RU2779296C1 - Method for redundancy in an optical communication network with wavelength division multiplexing systems - Google Patents

Abstract

FIELD: communication.

SUBSTANCE: invention relates to communication equipment and can be used in optical communication networks (OCN). To realise the technical result, the structure of an optical communication network is pre-formed as to provide at least three independent routes between each pair of corresponding nodes, with a minimum total consumption of the length of the optical cable; the maximum permissible value of the level of received optical signal, whereat a permissible error coefficient is provided, is set; the level of optical signal is monitored in the optical fibres of the lines of the optical communication network; if the optical power in the optical fibres of the optical cable included in the main route is reduced or lost, an automatic switch to one of the two remaining routes is performed in the fibres of optical cables with a sufficient level of the power of optical signal.

EFFECT: increase in the reliability of OSS due to the use of two other routes with a simultaneous reduction in the level of optical signal.

1 cl, 10 dwg

Description

The invention relates to optical systems, namely to communication networks, and can be used in existing and emerging optical communication networks (OSN).

A known method of redundant information flows (Shmalko A. V. Digital communication networks: the basics of planning and construction. - M.: ECO-TRENDS, 2001, pp. 186-188), which consists in installing backup optical fibers on each main optical fiber of the communication network and transmission of optical signals only over the main fiber. In the optical fiber, the level of optical power is measured and, when it is lost, a control signal is generated, at the command of which the signal is transmitted through the backup optical fiber.

The disadvantage of this method of redundancy is the high cost of its implementation due to the need for complete duplication of network elements.

A known method of redundancy in OSS with WDM systems (see US patent No. 6046832 dated 04/04/2000, IPC 7 H04V 10/20), in which a protection ring with two or more nodes of the communication network is isolated from a plurality of rings of a synchronous fiber-optic communication network , containing WDM terminals. In the presence of WDM terminals, each node of the protection ring installs a backup WDM terminal, an optical route switch, and an optical multiport switch. The outputs of the main and backup WDM terminals are connected to the inputs of the optical route switch, the outputs of which are connected to the optical fibers that make up the protection ring. The inputs of the optical multiport switch are connected to the outputs of the input-output multiplexers of the synchronous fiber-optic communication network. In addition, the inputs of the backup WDM terminals are connected to the outputs of the optical multiport switch.

During the operation of the communication network in optical fibers, the level of optical power is measured and, if it is lost, a control signal is generated, at the command of which the backup WDM terminals are turned on. The inputs of the backup WDM terminals located at the same node are interconnected through an optical multiport switch. Then the input of the optical route switch, which receives information traffic from the output of the WDM terminal, is switched to the backup optical fiber of the protection ring, through which this information traffic is transmitted to the communication network node on which the WDM terminal is installed.

This method of redundancy allows you to protect information flows transmitted over WDM lines, which generally increases the reliability of the communication network.

The disadvantage of this method is that its implementation requires high material costs. This is due to the need to use additional functional devices and redundant optical fibers in each section of the communication network when implementing this method.

The closest in technical essence to the claimed invention (prototype) is a redundancy method in a synchronous fiber-optic communication network with spectral division multiplexing systems (see RF patent No. 2307469, C1. class H04V 10/00, publ. . The known method consists in setting for each segment of the optical cable b _m , where m=1, 2, ..., M, and M is the total number of segments of the optical cable that unites the nodes of the communication network, the probability of failure-free operation p _m and allocate pairs of corresponding nodes Z _ij , where i=1, 2,…, N, j=1, 2,…, N, i≠j and set the maximum allowable value of the rank γ of _additional paths between them, allocate sets U ₁ , U ₂ , …, U _λ , …, U _Λ segments of the optical cable of the communication network, where Λ is the total number of sets, the failure of each of which breaks all paths between the corresponding nodes of the communication network and the total indicator of structural reliability of which P _λ <P _add , where P _add - the minimum allowable value of the total indicator of the structural reliability of a set of optical cable segments, moreover, the WDM terminals and optical route switches are installed at nodes incident to the segments of the optical cable of the communication network that make up the selected sets, measure the level and the power passing in the optical fibers, with the loss of optical power in one of the optical fibers of the optical cable segment included in one of the selected sets U _λ of the optical cable segments of the communication network, the generated control signal is transmitted to the control inputs of the WDM terminal of the optical route switch and the optical multiport switch, and according to the received control signals, the WDM terminal is switched on, the input of the optical route switch, which receives information traffic from the I / O multiplexer, is switched with the output connected to the input of the WDM terminal, the output of which is connected to the input of the optical multiport switch, the input which is then switched to a working optical fiber, through which this information traffic is transmitted to the communication node on which the WDM terminal is installed.

