WO2004010616A1

WO2004010616A1 - Method and arrangement in a communication system

Info

Publication number: WO2004010616A1
Application number: PCT/SE2003/001136
Authority: WO
Inventors: Gullik WEBJÖRN
Original assignee: Webjoern Gullik
Priority date: 2002-07-19
Filing date: 2003-06-30
Publication date: 2004-01-29
Also published as: AU2003243110A1; SE0202273D0

Abstract

The present invention relates to an optical fiber section, an optical network that comprises of a plurality of optical fiber sections and a method for controlling the signal power gain in such a fiber section. The optical fiber section comprises a first network element comprising means for receiving optical signals with a first power level and means for transmitting said optical signals with a second power level, said first network element is connected to a first end of a fiber strand, a second end of said fiber strand is connected to a second network element comprising means for receiving said optical signals, said fiber strand having an attenuation, wherein the fiber section comprises means for adjusting the power level of the output optical signals from the receiving means in the second network element to a power level substantial identical to the first power level, resulting in a gain of zero db over the entire section from an input of the first network element to an output of the second network element.

Description

METHOD AND ARRANGEMENT IN A COMMUNICATION SYSTEM

FIELD OF INVENTION

The present invention relates to an optical network, optical fiber sections within the optical network, a method and a computer program product for performing the steps of said method.

In particular, the invention relates to an improvement of the flexibility in an optical network comprising a plurality of optical fiber sections and a method for controlling the signal power gain in the fiber sections.

BACKGROUND OF THE INVENTION

Optical networks may be used for transmitting information signals, e.g., signals for

Internet communication, over long distances with a high transmission rate (e.g. over 40Gbit/ s). A typical optical network is shown in figure 1. A first terminal Tl is connected to second terminal T2 via optical fiber cables. The fiber cables are interrupted by several repeater stations Rl, R2, R3. Each repeater station comprises of the following components:

- A transponder using Optical-Electrical-Optical conversion to adapt the wavelength and power of the connecting optical signal to a wavelength and a power compatible with the optical transmission network.

- A wavelength multiplexor combining multiple transponder signals into a single fiber signal.

- Optionally one or more optical power amplifier(s) increasing the power of the composite signal for long haul transmission. - A wavelength demultiplexer splitting the composite optical signal on multiple receivers.

- Various other components for monitoring, synchronizing and creating redundant signal paths.

A typical topology for an optical network is a ring topology that ensures that there is always more than one path between source and destination as shown in figure 2. The fiber rings are interconnected at crossconnect points XI to X4. Frequently traffic from terminal Tl to T2 is crossconnected in the electrical or framing protocol level rather than the optical domain. Designing a circuit with redundancy from Tl to T2 involves detailed analysis of traffic and usage in all affected rings. However, some vendors have started to deploy other topologies, such as the "full mesh"- topology as shown in figure 3. A full mesh network implies that any node N can reach any other node N. Redundancy can be provided by selecting two different paths through the mesh. With electrical conversion of light in the fiber, complexity at the system level increases. If a mesh network is designed in the optical domain, using only fibers, optical amplifiers and light sources all and any protocols can be distributed without interaction.

The current typical optical network is constructed for a fixed scenario, and the optical amplifiers are used only as a gain component in a transmission system, where the topology is limited to or designed for the upper protocol layers such as SDH (Synchronous Digital Hierarchy) or SONET (Synchronous Optical NETwork) and various higher layer functions such as multiplexing and error correction. The flexibility of such a network is often restricted once a system from a specific vendor is chosen when the user wants to implement other types of formats and features that the equipment was not designed for.

There is a trend towards unbundling the various layers composing a network into separate entities that address only a single layer. The optical networks that focus more on the optical layer, that multiplexes optical wavelengths on a fiber, attempt to provide huge numbers of wavelengths on a single fiber pair. To handle the large amounts of wavelengths very expensive light sources are used (e.g. lasers) to enable extremely tight control of the wavelength.

