Method and apparatus for validating a path through a switched optical network

[0001] This application claims the benefit of Provisional application Ser. No. 60/210,211 filed June 8, 2000.

[0002] The present invention relates to switched optical networks. More specifically, the present invention relates to a method and apparatus for the validation and selection of light paths through a switched optical network.

[0003] Activity in the optics industry has been directed towards the development of a Wavelength-Division-Multiplexed (WDM) Automatically Switched Optical Network (ASON), which allows client network devices such as IP routers and ATM switches to automatically request bandwidth from the network when the need arises. Light-path (i.e., wavelength) connections in an ASON environment may be required to traverse a number of fiber links. These links may include optical amplifiers (OAs) and optical cross-connects (OXCs) which add noise and distort signal-bearing light pulses. Thus, the network is faced with the problem of delivering an acceptable level of performance for the connection.

[0004] One approach to address the quality of service for a connection is to architect a network topology such that every reasonable physically realizable light path delivers an adequate level of performance with little or no impact on every other light-path. Such a paradigm is known as “Islands-of-Transparency”, since the ASON is divided into a set of smaller sub-networks or “islands”. Within these sub-networks, paths are capable of delivering some universally acceptable quality of service. The sub-networks are required to be interconnected via optical-electronic-optical cross-connects (OEO-OXCs) which are capable of regenerating the signals in the electronic domain, thereby reducing undesirable optical transmission impairments. Network clients that reside in different sub-networks are required to establish paths through the OEO-OXCs regardless of their physical proximity. The large number of OEO-OXC devices that are required to support the “Island of Transparency” network topology make the topology costly to implement.

[0005] Thus, what is required is a method and apparatus for determining selecting and validating a path/connection through a switched optical network that is capable of delivering an acceptable level of performance that is both efficient and economical.

[0006] A method for validating a path through a switched optical network according to a first embodiment of the present invention is disclosed. A bit error rate for the path is determined. The path is validated or admitted into the network if the bit error rate is found to be within a predefined range.

[0007] A second method for validating a path route through a switched optical network according to a second embodiment of the present invention is disclosed. A cumulative optical signal to noise ration (SNR) at the output of every element along the path is determined while the path is being set up. The path's route is modified if the SNR at the output of any element on the path is outside a predefined range. The path is validated or admitted into the network if the bit error rate is within a predefined range.

[0008] A path validation unit is disclosed according to an embodiment of the present invention. The path validation unit includes a bit error rate processor that determines a bit error rate of a path through a switched optical network.

[0014] In the following description, for purposes of explanation, specific nomenclature is set forth to provide a thorough understanding of the present invention. However, it will be apparent to one skilled in the art that these specific details may not be required to practice the present invention. In other instances, well-known circuits and devices are shown in block diagram form to avoid obscuring the present invention unnecessarily.

[0015] FIG. 1 is a block diagram illustrating an optical network 100 utilizing an embodiment of the present invention. The optical network 100 is a switched optical network that includes a first sub-network 130 and a second sub-network 140. The first and second sub-networks 130 and 140 include a plurality of elements such as cross-connects, regenerators, optical amplifiers, wavelength multiplexors, optical fiber links and other network elements that transmit, route, and process optical signals. Blocks 131–134 and 141–143 may represent optically transparent photonic cross-connects (PXCs) or Add/Drop Multiplexors (OADMs) or Optical-Electronic-Optical cross-connects (OEO-OXCs) or Add/Drop Multiplexors (OEO-ADMs) in the optical network 100. Lines 151–157 and 161–164 may represent optical fiber links, regenerators, optical amplifiers or other devices in the optical network 100 that connect various OXCs and ADMs to one another. Blocks 110 and 120 represent clients that request a path connection through the optical network 100. The clients may be IP routers, ATM switches or other devices.

[0016] According to an embodiment of the optical network 100, blocks 134 and 141 represent OEO-OXCs or OEO-ADMs. Such OEO devices include OEO transducers that are used to regenerate an optical signal in electrical domain and, subsequently, convert it back to an optical signal that may be transmitted to the next device. By terminating an optical signal and converting it to an electrical signal, optical impairments in the optical signal may be reduced or eliminated. OEO-OXCs and OEO-ADMs, however, are more costly to implement than PXCs and OADMs, respectively. Thus, it is desirable to allocate OEO-OXC and OEO-ADM functionality in an economical manner.

