CN1185820C - NXN light exchanging structure in full photo exchanging nodal point - Google Patents

Abstract

An N×N optical switching structure used in an all-optical switching node, consisting of 2n stages, where n=log ₂ N, characterized in that a semiconductor optical amplifier is arranged between the nth stage and the n+1th stage On/off" gates form the 2n+1 level; different numbers of semiconductor optical amplifier "on/off" gates can also be set in any secondary optical switch according to specific needs. The invention adopts the reconfigured extended Benes optical switching structure, which has better restriction on crosstalk. In an N×N extended Benes optical switching structure, a semiconductor optical amplifier "on/off" gate (SOA gate) is added to change the routing equivalently. The present invention retains some excellent properties of the extended Benes optical switching structure, such as generalized non-blocking, hardware optimization, etc., and meanwhile it has a good effect on reducing crosstalk.

Description

A N×N Optical Switch Structure Used in All-Optical Switch Nodes

technical field

The invention belongs to the technical field of optical communication devices, and in particular relates to an N×N optical switching structure used in all-optical switching nodes, which is especially suitable for (DWDM) systems, and can solve the crosstalk problem.

Background technique

With the continuous development and wide application of IP technology, the capacity and rich and colorful business volume of optical communication are showing an exponential growth trend. The optical network based on dense wavelength division multiplexing (DWDM) technology has been recognized as the most effective way to alleviate the increasing communication capacity. The establishment of an optical network and the information exchange between channels will inevitably introduce all-optical switching nodes into the topology of the network. An all-optical switching node is generally composed of a wavelength division multiplexer (MUX), an optical switch (Optical Switch) and a demultiplexer (DMUX). In these optical passive devices, since it is impossible for the device to achieve complete isolation between optical channels, part of the power of the input signal leaks into other output channels that should not go, resulting in information crosstalk between optical channels (Cross- talk). As a kind of noise, crosstalk will become serious with the reduction of channel spacing and the expansion of optical switching scale. In addition, in the all-optical network of DWDM, in order to make up for various losses caused by devices and optical fibers in the optical network, linear optical amplifiers (such as erbium-doped fiber amplifier EDFA) are often used. While amplifying optical signals, the crosstalk between channels is also amplified. In this way, through the transmission of a large number of nodes in the network, the amount of crosstalk is continuously accumulated, which will have a very harmful impact on the DWDM optical fiber communication system, such as a decrease in the signal-to-noise ratio and an increase in the bit error rate, which may eventually lead to communication failure. For this reason, how to effectively eliminate the crosstalk is a problem that the optical switching nodes must solve.

Crosstalk in optical switching can be divided into interband crosstalk and intraband crosstalk. The so-called inter-band crosstalk means that the crosstalk component and the transmitted signal optical channel have different wavelengths; the intra-band crosstalk means that the crosstalk component has the same wavelength as the transmitted signal optical channel. Figure 1 is an optical switching structure composed of 2*2 optical switches (this optical switch can be any type of optical switch), which consists of three stages I, II, and III. In the figure, the actual path is indicated by a thick solid line, while the leaky path is indicated by a dashed line, and all solid line connections use the same narrowband wavelength. It can be seen from the figure that part of the signal power of the input signal S ₂ leaks to the output signal S ₁ , and since the wavelengths occupied by the signals S ₁ and S ₂ are different, inter-band crosstalk is formed. In addition, part of the power of S ₁ returns to the output port S ₁ after going through a wrong number of hops. Since S ₁ always occupies the same wavelength, an in-band crosstalk is formed.

