CN1694382A - Optical Protection Switching Method Based on Fast Tunable Coherent Receiver - Google Patents

Abstract

An optical protection exchange method of a quick adjustable coherent receiver used in fiber communication system and network includes the following steps: 1. A source node sends data to a working wavelength path and a protection wavelength path via two or more than two different wavelengths. 2. The wavelengths are routed to the adjustable receiver light port at the destination node. 3. Once the destination node detects the fault of the working path, the receiver is switched to the protection path at once to realize quick protection of 1+1.

Description

Optical Protection Switching Method Based on Fast Tunable Coherent Receiver

technical field

The invention relates to a method in the technical field of communication, in particular to a method for optical protection switching based on fast adjustable coherent receivers suitable for optical fiber communication systems and networks.

technical background

In a carrier-class network, due to the needs of its business, the requirement for network reliability must be able to reach more than 99.9999%. In this way, the survivability of the network has become an important issue in network design. The protection and recovery of optical paths play a key role in the survivability of the network. With the improvement of the speed of the optical transmission system, the data capacity carried by a single wavelength channel has also increased from the previous hundreds of megabits or several gigabits per second to tens of gigabits or even hundreds of gigabits per second. At this speed, even a relatively short-term network interruption will cause great data loss, causing very serious data loss and business interruption to the entire network. Therefore, higher requirements are put forward for the protection switching time. The 1+1 channel protection method used in the past can no longer meet the current needs. The protection and restoration of the nanosecond or even sub-nanosecond level is an inevitable choice to solve this problem. .

The existing 1+1 protection adopts two methods to switch between the working light path and the protection light path. One is to receive and demodulate both signals, and then switch them electrically. Although this method can increase the protection function of the receiving circuit failure, because the protection switching involves complex inter-board switching actions, it generally requires a switching time of more than milliseconds. In addition, this method also increases expensive electrical interfaces. Increased maintenance difficulty and expense. Another method does not use the electrical backup method, but uses an optical switch in front of the optical receiver. When a fault occurs, the switching action of the optical switch is used to switch to the protection optical path to receive signals. For example, the Chinese patent "Dynamic Optical Line Synchronous Switching Protection Device" (03254764.1) and the Chinese Patent "WDM Line Switching Ring Universal Node Protection Device" (01109025.1) use optical switches to complete optical path protection switching. Although this method overcomes some of the shortcomings of the electrical backup method, since the current commercial optical switches are mainly based on mechanical switches or micromechanical switches, their switching time can only be on the order of several milliseconds to tens of milliseconds. Even if fast photoelectric optical switches are used, they will eventually limit the transmission distance of the optical path and reduce the quality of signal transmission due to the additional insertion loss they introduce.

In order to overcome the shortcomings of these protection methods, a feasible method is to use adjustable reception at the receiving end to realize channel selection switching. The usual tunable receiver is not yet mature in technology, and there is no reliable nanosecond-level tunable filter commercially available, and the switching time still needs to be on the order of tens of nanoseconds. This invention proposes to adopt coherent adjustable receiving method for network protection. The coherent receiver has very good channel selection characteristics and fast switching characteristics. Our recent research shows that the channel switching time of coherent reception is mainly determined by the wavelength adjustment time of the local oscillator light source and the intermediate frequency bandwidth of the coherent receiver. By improving the fast adjustable The vibration light source has the ability to reach an order of magnitude less than 1 nanosecond, thereby realizing the fastest protection switching.

Contents of the invention

The purpose of the present invention is to overcome the deficiencies in the prior art, to provide a method for out-of-band optical protection switching based on fast adjustable coherent receivers, so that it can improve the flexibility and success rate of optical protection path establishment, while greatly improving the Reduce data loss and service interruption caused by switching, and finally realize error-free protection switching.

The present invention is achieved through the following technical solutions, and the present invention comprises the following steps:

① The source node sends the data to the working wavelength path and the protection wavelength path respectively through two or more different wavelengths;

②At the destination node, these wavelengths are simultaneously routed to the optical port of the adjustable receiver;

③Once the destination node detects the failure of the working path, the receiver immediately switches to the protection path to realize fast 1+1 protection.