[0017] According to an embodiment of the present invention, the optical network 100 includes a path validation unit (not shown) that checks a proposed path route through a network to determine whether an optical signal through the path may be provided with sufficient quality of service. The path validation unit calculates a bit error rate of the path through the optical network 100. If the bit error rate is within a specified range that is associated with an acceptable level of quality of service, the path is judged valid. If the optical network were operated according to the “Islands-of-Transparency” route selection rules, optical signals that do not originate and terminate in the same sub-network would be required to be terminated and converted to electric signals at an OEO-OXC or OEO regenerator or OEO-ADM before being transmitted between the sub-networks in order to ensure adequate quality of service. By determining the bit error rate of a proposed path, the path validation unit allows paths that travel between sub-networks that do not traverse an OEO device to be utilized without sacrificing quality of service. The path validation unit is capable of determining whether or not a given path between two network clients may be established transparently (without converting the optical signal to an electrical signal) with the desired quality of service. If no transparent path exists, a route may be created that traverses an OEO device, such as an OEO-OXC. The path validation unit, in effect, transforms a plurality of sub-networks into a single network with end-to-end transparent connectivity.

[0018] According to an embodiment of the present invention, the path validation unit may be implemented in a central connection controller or a plurality of central connection controllers (not shown) that is assigned to the network or each sub-network. Alternatively, the path validation unit may be implemented in any one or all of the cross-connects or ADMs in the optical network 100. FIG. 1 illustrates an exemplary optical network 100 with wavelength routing network elements (NEs) 131–134 and 141–143 and links 151–157 and 161–164. It should be appreciated that the present invention may be implemented in a network having any number of NEs and links.

[0019] FIG. 2 illustrates a block diagram of a path validation unit 200 according to an embodiment of the present invention. In a preferred embodiment of the present invention, the path validation unit 200 is implemented by software and resides in a memory (not shown) as a sequence of instructions. It should be appreciated that the path validation unit 200 may also be implemented using hardware or a combination of both hardware and software.

[0020] The path validation unit 200 includes a noise figure processor 210. The noise figure processor 210 determines a noise figure for each element in a path. Elements may include any NE or link in the path. According to an embodiment of the path validation unit 200, the noise figure for each element i is determined based on an equivalent average number of noise photons per bit at an input of the element n_i^eq, an average number of photons per bit at the output of the previous element <n_i−1>, an optical power gain of the element G_i, a constant V_i related to deviation of statistical noise distribution, and a number of modes M detectable by the receiver at an end of the path.

[0021] According to one embodiment of the present invention, n_i^eq is the equivalent average number of noise photons per bit at the input of the “i-th” link: n_i^eq≡<N_i>/G_i=n_i^sp+1/G_i with <N_i> as the per mode amplified average spontaneous emission photon number, and n_i^sp is commonly referred to as the OA's spontaneous emission factor, which is equal to 1 for an ideal amplifier (with complete medium inversion). Average number of photons at the “i-th” input, which equals the average number of photons at the “(i−1)st” output, may be computed recursively <n_i−1>=G_i−1(<n_i−2>+n_i−1^eq) with <n₁>=G₁(<n₀>+Mn₁^eq) where <n₀> is the average number of photons per bit out of the transmitting laser. In this embodiment, the noise figure for an element may be computed as follows NF_i^Poisson=2n_i^eq+1/G_i+[1+n_i^eqV_iG_i][Mn_i ^eq/(G_i<n_i−1>)].