For inter-band crosstalk, a tunable narrowband filter (TOF) can be set at the output to eliminate it, while intra-band crosstalk cannot be eliminated with a filter because it is on the same wavelength as the signal. Therefore, it is particularly important to eliminate in-band crosstalk and improve the communication quality of the DWDM system. The crosstalk between channels in the optical switching node is mostly caused by the imperfection of the demultiplexer and the optical switch. These crosstalk items can be classified by order, such as first order, second order, etc. Figures 2A and 2B show two working states of a 2×2 optical switch used in optical switching: parallel (bar) and cross (cross), which will introduce crosstalk due to their imperfections. In these two states, a part of the input signal power ξ (indicated by ∈ symbol in the figure) will leak into the wrong output port. The leakage parameter ξ is usually expressed in decibels as ξ _dB =10log ₁₀ ξ, which is called the unit crosstalk of the optical switch. If the crosstalk of each optical switch unit is the same, the high-order crosstalk will be the multiplication of the first-order crosstalk, because |ξ| is a number smaller than 1, so the high-order crosstalk is much smaller than the low-order crosstalk. The optical switch shown in Fig. 2A operates in a parallel state. It is precisely because of the imperfection of the optical switch that the crosstalk is caused. In the figure, the crosstalk is represented by a dotted line. The first-order crosstalk amount ∈ S ₂ in the signal S ₁ is part of the optical power in the signal S ₂ ; in the signal S ₂ The amount of first-order crosstalk ∈ S ₁ is part of the optical power in the signal S ₁ . FIG. 2B shows that the optical switch works in the cross state. Similarly, each signal contains first-order crosstalk. It is conceivable that in an optical switching node composed of many optical switches, as the number of optical switches increases, the amount of crosstalk will also increase due to continuous accumulation, which will inevitably lead to a decrease in the signal-to-noise ratio (SNR), resulting in bit errors The generation of the communication cannot be carried out, so certain methods must be used to reduce the amount of crosstalk. It was previously pointed out that the high-order crosstalk is much smaller than the low-order crosstalk, so a common method to eliminate crosstalk is to convert the crosstalk from low-order to high-order by expanding the optical switch at the cost of increasing the number of optical switches, as shown in Figure 3A and 3B These are the corresponding two working states obtained by respectively extending the 2×2 optical switches in FIGS. 2A and 2B : parallel and cross. The optical switch in this structure has only one input terminal activated (indicated by the thick solid line in the figure) at a certain time, while the other input terminal is not activated (indicated by the thin solid line in the figure), so that the first-order The crosstalk is properly routed into the unactivated line, thereby converting the first-order crosstalk to the second-order. Therefore, the first-order crosstalk ∈ ¹ in FIGS. 2A and 2B becomes the second-order crosstalk ∈ ² in FIGS. 3A and 3B , and the influence of the crosstalk is reduced by the first order. It can be seen from the figure that each output terminal also contains a second-order crosstalk amount ∈ ² in addition to its own signal. The extended Benes optical switching structure (DBN) can be formed by taking the 2×2 optical switch in Fig. 3A and 3B as its basic unit and adding the recursive decomposition of the Benes optical switching structure. As shown in Figure 4, it is a 4×4 DBN structure, which consists of 16 optical switches to form a 4×4 optical switching matrix, and the 4 rows are represented by (1), (2), (3), and (4). The 4 columns are also called the four levels and are represented by I, II, III, and IV. In Fig. 4, it is assumed that a connection is established between the input signals _S1 to _S4 and the corresponding output terminals, which are represented by thick solid lines in bold in the figure, and this connection is obtained by a ring algorithm. Table 1 can be used to analyze the signal-to-noise ratio of each output port of this optical switching structure. P ₁ , P ₂ , P ₃ , and P ₄ in the table represent the optical power of the four input ports S ₁ to S ₄ respectively, m represents the crosstalk coefficient of the optical switch, and for the convenience of discussion, the crosstalk higher than the second order and the optical switch loss.

