Method for pre-emphasizing an optical wavelength multiplex signal
The method adjusts signal powers at the transmitter using average power measurements and effective noise figures to balance signal-to-noise ratios in optical multiplex signals, addressing wavelength-dependent gain issues in WDM systems, ensuring fast and precise pre-emphasis without requiring planning tools.
Patent Information
- Application Number
- DE102004047693
- Authority / Receiving Office
- DE · DE
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2004-09-30
- Publication Date
- 2024-11-14
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Existing methods for pre-emphasizing optical multiplex signals in WDM or DWDM transmission systems with narrow channel spacing face challenges due to wavelength-dependent gain in optical amplifiers, leading to varying optical signal-to-noise ratios (OSNR) and performance differences among channels, which are not effectively addressed by current measurement and adjustment techniques.
A method that adjusts signal powers at the transmitter based on average power measurements at both transmitter and receiver, incorporating an effective noise figure, without requiring OSNR measurements, to achieve balanced signal-to-noise ratios across the transmission path, using a control module and software implementation.
This method enables fast, adaptive, and precise pre-emphasis of optical multiplex signals, independent of planning tools, ensuring consistent signal-to-noise ratios and reducing deviations, even with changing system conditions, thus improving transmission quality and eliminating the need for complex noise power measurements.
Smart Images

Figure 00000011_0000 
Figure 00000011_0001
Abstract
Description
[0001] The invention relates to a method for pre-emphasizing an optical multiplex signal according to the preamble of patent claim 1.
[0002] Optical amplifiers for broadband optical signals exhibit a wavelength-dependent gain that cannot be fully remedied by conventional smoothing filters. In WDM or DWDM transmission technology (WDM = Wavelength Division Multiplex; DWDM = Dense Wavelength Division Multiplex), the optical signal consists of several channels with different wavelengths, whose wavelength spacings can currently be below 100 GHz. Due to the wavelength dependence of the amplifier gain, power differences accumulate between the individual channels as they traverse an optical transmission link, resulting in significantly different optical signal-to-noise ratios (OSNR) and performance at the receivers.
[0003] In point-to-point connections, a method known as "pre-emphasis" is therefore often used to level the signal-to-noise ratios or OSNR values, at least at the end of the link, as described in AR Chraplyvy, JA Nagel and RW Tkach: "Equalization in Amplified WDM Lightwave Transmission Systems", IEEE Photonics Technology Letters, Vol. 4, No. 8, August 1992, pp. 920-922.
[0004] Based on an OSNR distribution measured at the end of the link, the transmitter-side channel power is adjusted in an iterative process until the same signal-to-noise ratios or OSNR values are obtained for all channels at the end of the link.
[0005] To determine the signal-to-noise ratio (OSNR), the amplified spontaneous emission (ASE) between the channels is often measured, and the noise power superimposed on the channels is calculated using interpolation. However, this is no longer possible if the amplified spontaneous emission (ASE) between the channels is attenuated by optical components depending on the wavelength. This is the case, for example, when additional modules such as add-drop modules or interleaver filters are connected in the transmission path.
[0006] All common measurement methods for OSNR distribution at the end of the link have in common that they are limited to channel spacings in a channel grid greater than or equal to 100 GHz. Furthermore, these methods are generally too slow (separate measurements for channel power and amplified spontaneous emission (ASE) power) to meet the time requirements of dynamic optical networks, e.g., a maximum of approximately 10 seconds for a channel upgrade.
[0007] DE 198 48 989 A1 discloses a method for adjusting transmit signal power levels on a channel-by-channel basis. In this method, if the transmit-side dynamic range is exceeded to an unacceptable extent, the individual transmit signal power levels are compressed in such a way that the total transmit signal power is kept approximately constant. This method is also used if the receive-side dynamic range is exceeded.
[0008] Since this method is based on measured OSNR values, the problems of OSNR measurement with small channel spacing described above also arise here.
[0009] From US 6 040 933 A a method for power equalization of channels in a wavelength division multiplexing system is known, in which the optical power of signals of different wavelengths is measured at the inputs of several optical amplifiers.
