TECHNICAL FIELD

The disclosure relates to the field of customer premises equipment devices being coupled via a digital subscriber line to a service provider delivering broadband services to a customer.

BACKGROUND

Residential gateways are widely used to connect devices in a home of a customer to the Internet or any other wide area network (WAN) . Residential gateways use for example digital subscriber line (DSL) technology that enables a high data rate transmission over copper lines. Over the years, several DSL standards have been established differing in data rates and in range, for example ADSL, ADSL2, VDSL and VDSL2, which are referred to in this context as xDSL.

Network operators, e.g. Internet service providers (ISP), are managing a large amount, up to millions, of residential gateways, and also other devices such as routers, switches, telephones and set-top boxes, which are understood in this context as customer premises equipment (CPE) devices.

FTTdp (fiber to the distribution point) is bringing the fiber optic connection of the service providers closer to the home of the customer. However the last few 100 meters of the broadband connection are still handled by the existing copper wire to the home (currently used for legacy technologies such as PSTN and xDSL) .

In order to match the increased speed capabilities that come with bringing the fiber nodes closer to the home, the technology on the copper wire is also evolving to higher speeds. This is the base for the introduction of G.fast as an access technology on the copper wire.

Figure 1 shows the migration from copper links to fiber links closer to the home. For G.fast, a Distribution Point Unit (DPU) provides the link between the optical fiber of the Central Office (CO) with the CPE of the customer. In order to enable a flexible deployment of the DPU close to the home of the customer, the G.fast standardization foresees reverse power feeding for the DPU. In this

scenario, the DPU is powered by the CPE, or a separate power injector located at the customer's premises. This enables a flexible placement of the DPU and allows for multiple subscribers to provide power for a DPU.

The benefits of using reverse power feeding lie mainly in flexibility and cost advantages, but several challenges arise. There are safety requirements that need to be met, for example since power is running on the telephone wires, the user might be exposed, or old legacy equipment that might still be connected to the telephone wire might be damaged by this power, therefore, the maximum voltage allowed needs to be lower than 60V. Other important aspects to be considered are that the CPE devices can be located at different ranges from the DPU and each cable might have different physical characteristics: diameter, impedance, power loss. This leads to the fact that the power

available at the PDU depends on distance and wire type and that the load from every home is different. A customer at a shorter distance will deliver more current and therefore pay more for electricity than a customer at a longer distance. Therefore some kind of fair power sharing method is needed to accommodate for this. US 8,601,289 Bl discloses an optical communications system including a plurality of CPE devices, each having a reverse power supply circuit and each being connected to a wire pair and configured to transmit and receive data and provide reverse power over the wire pair. A power

management circuit is connected to each of the

communication ports and configured to receive power and to provide power sharing and manage power consumption and power supply redundancy from the plurality of CPE devices through the communication ports. SUMMARY

A method for powering a distribution point unit by at least one customer premises equipment (CPE) device comprises: a first CPE device providing power (Pmax) for powering the DPU, the CPE device delivering said power via a copper line to the DPU, the DPU starting operation, and the DPU

determining the optimal power output (Popt) for the first CPE device. If Popt > Pmax, then the DPU makes the power allocation for that copper line temporarily permanent. The DPU checks available power from further copper lines, and the DPU selects a second available active CPE device for powering the DPU, The DPU determines Popt for the second CPE device, and if the delivered power from the first and the second CPE device is above an operating power of the DPU, then the DPU redistributes the power between the two CPE devices.

In preferred embodiment, the DPU selects one active CPE device after the other for powering the DPU, the DPU determines Popt for each further CPE device, and the DPU redistributes the power from all active CPE devices, to obtain from the CPE devices its operating power. In

particular, the DPU uses a water filling algorithm for distributing the power between the CPE devices, and each CPE device provides a maximum power for powering the DPU, when starting operation. In another aspect, the DPU determines Popt for each CPE device being connected to the DPU and stores Popt in a smart grid table. The DPU determines Popt for each CPE device by using a Dual Ended Line Test and/or a Single Ended Line Test.

In a further aspect, if a further CPE device starts

operation and delivers power via a further copper line, then the DPU selects the maximum possible available power as delivered by the further copper line, the DPU assignes temporarily the maximum possible available power for the further copper line, and determining Popt for the further CPE device by looking at the smart grid table. If the Popt is larger than the maximum possible available power, then the DPU makes the temporarily assigned power permanent for the further CPE device, removes the further copper line from the list of active lines, and continues with another active copper line. If Popt from the further copper line is smaller than the maximum possible available power, then the DPU redistributes the excessive power as provided by the further CPE device between a11 active CPE devices by using the water filling algorithm, and continues with another active copper line.