During the recent years, Semiconducter Optical Amplifiers (SOA) also referred to as linear optical amplifiers (LOA) are developed. A SOA is described in US A 6347104 and is used as a component in optical transmission equipment today, but constitute only a small fraction of the system cost. An introduction of SOAs in the optical networks facilitates a transmission that is only dependent on the optical layer, i.e. no higher layer protocols such as SDH or SONET is involved in the transmission. This development of new components and technologies in the optical layer creates new opportunities to design entirely new network topologies and functionality independent of legacy concepts. SOA comprises an Integrated Circuit. The SOA is linear which implies that all wavelengths is equally amplified which also reduces interference between neighbouring channels. A SOA provides e.g. a huge bandwidth and quantum speed. Furthermore, a SOA is less expensive, has a considerably lower power consumption, and requires less space than conventional EDFA (Erbium Doped Fiber Amplifiers) equipment.

There are several drawbacks associated with the current optical networks. Some drawbacks are listed below:

- An expansion or modification of current networks requires a re-evaluation and possibly redesign of network sections and parts far away from the point of interest.

- An addition of a tap or a branch to the network involves the signal budget of the network as designed, and can result in the need to add or move amplifiers, which in turn can cause a requirement for redesign of higher layers as well. Since addition or moving of amplifiers often also involves other functions, the resulting equipment might well affect the power and cooling distribution at the site in question.

- Adding a drop for a single wavelength in a specific location might result in requirement for installing equipment to regenerate or re-amplify other wavelengths because the total path loss is affected by the insertion loss of a new wavelength multiplexor/ demultiplexer pair. Since these functions were not needed before the modification they increase the production cost of existing connections. -Due to the high system cost, current optical network equipment rarely lights more than a single or double pair of fibers in a ring topology. Thus, most of the fiber cables in a fiber cable are usually unlit because of the expense.

EP Al 651 476 relates to a method that mitigates gain peaking by using a chain of fiber amplifiers. In one embodiment, the output power of the amplified signal after each amplified signal is normalized such that the output at the peak wavelength of each amplifier is set to 0 dB. It should be noted that the illustrated attenuators are not used for attenuating the signal to 0 dB but representing fiber or splitting loss. However, there is no disclosure or hint in this document that that embodiment should overcome the stated drawbacks that are solved by the present invention.

US Bl 6,356,383, describes an optical system including conventional erbium doped amplifiers configured to achieve maximum signal channel in a span downstream of the transmitter and amplifier site to decrease the interaction between the wavelengths at high signal powers. This is performed in order to accomplish an optical transmission system with an increased flexibility. However, this solution is based on conventional amplifiers and does not overcome the above-listed problems. The published US application US2002/ 0080438 Al discloses a system for solving problems arising when different signal levels in the network exist e.g. due to different attenuations for different paths. The problem is solved by applying a constant output signal level. This system does not control the input signal level.

US 6,392,769 Bl discloses a network that monitors and controls the signal strength per wave length at each node in a network. The signal strength is controlled by a central controller. The purpose is to maintain the signal strength within a predetermined range in order for the network to perform its own level stabilization which prevents destructive feedback oscillations.

US 6,400,479 Bl describes a Wavelength Division Multiplexed (WDM) optical communication network. The network is configured and operated to enable transmitter output power for a given wavelength channel to be adjusted to achieve a desired optical signal-to-noise-ratio (OSNR) for the channel independently of the power levels of other optical channels carried on the same path. Optical amplifiers in the optical links extending between the transmitter and an optical receiver are configured to operate with a constant gain over a specified range of input signal power and the links are configured such that the power level of the signal provided to each optical amplifier is within a specified range to prevent deep saturation of the optical amplifiers due to optical amplifier cascading. The purpose of the constant gain operation and input power control of the optical amplifiers, is that the OSNR of other signals carried on the path are not affected, which implies that it is not required to adjust the output power of other transmitters providing signals to the path. The constant gain operation is achieved locally in the amplifier. Thus, the input power level is compared with the output power of the transmitter, which results in that the network is not able to consider attenuation variations in the link.

Thus, one object of the present invention is to provide optical fiber sections and an optical network comprising of said fiber sections, that are flexible and allows network modification without requirement of extensive network redesign. A further object with the present invention is to provide a method for controlling the signal power gain of optical signals transmitted in the fiber sections according to the present invention and a computer program product for performing the steps of said method.

SUMMARY

The above-mentioned object is achieved by an optical fiber section, an optical network, a method, a computer program product set forth in the characterizing part of the independent claims.

Preferred embodiments are set forth in the depending claims.

