RU2310278C1 - Passive fiber-optic network - Google Patents

Abstract

FIELD: telecommunications; passive loop-architecture optical circuits.

SUBSTANCE: proposed passive fiber-optic network has unidirectional fiber-optic loop, N directional couplers disposed at different points of loop, communication center, and plurality of subscriber points. Communication-center transmitter is optically coupled with fiber-optic loop sending end and receiver, with its receiving end. Transmitter and receiver of each subscriber point are optically coupled with fiber-optic loop through respective directional coupler inserted in fiber-optic loop. Directional couplers, from first to last ones, have rising branching coefficients from sending to receiving ends of fiber-optic loop at transmitter radiation wavelength of communication center and reducing branching coefficients on transmitter radiation wavelength of any subscriber point. Transmission is conducted in network in two streams: from communication center to subscriber points and from the latter to communication center.

EFFECT: enhanced quantity of subscriber points in network, enlarged amount of hardware used in telecommunications.

4 cl, 5 dwg

Description

The invention relates to the field of telecommunications, to passive optical networks with a loop architecture. It can be used in broadcast telecommunication networks, as well as in local data exchange networks.

In the field of telecommunications, passive optical networks are understood to mean networks in which the transmission of an optical signal between a central node and a plurality of subscriber nodes is carried out by passive static components, without amplification, regeneration / relay, etc. active components, regardless of the architecture (or topology) of the network: bus (linear), loop or tree-like.

A typical passive optical network with a tree architecture contains a central node, a plurality of subscriber nodes and a transmission channel including a trunk optical fiber, a 1 × N splitter (star type) and N taps of optical fibers from a star. The transmitter and receiver of the central node are optically coupled to the trunk fiber through a bi-directional spectral multiplexer / splitter (WDM multiplexer). The transmitter and receiver of the subscriber unit is optically coupled through a WDM multiplexer of the subscriber unit to one of the taps of the “star” [1].

Transmission from the central node to the subscriber nodes is carried out at the same wavelength (downstream). Transmission from all subscriber nodes to the central node is carried out at a different wavelength (upstream).

The downstream is broadcast and organized by the method of temporary compression / separation of signals intended for different subscriber nodes.

The upstream is organized in accordance with the access protocol adopted in the network with time compression / separation of signals from different subscriber nodes, when each subscriber node is allocated a certain time interval for transmission. Separation of oncoming streams is carried out in network nodes by WDM-multiplexers.

It is believed that a network with a tree architecture has an advantage over other networks in saving optical fiber [2, p. 479]. However, in the case of a significant removal of subscriber nodes from the "star" and a large number of subscriber nodes (and the current level of technological development already allows you to include up to 64 nodes at gigabit transfer speeds [2]) - this network is less expensive than fiber networks with a unidirectional optical loop. In fact, if for the radius r of the loop we take the average length of the branch from the 1 × N splitter to the subscriber unit, the perimeter of the loop will be 2πr, and the total length of the branches N × r. From where, from the inequality 2πr <N × r, it follows that already for the number of subscribers N> 6, a network with a tree-like architecture (without taking into account the length of the main fiber) is inferior to a network with a loop in saving optical fiber. However, the known passive optical networks with a loop (or ring) architecture have a significant drawback: the limitation on the number of nodes. In a passive network with a star, the power of the optical signal at the receiver of the subscriber unit is proportional to N ^-1 (a 1 × N splitter divides the power into N parts). In a passive network with an optical loop and M nodes, the power of the optical signal at the receiver of the subscriber node is proportional to M ^-2 [3, 4]. Thus, with equal budgets, transmission speeds and acceptable BER error rates, a passive network with a “loop” √N times loses by the number of network nodes with a “tree”. We emphasize that the last statement is true only for fully passive networks with the transmission of information at the same operating wavelength. In the same work [4], this limitation can be circumvented by endowing the subscriber units with the function of repeaters and using dynamic couplers with a controlled branch coefficient in the loop, i.e. actually including active components in the transmission line.

Known passive fiber optic network containing a unidirectional fiber optic loop and N directional couplers located at different points of the loop. The network contains a central node and a plurality of subscriber nodes. The transmitter of the central node is optically coupled to the beginning of the fiber optic loop, and the receiver to its end. The transmitter and receiver of each subscriber unit are optically coupled to a fiber optic loop through an appropriate directional coupler [5].