Table 14×4 Crosstalk amount after each stage in extended Benes optical switching structure switch output after level one after level two After the third level After the fourth grade (1) superior _P1 P ₁ +m ² P ₂ P ₁ +m ² (P ₃ +P ₂ +P ₄ ) P ₁ +2m ² (P ₃ +P ₂ +P ₄ ) Down mP ₁ m(P ₁ +P ₂ ) m(P ₁ +P ₂ +P ₃ ) m(P ₁ +P ₂ +P ₃ +P ₄ ) (2) superior mP ₂ m(P ₃ +P ₄ ) P ₃ +m ² (P ₁ +P ₂ +P ₄ ) P ₂ +2m ² (P ₁ +P ₃ +P ₄ ) Down _P2 P ₃ +m ² P ₄ m(P ₁ +P ₂ +P ₃ ) m(P ₁ +P ₂ +P ₃ +P ₄ ) (3) superior P ₃ P ₂ +m ² P ₁ m(P ₂ +P ₃ +P ₄ ) P ₃ +2m ² (P ₁ +P ₂ +P ₄ ) Down _mP3 m(P ₁ +P ₂ ) P ₂ +m ² (P ₁ +P ₃ +P ₄ ) m(P ₁ +P ₂ +P ₃ +P ₄ ) (4) superior _mP4 m(P ₃ +P ₄ ) m(P ₁ +P ₂ +P ₄ ) P ₄ +2m ² (P ₁ +P ₂ +P ₃ ) Down P ₄ P ₄ +m ² P ₃ P ₄ +m ² (P ₁ +P ₂ +P ₃ ) m(P ₁ +P ₂ +P ₃ +P ₄ )

The semaphore and crosstalk of each output port can be obtained from the table. Take the signal S ₁ as an example, the signal S ₁ contains the amount of crosstalk from other ports, that is, the noise power is 2m ² (P ₃ +P ₂ +P ₄ ), so it can be obtained from the calculation formula of the signal-to-noise ratio The signal-to-noise ratio of its output signal is:

SNR=10log ₁₀ (P _s /P _n ), (1)

Among them, P _S is the signal power, and P _n is the noise power.

For convenience, it is assumed that the signal power of each input port is the same and is P, so it can be seen from Table 1 that the signal-to-noise ratio of each output port is the same and is:

$\mathrm{<span\; class="notranslate">\; SNR</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 10</span>}{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}}{\mathrm{<span\; class="notranslate">\; 6</span>}{\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; P</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

Through the analysis of the 4×4 extended Benes optical switching structure, it is not difficult to obtain the following rules for an N×N extended Benes optical switching structure:

●After the second stage, there is a second-order crosstalk of the input port that crosses into the signal;

●After the third stage, the second-order crosstalk of two input ports is inserted into the signal;

●Through the fourth stage, the second-order crosstalk of three input ports enters into the signal;

●After the Nth stage, the second-order crosstalk of N-1 input ports is crosstalked into the signal.

Can calculate the signal-to-noise ratio of the extended Benes mere switching structure of arbitrary N*N through this rule. For the extended Benes optical switching structure of N×N, its number of stages is 2K (wherein K=log ₂ N), because each stage has N optical switches, so the total number of optical switches it needs is N×2K, so any The amount of crosstalk at the output port is the noise power:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; P</span>}\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; K</span>}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span>$

The signal-to-noise ratio is thus:

$\mathrm{<span\; class="notranslate">\; SNR</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 10</span>}{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}}{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}}<span\; class="notranslate">\; )</span>$

$<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; 10</span>}\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; P</span>}}{\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; P</span>}}<span\; class="notranslate">\; )</span>$

$<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 10</span>}\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

In the formula $\left|x\right|=10{\log }_{10}\left(\frac{1}{m}\right).$

It can be seen from (4) that with the increase of the K value, that is, the increase of the number of optical switches and the expansion of the switching scale, the signal-to-noise ratio will decrease rapidly, so the elimination of crosstalk by extending the Benes optical switching structure is very limited. It is necessary to take measures to further reduce the crosstalk and improve the signal-to-noise ratio.