[0022] The path validation unit 200 also includes a total noise figure processor 220. The total noise figure processor 220 receives the noise figures for elements on the path and generates a total noise figure for those elements. According to an embodiment of the validation unit 200, the total noise figure processor 220 generates a total noise figure based on the noise figures of all individual element on the path. According to an embodiment of the present invention the total noise figure for a path may be computed as follows NF_i^Total=NF_i−1^Total+A_i−1(NF_i^Poisson−1), where A_i≡A_{i −1}(g_i/G_i) is also defined recursively with excess noise factor g_i=1+M(n^ieq/<n_i−1>). Initially, NF^Total₀=1 and A₀=1/F₀, where F₀ is the Fano factor of light out of the transmitting laser (A₀=1 for a shot-noise limited source). The optical SNR at the receiver is given by SNR_k=NF_k^Total/SNR₀ for a path containing a chain of k elements. In this embodiment, the entire signal path from the source transmitter (client 110 shown in FIG. 1) to the destination receiver (client 120 shown in FIG. 1) is viewed as a chain of lumped or distributed elements whose emission and absorption coefficients follow periodic distribution along the path. Power loss is incorporated in this model by allowing the element's optical power gain G_i to dip below the value of one (“i” indicates the element's position in the chain). The net contribution of all the nonlinear interactions (and the spontaneous emissions inside elements) is reflected in the degradation of the optical SNR. The noise propagation problem is solved via linearization approximation around the average powers that are computed by routing software for components in the path. Since in a WDM-enabled ASON a new circuit, which traverses the same physical link, may be turned on at any time, the worst case assumptions is used regarding presence of other WDM-channels on every link of interest. In such a linearized model the noise contribution of every transmission or switching element is summarized by its optical Noise Figure NF_i^Poisson=SNR_in^Poisson/SNR_i, where SNR_i is the SNR at the output of the “i-th” element. Since the optical SNR evolves along the path, NF_i is defined with respect to a consistent state of noise on the input, thus making NF_i independent of its position in the chain, i.e., a true device parameter. A shot-noise limited (i.e., Poissonian) input signal is assumed in the definition of SNR_in^Poisson, which is equal to the average number of photons per bit at the input to element “i”. With this definition a modified Friis' formula for a cumulative noise figure, which summarizes noise performance of a series connection of optical elements “1” through “i” is derived. This may be written recursively as shown above.

[0023] The path validation unit 200 includes a transmitter parameter unit 230. The transmitter parameter unit 230 contains or obtains information in connection with the transmitting clients connected to the network, such as the transmitter optical SNR (i.e., SNR₀) and Fano factor, transmitter signal power <n₀>, and other relevant parameters.

[0024] The path validation unit 200 includes a receiver parameter unit 240. The receiver parameter unit 240 contains or obtains information in connection with the receiving clients connected to the network such as the required power of a signal at the receiving client, receiver bandwidth and responsivity, number of modes M, thermal noise, decision threshold and other relevant parameters.

[0025] The path validation unit 200 includes a bit error rate processor 250. The bit error rate processor 250 is coupled to the total noise figure processor 220, the transmitter parameter unit 230, and the receiver parameter unit 240. The bit error rate processor 250 generates an estimated optical SNR (i.e., SNR_k) and bit error rate at the receiver for a path route based on the total noise figure received from the total noise figure processor 220, transmitter parameters received from the transmitter parameter unit 230, and detection parameters received from the receiver parameter unit 240.

[0026] The optical network 100 shown in FIG. 1 may utilize a topology distribution protocol, such as OSPF or IS—IS that are in common use in IP networks and are proposed for use in optical networks as well, in routing paths through the optical network 100. In this embodiment, physical parameters for elements in the network, such as the equivalent average number of noise photons per bit at an input of an element n_i^eq, the optical power gain of an element G_i, and the constant V_i related to deviation of the statistical noise distribution, etc. of an element, may all reside in a central connection controller that resides in any one or a plurality of NEs in the optical network 100 or in one or more centrally-located connection controllers that reside external to any NE. The connection controller incorporates the functionality of the path validation unit 200, as described above.

[0027] Alternatively, the optical network 100 may utilize a signaling protocol such as RSVP or CR-LDP that are in common use in IP networks and are proposed for use in optical networks as well in establishing paths through the optical network 100. In this second embodiment, characteristics for an element, such as the equivalent average number of noise photons per bit at an input of the element n_i^eq, the optical power gain of the element G_i, and the constant V_i related to deviation of the statistical noise distribution of the element, etc., are not distributed to a centralized connection controller in the network. Instead, each NE includes a path validation unit that calculates a noise figure for itself and, perhaps, any component connected to it that is not capable of calculating its noise figure. In this embodiment, a total noise figure for a path may be calculated by each NE on a path route using cumulative noise-related information received from a previous element on the path. If the resulting optical SNR or the bit error rate at an output of an element along the path falls outside a predefined range, the path's route may be modified during the path establishment phase.

