersion, dispersion compensation schemes are applied with the aim that the residual dispersion or dispersion at the end of the transmission system is as small as possible.

To compensate for chromatic dispersion, what is proposed is to use the strategy of pre-compensation (39), online compensation (40), and post-compensation (41). For the pre-compensation stage, we compensate 30% of the dispersion of a single-mode fiber section at the beginning and the remaining 70% at the end in the post-compensation stage. The online compensation sections compensate 100% of the dispersion of the single mode fiber. Figure 20 shows the dispersion map strategy used in the present invention. As you can see in the figure, the objective of this dispersion map is to make the residual dispersion zero at the end.

This dispersion compensation scheme has been widely used by different technologies, however, it only cancels the dispersion for the central channel, the channels that are at the ends of a multi-channel system see a non-zero dispersion. The dispersion is greater the greater the distance (in frequency) to the central channel and also the dispersion is greater the greater the transmission distance as shown in figure 21. In figure 21, the residual dispersion is represented or cumulative for different channels in the same optical fiber, with the central curve corresponding to the central channel and the remaining corresponding to the channels at the ends.

There are several strategies to solve this residual dispersion problem for channels that are located away from the center frequency in a multi-channel communication system. One solution for high-speed systems is to use variable dispersion compensation per channel for the affected channels. This scheme adds more cost to the transmission system and we must try to avoid them, however there are cases where it is necessary to implement it.

Fibers used: The fibers used are presented in table 2. There is information about single-mode transmission fiber and also about the compensating fiber.

Table 2. Basic parameters of the fibers used Channel bandwidth and channel separation To test the feasibility of the present invention we use nine transmission channels placed at the transmission frequencies recommended by the ITU. The separation between channels that we used was 50 GHz. The optical filters that we use in the simulations for the multiplexers and demultiplexers have a third order Gaussian function. The bandwidth of these filters was optimized for the different modulation formats used and it was ensured that the filters did not produce a interference between channels greater than 25 dBm. In other words, it was sought that the bandwidth of the filters first met the separation between channels and also that the optical filter was narrow enough so as not to cause significant interference to the adjacent channels.

Figure 22 (a) and Figure 22 (b) show results of the opening of the signal eye for the DQPSK modulation format. The bit rate used is 66 Gbit / s. The separation between channels is 50 GHz and then the system has a spectral efficiency of 1.2 bits / s / Hz. The filters used in the simulation for the multiplexer and demultiplexer have a Gaussian transfer function of the third order and a bandwidth of 32 GHz to 3 dB with respect to the peak. Figure 22 (a) is the eye opening of a signal without the regenerator and Figure 22 (b) is the eye opening of the signal with the regenerator running. Note that the opening of the eye when the regenerator is used is better compared to the eye opening when the regenerator is not used. To obtain these results, a transmission distance of 600km was used. The power injected to the transmission fiber is 2 dBm and the power injected to the dispersion compensating fiber is -5 dBm. A pseudo random sequence of 23 1 bits was used in the simulation. It was also checked that the output signal of the decoders was equal to the input signals.

Figure 23 (a) shows results of the Q factor versus the Optical Signal to Noise Ratio (OSNR) parameter of the system with regeneration and without regeneration when white noise is added to the system from optical amplifiers doped with erbium (Erbium Dopped Fiber Amplifiers). The transmission distance is 600km. To obtain different values for the optical signal to noise ratio, what was done was to change the noise figure factor of the last amplifier. By changing that noise figure factor one can control the amount of white noise that an amplifier generates and thus vary the parameter of the signal to noise ratio OSNR. The Q factor was calculated using the following equation Q [dB] = 20log [^ | -p.o) / s, + s?] Where μ, y μ? are the average voltages for an l 's and 0's, also s, and s? are the standard deviations of the voltages of l 's and O's, respectively. Figures 23 (a), 23 (b) 23 (c) and 23 (d) show results when not used and when the regenerator is used, indicated with a 2R. The reason for putting a 2R is because the regenerator presented in this invention performs two types of regeneration. The first is that the phase of the signal is regenerated and the second regeneration is with respect to which the duration of the symbol is also regenerated. where the signal contains the same phase. Note that the results, of the Q factor, using the regenerator are 2 dB higher with respect to the results when the regenerator is not used. This improvement in performance is one of the main advantages of the use of the phase regenerator. The presented regenerator improves the signal even when the noise of the optical amplifiers is high.