[0010] DE 10 2004 018 166 A1 discloses a method for the fast and simple pre-emphasis of an optical multiplex signal transmitted from a transmitter to a receiver. Signal-to-noise ratios are balanced at least at the receiver without measuring noise power or signal-to-noise ratios, but rather by simply measuring or readjusting signal power across the bandwidth of the optical multiplex signal. This method is based on a signal-to-noise ratio balance permitted by a transmission system, taking into account spectral influences from gain profiles, noise effects, and attenuation.
[0011] The object of the invention is to provide a novel method that enables simple pre-emphasis of an optical multiplex signal. The method should be suitable for WDM transmission along a defined optical transmission path with arbitrarily narrow channel spacing.
[0012] The problem is solved with regard to its method aspect by a method having the respective features of each of the patent claims 1, 6, 7 and 8.
[0013] A method is described for pre-emphasizing an optical multiplex signal, which has several signals of different wavelengths as channels, transmitted from transmitters to receivers. In this method, the signal powers are adjusted at the transmitter and measured at the receiver. This eliminates the need to determine the signal-to-noise ratio (OSNR) at the receiver. Instead, an average signal power is determined at the transmitter, and then new signal powers are adjusted at the transmitter end based on the current signal powers at the transmitter and receiver, and the average power at the transmitter, such that the signal-to-noise ratios at the receiver are approximately leveled or adjusted.According to the invention, a significant improvement in pre-emphasis is achieved by including an effective noise figure in the determination of the new signal power values, by which the multiplex signal is impaired during the entire transmission, by optical components, preferably by attenuation, and possibly, for example, in optical amplifiers.
[0014] This method advantageously eliminates the need for a planning tool, for example, to select an exponent parameter, as described in the prior art, when calculating a gain to determine new signal powers. The method enables automatic and iterative pre-emphasis, which, based on a precise characterization of gain values and noise figures of components such as amplifiers, fibers, etc., of one or more sections of a corresponding transmission system, is precisely calculated with respect to
[0015] The power of the transmitting signals is readjusted. This allows pre-emphasis to be performed autonomously, without the need for additional planning tools to control it.
[0016] To control the amplifiers in a transmission system, the total power at the amplifier input and output is measured. From this data, the average gains of the individual amplifiers and the attenuation of the fiber links can be calculated. By using this data, which is already known in the system, it is possible to design a pre-emphasis in such a way that the determination of an optimal exponent parameter using a planning tool, as described in the state of the art, is no longer necessary. This aspect is described in detail in the subsequent sections of the description.
[0017] A further advantage of the invention is that no measurement of the signal-to-noise ratios or noise powers is required, but only level measurements of the signals. Because the powers at a transmitter are measured and readjusted using a simple measurement of signal powers at a receiver, the pre-emphasis according to the inventive control formula is much faster than a pre-emphasis based on signal-to-noise ratios. This also eliminates the need for system-specific and therefore complex measurements of the noise powers of the signals. A complicated measurement of noise powers between the channels or even a direct and technically very complex measurement of the amplified spontaneous emission ASE superimposed on the channels to determine the signal-to-noise ratios OSNR is no longer necessary. The method is therefore ideally suited for any small wavelength spacing between the channels.
[0018] Of course, it is possible to combine this procedure with a subsequent pre-emphasis based on a measurement of the signal-to-noise ratio, which leads to an optimal setting of the channel power at the transmitter.
[0019] The fact that this requires significantly more time does not have a negative impact on transmission quality. However, the method according to the invention no longer requires such a known pre-emphasis to meet approximately the same requirements. This advantageous aspect has been proven theoretically and experimentally in the laboratory. This eliminates the need for costly spectrally resolving measuring instruments such as optical spectrum analyzers.
[0020] A significant advantage of the invention is also that the described method is insensitive to an existing tilt or to a further existing uneven spectral distribution of the powers and / or the signal-to-noise ratios OSNR at the transmitter.