The disadvantage of the prototype method is the high material costs for building a communication network while maintaining its reliability. This is due to the fact that the nodes in the communication network Z _ij are mutually corresponding and in the event of a loss of optical power in one or more optical fibers of the optical cable segments included or included in one or more of the selected sets U _λ , switching to another working fiber may be difficult in view of the lack of connection to optical route switches of other optical fibers of optical cable segments included in independent routes for delivering traffic between communication nodes. In addition, if other optical fibers are connected to optical route switches, optical cable segments can be difficult to switch to another working fiber due to the transmission of traffic at established wavelengths between other corresponding nodes in the network.

The technical result of using the claimed redundancy method in an optical communication network with WDM systems is to reduce material costs for building a communication network while maintaining its reliability.

The technical result is achieved by the fact that in the known method of reserving communication between corresponding nodes in an optical communication network containing N> 4 nodes connected by fiber optic lines with spectral division of channels, which consists in the fact that pairs of corresponding nodes Z _ij are selected, where i= 1, 2, (,N, j=1, 2, ..., N, i≠j, measure the power levels coming in optical fibers, additionally pre-form the structure of the optical communication network so that between each pair of corresponding nodes Z _ij , there were at least three independent routes with a minimum total consumption of the length of the optical cable.Set the maximum allowable value of the level of the received optical signal p _c >p _add , where p _add is the minimum allowable value of the optical signal level, at which the allowable error rate is provided.Monitor the level of the optical signal in the optical fibers of the lines of the optical communication network.In the event of a decrease or loss of op optical power in the optical fibers of the optical cable p _c <p _ext included in the main route, automatically switch to one of the two remaining routes, in the fibers of the optical cables the power level of the optical signal which ensures the condition P _c >P _ext .

Thanks to a new set of essential features, the method implements the possibility of automatically switching to one of the two remaining traffic delivery routes between the corresponding nodes in an optical communication network built with a minimum total consumption of the optical cable in the event of a decrease or loss of optical power in the optical fibers of the optical cable included in the main route, which achieves a reduction in material costs for building a communication network while maintaining its reliability.

The claimed method is illustrated by drawings, which show: figure 1 - typical structures of the OSS: annular and radial nodal; figure 2 - the structure of the OSS, built by the union of the annular, radial-nodal structures and a separate edge ρ(i -k);

fig. 3 - ring structure of the OSS fragment without nodes i and k ensures the formation of one independent route between each pair of nodes; fig. 4 - radial-nodal structure of the OSS fragment without node j provides the formation of two independent routes between each pair of nodes;

fig. 5 - the ring structure of the OSS fragment without the edge ρ(i - k) provides the formation of two routes between node i and all other nodes i-j and j - k formed along one route π(i - j) and π(i-k);

fig. 6 - fragment of the OSS provides the formation of two routes between nodes i and j π ₁ (i - k- j) and π ₂ (i - (i -1) - j);

fig. 7 - fragment of the OSS provides the formation of two routes between nodes j and k π ₁ (jik) and π ₂ (j-(k+i)-k) and one route between nodes i and k ((i - j -k);

fig. 8 - the structure of the OSS, formed by combining the fragments of the network presented in Fig.3 ... Fig.7;

fig. 9 - an example of the structure of the OSS with an indication of the number of optical wavelengths that should be implemented on each of its edges and determined through the number of network nodes N=10;

fig. 10 - a set of fragments of the structure of the resulting OSS explaining the essence of the formation of the number of optical wavelengths in each of the components.