Another important parameter to evaluate in a communication system is the impact of chromatic dispersion. It is very important that a modulation format is tolerant to the chromatic dispersion introduced by the optical fiber and optical filters of the ultiplexers and demultiplexers of a communication system. Figure 23 (a) to Figure 23 (c) show results of the DQPS modulation format of the center channel of a nine-channel system after a transmission distance of 600 km of the nine channels. The power used for the single-mode transmission fiber is 2 dBm. If, for example, a tolerance of 0.5 dB is assumed as a limit for the residual dispersion, then when the regenerator is used, there is a residual tolerance of 50 ps / nm. It is worth mentioning that the eye opening penalty (Eye Opening Penalty) is defined as EOP = -20log (Et / Eb). Where Et is the eye penalty in the destination and E is the eye penalty in the origin of the transmission. Figure 23 (c) shows results of the dispersion by polarization mode. The results are presented based on the differential by group retracement. To obtain these results, we considered the case where the light of the signal is equally divided into the two main polarization states of the optical fiber. In Figure 23 (d) the eye penalty against transmission power is presented to analyze the non-linear effects. As is known, non-linear effects increase with the transmission power used in the channels of the optical system. The eye penalty shown is for the central channel of a nine-channel system. To obtain these results I consider a 600 km system. The results include the non-linear effects known as phase-to-itself modulation, cross-phase modulation, and mixing of four wavelengths. To obtain results of the effects of residual dispersion, dispersion by polarization mode, and non-linear effects in an isolated manner, the white noise of the amplifiers was suppressed to obtain these results.

Figures 24 (a), 24 (b) and 24 (c) show the results of the D8PS modulation format. Note that this is a different format than the DQPSK. In D8PSK, 3 information bits are transmitted per symbol. For example, if we consider that the symbol ratio is 21.5 GHz, then the bit ratio is 3x21.5 = 64.5 Gbit / s. D8PS is a very recent modulation format. One of the advantages of this format is that it can transmit more bits of information per symbol, however that warrants 8 possible phase changes. By having more levels for the phase, the non-linear effects of the transmission system mainly affect this modulation format In these figures 24 (a), 24 (b) and 24 (c) it can be seen that the regenerator makes possible transmission distances of 400 km, figure 24 (b) and 600 km, figure 24 (c) with a signal eye completely open. If the regenerator is not used, the eye of the signal appears completely closed at a distance of 400km, however with the regenerator it is possible to obtain a completely open eye.

Figures 25 (a), 25 (b), 25 (c) and 25 (d) show the results of the D8PS modulation format at 64.5 Gbit / s. Note that Fig. 25 (a) shows results of the three output signals obtained after remodulating the D8PSK signal. This figure shows results of the Q factor as a function of the optical signal-to-noise ratio (OSNR for its acronym in English of Optical signal to noise ratio). Note that the results show an excellent Q factor when the OSNR is 13. 1 dB. However, it can be seen in the eye diagram that there are lines that cause the eye to close to this same OSNR of 13.1 dB for the third signal. Note that most of the high-level lines or "1" open the eye very well, however, there are 2 lines that degrade the eye of the signal. This has as a consequence that the The signal eye is not 100% reliable and a OSNR limit of 15 dB OSNR has to be set where it is shown that the eye is completely open. Figure 25 (b) shows the results of the impact of chromatic dispersion.

To obtain these results, a transmission distance of 600 km and a transmission power of -2 dBm were used for the monomode transmission fiber. If an eye penalty of 0.5 dB is considered as a limit, then the third signal has a range for the residual dispersion of 35 ps / nm. Results of the dispersion by polarization mode is another parameter to analyze. Figure 25 (c) shows results for the three output signals of the modulators. The third signal and the second give the worst results because it requires more electronic processing to obtain them. Results of regenerator performance with respect to non-linearities of the fiber are shown in Figure 25 (d) where we have the eye penalty against the signal strength. This figure presents results of the central channel of a nine-channel system. As is known, the central channel in a multi-channel system suffers the most from the four-wave mixing effect and that is why only the central channel results are presented, that is, the worst case.