[0021] Throughout the invention, the terms "transmitter" and "receiver" are used for simplicity of illustration. It should be clarified here that these terms refer to any point along a transmission path at which the pre-emphasis according to the invention can be implemented, e.g., at optical amplifiers, at multiplexers and demultiplexers, at spectrally controllable filters, etc. For this purpose, a first control and measurement module intended for the power spectrum must be present at least at a "transmitter" point, and a second measurement module intended for the power spectrum must be present at a "receiver" point.
[0022] The inventive method for pre-emphasis and its implementation are suitable for an optical transmission link with multiple amplification sections. The amplification section refers to at least one amplifier and one fiber, as well as optional components such as add-drop modules, filters, isolators, etc. This transmission link could be part of a more complex optical network in which at least WDM signals are transmitted. The inventive pre-emphasis can be implemented in software in a state-of-the-art system without any hardware upgrades. For this purpose, a data carrier with a program is used that can be loaded into a control module (usually the network management system). The control module executes the inventive method according to all technical aspects and effects of the invention when the said program is executed.Advantageous further developments of the invention are specified in the subclaims.
[0023] Embodiments of the invention are explained in more detail below with reference to the drawing.
[0024] Showing: Fig. 1 an optical transmission system for applying the method according to the invention, Fig. 2 a deviation of the signal-to-noise ratios between the present method and a method with the required planning tool.
[0025] Fig. 1 shows an optical transmission system for applying the method according to the invention, into which optical signals are fed as channels of a WDM signal by means of a multiplexer MUX on the transmit side. A first amplifier V1 (booster) is arranged at the input of the transmission system. At the input of the first amplifier V1, the power P_Tx(λ) of the signals is measured, for example, using a power meter. A first transmission fiber F1 is connected downstream of the first amplifier V1. The first amplifier V1 and the first transmission fiber F1 form a first amplification section. Two further optical amplification sections connected in series, comprising a second amplifier V2, a second transmission fiber F2, a third amplifier V3, and a third transmission fiber F3, are also connected downstream of the first transmission fiber F1.A final amplifier V4 is connected to the output of the last transmission fiber F3, at which the power P_Rx(λ) of the transmitted signals is measured. The term "power" here refers to signal power as opposed to noise power or signal-to-noise ratios. The fourth amplifier forms the final stage of the transmission system, and its output is connected to a demultiplexer DEMUX for re-separating the multiplexed signals. This "point-to-point" link can also be part of a complex network in which the inventive method can be applied to all "point-to-point" links formed as a plurality of serially connected amplification sections.
[0026] In general, such "point-to-point" links nowadays have pre-emphasis adjustment means for adjusting the power P_Ťx(λ) of the signals at the input of the first amplifier V1, e.g., such that a desired spectrum profile of power P_Rx(λ), signal-to-noise ratios, or bit error rates is generated at the output of the transmission system. Such pre-emphasis adjustment means can also be provided section by section for specific amplification sections, e.g., so that power tilts are better distributed along the link, so that a required high tilt at the input of the entire transmission system and thus possible signal distortions due to non-linearities in high-power channels do not occur. For the method according to the invention, a division into several pre-emphasises can also be used for one or a group of amplification sections (in the case of Fig. 1, three amplification sections are defined with the receiving-side amplifier V4).
[0027] In the following, the method according to the invention is described in detail, but first it is mentioned why the method is considerably simpler and cheaper than a method with the required planning tool.
[0028] A known method is to first measure the corresponding channel power distribution P_Rx(λ) at the output of the transmission system for any channel power distribution P_Tx(λ) at the input of the transmission system. In a second step, new input powers P_Tx(λ) are applied for pre-emphasis. NEW according to the regulation: P_Tx(λ)NEW=〈P_Tx(λ)〉⋅(P_Tx(λ) / P_Rx(λ))0.5〈(P_Tx(λ) / P_Rx(λ))0.5〉 , where square brackets () surrounding an argument represent the average of the argument over the wavelength. This allows for approximately identical OSNR signal-to-noise ratios to be achieved for all output channels, but still results in deviations from a pure OSNR preemphasis (i.e., a preemphasis based on OSNR measurements to achieve approximately identical OSNR signal-to-noise ratios).