[0028] It should be appreciated that in evaluating a potential path through the network, the path validation unit 200 will also consider other criteria such as criteria that affect the shape of propagating optical pulses. The path validation unit 200 may consider, for example, self-phase modulation, group velocity dispersion, polarization mode dispersion, optical amplification transient-suppression performance, and other criteria that affect the shape of the optical signal pulses.

[0029] FIG. 3 is a flow chart illustrating a method for validating a path through a switched optical network according to a first embodiment of the present invention.

[0030] At step 301, parameters at the transmitter are determined. According to this embodiment of the present invention, transmitter's average number of photons per bit <n₀> and optical SNR₀ and Fano factor are determined in addition to pulse-shape related information, which may include modulation format, pulse chirp, etc.

[0031] At step 302, parameters at the receiver are determined. According to an embodiment of the present invention, the parameters at the receiver include the required optical SNR (and the corresponding bit error rate, which depends on pulse-shape integrity and power, detector bandwidth and thermal noise properties, decision threshold, etc.) and the number of detected modes M for a given bit rate connection in addition to pulse-shape related requirements, which include signal power, modulation format, pulse-shape integrity, etc. At step 303, a path route through a switched optical network is generated. The path route may be generated using any known route-generating algorithm. If no more paths can be generated the path connection fails (not shown in FIG. 3).

[0032] At step 304, a noise figure is determined for each element on the path. According to an embodiment of the present invention, the noise figure for each element i is determined based on physical parameters: equivalent average number of noise photons per bit at an input of the element n_i^eq, an average number of photons per bit at an output of the previous element <n_i−1>, an optical power gain of the element G_i, a constant V_i related to deviation of the statistical noise distribution, and a number of modes M detectable at an end of the path. According to an embodiment of the present invention, the noise figure for an element may be computed as NF_i^Poisson=2n_i^eq+1/G_i+[1+n_i^eqV_iG_i][Mn_i^eq/(G_i<n_i−1>)].

[0033] At step 305, a total noise figure is determined for the path route based on the noise figure for each element of the route. According to an embodiment of the present invention, the total noise figure is computed as NF_i^Total=NF_i−1^Total+A_i−1(NF_i^Poisson−1), where A_i≡A_i−1(g_i/G_i) is also defined recursively with excess noise factor g₁=1+M(n_i^eq/<n_i−1>). Initially, NF_Total0=1 and A₀=1/F₀, where F₀ is the Fano factor of light out of the transmitting laser (A₀=1 for a shot-noise limited source).

[0034] At step 306, an optical SNR at the receiver is determined. The optical SNR at the receiver is given by SNR_k=NF_k^Total/SNR₀ for a path containing a chain of k elements.

[0035] At step 307, it is determined whether the optical SNR for the path route that was determined in the previous step is equal to or greater than the one required by the receiver. If not, control proceeds to step 303, otherwise control proceeds to step 308.

[0036] At step 308, a bit error rate is determined based on the optical SNR determined in step 306, parameters of the transmitter, and parameters of the receiver.

[0037] At step 309, it is determined whether the bit error rate for the path is within a predefined range. If the bit error rate for the path is outside the predefined range, control proceeds to step 303. If the bit error rate is within the predefined range, control proceeds to step 310. At step 310, the path is validated.

[0038] FIG. 4 is a flow chart illustrating a method for validating a path through a switched optical network according to the second embodiment of the present invention.

[0039] At step 401, parameters at the transmitter are determined. According to this embodiment of the present invention, transmitter's average number of photons per bit <n₀> and optical SNR₀ and Fano factor are determined.

[0040] At step 402, parameters at the receiver are determined. According to an embodiment of the present invention, the parameters at the receiver include the required optical SNR (that corresponds to a required bit error rate, which also depends on pulse-shape integrity and power, detector bandwidth and thermal noise properties, decision threshold, etc.) and the number of detected modes M for a given bit rate. The list of the prohibited “nearest neighbor NEs” is empty following this step.

[0041] At step 403, the next “(i+1)-th” NE that can be reached from the present “i-th” NE (or the original transmitter) by a single hop along a tentative path route through a switched optical network is generated. The single hop route to the next NE may be generated using any known route-generating algorithm. It is possible to specify to a procedure in this step (not shown on FIG. 4) a list of “nearest neighbor NEs” that should not be picked as the next NE (this is done to prevent the hop-by-hop routing procedure from running into loops). The next step in this algorithm will be performed at the next NE (i.e., “i” is incremented).