The network independently routes these optical wavelengths to the destination node.

The destination node uses an adjustable coherent receiver to receive data. When the working path is normal, the receiver receives data from the working path. When the destination node detects that the working path fails, the adjustable receiver quickly changes the wavelength of the local oscillator. Switch to the protection channel to receive data and complete the protection exchange. The destination node includes a high-speed fault detection circuit, which will quickly trigger the switching action of the receiver after detecting a fault.

In the adjustable receiver, the local oscillator optical signal source is composed of an adjustable multi-wavelength laser, and its output wavelength can be quickly selected through a fast electrical gating signal.

In the 1+1 protection, the source node sends data to the working wavelength path and the protection wavelength path respectively through two or more different wavelengths: when there is one protection wavelength, it is 1+1 protection; when the protection wavelength is N , it is N+1 protection. These different wavelengths can be routed separately without requiring a common edge between them.

The source node uses a multi-frequency laser at its transmitting end, and uses the internal modulation of the laser or an external modulator, such as Mach-Zehnder or an electroabsorption modulator, to simultaneously modulate data to these multiple wavelengths. One is transmitted to the working path as the working wavelength, and the remaining wavelengths are transmitted to the protection path as the protection wavelength. After the wavelength is modulated, these wavelengths are separated in space by a wavelength splitting device, such as a thin dielectric film filter or pull Tapered fiber splitters, etc., arrayed waveguide gratings can be used when there are many wavelengths, and these wavelengths can be independently routed to the destination node by the optical network. Since the protection path and the working path use different wavelengths, these wavelengths can pass through the same optical fiber link in the network without interfering with each other.

Based on this difference in wavelength, it is called out-of-band protection. The in-band protection method is as follows: use the same wavelength for protection, and the wavelengths of the protection path and the working path are mutually exclusive and cannot pass through the same optical fiber link. It also requires at least There are two free fiber optic links connected to the source and destination nodes, so-called dual-homing connections. Therefore, the out-of-band protection method of the present invention can avoid these limitations, thereby increasing the flexibility of protection path routing.

At the destination node, the optical signal of the working path and the optical signal of the protection path are simultaneously input to the input end of the fast adjustable receiver, and the adjustable receiver adopts a coherent receiving method. A typical coherent reception method can adopt heterodyne reception or homodyne reception. The coherent receiving method avoids the use of a mechanical optical switch in front of the receiver, and its typical switching time can only reach the order of 10ms. Using the coherent receiving method, the switching between channels is realized by the wavelength switching of the local oscillator light source. Typically, nanosecond-level tunable lasers, electro-absorption multi-frequency lasers, or array lasers controlled by fast photoelectric effect optical switches can be used. Typical switching of these lasers The time is less than 1ns.

The time for receiving and switching the final optical channel is determined by the switching speed of the local oscillator and the IF filtering bandwidth of the receiver, which is less than 1 ns. Local oscillation switching is triggered by a high-speed fault detection circuit, which monitors the quality of the optical signal and uses it as a threshold trigger. The threshold is set near the sensitivity of the receiver. Once the signal quality is lower than the threshold, a switching signal is sent to the adjustable receiver immediately.

The invention uses the out-of-band protection method, which can improve the flexibility and success rate of optical protection path establishment. At the same time, the adoption of the fast switching method greatly improves the protection switching time, thereby greatly reducing the data loss and service interruption caused by the switching, and finally has the ability to realize the protection switching without error codes. The present invention utilizes the transmission distance difference between the two paths (assuming that the protection path is longer than the working path) to realize protection switching within the transmission delay difference, thereby realizing reliable error-free protection switching, which is helpful for improving network reliability and providing high service quality business is of great significance.