[0029] It can further be shown that the magnitude of these deviations can be significantly reduced by introducing a variable exponent k with values in the range from 0 to 1 – compared to 0.5 in the above equation – so that the method comes closer to the optimal results of an OSNR pre-emphasis. The transmit-side powers P_Tx(λ) are now calculated according to the rule P_Tx(λ)NEW=〈P_Tx(λ)〉⋅(P_Tx(λ) / P_Rx(λ))k〈(P_Tx(λ) / P_Rx(λ))k〉 Since the transfer function of the link can change when the input spectrum changes, it is advisable to apply this procedure several times.
[0030] The task now arises of determining an optimal setting for the exponent k. To do this, the signal-to-noise ratios (OSNR) must be determined as a function of the exponent k. This function shows the minimum occurring signal-to-noise ratios (OSNR) as a function of the exponent k for a specific transmission system configuration. An optimum value for the exponent k—e.g., at approximately 0.45—is sought so that the signal-to-noise ratios (OSNR) exhibit maximum values. If transmission parameters change, a planning tool must take this into account in order to readjust the value of the exponent k.
[0031] The method according to the invention advantageously eliminates the need for a planning tool or a new search for an optimum. This results in a simpler and faster pre-emphasis. Furthermore, it is shown that the pre-emphasis is also more accurate.
[0032] New procedural steps for readjusting the power of the transmitting signals (here by power tilting) can be defined as follows: ►In a first step, the amplifiers are adjusted so that a tilt (also called "power tilt") at the input of the subsequent amplifier and at the output of the last amplifier V4 (preamplifier) at the end of the line disappears (power tilt := 0). ►From the linear power spectra P_Tx(λ), P_Rx(λ) measured at the input of the booster V1 and at the output of the preamplifier V4, a gain G link(λ) is calculated. Its mean value is then subtracted from this. ►The function Q(λ) for describing the newly adjusted power spectrum P_Tx(λ) at the input of the transmission system with N amplification sections is now calculated according to the rule Q(λ)=∑i=0NGlink−t / (N+1)(λ)⋅〈Gi(λ)〉⋅∏j=i+tN〈Gj(λ)〉⋅〈aj(λ)〉 with Glink(λ)=P_Rx(λ)P_Tx(λ) calculated.
[0033] The sizes G i (λ), a i (λ) and their mean values 〈G i (λ)〉, 〈a i (λ)〉 denotes the profit of the i tenThe gain of an amplifier section is the ratio of the output power spectrum to the input power spectrum of an amplifier section, or the attenuation value of an amplifier section. Depending on the situation, these quantities can also be the gain of an amplifier and the attenuation value of a fiber connected downstream of the amplifier, instead of values for an entire amplifier section. If the gain differences between the individual channels are small compared to the average gain, it is permissible to approximate the average values using the quotients (P_Rx(λ)) / (P_Tx(λ)) of the average total output power and the average total input power.
[0034] A spectral effective noise figure F eff(λ) can also be calculated directly by the system if the properties of all amplifiers (and fibers) have been measured beforehand or, for example, specified by the supplier. Alternatively, a typical value can also be used. If this is not possible, a constant value is assumed. ► The power distribution P_Tx(λ) at the input of the transmission line is thus P_Tx(λ)=〈P_Tx(λ)Start〉〈Q(λ)⋅Fcff(λ)⋅〉⋅Q(λ)⋅Feff(λ), where the fraction ensures that an average input power 〈P_Tx(λ) Beginn 〉, which existed at the beginning of the procedure, does not change and thus the input level dependent amplifier characteristics remain approximately the same. ►The tilt at the input of the amplifiers, called “power tilt”, is now adjusted so that it corresponds to the tilt of the product Q(λ)·F eff(λ). The tilt of the function Q(λ) (in dB / THz) and the tilt of the effective noise figure F eff (λ) (also in dB / THz) is added to the resulting power tilt. ►After the previous step, the process is repeated iteratively starting with the second step until the power changes of the signals at the input of the transmission path from one iteration cycle to the next are below a specified limit.