The network implements the transmission of information in two streams: from the central node to the subscriber nodes and from the subscriber nodes to the central node in one direction in the loop.

This technical solution is taken as a prototype.

The prototype has the same drawback as passive networks - analogues with a unidirectional optical loop: a limitation on the number of subscriber nodes in the network.

If the α-coefficient of the branches is the same in both transmission streams, then the attenuation introduced into the optical signal only by the couplers leads to the following expression for the transmission coefficient k in the loop:

Expression (1) takes into account the worst cases (in terms of insertion loss) of the propagation of an optical signal in a loop: from the transmitter of the central node through (N-1) taps and the branch to the receiver of the latter in the loop of the subscriber unit; or from the transmitter of the first subscriber unit to the receiver of the central node. For the transmission coefficient k, the following estimate is true:

In turn, the right-hand side of inequality (2) has a maximum at α = 1 / (N-1).

T.O.

Now let P ₁ be the transmitter power (more precisely, the power introduced into the optical fiber from the transmitter laser), and P _{2 the} minimum detectable power of the receiver (for a given level of the error coefficient BER); then from the equality P ₂ = kP ₁ and inequality (3) it follows

The ratio P ₁ / P ₂ in inequality (4) is known as the network budget or the energy potential of the network. The whole part [P ₁ / P ₂ ] is equal to the limit number of subscriber nodes, theoretically possible in passive networks. From inequality (4) it follows that the prototype is almost three times inferior to the theoretical limit in the number of subscriber nodes.

The present invention solves the problem of increasing the number of subscriber nodes in the network and expanding the arsenal of technical means in the field of telecommunications.

To achieve this technical result, a passive fiber optic network comprises a unidirectional fiber optic loop and N directional couplers located at different points of the loop, a central node and a plurality of subscriber nodes. The transmitter of the central node is optically coupled to the beginning of the fiber optic loop, and the receiver to its end. The transmitter and receiver of each subscriber unit are optically coupled to the fiber optic loop through an appropriate directional coupler. Unlike the prototype, directional couplers from the first to the last, in the order of their arrangement from the beginning to the end of the fiber optic loop, are made with increasing branch coefficients at the radiation wavelength of the transmitter of the central node and with decreasing branch coefficients at the radiation wavelength of the transmitter of any subscriber node.

According to a special case of execution, directional couplers are made with increasing branch coefficients in the sequence: 1 / N, 1 / (N-1), ..., 1/3, 1/2, 1 at the radiation wavelength of the transmitter of the central node and with decreasing coefficients branches in the sequence: 1, 1/2, 1/3, ..., 1 / (N-1), 1 / N at the radiation wavelength of the transmitter of any subscriber unit.

According to a special case of execution, each directional coupler is made on two connected optical waveguides, the ends of which form two pairs of input and output ports, one couple of ports the coupler is included in the gap of the optical fiber loop, and the other pair of ports is designed to connect to the transmitter and receiver of the subscriber unit.

According to a special case of execution, the receiver of each subscriber unit is equipped with a selective filter for one or another radiation wavelength of the transmitter of the central node, and the receiver of the central node is equipped with a blocking filter for the same wavelengths.

The invention is illustrated by drawings, Fig.1-5.

Figure 1 shows a block diagram of a passive fiber optic network according to the invention.

Figure 2 shows a functional diagram of a spectrally dependent directional coupler.

Figure 3 shows a schematic representation of a directional coupler on two connected optical waveguides.

Figure 4 shows graphs of the beat of power along homogeneous coupled waveguides for two wavelengths.

Figure 5 shows a block diagram of a network with a backup transmission line.

Passive fiber optic network - Figure 1 - contains a unidirectional fiber optic loop 1 with one optical fiber, a central node 7 and many subscriber nodes 8, 9, ..., 12. At different points of the loop 1 are N directional couplers 2, 3, ..., 6. The transmitter 13 of the central node 7 is optically connected to the beginning 15 of the loop 1, and the receiver 14 is connected to the end 16 of the loop 1.