Claims (29)

1. CLAIMS Having described my invention enough, I consider it as a novelty and therefore claim as my exclusive property, what is contained in the following clauses: 1 . Phase Regenerator of Optical Signals characterized by processing two identical signals or replicas and three identical signals or replicas using Differential modulation with Phase Changes where for 2 replicas it uses: a) an amplitude modulator having as inputs an optical signal and an electrical control signal, where the electrical signal is composed of a clock signal with a time greater than T / 2 and a time delay. The optical signal is transmitted to an optical coupler, which in turn is connected to an optical time retarder and an optical phase shifter with a 180 ° phase shift, subsequently passing through an optical coupler and an optical amplifier. The output signal of the optical amplifier is sent in turn through optical fiber to the amplitude modulator and to a PIN detector photo whose function is to convert the optical signal to an electrical signal, which passes through a level inverter. This electrical signal is used to control the amplitude modulator where the optical signal passes and this modulates the signal to eliminate the power transients of the optical signal resulting in a regenerated optical signal with differential modulation with phase changes. Where for 3 replicas you use: b) an amplitude modulator at the input which is modulated by means of a clock signal whose duration is T / 3. The optical signal is sent to the first coupler which produces two replicas. The second replica is simultaneously sent to a second coupler which generates a third replica in addition to passing the second input replica. After the reconstruction of the signal is done by two cascade couplers, where the three replicas are added and one of the replicas is not delayed; another one is delayed third of the symbol period and the third replication is delayed by two thirds of the period. The signal reconstructed by the amplifiers in cascade is amplified and this signal is divided into two where a part of the signal is sent to a PIN detector whose function is to convert the optical signal to an electrical signal, which passes through a level inverter . This electrical signal is used to control the amplitude modulator where the optical signal passes and this modulates the signal to eliminate the power transients of the optical signal resulting in a regenerated optical signal with differential modulation with phase changes. 2. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) characterized in that the signal coming from a multiplexer of a multi-channel optical transmission system is introduced. The signal is input to an amplitude modulator controlled by a clock signal with duration of T / 2. where T is the period of a symbol and to obtain the desired result, it is necessary to synchronize the input clock to the amplitude modulator with the input optical signal so that the high level interval of the clock coincides with the center of the symbol. 3. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) characterized in that each of the replicas is processed separately, and the signal obtained in the direct port is displaced a time equivalent to half period of symbol T / 2. and the signal that leaves the coupling port of the first coupler, is 180 ° out of phase. 4. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) characterized in that each optical coupler introduces a phase shift of 90 ° so that the total phase shift must be 1 80 °. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) characterized in that both processed signals are introduced to the coupler inputs to obtain a complete period in the direct port of the second coupler. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1, characterized in that the second coupler adds the two input signals together. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 b) characterized in that the first replica does not phase out, the other two created replicas, crosses the four couplers so it has a loss of 12 dBs (approximately 3 dBs for each coupler) and for the configuration used only in two of them is offset, that is, a phase shift of 1 80 ° .. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 b) characterized in that the third of the signals passes through the four couplers and suffers 12 dB-s of losses, where the four couplers introduce a phase shift of 90 ° each, with which the total phase shift is 360 ° and, therefore, the third replica of the signal will not have to be compensated with an extra phase shift. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 b) characterized in that both the second and third replicas pass through four couplers while the first crosses only two. This implies that the first of the replicas suffers 6 dBs less losses than the other two. Therefore, it must be attenuated so that the three signals have the same power level at the output of the couplers. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 b) characterized in that for 3 replicas the output signal must be reconstructed correctly by adding the three replicas where: one of the replicas is not delayed; another one third of symbol period is delayed; the third reply is delayed by two thirds of the period. Where the first replica does not get out of phase because it crosses two couplers and in none of them there is a gap, because in both it enters through the entrance port and exits through the direct and the second replica, it crosses the four couplers so that it has a loss of 12 dBs (approximately 3 dBs for each coupler) and for the configuration used only in two of them it is phase-shifted, giving a phase shift of 1 80 °. and the third of the signals crosses the four couplers suffering a loss of 12 dBs. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 b) and characterized in that the four couplers introduce a phase shift of 90 ° each, so that the total phase shift is 360 ° and, therefore, the third signal will not have to be compensated with an extra phase shift, and both the second and third replicas pass through four couplers while the first one crosses only two. This implies that the first of the replicas suffers 6 dBs less losses than the other two. Therefore, it must be attenuated so that the three signals have the same power level at the output of the couplers. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) for two replicas as 1 b) for three replicas, where the synchronization process of the clock signal controlling the first modulator will be controlled by an electronic circuit that measures the quality of the signal received in the receiver and sends a signal to the retarder of the clock signal so that the high clock level matches the maximum value of the envelope and with the most stable value of the phase for each symbol. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b) characterized in that the electronic control is composed of a scale element and a limited level inverter (1 / x) so that low powers do not cause excessively high peaks. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b) characterized in that the regenerator must be placed after the multiplexer of a communications system and before the demodulators or decoders (receivers ) Of the signal. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b) characterized in that the input signal to the regenerator, the envelope / phase representation of the signal, it is observed that the level Clock height matches the maximum value of the envelope and the most stable value of the phase for each symbol. Optical Signal Phase Regenerator for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b) characterized in that to increase the power level an optical amplifier doped with erbium before the last amplitude modulator is used. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b) characterized in that after amplifying the optical signal it is divided into two and one of the outputs of the Optical splitter is connected to a photo-detector and this gives an electrical signal proportional to the envelope of the optical signal of the input. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b) characterized in that the received electrical signal is processed to obtain the signal to be introduced to the corresponding amplitude modulator that reduces the transients of power of the optical signal. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b) characterized in that the amplitude modulator is a passive element and the control signal can not exceed the unit value so that the inverter must be limited so that it never gives an excessive value and to compensate for the delay a fiber section is introduced which delays the electronic processing time so that the control and optical signal are synchronized in the second modulator of the signal. amplitude. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b), characterized in that the combination process of the replicas will have no effect on the output since differential modulation is being used and the resulting signal is obtained from the phase difference between two symbols so that the discontinuity in the phase, produced by the combination of the replicas in the couplers, of one symbol is compensated by the discontinuity of the subsequent symbol. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b) characterized by being used to regenerate the phase and the phase period for any differential modulation format with phase changes preferably DPS, DQPS, RZ-DQPSK, RZ-D8PSK, D8PS, RZ-D 16PS, D 16PS .. 22. Phase Regenerator of Optical Signals for Modulation Formats Differential with Phase Changes according to claim la) and 1 b) characterized by enabling the use of high spectral efficiency formats such as DPS, DQPS, RZ-DQPS, RZ-D8PS , D8PS, RZ-D 16PS, D 16PSK for transmission distances in the order greater than 600 km. 23. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim la) and 1 b) characterized in that it is capable of regenerating the phase and the period of differential modulation formats with phase changes such as DPSK, DQPSK , RZ-DQPSK, RZ-D8PSK, D8PSK, RZ-D 16PSK, D 16PSK that operate at high transmission rates of the order of 66 Gbit / s or greater. 24. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b) characterized in that it is capable of regenerating the phase and the period of differential modulation formats with phase changes as DPSK, DQPSK, RZ-DQPSK, RZ-D8PSK, D8PSK, RZ-D I 6PSK, D 16PSK operating at high spectral efficiencies of 1.2 bits / s Hz. 25. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase changes according to claim 1 a) and 1 b) characterized in that using the regenerator improves the Q factor by 2 dB with respect to the results when the regenerator is not used for the DQPSK format and for a transmission distance of 600 km. This improvement in performance is one of the main advantages of the use of the phase regenerator. The presented regenerator improves the signal even when the noise of the optical amplifiers is high. 26. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b) characterized because using the regenerator improves the Q factor significantly, so that 600 km of transmission can be achieved, with respect to the results when the regenerator is not used for the D8PSK format. This improvement in performance is one of the main advantages of the use of the phase regenerator. The presented regenerator improves the signal even when the noise of the optical amplifiers is high. 27. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b) characterized in that using the regenerator improves the Signal Eye significantly, so that the 600 Km can be reached of transmission, with respect to the results when the regenerator is not used for the D8PSK format. This improvement in performance is one of the main advantages of the use of the phase regenerator. The regenerator presented improves Signal Eye significantly in such a way that can reach at least 600 km. 28. Phase Regenerator of Optical Signals for Differential Modulation Formats with Phase Changes according to claim 1 a) and 1 b) characterized in that using the regenerator improves the quality of the signal when there is significant residual dispersion, at 600 km of transmission, the regenerator has a residual tolerance of 50 ps / nm and 35 ps / nm when using DQPSK and D8PSK respecty. The presented regenerator improves the signal even when the residual dispersion is significant. 29. A differential modulation system with phase changes characterized in that it is formed of a phase regenerator according to any of claims 1 to 28.