[0035] In practice, the quantities appearing in the above equations must be used on a linear scale.
[0036] A significant advantage of the described modification is that no specifications from a planning system are required for the pre-emphasis with power tilt and the method reacts adaptively to changes in the transmission system (e.g., an increase in the attenuation values due to aging).
[0037] To verify the present method, simulations were carried out with different configurations - with a conventional planning system, e.g. for determining an exponent k and with the process steps described here - the results of which are presented in Fig. 2 are shown.
[0038] For a given number of amplification sections SPAN_NR = 1, 2, ..., 20, the resulting deviations DEV_OSNR (see curve pairs A, B) of signal-to-noise ratios OSNR after OSNR pre-emphasis were plotted with pre-emphasis using an exponential rule (with a planning tool) and after OSNR pre-emphasis using the inventive method. Initially, it was assumed that all amplification sections had an attenuation of 21 dB. Subsequently, the simulations were repeated for all possible cases, each with an attenuation of 27 dB. The maximum deviation between the OSNR pre-emphasis result considered as the reference and each variant of the present pre-emphasis is shown in the above illustration (see upper curve for each curve pair A, B) as a function of the number of transmission sections and marked with a dot.The simulations were repeated using the same procedure, with the transmission section with different attenuation now exhibiting a lower attenuation, namely 15 dB. The resulting results were marked by star points (see lower curve for each pair of curves A, B).
[0039] The Fig. Figure 2 shows that, in addition to the advantage of independence from a planning tool, the invention also delivers an OSNR improvement of up to 0.2 dB (see the difference between the corresponding curves of the curve pair A and B at SPAN_NR=20). Thus, in addition to being simple and fast, the method enables more precise pre-emphasis of the signals, which results in a better range for the transmitted signals.
[0040] In the following, further interesting aspects of the pre-emphasis according to the invention are explained.
[0041] In the previous equation for calculating the function Q(λ), it was assumed that all amplification sections are identical in terms of gain. This means that for N identical amplification sections, a gain value G i (λ) i= 1, 2, ..., N is determined as follows: Gi(λ)=g(λ)⋅〈Gi(λ)〉,∀i=1,2,…,N where g(λ) denotes a spectral gain factor common to the identical amplification sections to simplify the determination of the function Q(λ).
[0042] Likewise, for N identical amplification sections, an attenuation value a i (λ) is determined as follows: ai(λ)=g(λ)⋅〈ai(λ)〉,∀i=1,2,…,N where a(λ) denotes a spectral attenuation factor common to the identical amplification sections to simplify the determination of the function Q(λ).
[0043] However, this is only approximate. The following describes possible improvements that require knowledge of the individual amplifiers, either based solely on the type used or through measurements, for example, at amplifier outputs.
[0044] First, it is assumed that a typical extent of the deviation of the profit spectra from the mean of a profit spectrum is known.
[0045] In a transmission system, various amplifier types are used in the amplification sections, each with a gain profile that differs in magnitude from the mean value of a defined reference gain spectrum. Depending on its amplifier type, each optical amplifier is assigned a relative characteristic value k. iwhich relates the extent of its typical gain variations to those of a reference amplifier. The characteristic function Q(λ) for calculating the power distribution at the input is thus Q(λ)=∑i=0NGlink(λ)−(∑j=0i−1kj) / (∑j=0Nkj)⋅〈Gi(λ)〉⋅∏j=i+1N〈Gj(λ)〉⋅〈aj(λ)〉 where for N different gain sections a gain value G i (λ) of one of the gain sections relative to a mean value 〈G j (λ)〉 j=1, 2, ..., N (or j=0 for a completely theoretical mean) of a chosen profit using the relative key figure (k i ) is calculated.
[0046] Furthermore, it is assumed that a typical deviation of the gain spectra from the mean value of a gain spectrum is known as a function of wavelength.