Each CPE device is designed in particular for an operation with the DPU in accordance with a G.fast standard.

A DPU comprises a processor and a memory for performing the method.

A non-transitory program storage medium, being readable by a computer, comprises executable program code for

performing the method.

BRIEF DESCRIPTION OF THE DRAWINGS Preferred embodiments of the disclosure are explained in more detail below by way of example with reference to schematic drawings, which show:

Fig. 1 a migration of broadband services from copper links to fiber links closer to the home,

Fig. 2 a distribution point unit being adapted to

operate with a multitude of CPE devices, and Fig. 3 a method for powering the distribution point unit of figure 2.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS In the following description, example methods for powering a distribution point unit (DPU) by at least one customer premises equipment device (CPE) device are described. For purposes of explanation, various specific details are set forth in order to provide a thorough understanding of preferred embodiments. It will be evident, however, to one skilled in the art that the present disclosure may be practiced without these specific details.

The CPE device includes in a preferred embodiment at least a controller, e.g. a microprocessor, a memory, in which an operating system is stored for the operation of the CPE device, a CPE power supply for powering the CPE device and the DPU, and a port for a broadband xDSL connection, in particular for a G.fast DSL connection. A CPE device of this kind is for example an access gateway, a residential gateway, a business gateway, a router or an Internet switch .

The DPU includes in a preferred embodiment at least a controller, e.g. a microprocessor, a memory, in which an operating system is stored for the operation of the DPU, a DPU power supply receiving energy from a CPE device and powering the DPU, a port for an optical fiber connection, e.g. for a connection with a central office, and ports for a broadband xDSL connection, in particular for a G.fast DSL connection.

A DPU 1, being connected via a fiber optical line with a central office 2 and being connected with n CPE devices Ci, i = 1 - n, is schematically illustrated in figure 2. The value "n" represents the maximum number of CPE devices that the DPU 1 supports. Each CPE device Ci is connected via a respective copper line Li, i = 1 - n, the traditional telephone copper wire, with the DPU 1. From the n CPE devices, m CPE devices are active and provide energy for powering the DPU 1.

In a preferred embodiment, for a reverse powering of the DPU 1, three basic parts are provided: - A power source equipment (PSE) including the CPE power supply, being located at the end user side, e.g.

within the CPE device, that is responsible to inject the power into a copper line, e.g. copper line Li. Each copper line of the DPU is connected with a power extractor for receiving power from a CPE device. The

DPU power supply is present on the DPU to merge the power extracted from each line and to convert it into proper voltage levels. A reverse power management system coordinates the power functions over all active lines within the DPU.

A power management protocol for the communication between each PSE and the Reverse Power Management System of the DPU. The powering method works as follows: "n" is the maximum number of CPE devices that the DPU supports, as depicted in figure 2. At any point in time there will be m CPE devices active, where m is 0<=m<=n. First, during DSL

initialization and exchanging measurement information that are done during DSL training and channel discovery, it is possible for the microprocessor of each CPE device, to determine the distance at which the CPE device is

operating. Using DELT (Dual Ended Line Test) or SELT

(Single Ended Line Test) primitives, it is also possible to identify the type of cable characteristic. By feeding these data to the microprocessor of the DPU via the power

management protocol, it is possible for the DPU 1 to estimate the power loss for each copper line of each active CPE device. Because figure 2 depicts a multiuser environment, DSL vectoring is a mandatory requirement for FTTdp and the G.fast operation. Therefore, the number of active CPE devices m can be estimated by the DPU by looking at a crosstalk transfer function being used for the G.fast transmission. By exchanging all these information between the active CPE devices and the DPU 1 using the power management protocol, it is possible to build a smart grid table that identifies for each active CPE device the maximum amount of power, taking also into account the power loss due to cable type and distance, that can be delivered on that line. This smart grid table is basically a "n x m" matrix with P _n;m matrix elements, in which each diagonal element represents the case when there is only one single CPE device active and in that case the element P _n , _m is equal to the total operating power needed to power the DPU, P _tot , while each of the non-diagonal elements needs to be

determined such that:

∑ ^m p = P

m=l m,n Cat .