где М - общее число ребер сети. Расчет структуры сети начинается с составления матрицы расстояний |R|, в которой записывают измеренные расстояния между всеми узлами. Если между какими-либо узлами измерения и прокладку линии выполнить невозможно (имеются естественные или искусственные преграды), то длину такой линии в матрице расстояний принимают много больше самой длинной линии. Каждую из подсетей (кольцевая без ребра ρ(k-i) и радиально-узловая) рассчитывают с использованием отдельных оптимальных алгоритмов. Расчет кольцевой структуры выполняют с помощью алгоритма «Коммивояжер» [Гэри М., Джонсон Д. Вычислительные машины и труднорешаемые задачи: Пер. с англ. - М.: Мир, 1982. - 416 с., ил.]. Данный алгоритм относится к классу NP и позволяет выполнять расчет в рамках заданного критерия с определенной точностью, зависящей от его сложности. Радиально-узловую структуру рассчитывают с помощью алгоритма «минимальная медиана графа» [Кристофидиес Н. Теория графов. Алгоритмический подход. М.: Мир, 1978. Ху Т. Целочисленное программирование и потоки в сетях. - М.: Мир, 1974]. Он относится к классу полиномиальных алгоритмов и обеспечивает точный результат за приемлемое время работы компьютера.When creating communication networks for various purposes, typical structures are used: ring and radial-nodal (figure 1) [Gary M., Johnson D. Computers and difficult tasks: Per. from English. - M.: Mir, 1982. - 416 p., ill. Adelson-Velsky G. M., Dinits E. A., Karzanov A. V. Stream algorithms. - M.: Nauka, 1975]. The structure of an optical communication network with WDM systems and its graph G(N, M) is a combination of two typical structures, ring and radial-nodal, into a single network with the addition of a separate edge (figure 2). A separate edge is always connected between the same nodes k, i and is included in each of the two ring subnets (figure 2). This edge plays an important role in the formation of three independent routes between any two nodes and allows independent transmission of optical signals along each of the routes. These three subnets are built using the criterion of the minimum total amount of optical cable required for their construction

where M is the total number of network edges. The calculation of the network structure begins with the compilation of the distance matrix |R|, in which the measured distances between all nodes are recorded. If it is impossible to measure and lay a line between any nodes (there are natural or artificial barriers), then the length of such a line in the distance matrix is assumed to be much larger than the longest line. Each of the subnets (ring without an edge ρ(ki) and radial-nodal) is calculated using separate optimal algorithms. The calculation of the ring structure is performed using the "Traveling Salesman" algorithm [Gary M., Johnson D. Computers and difficult tasks: TRANS. from English. - M.: Mir, 1982. - 416 p., ill.]. This algorithm belongs to the NP class and allows you to perform a calculation within a given criterion with a certain accuracy depending on its complexity. The radial-nodal structure is calculated using the algorithm "the minimum median of the graph" [Christofides N. Theory of graphs. Algorithmic approach. M.: Mir, 1978. Hu T. Integer programming and flows in networks. - M.: Mir, 1974]. It belongs to the class of polynomial algorithms and provides an accurate result in an acceptable computer time.