An advantage with the optical network according to the present invention is that it circumvents redundancy and reliability problems that current optical networks suffer from. By using the network, entire fiber cables, comprising a plurality of fiber strands, can be lit up, since the linear optic amplifier consumes less energy than conventional technique and bandwidth bottlenecks can be avoided. 1 Watt or less is required of one amplifier. By using the low power consuming amplifiers, it is possible to power the amplifier by using e.g. solar or wind energy. With very large amounts of lit fiber multiplied by many wavelengths a large freedom becomes available when pre-configuring the optical network.

A further advantage is that the resulting optical network built with large amounts of amplifiers, with each section unified to e.g. substantially 0 dB gain can be extended, expanded, reconfigured and without reengineering each path. Since the gain/loss levels within the optical network are unified, the cost of optics in connecting equipment is significantly reduced.

Yet another advantage is that lighting massive amounts of fiber is facilitated by combining existing technology and the simple but yet powerful network elements in accordance with the present invention. Moreover, since the optical network is compatible with existing systems, it enables a graceful migration. Yet another advantage is that the interface requirement for connecting equipment is limited to factors that address only the optical (i.e. physical) portion of the protocol stack, i.e. wavelength and power level range.

Yet another advantage with the present invention is that any terminating optical equipment will work transparently over an entire optical network if each fiber section has a unified amplification of e.g. substantially 0 dB.

Yet another advantage is that any desired topology can be built, such as sections, rings or mesh topologies, without the traditional constraints.

Furthermore, a traditional repeater station has a size of approximately 50-200 m², by using a semiconducter optical amplifier, the space requirement will be about 1 m³ or less.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 shows an optical network according to prior art. Figure 2 shows a network with a ring topology.

Figure 3 shows a network with a full mesh topology.

Figure 4 shows a network element for use in an optical network according to the present invention.

Figure 5 shows a part of an optical network according to the present invention.

Figure 6 shows a first scenario according to the present invention.

Figure 7 shows a second scenario according to the present invention.

Figure 8 shows an optical fiber section according to the present invention.

Figure 9 illustrates a network element that may be implemented within a fiber section according to the present invention.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS The solution according to the present invention is to design an optical network comprising of optical fiber sections 800, 808 illustrated in figure 8, where input optical signals S (at point 1) transmitted to a fiber section 800 has a substantial identical power level as the output optical signals S (at point 2) from said fiber section 800. Each fiber section 800 comprises a transmitting network element 802 connected to a first end of a fiber strand 804, the second end of the fiber strand 804 is connected to a receiving network element 806. The receiving network element is further connectable to a transmitting network element 810 belonging to a further fiber section 808, but also to other equipment such as local terminals or legacy network . Thanks to the fact that each fiber section according to the present invention is adapted to provide a unity gain, i.e. the power level of the input signals S at point 1 is substantial identical to the power level of the output signals S at point 2 as the power level of the input signals S at point 2 is substantial identical to the power level of the output signals S at point 3, it is possible to connect further fiber sections to an existing fiber section at a point P that has a unity gain, i.e. at the output from the receiving network element, without the requirement of an extensive network redesign. Thus, the amplification of the signals S at a point between the output of a receiving network element 806 and the input of a transmitting network element 806 is substantially zero dB.

In a preferred embodiment of the present invention, a network element 400 is adapted for bidirectional transmission as illustrated in figure 4. The network element in the preferred embodiment comprises one optical amplifier 406 for at least a first direction, one attenuator 404 for at least a second direction, one user port 410, one trunk port 412 and preferable one management unit 408 for each fiber pair (or for each single fiber strand if the fiber is adapted to bidirectional transmission) . The amplifier and the attenuator, respectively, may affect one or more wavelength. The user port 410 can be connected to terminals or/ and other user ports located on similar network elements 402.

A connection between two sites A and B, comprising of a network element 502 at Site A and a network element 504 at Site B connected by fiber strands 506 as depicted in the fiber section 500 is illustrated in figure 5. The network elements are adapted for bidirectional transmission as in the preferred embodiment above. The signals S having an original power level enter the network element 502 at site A through its user port 516, is amplified in an optical amplifier 508 with X dB and exits through a trunk port 518 of said network element 502. The trunk port is connected via the strand of fiber 506, which has an attenuation of Y dB, from Site A to Site B where it is connected to a trunk port 520 of network element 504 at site B. The signals S enter the attenuator 510, and the attenuator adjusts its attenuation in order to provide signals S with a power level that is substantial identical to the original power level of the signals S. When the signals are attenuated they exit via the user port 522. According to the present invention, the total gain of the signals S from site A to site B is zero dB.