[0047] In contrast to the case described above, the typical deviations gjtyp(λ) to the gain factor g(λ) as a function of wavelength λ. These deviations are now combined with an exponent to calculate a total gain G gu (λ) of the transmission system. In this case, N different amplification sections with known gain factors Gi(λ)=gityp(λ)⋅〈G0(λ)〉, ∀i = 1,2,...,N the total profit G ges (λ) between transmitter and receiver is calculated as follows: Gges(λ)=〈G0(λ)〉⋅[∏j=1N〈Gj(λ)〉⋅〈aj(λ)〉]⋅∏i=0N(gityp)mi where m i as an exponent of profit factors gjtyp(λ) is chosen in such a way that differences between this total profit G ges (λ) and the measured gain G lunk (λ) (= ratio between transmitting and receiving side measured power of the signals) must be minimized.
[0048] The corresponding function Q(λ) can then be calculated as follows: Q(λ)=∑i=0N〈Gi(λ)〉⋅[∏j=i+1N〈Gj(λ)〉⋅〈aj(λ)〉⋅∏j=1N(gjtyp)mj]∏i=0N(gityp)mi
[0049] In the following, it is assumed that the spectral power distribution at points within the transmission system is known through a measurement. This is often the case when signal power monitoring devices are present at one or more amplification sections.
[0050] The prerequisite for the realization of an optimized procedure is that a system control of the pre-emphasis all damping a i (λ) and profits G i(λ) of the gain sections are known as a function of wavelength λ. Since both quantities (gain and attenuation) can change due to aging or replacement of components, a continuous transmission of the current quantities using appropriate measuring devices and data transmission, e.g., via a monitoring channel to the system control system, is advisable. For cost reasons, however, both quantities are only measured at selected points. i (λ), G i (λ) is recorded spectrally and not only in the mean value, so that only the partial transfer function: g(γ)km=∏j=kmGj(λ)⋅aj(λ)⋅aj(λ) a group of m-k+1 amplification sections (amplifier with transmission fiber) as well as all mean values 〈G j (λ)〉,〈a j (λ)〉 are known.
[0051] For the group of 1< m-k+1< N of the N amplification sections, the partial transfer function g(λ)km=∏j=kmGj(λ)⋅aj(λ) used to calculate the function Q(λ).
[0052] The profit G j (λ) and the damping a j (λ) and their mean values 〈G j (λ)〉, a j (λ)〉, preferably by a measurement or by external information, are made known as spectral functions taking into account the aging of components.
[0053] For a group with 1< m-k+1< N completely known gain sections and identical gain sections, the function Q(λ) is defined as follows: Q(λ)=∑i=0N∏k=0iGk−1(λ)⋅〈Gi(λ)〉⋅∏j=i+1N〈Gj(λ)〉⋅〈aj(λ)〉. The implementation of the new method can be carried out by means of a control module in which a data carrier with a program that can be loaded into the control module, wherein the control module executes the inventive method according to all technical aspects and effects of the invention when the said program is executed.
[0054] As an example, the control module may comprise a plurality of local control units at each amplification section, which are for controlling or detecting control signals in connection with at least a regulation of the new power values of the transmitter.
[0055] The control module can also be part of a (central) network management system that monitors the entire transmission system.