There are many ways to eliminate crosstalk. At present, the methods of eliminating crosstalk at home and abroad mainly include: (1) Through the design of the network node topology, the amount of crosstalk is converted from low-order to high-order, thereby eliminating the impact of crosstalk on the transmission system to a certain extent. However, it will inevitably cause the complexity of network nodes and the increase of the number of passive devices, and the cost of a single network node will also increase. (2) Reduce the crosstalk by concentrating on improving the performance of passive devices in the network nodes, which will inevitably increase the cost of the devices, and at the same time, the crosstalk cannot be well eliminated when the scale of the switching node is large. (3) By scrambling the transmitted optical signal and its phase, it will lead to the complexity of the system, and it is only suitable for a local network. Thereby above-mentioned method all can not thoroughly eliminate the crosstalk in the DWDM optical communication system, and its method for eliminating crosstalk mostly considers from one side: on the one hand they only consider the topological structure of the network; device performance to reduce crosstalk. But none of them have been taken into account with the DWDM system, so the degree of reduction of crosstalk is also limited.

Contents of the invention

The object of the present invention is to provide a kind of N * N optical switching structure that is used in the all-optical switching node that overcomes above-mentioned defect, this all-optical switching structure considers the characteristics of network topology and DWDM system comprehensively, makes in the optical switching structure The device's tolerance to crosstalk is improved, and the crosstalk is improved or even eliminated.

In order to realize the purpose of the above invention, an N×N optical switching structure used in all-optical switching nodes is composed of a 2n-level optical switch matrix, wherein n=log ₂ N, and is characterized in that: the nth level and the n+th A semiconductor optical amplifier "on/off" gate is arranged between the first stages to form the 2n+1st stage.

According to specific routing requirements, different numbers of semiconductor optical amplifier "on/off" gates can be set between any two of the above-mentioned stages.

The invention adopts the reconfigured extended Benes optical switching structure, which has better limitation on crosstalk. In an N×N extended Benes optical switching structure, adding a semiconductor optical amplifier "on/off" gate (that is, SOAgate), only equivalent changes are made to the routing. Therefore, the present invention not only retains some excellent properties of the extended Benes optical switching structure, such as generalized non-blocking property, hardware optimization, etc., but also greatly reduces crosstalk. The technical effect of the present invention will be described in detail in the specific embodiment section.

Description of drawings

Figure 1 is a schematic diagram of inter-band and intra-band crosstalk in all-optical optical switching;

2A and 2B are schematic diagrams of the two working states of the 2×2 optical switch, namely: parallel and crossover;

Figures 3A and 3B respectively show the two states of the 2×2 extended Benes optical switch structure: parallel and crossover;

Fig. 4 is a schematic diagram of a 4×4 extended Benes optical switch structure;

Fig. 5 is a structural schematic diagram of a specific embodiment of the optical switching structure of the present invention;

Fig. 6 is the recursive algorithm schematic diagram of expanding Benes structure;

FIG. 7 is a schematic structural diagram of an all-optical switching node used in a DWDM system;

Fig. 8 is the schematic diagram that demultiplexer produces crosstalk;

Fig. 9 is a structural schematic diagram of another specific embodiment of the optical switching structure of the present invention;

Fig. 10 is a schematic diagram of two-wavelength switching of an all-optical switching node in a DWDM system adopting the present invention.

specific implementation

光交换结构，所以图5和图4的区别仅为在递归分解的第II级与第III级之间加入半导体光放大器“开／关”门2，形成第V级。这样加入半导体光放大器“开／关”门2后只是对路由进行等价的变化因而不仅没有改变扩展Benes光交换结构的一些优良的性质如：广义无阻塞性、硬件的最优性等，同时它还对串扰有很好的降低作用。The structure of the present invention is shown in Fig. 5, which is called a reconstructed extended Benes optical switching structure, which can greatly reduce crosstalk. For an N×N extended Benes optical switching structure, the recursive decomposition shown in Figure 6 can be used, and 3 in the figure represents the structure composed of optical switch 1

Optical switching structure, so the difference between Fig. 5 and Fig. 4 is only to add a semiconductor optical amplifier "on/off" gate 2 between the recursively decomposed stages II and III to form the V stage. In this way, after adding the "on/off" gate 2 of the semiconductor optical amplifier, it only changes the routing equivalently, so not only does not change some excellent properties of the extended Benes optical switching structure, such as: generalized non-blocking, hardware optimality, etc., but also It also does a good job of reducing crosstalk.