The transmitter 17 and receiver 18 of each subscriber unit 8, 9 ..., 12 are optically coupled to the fiber optic loop 1, respectively, through couplers 2, 3, ..., 6. Directional couplers 2, 3, ..., 6 in one pair from the input 19 and output 20 ports are included in the breaks of the optical fiber loop 1 at different points of the loop. Another pair of input 21 and output 22 ports couplers 2, 3, ..., 6 are connected by two-fiber communication lines 23, respectively, with a transmitter 17 and a receiver 18 of the subscriber units 8, 9, ..., 12. Receivers 18 of the subscriber units 8, 9 , ..., 12 are equipped with selective filters 24, 25, ..., 28 for the radiation wavelength of the transmitter 13 of the central node 7. The transmitter 13 contains a laser (not shown) with a fixed or tunable radiation wavelength.

Filters 24, 25, ..., 28 are tuned either to the same wavelength λ ₁ or to different wavelengths λ ₁₁ , λ ₁₂ , ..., λ _1N of the laser radiation of the transmitter 13.

The receiver 14 of the Central node 7 is equipped with a blocking filter 25 for the radiation wavelengths of the transmitter 13.

Directional couplers 2, 3, ..., 6, the functional diagram of which is shown in FIG. 2, have a spectrally dependent branch coefficient α _i (λ), where i is the serial number of the location of the coupler in loop 1 when looping in the direction of transmission from beginning 15 to end 16 (Figure 1)

The branch coefficients α _i (λ) increase with increasing number i of the coupler at the radiation wavelength λ ₁ of the transmitter 13 and decrease at the wavelength λ _{2 of the} radiation of the transmitter 17 of any subscriber unit 8, 9, ..., 12.

The table shows the optimal branch coefficients α _i for taps 2, 3, ..., 6, allowing you to connect the maximum possible number of subscriber nodes to the fiber optic loop 1.

Pos. in figure 1 2 3 four ... 5 6 i one 2 3 ... N-1 N α _i (λ ₁ ) 1 / N 1 / (N-1) 1 / (N-2) ... 1/2 one α _i (λ ₂ ) one 1/2 1/3 ... 1 / (N-1) 1 / N

Moreover, the number N is equal to the theoretical limit determined by the network budget

Below we prove the last statement.

An optical signal at a wavelength λ _{1 of} transmitter 13 is sequentially extracted from loop 1 to receivers 18 of subscriber units 8, 9, ..., 12. In accordance with Table 1, the first coupler 2 in loop 1 transfers power P ₁ / N. The signal to the second coupler 3 is attenuated (1-1 / N) P ₁ and the coupler 3 diverts power equal to (1 / (N-1)) (1-1 / N) P ₁ = P ₁ / N, etc. . T.O. a chain of taps 2, 3, ..., 6, included in loop 1, selects 18 subscriber units 8, 9, ..., 12 at the receivers exactly in P ₁ / N optical power. Thereby, the theoretical limit (5) is achieved in the first transmission stream: from the central node to the subscriber nodes. An optical signal at a wavelength of λ ₂ from the transmitters 17 of the subscriber units 8, 9, ..., 12 is introduced by the couplers 2, 3, ..., 6 into loop 1. In accordance with the table, the ith coupler introduces power P into line 1 ₁ / i. The optical signal propagating in loop 1 is sequentially attenuated by couplers with numbers (i + 1), (i + 2), ..., (N-1), N and is fed to the receiver 14 of the central node 7 weakened by N times, independently from number i of the sending subscriber:

T.O. and in the second transmission stream (from subscriber nodes to the central node), the number of subscriber nodes also reaches the theoretical limit (5).

Among the known directional couplers that meet the requirements of the invention, we should single out the couplers arranged according to the principle of coupled waveguides. A diagram of such a coupler is shown in FIG. 3. Optical waveguides 31, 32 located parallel to each other interact with each other by decreasing external fields. The interaction of the waveguides leads to the fact that the power of the mode of one waveguide is partially transferred to the mode of another waveguide. The power transmitted to the mode of the waveguide 32 has the form [6, p.231-236].

where F ₁ is the power at the input port 33 of the waveguide 31; Δβ = (β ₁ -β ₂ ) / 2; β ₁ , β ₂ are the phase propagation constants of the modes of the waveguides 31, 32, respectively; c is the coupling coefficient between the waveguide modes.

The coupling coefficient c has an inverse exponential dependence on the distance d between the waveguides and inversely proportional to the wavelength λ.