Routing in OSS with WDM systems is performed by dividing the network into separate subnets, followed by solving the routing problem in each of them. At the end, all subnets are added, and the routes are summed on each edge of the network. Since the graphs of the proposed networks (figure 2) are invariant (the network dimension is different, but the structure is the same), the routing problem for each of them has the same type of solution, which differs only in the network parameter N. This allows for a network with an arbitrary value of N, the routing problem is not solved again , but only modify it taking into account the value of N given for the given network. The division of the OSS with spectral multiplexing systems occurs into two annular parts and a radial-nodal part. Routes in each of the subnets are found in compliance with the following rule: each node is connected to all other nodes by independent routes, i.e. Each route must not contain common nodes). At the end, the network fragments are combined and the total number of routes (S) passing through each of its edges is determined. The number S cannot be greater than the value N(N - 1)/2, which is determined by the enumeration problem, when each element of the set is associated with all its other elements. However, taking into account the presence of three independent routes, the number S is tripled. The first ring subnetwork is shown in Fig.3, in which there is no node y and it includes the edge ρ(i - k). Figure 3 shows two routes between nodes i and (k-1) (the first route and the second route). Similarly, two independent routes are formed between any pair of network nodes. Their total number is defined as (N-1)(N-2), and the total number of routes passing on each edge of the network will be half as much, i.e. (N-1)(N-2)/2. Figure 4 shows a radial-nodal subnet, in which there are no nodes i and k. It forms one route between all pairs of OSS nodes. Routes are formed by a common node j and two separate edges, and between the central node j and all other nodes using one edge connected to this node. The total number of routes is defined as (N-2)(N-3)/2. The number of routes passing through each edge of the subnet is A - 3. Figure 5 shows a fragment of the annular part of the OSS, in which the edge ρ(i-k) is excluded. In this part of the network, two routes are formed between node j and all other nodes of the network (for example, the formation of two routes between nodes j and r). The number of such routes is equal to (N-1)×2, and the total number of routes on each edge is two times less, i.e. (N - 1). Between nodes i, j and j, to form one route of the form π(i - j), π(j - k). Two other routes between these nodes are formed in the next two fragments of the network (Fig.6 and Fig.7) Fragment (Fig.6) provides the formation of two routes between nodes i and j [π(i - k -j); ρ(i - (i - 1) -j)]. Fragment (Fig.7) provides the formation of two routes between nodes j and k [π(j - i - k); π(j-(k+1) - k)] and one route between nodes i and k π[(i -j - k)], (two other routes between these nodes were formed in the subnet, Fig.3). The process of forming three independent routes between all pairs of nodes of the original network (Fig.3...Fig.7) is completed by summing the number of routes of all subnets on each edge (Fig.8). This makes it possible to solve the problem of choosing the parameters of the DWDM technology, the composition and number of optical equipment at each switching center and on the communication lines connected to it. Independent routes between each two switching centers allow the transmission of optical signals without the use of optical converters (transformers and transformers). In the redundancy method in OSS with WDM systems, this is achievable between all pairs of the network without limiting its dimension using only optical multiplexers. Moreover, the number of optical wavelengths transmitted over three independent routes can be increased if the capacity of optical cables and the capabilities of multiplexers are not limited. The principle of forming routes for the OSS structure remains unchanged for any of its dimensions and different load values.

The task of load distribution in the redundancy method in OSS with WDM systems is determined depending on the purpose of the network, the number of its switching centers and other parameters. When distributing the load, the load matrix |Z| _h , which prescribes how many optical wavelengths and between which nodes should be installed (implemented). The task of load distribution in the theory of algorithms refers to the NP cash register and can only have an approximate solution. When solving it, to obtain an acceptable result, it can be repeated several times. In the redundancy method in OSS with WDM systems, taking into account the network structure and route formation features (Fig.3...Fig.7), the load between any two centers is determined only by changing the flow between these centers along one of three independent routes. A distinctive feature of the network structure is that it is not required to solve the problem of routing and load distribution every time. The routing problem always has the same solution, depending only on the number of network nodes N, and the load is distributed over three routes, regardless of its size and transmission methods.