For a corresponding signal in the opposite direction, the same scenario happens but starts in the user port 522 of network element at B, amplified at the optical amplifier 514, attenuated at the attenuator 512 and exiting in the user port 516 of network element at A.

However, as illustrated in figure 9, other elements 906 may be connected between the attenuator 906 and the user port 908. In such cases, the attenuator must compensate for possible losses caused by the other elements 906. I.e., if the other element has a loss of 2 dB the attenuator should attenuate the signals to a level of 2 dB above the original gain. It should be noticed that the management unit is not shown in figure 5.

In the preferred embodiment shown in figure 5, the unity gain of the fiber section is provided by adjusting the attenuator to a suitable attenuation level. The unity gain may also be provided by adjusting the gain of the amplifier in the transmitting network element or by adjusting both said amplifier and utilizing an attenuator in the receiving network element.

In the preferred embodiment of the present invention the management unit 408 controls the attenuation level of the attenuator 404, the adjustment of the attenuation level may however also be performed manually. It is possible to perform a fine tuning of the amplification on the optical amplifier 406. The management unit is connected to the input from the fiber cable and the amplified output signals respectively. Each pair of amplifier/ attenuator may comprise a separate monitoring/ management unit. The monitoring/management signal is preferably not amplified, and may be transmitted on a separate wavelength.

Management units controlling different information signals are interconnected using a switch function where multiple management signals are combined and routed using simple protocols and algorithms, creating with regard to the management signal e.g. an automatic discovery of remote management unit, an automatic topology repair function, automatic path selection between management units and participation in the network without prior configuration.

The network elements comprise semiconductor optical amplifier in a preferred embodiment in the present invention. However other optical amplifiers may also be used. A network element with a semiconductor optical amplifier is substantially less expensive than network elements according to prior art, which facilitates frequent connection of network elements. That is not possible with current technology, since the corresponding equipment used today is too expensive and bulky. The semiconducter optical amplifier within the network element addresses substantially only the optical layer, which implies that the transmission optical network is optic only, and is transparent to transmitted wavelengths within the allowed bandwidth. The bandwidth is independent for each wavelength, and transparent to upper layer or convergence protocols.

An optical network according to the present invention comprises multiple network elements, which are connectable via optical fibers. By deploying such an optical network where the output signal from substantially each fiber section has a unity gain preferably substantially zero dB, it is possible to connect additional network sections from the point with the unity gain without redesigning the current network. Thus, it is possible to provide a more flexible network by using the present invention.

In a preferred embodiment of the present invention, a network element comprises several Semiconducter Optical Amplifiers (SOA), e.g. 48, which means that several interfaces to optical fibers are provided, two SOAs per fiber pair. Each path between two network elements has bidirectional optic transmission.

The power levels of the signals are measured at network installation and stored in an onboard memory in the management unit. Furthermore, in order to maintain the unified gain, the signal level of the signal entering the user port of network element A is measured, as the signal after amplification. The received signal at network element B is thus measured just before exiting through the user port in order to measure the attenuation of the fiber strands. When the fiber section with the unified gain is working properly, the input power level at the user port of network element A should be the same as the output level exiting the user port of network element B. The intermediate measurement in network element A after the amplifier can serve as an indicator of whether any malfunction of the section is due to problems with the amplifier, or problems with the fiber section connecting the trunk ports.

The signal information of all points can be retrieved and stored in the management unit for diagnosis and registration of ageing and temperature variations and other functions associated with operating the entire network. The amplifier 406 gain can also be adjusted slightly in order to compensate for such variations. The power values may be transmitted by using inband or "slightly" out of band signalling. All values, such as input and output power levels on a path are available to one or more network management systems via the management unit's signal. The network element comprises units, e.g. a photo detector and an A/D converter, for measuring the transmitted optical power and controlling the transmitted power and the receiver attenuation. The management unit controls the attenuation level of the attenuator, when the attenuator is electrical adjustable.