Figure 4 presents graphs of the runout of power F ₂ along the waveguide for two wavelengths λ ₁ = 1550 nm and λ ₂ = 1310 nm for the normalized power F ₁ = 1. The beat lengths l ₁ , l ₂ at wavelengths λ ₁ , λ _{2 are} related as follows: l ₁ : l ₂ ≈λ ₁ : λ ₂ = 1550 / 1310≈1.28. The graphs clearly show that by varying the values of c, Δβ and the bond length L (Figure 3), it is always possible to achieve branch coefficients α _i (λ ₁ ) and α _i (λ ₂ ) corresponding to the table. Note that the couplers considered are broadband devices, and ratios close to those indicated in the table will be fulfilled for two bands: 1550 ± 40 nm and 1310 ± 40 nm.

Spectrally dependent couplers are used in frequency multiplexing / separation devices (WDM-multiplexers), for which they tend to have branch coefficients α (λ ₁ ) = 1, α (λ ₂ ) = 0. As follows from expression (6) and the graphs, this is possible only if phase synchronism is observed Δβ = 0. The latter is difficult to implement in practice, and therefore WDM-multiplexers always have excess losses that reduce the level of the optical signal at the receivers of networks using WDM-multiplexing. And, as a result, the number of subscriber nodes is reduced in proportion to the losses introduced. So, the use of α (λ ₁ ) = 1/2 in the analogue of [1] leads, at least, to halving the number of subscriber nodes. At the same time, the indicated drawback of the couplers does not significantly affect the total number of subscriber nodes in the invention, since it leads to the reduction of only two extreme subscriber nodes in the loop (Fig. 1) (see Table 1).

The operation of the claimed network Figure 1 is based on the same principle as the operation of networks known as PON (Passive Optical Networks) [2, pp. 469-478].

The central node OLT (Optical Line Terminal) 7 receives data from the backbone networks through the connection interfaces SNI (Service Node Interfaces) and forms the first stream (similar to the downstream in PON) to the subscriber nodes ONU (Optical Network Unit) 8, 9, .. ., 12 by loop 1. In this case, any of the methods of flow formation known in PON technologies is used, for example, synchronous transmission with time division TDM (Time Division Multiplexing) signals intended for different subscriber units 8, 9, ..., 12. Transmitter 13 in this case operates at a fixed wavelength λ ₁ = 1550 nm, and e filters 24, 25, ..., 28 are set to the wave length λ _1. The second stream (similar to the upstream in PON) from the subscriber nodes 8, 9, ..., 12 to the receiver 14 of the central node 7 is formed by synchronous access with time division TDMA (Time Division Multiplexing Access) signals from the transmitters 18 of the subscriber nodes 8, 9 , ..., 12. In this case, the transmission is carried out at a wavelength of λ ₂ = 1310 nm and an individual schedule is set for each subscriber unit 8, 9, ..., 12, taking into account the distance of the node from the receiver 14. Selective filters 24, 25,. .., 28 at a wavelength of λ ₁ and a blocking filter 29 at the same wavelength of λ ₁ perform the function in the network separation of flows propagating in loop 1 in one direction (in FIG. 1 by clock). Filters 24, ..., 28 and 29 in this case can be of the simplest design, for example, interference in the form of multilayer coatings directly on the input windows of the receivers 18.

In a network without filters 24, 25, ..., 28 and 29, transmission is carried out by alternating time cycles of the first and second flows (low-speed network operation mode).

In the claimed network, it is also possible to transmit frequency-time packets of signals [7, p. 250]. In the first stream, the temporary positions of the signals intended for different subscriber units 8, 9, ..., 12 are transmitted at different wavelengths λ ₁₁ , λ ₁₂ , ..., λ _{1N of the} tunable transmitter laser 13. Filters 24, 25, .. ., 28 are tuned to wavelengths λ ₁₁ , λ ₁₂ , ..., λ _1N, respectively, and are structurally designed as Fabry-Perot filters. In the second stream, transmission from subscriber nodes 8, 9 ..., 12 to the receiver 14 of the central node 7 is carried out at the same wavelength λ ₂ by the method of synchronous access with time division TDMA.