Calculation of the number of wavelengths formed on each edge of the optical network providing the implementation of a given traffic. With the number of nodes in the network N, the number of routes is A=[N ((N-1)/2]×3. The total number of routes passing on each edge of the network is different and is found by adding the subnets and the number of their routes. Figure 8 shows the result combining subnets (fig.3...fig.6) and the total value of routes on each edge of the proposed network

In the annular part, two independent routes are implemented and one in the radial-nodal one. The differences in their formation lies in the fact that in the radial nodal part they form one route (one wavelength is transmitted), and in the annular part of two routes (two wavelengths are transmitted). The wavelengths of all routes combine what happens within one waveguide, one cable or cables laid on a given edge of the network. The required number of wavelengths in the radial part of the network required to form the load matrix |z| when the number of nodes N, is defined as D=N(N-1)/2. The ratio between the required number of waves, the number of fibers and the selected technology (the number of wavelengths implemented in each fiber) is D≤L×S, where L is the number of fibers in the optical cable(s), S is the number of wavelengths implemented in each optical waveguide. In the radial-nodal part of the network, two separate lines are laid between any two nodes, connected at the node y, which provide the formation of the third wavelength (as an example, it is assumed that between any two network nodes three optical signal lengths are realized). In addition, by increasing the number of wavelengths in each optical fiber of this part of the network, a significant increase in its reliability and throughput of communication directions is achieved.

An example of calculating the parameters of the OSS with DWDM technology with the initial data: the number of nodes (switching centers) of the network N=10; network load - three wavelengths of an optical signal must be transmitted between each pair of nodes; Signals of the same wavelength are transmitted over three independent paths.

It is required to define: routes between each pair of switching centers; distribution of the required number of optical signals (wavelengths) passing along these routes; required total signal capacities on each edge of the network; the number of optical waveguides that make up the cable, as well as a type of DWDM technology that allows you to provide the required number of wavelengths in each section of the network. Determining routes in a network can be done using routing algorithms. However, since the graph of the proposed network is invariant and the routing solution is found (Fig.8), it makes sense to bring the network graph (Fig.9) indicating the total wavelengths on each edge of the network. An analysis of the distribution of wavelengths along the edges shows that it is not uniform and differs significantly in its value (7 is the minimum, 46 is the maximum). The choice of the capacity of the optical cable (the number of its fibers), as well as the features of the DWDM technology, is determined using the ratio D≤L (S. It shows that the cable can contain only one optical waveguide (L) and this is enough to build a network (DWDM technology allows the formation (S>46)) It is possible when L=5, and S>10. Then a cable with a different number of fibers will be laid in different sections of the network. In addition, to build an OSS, you will need to have three optical multiplexers at each switching center ( The number of communication directions in node 1 is 9, which will require the center to have the same number of optical multiplexers.

Monitoring the level of the optical signal in the optical fibers of the OSS lines and assessing the quality of communication can be carried out in various ways, for example, as described in the "Method for monitoring a communication channel and a transmitting device" (see RF patent No. 2460223, C1. class H04L 12/56, publ. . 27.08.2012), and each time a new measurement channel is selected, the first transmitter and the second transmitter sequentially transmit a plurality of test signals with different optical wavelengths to a new measurement channel.

The use of the proposed method ensures the formation of at least three independent routes between all communication nodes, which significantly increases the reliability of the OSS while reducing the material costs of its construction by using two other routes if the first one for some reason has reduced the level of the optical signal or disappeared altogether.

Claims (1)

A method for reserving communication between corresponding nodes in an optical communication network containing N≥4 nodes connected by fiber-optic lines with spectral division of channels, which consists in the fact that pairs of corresponding nodes Z _ij are selected, where i=1, 2, ..., N, j=1, 2,…, N, i≠j, measure the power levels coming in the optical fibers, characterized in that the structure of the optical communication network is preliminarily formed in such a way that there are at least three independent routes between each pair of corresponding nodes Z _ij at the minimum total consumption of the length of the optical cable, set the maximum allowable value of the level of the received optical signal p _c >p _add , where p _add is the minimum allowable value of the optical signal level at which the allowable error rate is ensured, monitor the level of the optical signal in the optical fibers of the optical lines communication networks, in the event of a decrease or loss of optical power in the optical fibers of the optical cable p _with <p _ext included in the main route, carry out automatic switching to one of the two remaining routes, in the fibers of optical cables, the power level of the optical signal which ensures the condition p _with >p _ext .

Images