The present invention will be considered in conjunction with DWDM systems. Using the structure of Figure 5 in the all-optical switching node of the DWDM system has obvious changes in the reduction of crosstalk compared to the expanded Benes structure. Fig. 7 is an all-optical switching node structure used in the DWDM system adopted in the present invention, which consists of three parts I, II and III. The first part is the demultiplexer 4; the second part is the optical switching fabric 5; the third part is the multiplexer 6. This structure groups the channels of different wavelengths under the condition that the number m of wavelengths multiplexed to each optical fiber is greater than the number N of optical fibers input to the node, so that the channels of different wavelengths of each all-optical switching unit can be spatially exchanged. On the one hand, Avoid causing wavelength blocking in the optical switching unit; on the other hand, reduce the amount of crosstalk in the band. The optical switching structure 5 used in FIG. 7 is shown in FIG. 5 . Adopting the switching structure shown in Fig. 7 can reduce the crosstalk to the third order so as to completely eliminate the crosstalk lower than the third order. Since the channels of different wavelengths are grouped under certain conditions in Fig. 7, the channels of different wavelengths of each all-optical switching unit are spatially switched. Different wavelengths are exchanged, so there will be no intra-band crosstalk, and the inter-band crosstalk can be eliminated by a tunable narrow-band optical filter (TOF). But in reality, due to the imperfection of the multiplexer and demultiplexer, there will also be crosstalk between channels (as shown in Figure 8), so in the all-optical switching node shown in Figure 7, although the channels switched in the same switching unit are different wavelengths, but they also have in-band crosstalk, so the structure in Figure 5 is used to reduce the in-band crosstalk. For illustration, in Fig. 8, the demultiplexer 8 composed of multilayer dielectric film interference filter is selected for demultiplexing, and the signal light of 8 wavelengths enters the demultiplexer after passing through the collimating lens 7, and the demultiplexer After use, it is output through a collimating lens. Assuming that the crosstalk coefficient between channels is m and ignoring the second-order channel crosstalk, it is easy to obtain that the first-order crosstalk power ∈P _i in the λ _j channel contains the λ _i channel as:

In the formula, P _i is the optical power in the channel λ _i , and N is the total number of wavelengths input to the demultiplexer. Due to the introduction of the "on/off" gate of the semiconductor optical amplifier, when it works "on", it will completely isolate the crosstalk from different information of the same carrier wavelength, so this will eliminate the crosstalk in the band. When it works in "off", it is equivalent to a light absorber, which makes the light have a large loss and cannot pass through at all, so that the spread of crosstalk can be well limited. In order to analyze the limitation of the crosstalk of the "on/off" gate of the semiconductor optical amplifier, it is assumed that it is in the worst case: that is, the last wavelength exchanged has its first order in other channels after passing through the demultiplexer The amount of crosstalk ∈P _N . For the convenience of analysis, only the extended Benes structure of 4×4 reconstruction is illustrated.

Table 2 shows the amount of in-band crosstalk after each stage of the optical switching structure shown in Figure 5 switch output after level one after level two After the third level After the fourth grade (1) superior P ₄ P ₄ +m ³ P ₄ P ₄ +m ³ P ₄ P ₄ +2m ³ P ₄ Down _mP4 mP ₄ +m ² P _{4 mP4} mP ₄ +m ² P ₄ (2) superior m ² P ₄ 2m ² P ₄ mP ₄ +m ³ P ₄ mP ₄ +2m ² P ₄ Down _mP4 mP ₄ +m ³ P ₄ m ² P ₄ mP ₄ +m ² P ₄ (3) superior _mP4 mP ₄ +m ² P ₄ m ² P ₄ mP ₄ +2m ³ P ₄ Down m ² P ₄ mP ₄ +m ² P ₄ mP ₄ +m ² P ₄ 2m ² P ₄ (4) superior m ² P ₄ 2m ² P ₄ m ² P ₄ mP ₄ +2m ³ P ₄ Down _mP4 mP ₄ +m ³ P ₄ mP ₄ +m ³ P ₄ 2m ² P ₄