[0042] If the list of the prohibited “nearest neighbor NEs” includes every plausible nearest neighbor, “i” is decremented and the same step is performed at the previous NE. If the previous NE is the transmitter itself, the control passes to step 413 (not shown on FIG. 4).

[0043] At step 404, a noise figure NF_i^Poisson is determined for the current NE on the tentative path (including calculation on behalf of any attached links and devices that are incapable of performing the calculation). According to this embodiment of the present invention, the noise figure for the NE may be stored or computed inside the NE based on vendor's proprietary algorithm that may or may not take into account the average number of photons per bit (and maximum permitted average power or number of channels) at the NE's input <n_i−1> and the mode parameter M at the receiver, which were passed from the previous NE by the signaling protocol. Alternatively, NF_i^Poisson may be computed based on parameters n_i^eq, G_i, V_i that are associated with every device in addition to <n_i−1> and M, as stated in the first embodiment, according to a formula:
NF_i^Poisson=2n_i^eq+1/G_i+[1+n_i^eqV_iG_i][Mn_i^eq/(G_i<n_i−1>)].

[0044] At step 405, a total noise figure is determined for the path route up to and including the present NE based on the total noise figure up to and including the previous element NF_i−1^Total, average number of photons <n_i−1>, and the cumulative products of excess noise coefficients and gains: g₁. . . g_i−1 and G₁ . . . G_i−1. Using these four numbers and the M-parameter, the NF_i^Total, <n_i>, g₁ . . . g_i−1 and G₁ . . . G_i can be determined, which will be passed on to the next element along the tentative path route by the signaling protocol. According to an embodiment of the present invention, the total noise figure is computed as NF_i^Total=NF_i−1^Total+A_i−1(NF_i^Poisson−1), where A_i≡A_i−1(g_i/G_i) is also defined recursively with excess noise factor g_i=1+M(n_i^eq/ <n_i−1>) as stated in the first embodiment. Here, NF_i^Poisson and n₁^eq are as determined in the previous step.

[0045] At step 406, an optical SNR at the output of the present element is determined. The optical SNR is given by SNR_i=NF_i^Total/SNR₀ for a path containing a chain of i elements and a transmitter with SNR₀.

[0046] At step 407, it is determined whether the optical SNR for the path route that was determined in the previous step is equal to or greater than the one required by the receiver. If not, the control passes to step 414, otherwise, control proceeds to step 408.

[0047] At step 414, the current element is added to the list of prohibited “nearest neighbor NEs”, “i” is decremented and control returns to step 403, which will now be performed at the previous NE.

[0048] At step 408, it is determined if the intended receiver has been reached. If not, control returns to step 403, otherwise it proceeds to step 409.

[0049] At step 409, a bit error rate is determined by the receiver based on the optical SNR determined in step 406, parameters of the transmitter, and parameters of the receiver.

[0050] At step 410, it is determined whether the bit error rate for the path is within a predefined range. If the bit error rate for the path is outside the predefined range, control proceeds to step 411, otherwise control proceeds to step 412.

[0051] At step 411, signaling control commands may be issued by the receiver that go back to all or some elements along the tentative path to adjust the transmission parameters. Examples of such commands may be directives to change transmission power (i.e., <no>) or chirp, or adjust dispersion compensating devices inside those elements that have such capabilities. If no such capabilities are present in any of the elements, the control passes to step 413. If, however, such capabilities are present on all or some of the elements, the relevant transmission parameters are altered and control passes back to step 401.

[0052] At step 412, path connection is validated.

[0053] At step 413, path connection fails.

[0054] FIGS. 3 and 4 illustrate flow charts describing a method for managing and validating paths through a switched optical network according to embodiments of the present invention. Some of the steps illustrated in these figures may be performed sequentially, in parallel or in an order other than that which is described. It should be appreciated that not all of the steps described are required to be performed, that additional steps may be added, and that some of the illustrated steps may be substituted with other steps.

[0055] In the foregoing specification the invention has been described with reference to specific exemplary embodiments thereof. It will, however, be evident that various modifications and changes may be made thereto without departing from the broader spirit and scope of the invention. The specification and drawings are, accordingly, to be regarded in an illustrative rather than restrictive sense.