In the preferred embodiment, a managing unit for remote control of all adjustable parameters such as fine tuning of amplifier gain, and adjustment of receiving attenuation is included within the optical network. Units for implementing automatic discovery of neighbouring network element (each network element is provided with an identity) that facilitate a visual representation of the resulting topology and units for management path establishment (i.e., automatic discovery of an added network element and automatic control and monitoring of network elements) may also be included within the network element according to the present invention or implemented in a separate element.

In a first scenario, according to the optical network 600 shown in figure 6, a path from a first location A via a second location B to a third location C is connected with fiber strands. At site A, a network element N2 terminates some or all of the fiber strands, the network element N2 is connected to a network element N3 at site B via the fiber cable. In location B the two network elements N3 and N4 connect the cable from A and the cable from C in accordance with figure 5. However, one single network element may also be used if the number of optical amplifiers and attenuators is sufficient. Each pair of amplifier/ attenuator terminates one optical path, i.e., either one single fiber with two circulators or one fiber pair. In the preferred embodiment, one network element terminates an entire fiber cable, or at least a substantial part of the fiber cable. Next, at site C the cable from B is terminated in a network element N5. Optical signals located between network element Nl and Network element N2 have substantial identical power levels, like the signals between network elements N3 and N4. Thus, the amplification of the signals S at the points between the output of a receiving network element Nl and the input of a transmitting network element N2, preferrably located at the same site, is substantially zero dB, i.e., between Nl and N2 and between N3, N4 and N6.

When each fiber section (e.g. the path between A and B) is adjusted for the unity gain, any output port 522 of a network element can be interconnected to any other network element input port 516 while the total optical path (e.g. the path between A and C) still having 0 dB gain or loss as in a second scenario described below.

In the second scenario, according to the optical network 700 shown in figure 7, a fourth location D is connected via a cable to the site B by means of a further network element N6. The gain between the network elements N3 and N4 is still substantially 0 dB, which facilitates the additional connection of Site D by using the network element N6 without a redesign of the entire network. The fiber cable from B is terminated in a network element N7 at site D. All or any optical paths can be directed to either A or C without readjustments.

Any fiber section in a network with, e.g. a mesh topology, that comprises the network elements used in the present invention having the unity gain at substantially each fiber section, can thus be extended or reconfigured without regard to adjustment of existing paths.

Thus, the method for controlling the power gain of optical signals in an optical network comprising multiple fiber sections 800, 808, comprises the steps of:

-receiving optical signals S having a first power level at the first network element

802

- transmitting said optical signals S having a second power level on the fiber strand having an attenuation, and - adjusting the power level of the output optical signals from the receiving means

806, 812 to a power level substantial identical to the first power level. The method may be implemented by means of a computer program product comprising the software code means for performing the steps of the method. The computer program product is run on processing means within a network element, or in a separate element connected to a network element. The computer program is loaded directly or from a computer usable medium.

The present invention is not limited to the above-described preferred embodiments. Various alternatives, modifications and equivalents may be used. Therefore, the above embodiments should not be taken as limiting the scope of the invention, which is defined by the appending claims.

Claims

1. Optical fiber section (800;808;500) comprises a first network element (802) comprising means for receiving optical signals (S) with a first power level and means for transmitting said optical signals with a second power level, said first network element (802) is connected to a first end of a fiber strand (804), a second end of said fiber strand (804) is connected to a second network element (806) comprising means for receiving said optical signals (S), said fiber strand, characterised in that the fiber section (800;808;500) comprises means for adjusting the power level of the output optical signals from the second network element (806) to a power level substantial identical to the first power level, resulting in gain of substantially zero dB over the entire fiber section (800;808;500).

2. Optical fiber section (800;808;500) according to claim 1, wherein the means for adjusting the power level of the output optical signals is an attenuator

(510) located within the second network element (504).

3. Optical fiber section (800;808;500) according to claim 1, wherein the means for adjusting the power level of the output optical signals is an optical amplifier (508) located within the first network element (502).

4. Optical fiber section (800;808;500) according to claim 1, wherein the means for adjusting the power level of the output optical signals is an attenuator (510) located within the second network element (504) and an optical amplifier (508) located within the first network element (502) .

5. Optical fiber section (800;808;500) according to any of claims 2 and 4, wherein the attenuator (510) is electrically adjustable.