The exchange situation is that the four input ports in Fig. 5 are P ₄ , mP ₄ , mP ₄ , mP ₄ , and mP ₄ (P ₄ represents the optical power of the fourth channel). With the introduction of semiconductor optical amplifier "on/off" gates, there will be no crosstalk in stages II and III. For the convenience of discussion, it is assumed that the crosstalk coefficient of the optical switch is also m, then the crosstalk in the band after each level of switching is shown in Table 2 (ignoring the crosstalk higher than the third order), which is compared with each output in Table 1 Crosstalk at both ends has been reduced. The same method can be used to analyze other switching wavelengths, such as N×N optical switching (always for the worst case, that is, switching the last wavelength): for N×N all-optical switching nodes, it has a total of 2K The +1 (K=log ₂ N) level is the imported intermediate SOA gate level and the K levels of each edge. Due to the function of the "on/off" gate of the semiconductor optical amplifier, there is no crosstalk in the middle stage K+1 and the 1st stage and the K+2 stage. For stages 2 and K+3, third-order crosstalk in one channel is added to the channel being transmitted, and for stages 3 and K+4, third-order crosstalk in two channels is added to the channel being transmitted In the transmission channel, for the Kth level and 2K+1 level, the third-order crosstalk of K-1 channels is added to the channel being transmitted, which is equivalent to the addition of 1, 2...2K-1 natural numbers by After the introduction of SOA, it is divided into two groups of 1, 2...K-1 natural numbers are added, which naturally reduces the sum of them, that is, the influence of crosstalk is reduced, so the total amount of crosstalk for each output port is the noise power It can be expressed as:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; 3</span>}}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; N</span>}}\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; 3</span>}}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

Among them, P _N represents the optical power of the last optical signal being exchanged, and P _N represents its corresponding noise power.

$\mathrm{<span\; class="notranslate">\; SNR</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 10</span>}{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}\frac{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; N</span>}}}{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}}$

$<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 10</span>}{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}\frac{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; N</span>}}}{\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; 3</span>}}{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; N</span>}}}$

$<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 10</span>}{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$

Compared with the extended Benes optical switching structure (4), its signal-to-noise ratio has more than doubled, and its dependence on K has also been reduced a lot. It can also be seen that as long as the crosstalk coefficient m in the demultiplexer is increased , you can eliminate the crosstalk in the band very well without changing the routing of the optical switch, which will be very useful for the DWDM system. Of course, adding the "on/off" gate of the semiconductor optical amplifier also has a good effect on limiting the crosstalk between the bands. For inter-band crosstalk, it is assumed that in the worst case, each input port of the optical switching structure in Fig. 7 is activated. For the 4×4 case, that is, the four input ports of S ₁ to S ₄ in Fig. 5 are respectively P ₁ , P ₂ , P ₃ , P ₄ , which correspond to the powers of the signals S ₁ to S ₄ respectively. In the same analysis, for the case of N×N, assuming that the signal optical power of each port is P, the crosstalk amount of each output port, that is, the noise power can be obtained as:

${\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}{\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; P</span>}\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; P</span>}$

So the corresponding output port signal-to-noise ratio is:

$\mathrm{<span\; class="notranslate">\; SNR</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 10</span>}{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}\frac{\mathrm{<span\; class="notranslate">\; P</span>}}{{\mathrm{<span\; class="notranslate">\; P</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}}$

${<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 10</span>}\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}\frac{\mathrm{<span\; class="notranslate">\; P</span>}}{\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; m</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; P</span>}}$

$<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 10</span>}{\mathrm{<span\; class="notranslate">\; log</span>}}_{\mathrm{<span\; class="notranslate">\; 10</span>}}\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; K</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; .</span>$