To solve this, the active CPE devices are handled in a Round-Robin fashion: when the first CPE device becomes active, all the power budget is allocated to that CPE device. Since all the power is allocated to a single CPE device, the total power needed from that CPE device reduces when another line becomes active.

When a new CPE device becomes active, the power budget allocated to the current CPE device will be redistributed first in an evenly manner. But since the two active CPE devices may have two different power losses due to

different loop length, this might not be the most optimal configuration. Therefore, a water filling method is used to optimize it.

One of the advantages of this solution is that it uses an iterative multiuser water filling algorithm to optimize the power distribution among the different lines between the CPE devices and the DPU. The solution proposed also

foresees a direct link communication between the PSE of the DPU and the microprocessor of the CPE devices to share power consumption information. The CPE devices and the DPU will also use standard initialization messages to estimate the number of active users, based e.g. on crosstalk

information, and the loop length over which each user is located. A dedicated message is used between the DPU and the CPE devices to exchange power consumption information, actual available power budget, and guaranteed available total power.

A preferred method for powering the DPU 1 by at least one CPE device Cm is illustrated in more detail as a flowchart in figure 3. For starting the DPU 1, a first CPE device, for example CPE device CI, provides a power Pmax via the copper line LI to the DPU 1, for powering the DPU 1. The power Pmax from the single CPE device CI is sufficient for starting the operation of the DPU 1. Should the power Pmax be larger than the power Ptot as required by the DPU 1, then the power Pmax as received from the CPE device CI is reduced to the power Ptot by using the power management protocol . The DPU 1 then looks for a further active copper line, step S10. In case an active copper line is detected, e.g. the copper line L2, then the maximum possible available power Pmax as provided by the copper line L2 is selected, step S12. The power Pmax as delivered by the CPE device C2 is then assigned temporarily for the copper line L2, S14. In a further step, the optimum power Popt for the CPE device C2 is determined by looking at the smart grid table, S16. Is the power Popt larger than the power Pmax, S18, then the temporarily assigned power Pmax for the CPE device C2 is made permanent, S20, and the copper line L2 is removed from the list of active copper lines, S22. The method continues then by looking for a further active copper line, S24, and returns to step S12, if a further copper line is available not being optimized yet.

If the power Popt is smaller than Pmax, S18, then the excessive power as provided by the CPE device C2 is

redistributed among all active CPE devices, S26, e.g. by using the water filling algorithm. The method continues then with step 24, and returns to step S12, if a further copper line is available not being optimized yet, or, if the list of active, not optimized copper lines is empty, S28, returns to step S10 and begins with the procedure from the beginning.

When the CPE device Cm starts operation, the DPU 1 selects the maximum possible available power Pmax as delivered by the copper line Lm. The power Pmax from the CPE device Cm is then assigned temporarily for the copper line Lm, S14. In a further step, the optimum power Popt for the CPE device Cm is determined by looking at the smart grid table, S16. Is the power Popt larger than the power Pmax, S18, then the temporarily assigned power Pmax for the CPE device Cm is made permanent, S20, and the copper line Lm is removed from the list of active lines, S22. The method continues then by looking for a further active copper line, S24, and returns to step S12, if a further copper line is available not being optimized yet.

If the power Popt is smaller than Pmax, S18, then the excessive power as provided by the CPE device Cm is

redistributed among all active CPE devices, S26, e.g. by using the water filling algorithm. The method continues then with step 24, and returns to step S12, if a further copper line is available not being optimized yet, or, if the list of active, not optimized copper lines is empty,

S28, returns to step S10 and begins with the procedure from the beginning. After step S28, the power from all active CPE devices is distributed in an optimized manner according to the water filling algorithm.

The method as described with regard to figure 3 defines therefore an algorithm that allows the DPU 1 to work with any number m of active CPE devices, at any point in time, even with a single CPE device. In case of a multitude of CPE devices, the reverse power management system of the DPU makes sure to have a fair distribution of power over all active CPE devices even in case of different lengths of the copper lines, by also considering the actual power state and DPU load of each copper line. In case of all copper lines being active, the power management protocol also ensures that the total power dissipation remains within the thermal limits of the passive cooling imposed by the DPU form factor and the environmental conditions, in which the DPU is deployed, buried or aerial, and therefore, the DPU power consumption scales with the number of active copper lines . Also other embodiments may be utilized by one skilled in the art without departing from the scope of the present disclosure. The method for powering the DPU is in

particular not limited to reverse power from gateways, but also any other CPE device may be used for powering the DPU. The disclosure resides therefore in the claims herein after appended .