Description of drawings:

Fig. 1 schematic diagram of fast optical protection switching method of the present invention

Figure 2 Schematic diagram of the implementation method of the multi-wavelength transmitter

Figure 3 is a schematic diagram of a method for implementing fast protection switching at the receiving end

Figure 4 Schematic diagram of the implementation method of fast tunable coherent receiver

Detailed ways

In an 8-node optical network, as shown in Figure 1, the optical nodes are connected to each other through optical fiber links to form a network. An optical path and its protection optical path need to be established between the source node and the destination node. The source node uses a multi-wavelength transmitter to send data to multiple wavelengths, one of which is the working wavelength, and the other wavelengths are protection wavelengths. According to the number of protection wavelengths N, the protection method is called N+1 protection. The figure shows 1+1 protection mode.

An implementation example of a multi-wavelength transmitter is shown in Figure 2. The output of the laser array is connected to the wavelength combiner, the output of the combiner is connected to the external modulator, and the output of the external modulator is connected to the wavelength splitter. The laser array emits continuous light waves of N+1 wavelengths (two lasers JDSU CQF935/208 and SUMITOMOSLT5411 are used for 1+1 protection), which are combined by the wavelength combiner and sent to the external modulator (using fast lithium niobate Mach-Zeng Del-type optical switch JDSU 15Gb/s Modulator). External modulation modulates data onto these N+1 wavelengths simultaneously. The modulated optical waves are spatially separated by wavelength splitters to be sent to different optical paths.

After the two wavelengths exit the transmitter, the working wavelength is sent to the working optical path for transmission, and the protection wavelength is sent to the protection optical path for transmission. The two paths are jointly routed to the destination node via different intermediate nodes, and path overlap can be allowed between them. Generally speaking, in order to improve the protection efficiency, the two paths should try to reach the destination node through different fiber links. Therefore, traditional protection usually uses the same wavelength for protection transmission, but this method also imposes great restrictions on the routing of these paths, such as no common edge, and the source and destination nodes have at least two optical links and network connect and so on. These constraints are not easily met when the mesh network topology, especially when the logical topology has changed greatly from the original design under dynamic bandwidth supply conditions. The present invention uses different wavelengths for protection transmission, which is called an out-of-band protection method. When it cannot find an optical path that is not on the same side as the working path, it can allow them to overlap in low-risk optical fiber links, thereby improving the protection path establishment. flexibility and success.

In order to increase the switching speed of protection switching, the present invention adopts a fast adjustable coherent receiver at the receiving end. Fig. 3 is an implementation example of realizing fast protection switching at the receiving end. The two-path optical signals are connected to the optical splitter, the main optical power output end of the splitter is connected to the optical combiner, the output of the optical combiner is connected to the optical one input end of the optical mixer, and the secondary optical power output end of the splitter is connected to the optical The signal monitoring module is connected, and its output is connected to a fast laser source. The output of the adjustable laser source is connected to the optical mixer, and the frequency mixing output is connected to the optical coherent receiver. At the destination node, the working optical path signal and the protection optical path signal are combined by an optical combiner and sent to the input end of the receiver. A small part of the optical power is separated from the two optical signals by the optical splitter, and sent to the optical signal monitoring module to monitor the change of the signal quality in the optical line in real time, so as to judge the occurrence of the fault. The realization method of fast fault monitoring is to adopt the threshold value triggering method, the threshold value is set near the sensitivity of the receiver, once the signal quality is lower than the threshold value, the wavelength switching signal is sent to the fast adjustable local oscillator light source of the adjustable receiver immediately. The output of the fast adjustable local oscillator light source and the input signal light are optically mixed through the optical mixer and then sent to the optical coherent receiver. The optical coherent receiver only detects the mixed-frequency optical signal in the same wavelength channel as the local oscillator source. When the fault of the working path is detected by the optical signal monitoring module, the fast adjustable light source will receive the protection switching signal, and the adjustable light source will quickly respond to adjust the output optical wavelength to the protection wavelength channel, so that the coherent receiver will switch from the corresponding protection channel Receive data to realize high-speed protection switching.