Comparing the formula with formula (4), it can be seen that the dependence of the signal-to-noise ratio on K is greatly reduced, that is to say, the crosstalk between bands is reduced. It can also be seen from the above analysis that if the signal-to-noise ratio is to be further improved, a semiconductor optical amplifier "on/off" gate can be introduced between every two stages, which is equivalent to the natural number of 1, 2...2K-1 The addition becomes the addition of the equal number of units through continuous dichotomy (the number of units is often very small), so that the sum is naturally reduced, that is, the amount of crosstalk is reduced. In order to facilitate integration, it can also be made into a new 2×2 low crosstalk optical switch, the structure of which is shown in Figure 9. It can be seen from FIG. 9 that this optical switch can completely eliminate the crosstalk in the existing 2×2 optical switch. Using the switching unit shown in Figure 9 to replace the 2×2 optical switch in the extended Benes structure will result in lower crosstalk and higher signal-to-noise ratio. The crosstalk in the all-optical switching node shown in Figure 7 can be greatly reduced by recursively adding the "on/off" gate of the semiconductor optical amplifier.

Another issue that must be considered in optical switching nodes is the issue of optical loss. With the expansion of the scale of optical switching nodes, the devices (such as optical switches) that light passes through will naturally increase. Due to the existence of the insertion loss of the optical switch, most of the useful optical energy is lost. If the optical loss is not compensated, it will also have a great impact on the communication quality. As mentioned in the background art, in order to make up for the loss caused by devices and optical fibers in the optical network, linear optical amplifiers (such as erbium-doped fiber amplifiers EDFA) are often used, but when they amplify optical signals, crosstalk between channels Amplification is also obtained, which of course is not what we would like. However, after the introduction of the "on/off" gate of the semiconductor optical amplifier, this problem will be completely solved, because the "on/off" gate of the semiconductor optical amplifier has the following advantages:

●It can amplify the light to make up for the loss while reducing the crosstalk;

●Higher extinction ratio;

●Low insertion loss;

●It has a very low switching time, which is on the order of nanoseconds;

●has low polarization sensitivity;

●Small size, light weight, easy to integrate.

Because of these unique advantages of the "on/off" gate of the semiconductor optical amplifier, its introduction can solve the problem of crosstalk and loss in the optical switching node, which undoubtedly has a good effect on improving the quality of optical communication and is more conducive to Large scale integration of optical switching fabrics.

The present invention can be implemented using the routes shown in FIG. 5 or FIG. 7 . In order to illustrate its principle in more detail, an all-optical switching node with two wavelengths λ ₁ and λ ₂ as shown in Fig. 10 is selected for illustration. The two wavelengths are coupled into the same optical fiber, and the demultiplexer 4 is used to separate the two wavelengths at the all-optical switching node, and then the different wavelengths are switched in the same switching unit through a certain combination. Due to the introduction of the "on/off" gate of the semiconductor optical amplifier with strong anti-crosstalk ability, the in-band crosstalk is controlled above the third order, so that the in-band crosstalk is well eliminated. For inter-band crosstalk, a narrow-band tunable optical filter (TOF) 9 can also be added at the output end to filter out the crosstalk. What is shown in the figure is just a routing situation, and different routings can be selected according to the needs. In addition, the 2×2 optical switch shown in FIG. 9 can be used as a unit to expand an all-optical switching structure with any number of levels and no crosstalk.

It has long been proved that the extended Benes structure is a non-blocking optical switching structure. In this patent application, the "on/off" gate of the semiconductor optical amplifier with high switching speed and high extinction ratio is introduced into the extended Benes structure, and through reasonable The DWDM channel packet switching method eliminates all kinds of crosstalk in the DWDM optical communication system.

It is not difficult to see that the above are only examples of some specific embodiments of the present invention, and those skilled in the art can implement them in other ways according to the above description.

Claims (2)

1. N * N the optical exchange structure that is used for full Optical Switch Node is made of the 2n level, wherein n=log ₂N is characterized in that: be provided with semiconductor optical amplifier " ON/OFF " door between above-mentioned n level and n+1 level, form the 2n+1 level.

2. optical exchange structure according to claim 1 is characterized in that: be provided with semiconductor optical amplifier " ON/OFF " door between other two-stage.

Images