Claims
[1] Method for pre-emphasizing an optical multiplex signal (OS) comprising a plurality of signals with different wavelengths, which are transmitted from a transmitter via a plurality of N amplification sections (i = 1, 2, ... N), each comprising an amplifier and a transmission fiber, and via an additional amplifier (i = 0) to a receiver, for equalizing signal-to-noise ratios of all signals at the receiver, characterized by following steps: - Measuring performance (P_Rx( λ ) Beginn ) of the signals at the receiver, - Determine an average power (〈P_Tx(λ) Beginn 〉) of the signals at the transmitter, - Calculate a spectral function Q(λ) according to the rule Q(λ)=∑i=0NGlink−1 / (N+1)(λ)⋅〈Gi(λ)〉⋅∏j=i+1N〈Gj(λ)〉⋅〈aj(λ)〉, where 〈G i (λ)〉, with i = 0, 1, 2, ... N, denote the average gains of the individual amplifiers, where 〈a i(λ)〉, with i = 1, 2, ... N, denote the average attenuation of the individual transmission fibers and where G link (λ) denotes the gain between transmitter and receiver, - Calculate new signal power values P_Tx(λ) taking into account an effective noise figure F eff (λ), by which the multiplex signal (OS) is affected during the entire transmission, as follows P_Tx(λ)=Q(λ)⋅Feff(λ)〈Q(λ)⋅Feff(λ)〉⋅〈P_Tx(λ)Start〉, - setting the calculated new signal power values (P_Tx(λ)) on the transmit side, where the square brackets < > denote the mean value of the argument over the wavelength. [2] Method according to claim 1, characterized by that the profit G link (λ) between transmitter and receiver is calculated as the ratio between the transmitting and receiving side measured power of the signals. [3] Method according to claim 1, characterized bythat the profit between sender and receiver consists of profit values G i (λ) (i = 0, 1, 2, ..., N) of the individual amplifiers and from attenuation values a i (λ) (i = 1, 2, ..., N) of the individual transmission fibers is calculated. [4] Method according to claim 3, characterized by that for N identical amplification sections the gain values G i (λ) of the individual amplifiers can be determined as follows: Gi(λ)=g(λ)⋅〈Gi(λ)〉,∀i=1,2,…,N where g(λ) denotes a spectral gain factor common to the identical amplification sections to simplify the determination of the function Q(λ). [5] Method according to claim 3 or 4, characterized by that for N identical amplification sections the attenuation values a i (λ) can be determined as follows: ai(λ)=a(λ)⋅〈ai(λ)〉,∀i=1,2,…,N where a(λ) denotes a spectral attenuation factor common to the identical amplification sections to simplify the determination of the function Q(λ). [6] Method for pre-emphasizing an optical multiplex signal (OS) comprising a plurality of signals with different wavelengths, which are transmitted from a transmitter via a plurality of N amplification sections (i = 1, 2, ... N), each comprising an amplifier and a transmission fiber, and via an additional amplifier (i = 0) to a receiver, wherein different types of amplifiers are used to equalize signal-to-noise ratios of all signals at the receiver, characterized by following steps: - Measuring the power (P_Rx(λ) Beginn ) of the signals at the receiver, - Determine an average power (P_Tx(λ) Beginn ) of the signals at the transmitter, - Assign a relative key figure k ito each of the amplifiers according to its amplifier type, where the relative characteristic k i indicates a relative extent of gain variation of the amplifier type, - Determine a spectral function Q(λ) as follows Q(λ)=∑i=0NGlink(λ)−(∑j=0i−1kj) / (∑j=0i−1kj)⋅〈Gi(λ)〉⋅∏j=i+1N〈Gj(λ)〉⋅〈aj(λ)〉, where 〈G i (λ)〉, with i = 0, 1, 2, ... N, denote the average gains of the individual amplifiers, where 〈a i (λ)〉, with i = 1, 2, ... N, denote the average attenuation of the individual transmission fibers and where G link (λ) denotes the gain between transmitter and receiver, - Calculate new signal power values P_Tx(λ) taking into account an effective noise figure F eff (λ), by which the multiplex signal (OS) is affected during the entire transmission, as follows P_Tx(λ)=Q(λ)⋅Feff(λ)〈Q(λ)⋅Feff(λ)〉⋅〈P_Tx(λ)Start〉, - setting the calculated new signal power values (P_Tx(λ)) on the transmit side, where the square brackets < > denote the mean value of the argument over the wavelength. [7] Method for pre-emphasizing an optical multiplex signal (OS) comprising a plurality of signals with different wavelengths, which are transmitted from a transmitter via a plurality of N amplification sections (i = 1, 2, ... N), each comprising an amplifier and a transmission fiber, and via an additional amplifier (i = 0) to a receiver, wherein different types of amplifiers are used to equalize signal-to-noise ratios of all signals at the receiver, characterized by following steps: - Measuring the power (P_Rx(λ) Beginn ) of the signals at the receiver, - Determine an average power (〈P_Tx(λ) Beginn 〉) of the signals at the transmitter, - Calculate a total profit G ges(λ) between sender and receiver using known gain factors gjtyp(λ) the amplifier as follows: Gges(λ)=〈G0(λ)〉⋅[∏j=1N〈Gj(λ)〉⋅(aj(λ))]⋅∏i=0N(gityp)mi, where the profits G i (λ) the amplifier from Gi(λ)=gityp(λ)⋅〈G0(λ)〉, ∀i = 0,1,..., N, where m i as an exponent of the profit factors gjtyp(λ) is chosen in such a way that differences between the calculated total profit G ges (λ) and a profit (G link (λ)), which corresponds to the ratio between the measured power of the signals at the transmitter and receiver sides, where 〈G0(λ)〉 is the average gain of the additional amplifier mentioned and where 〈a i (λ)〉, with i = 1, 2, ... N, denote the average attenuation of the individual transmission fibers, - Calculate a function Q(λ) as follows Q(λ)=∑i=0N〈Gi(λ)〉⋅[∏j=i+1N〈Gj(λ)〉⋅〈aj(λ)〉]⋅∏j=iN(gjtyp)mi∏i=0N(gityp)m, - Calculate new signal power values P_Tx(λ) taking into account an effective noise figure F eff (λ), by which the multiplex signal (OS) is affected during the entire transmission, as follows P_Tx(λ)=Q(λ)⋅Feff(λ)〈Q(λ)⋅Feff(λ)〉⋅〈P_Tx(λ)Start〉, - setting the calculated new signal power values (P_Tx(λ)) on the transmit side, where the square brackets < > denote the mean value of the argument over the wavelength. [8] Method for pre-emphasizing an optical multiplex signal (OS) comprising a plurality of signals with different wavelengths, which are transmitted from a transmitter via a plurality of N amplification sections (i = 1, 2, ... N), each comprising an amplifier and a transmission fiber, and via an additional amplifier (i = 0) to a receiver, wherein a group of 1< m-k+1< N of the N amplification sections has identical amplification sections and for this group a partial transfer function (g(λ) k:m ) and is used to calculate a function Q(λ) to adjust the signal-to-noise ratios of all signals at the receiver, characterized by following steps: - Measuring the power (P_Rx(λ) Beginn ) of the signals at the receiver, - Determine an average power (〈P_Tx(λ) Beginn 〉) of the signals at the transmitter, - Calculate a function Q(λ) as follows Q(λ)=∑i=0N∏l=0iGl−1(λ)⋅〈Gi(λ)〉⋅〈aj(λ)〉, where the profits G i (λ) of the individual amplifiers, with i = 0, 1, 2, ... N, the attenuations a i (λ) of the individual transmission fibers, with i = 1, 2, ..., N, and their mean values 〈G i (λ)〉, a i (λ)〉, by a measurement or by external information, are known as spectral functions taking into account the aging of components, - Calculate new signal power values P_Tx(λ) taking into account an effective noise figure F eff (λ), by which the multiplex signal (OS) is affected during the entire transmission, as follows P_Tx(λ)=Q(λ)⋅Feff(λ)〈Q(λ)⋅Feff(λ)〉⋅〈P_Tx(λ)Start〉, - setting the calculated new signal power values (P_Tx(λ)) on the transmit side, where the square brackets < > denote the mean value of the argument over the wavelength.
Citation Information
Patent Citations
Pre-emphasis method for optical multiplex signal in optical transmission system has new signal values calculated from actual signal powers at transmitter and receiver and mean power at transmission side
DE102004018166A1
Method for channel-by-channel setting of transmission signal powers of a wavelength-division multiplex transmission system
DE19848989A1
Method and apparatus for channel equalization in wavelength division multiplexed systems
US6040933A