Figure 4 shows a specific implementation method of the fast tunable coherent receiver. It is a specific embodiment of the adjustable local oscillator light source and optical coherent receiver in FIG. 2 . The adjustable local oscillator light source receives the signal from the optical signal monitoring module, and the channel selection module is connected to the fast electro-optical switch (the example uses a fast lithium niobate Mach-Zehnder type optical switch JDSU 15Gb/s Modulator) realizes the control. The output of the working channel and protection channel local oscillator light source (JDSU CQF935/208 and SUMITOMO SLT5411) is connected to the optical input of the optical switch, and the optical output of the optical switch is connected to the optical mixer. Its other input receives optical signals. The output of the optical mixer is connected to the direct light intensity receiver (Multiplex MTRX192L BW~10GHz), converted into an electrical signal and then connected to the amplifier (JDSU H301 bandwidth 75kHz-10GHz gain~26dB), the output of the amplifier and the intermediate frequency filter (bandwidth 6.5-15GHz) connected, the output electrical signal is divided into two ways to connect to the electric mixer (IF bandwidth 6-18GHz, output DC-3GHz), the output of the mixer is connected to a low-pass filter (bandwidth DC-4GHz), Finally output to the clock pick-up and decision circuit. The adjustable receiver works according to the following method. When the working channel needs to be selected for reception, the switch of the working channel is in the open state, and the protection channel is in the off state. When the data reception of the protection channel needs to be selected, the opposite is true. Select to output the local oscillator light of a certain channel. The direct optical intensity receiver receives the mixed optical signal, and the example adopts the optical intensity receiver integrated with the amplifier. The received electrical radio frequency signal is amplified to an appropriate level by the amplifier. The redundant low-frequency signal is filtered out by the intermediate frequency filter, and only the intermediate frequency signal to be received is passed. The output of the intermediate frequency filter is passed through the mixer to obtain the low frequency oscillation component. Then the low-pass filter filters out the interfering intermediate frequency and high-frequency oscillation to obtain the demodulated low-frequency signal. The clock pick-up and judgment circuit correctly interprets the analog waveform as a digital bit stream composed of 0 and 1. When performing protection switching, the phases of the bit streams of the two channels are usually not synchronized, so it is necessary to use a fast clock pickup circuit to achieve ultra-fast phase locking to achieve correct data demodulation. In our experiment, the rising and falling edges of the local oscillator switching optical signal are 0.6ns, and the channel switching time is <0.8ns after intermediate frequency filtering. In the experiment, we always keep the phase synchronization of the two signals.

The physical switching speed of the two channels is ultimately limited by the IF bandwidth of the receiver. Because there is a random phase mutation at the junction of the switching of the two intermediate frequency signals. After IF filtering, a transition zone that is inversely proportional to the bandwidth will be produced. Suppose the amplitude of the two intermediate frequencies (IF) is A ₁ =A ₂ =A, the frequency is ω ₁ =ω ₂ =ω _c , and the phases are  ₁ and  ₂ , namely:

IF ₁ (t)＝Aexp[j(ω _c t+ ₁ )]

IF ₂ (t)＝Aexp[j(ω _c t+ ₂ )]

The time function of the oscillation can be written as:

f(t)＝Aexp(jω _c t)u(t)(expj ₁ -expj ₂ )+Aexp[j(ω _c t+ ₂ )]

Let K=A(expj ₁ -expj ₂ ), its Fourier transform is:

u(t) is a step function: $u\left(t\right)=\left\{\begin{array}{cc}1& t>0\\ 0& t<0\end{array}\right.$

After being filtered by the intermediate frequency filter H(ω), ( $h\left(\mathrm{\&omega;}\right)=\left\{\begin{array}{cc}1& {\mathrm{\&omega;}}_c-{\mathrm{\&omega;}}_f<\mathrm{\&omega;}<{\mathrm{\&omega;}}_c+{\mathrm{\&omega;}}_f\\ 0& \mathrm{others}\end{array}\right.,$ ω _f is half of the cut-off bandwidth of the filter), after Fourier inverse transformation, the final time response r(t) can be obtained:

function $\mathrm{Si}\left(t\right)=\underset{0}{\overset{t}{\&Integral;}}\frac{\sin x}{x}\mathrm{dx},$ When t→±∞, $\mathrm{Si}\left[{\mathrm{\&omega;}}_f(t-t_0)\right]\&Right\; Arrow;\&PlusMinus;\frac{\mathrm{\&pi;}}{2}.$ Its rising edge is 2π.