6. Optical fiber section (800;808;500) according to any of claims 3 and 4, wherein the optical amplifier (508) is a semiconductor amplifier.

7. Optical fiber section (800;808;500) according to claim 1, wherein it comprises means for measuring the first and second power level of the signals, and means for transmitting a result of said measurements to any of said first (502) and second network elements (504) in order to maintain the power level of the output signals substantial identical to the first power level.

8. Method for controlling the power gain of optical signals in an optical network, said optical network comprises multiple fiber sections (800;808;500), wherein each fiber section (800;808;500) comprises a first network element (802) that is connected to a first end of a fiber strand (804), a second end of said fiber strand (804; 814) is connected to a second network element (806), said method comprises the steps of:

-receiving optical signals (S) having a first power level at the first network element (802)

- transmitting said optical signals (S) having a second power level on the fiber strand, characterised in that the method comprises the further step of: - adjusting the power level of the output optical signals (S) from the second network element (806) to a power level substantial identical to the first power level, resulting in a gain of substantially zero dB over each fiber section (800;808;500).

9. Method according to claim 8, wherein the step of adjusting the power level of the output optical signals is performed by an attenuator (510) located within the second network element (504) .

10. Method according to claim 8, wherein the step of adjusting the power level of the output optical signals is performed by an optical amplifier (508) located within the first network element (502).

11. Method according to claim 8, wherein the step of adjusting of the power level of the output optical signals is performed by an attenuator (510) located within the second network element (504) and by an optical amplifier (508) located within the first network element (502).

12. Method according to any of claims 9 and 11, wherein the attenuator (510) is electrically adjustable.

13. Method according to any of claims 10 and 11, wherein the optical amplifier (508) is a semiconductor amplifier.

14. Method according to claim 9, wherein the method comprises the further step of:

-measuring the first and second power level of the signals (S) , and -transmitting a result of said measurements to the attenuator (510) in order to maintain the power level of the output signals (S) substantial identical to the first power level.

15. Method according to claim 10, wherein the method comprises the further step of:

-measuring the first and second power level of the signals (S), and -transmitting a result of said measurements to the amplifier (508) in order to maintain the power level of the output signals (S) substantial identical to the first power level.

16. Method according to claim 11, wherein the method comprises the further step of: -measuring the first and second power level of the signals (S) , and

-transmitting a result of said measurements to the attenuator (510) and the amplifier (508) respectively in order to maintain the power level of the output signals (S) substantial identical to the first power level.

17. A computer program product directly loadable into a network element, or in a separate element connected to a network element within an optical network, comprising the software code portions for performing the steps of any of claims 8-16.

18. A computer program product stored on a computer usable medium, comprising readable program for causing a processing means within a network element, or in a separate element connected to a network element within an optical network to control the execution of the steps of any of claims 8-16.

19. Optical network (700) comprising a plurality of optical fiber sections (800;808;500), said sections (800;808;500) comprises a first network element (802) comprising means for receiving optical signals (S) with a first power level and means for transmitting said optical signals with a second power level, said first network element (802;810) is connected to a first end of a fiber strand (804), a second end of said fiber strand (804) is connected to a second network element (806) comprising means for receiving said optical signals (S), said fiber strand is characterised in that each fiber section (800;808;500) comprises means for adjusting the power level of the output optical signals from the second network element (806) to a power level substantial identical to the first power level.

20. Optical network (700) according to claim 19, wherein the means for adjusting the power level of the output optical signals is an attenuator (510) located within the second network element (504).

21. Optical network (700) according to claim 19, wherein the means for adjusting the power level of the output optical signals is an optical amplifier (508) located within the first network element (502).

22. Optical network (700) according to claim 19, wherein the means for adjusting the power level of the output optical signals is an attenuator (510) located within the second network element (504) and an optical amplifier (508) located within the first network element (502).

23. Optical network (700) according to any of claims 20 and 22, wherein the attenuator (510) is electrically adjustable.

24. Optical network (700) according to any of claims 21 and 22, wherein the optical amplifier (508) is a semiconductor amplifier.

25. Optical network (700) according to claim 19, wherein it comprises means for measuring the first and second power level of the signals, and means for transmitting a result of said measurements to any of said first (502) and second network elements (504) in order to maintain the power level of the output signals substantial identical to the first power level.