It is known that a loopback (ring) network architecture with a backup transmission line provides reliable protection for the schedule. The solution to the backup task in the claimed network can be implemented in different ways. The simplest solution is to completely duplicate the main network with a backup segment with the opposite direction of transmission in the loop. It is also possible duplication of only the passive loop, or, at the same time, a partial duplication of active network components, such as transmitters in nodes. Such a solution is schematically represented in FIG. 5. The passive fiber optic network comprises a main loop 51 and a backup loop 52 with opposite transmission directions in the loop. The central node 53 includes a transmitter 57 and a receiver 58 connected to the main loop 51; and a transmitter 59 and a receiver 60 associated with the backup loop 52. Each subscriber unit 54, 55, 56 comprises a first transmitter 67, a second transmitter 68, and a receiver 69. The transmitter 67 is optically coupled to a main transmission line through a corresponding coupler 61, 62, or 63 51. The transmitter 68 through an appropriate coupler 64, 65 or 66 is optically connected to the backup line 52. The receiver 69 of each subscriber unit 54, 55, 56 is optically connected simultaneously with the main and backup loops 51, 52.

The taps 61, 62, 63 of the main loop 51 and the taps 64, 65, 66 of the backup loop 52 have spectrally dependent branch coefficients in accordance with the table.

If the optical fiber is broken in the main loop 51, the transmission is carried out through the backup loop 52 for the repair time. In the case of a simultaneous breakage of the optical fibers in loops 51, 52 (in Fig. 5 at points 70, 71), the schedule is completely restored by transmission in the “tails remaining from the loop” "As in dual-fiber linear buses. Transmitters 67 of subscriber units operate in the left “tail”, transmitters 68 in the right “tail” (in Fig. 5, the transmission route is indicated by a dotted line).

The claimed network, despite the loopback architecture, can be considered as a variant of the PON network. That is, it is fully compatible with hardware and software products released by the industry for PON technology.

On the other hand, the invention has an advantage over the well-known PON networks with a tree architecture in saving optical fiber, ease of monitoring during network operation and availability (traffic protection). Therefore, the industrial application of the invention will be a cost-effective alternative to known networks.

Literature

1. Patent US 5574584, CL 359-125, 1996 publ.

2. Freeman R. Fiber-optic communication systems. 2nd additional edition: Per. from English under the editorship of N.N.Slepova. - M .: Technosphere, 2004

3. Patent RU 2264692, Н04В 10/12, 2005, publ.

4. Patent GB 2154091, Н04В 9/00, 1985, publ.

5. Patent US 5576875, cl. 359-125, 1996 publ.

6. Unger H.-G. Planar and fiber optical waveguides. / Per. from English V.V.Shevchenko. - M .: Mir, 1980

7. Telecommunication systems and networks: Textbook in 3 volumes. Volume 3. - Multiservice networks; under the editorship of prof. V.P. Shuvalova. - M .: Hot line - Telecom, 2005

Claims (4)

1. A passive fiber optic network containing a unidirectional fiber optic loop and N directional couplers located at different points of the loop, a central node and a plurality of subscriber nodes, the transmitter of the central node is optically connected to the beginning of the fiber optic loop, and the receiver to its end , the transmitter and receiver of each subscriber unit are optically connected to the fiber optic loop through a corresponding directional coupler, each coupler is included in one of the input and output ports a window-optical loop, and another pair of input and output ports is designed to connect respectively to the transmitter and receiver of the subscriber unit, characterized in that the directional couplers from the first to the last, in the order of their arrangement from the beginning to the end of the fiber-optic loop, are made with an increase in the branch coefficients at the radiation wavelength of the transmitter of the central node and with a decrease in the branch coefficients at the radiation wavelength of the transmitter of any subscriber unit. 2. The network according to claim 1, characterized in that the directional couplers are made with increasing branch coefficients in the sequence: 1 / N, 1 / (N-1), ..., 1/3, 1/2, 1 at a wavelength radiation of the transmitter of the central node and with decreasing branch coefficients in the sequence: 1, 1/2, 1/3, ..., 1 / (N-1), 1 / N at the radiation wavelength of the transmitter of any subscriber unit. 3. The network according to claim 1 or 2, characterized in that each directional coupler is made on two connected optical waveguides, the ends of which form two pairs of input and output ports. 4. The network according to claim 1, characterized in that the receiver of each subscriber unit is equipped with a selective filter for one or another radiation wavelength of the transmitter of the central node, and the receiver of the central node is equipped with a blocking filter for the same wavelengths.

Images