When the phase of the two intermediate frequencies suddenly reaches the maximum value, such as  ₁ = 0,  ₂ = π, then K=2A, the amplitude of r(t) will experience a width of $\frac{2\mathrm{\&pi;}}{{\mathrm{\&omega;}}_f}=2\frac{2\mathrm{\&pi;}}{2{\mathrm{\&omega;}}_f}=\frac{2}{B}$ The exchange transition area, where B is the IF filter bandwidth. Assuming that the IF bandwidth is 4GHz, the rise time is t _r =0.5ns. If the intermediate frequency signal itself has a rising and falling edge, which is equal to the finite switching time t _s of the local oscillator, then the final channel switching time is jointly determined by these two times T<t _s +t _r . Therefore, the switching time of optical path protection can be reduced by increasing the switching time of the local oscillator, but it will be limited by the limited IF filtering bandwidth in the end.

Claims (9)

1. A method for out-of-band optical protection switching based on fast tunable coherent receiver, characterized in that, comprising the following steps: ① The source node sends the data to the working wavelength path and the protection wavelength path respectively through two or more different wavelengths; ②At the destination node, these wavelengths are simultaneously routed to the optical port of the adjustable receiver; ③ Once the destination node detects the failure of the working path, the receiver immediately switches to the protection path to realize fast 1+1 protection. 2. The method for out-of-band optical protection switching based on fast tunable coherent receivers according to claim 1, characterized in that, for the wavelengths, the network independently routes these optical wavelengths to the destination node. 3. The method for out-of-band optical protection switching based on fast adjustable coherent receivers as claimed in claim 1 or 2, wherein the destination node adopts adjustable coherent receivers to receive data, and when the working path Normally, the receiver receives from the working path. When the destination node detects that the working path fails, the adjustable receiver quickly switches to the protection channel to receive data by changing the wavelength of the local oscillator light, and completes the protection switching. 4. The method for out-of-band optical protection switching based on a fast tunable coherent receiver as claimed in claim 1 or 2, wherein, at the destination node, the optical signal of the working path and the optical signal of the protection path are simultaneously Input to the input of the fast tuneable receiver, the tuneable receiver adopts the coherent reception method. 5. The method for out-of-band optical protection switching based on fast tunable coherent receivers as claimed in claim 1 or 3, characterized in that, in the tunable receiver, its local oscillator optical signal source is composed of tunable multi-wavelength It consists of a laser whose output wavelength can be quickly selected by a fast electrical gating signal. 6. The method for out-of-band optical protection switching based on fast tunable coherent receivers as claimed in claim 1, characterized in that, in the 1+1 protection, the source node passes data through two or more different wavelengths Send to the working wavelength path and the protection wavelength path respectively: when there is one protection wavelength, it is 1+1 protection. 7. The method for out-of-band optical protection switching based on a fast tunable coherent receiver as claimed in claim 1 or 6, wherein the source node adopts a multi-frequency laser at its transmitting end, and uses internal modulation of the laser or Using an external modulator, the data is modulated onto these multiple wavelengths at the same time, one of these wavelengths is transmitted to the working path as the working wavelength, and the remaining wavelengths are transmitted to the protection path as the protection wavelength. After the wavelength is modulated, it is divided by the wavelength The device separates these wavelengths in space. Since the protection path and the working path use different wavelengths, these wavelengths can pass through the same optical fiber link in the network without interfering with each other. 8. The method for out-of-band optical protection switching based on fast tunable coherent receivers as claimed in claim 1 or 3, characterized in that, for said switching, the final optical channel reception switching time is determined by the switching speed of local oscillator light It is determined by the IF filter bandwidth of the receiver, less than 1ns. 9. The method for out-of-band optical protection switching based on fast tunable coherent receivers as claimed in claim 8, wherein the switching of the local oscillator is triggered by a high-speed fault detection circuit to monitor the optical signal quality in real time, And set the threshold value, once the signal quality is lower than the threshold value, the switch signal is sent to the adjustable receiver immediately.

Images