Method for configuring a mesh topology Ethernet optical network

Abstract

The invention relates to a method (100) for configuring a mesh topology Ethernet optical network connecting a plurality of devices and switches. Each switch includes at least one transceiver. Each transceiver has a transmitting section and a receiving section. Each transceiver section is associated with a virtual port. The method (100) includes an initialization (101), a network mapping (103) including the identification of possible links enabling communication between two devices, a possible link being a sequence of virtual ports enabling communication. The method (100) then includes a definition (104) of a network configuration including selected links, a deployment (105) of the network configuration including the disabling of virtual ports not used by the selected links, and a power-off (106) of the transceiver sections equipping each disabled virtual port.Figure for the abbreviation: [Fig. 1].

Description

Title of the invention: Method for configuring a mesh topology Ethernet optical network Technical field of the invention

[0001] The invention relates to a method for configuring a mesh topology Ethernet optical network that reduces the energy consumption of the network.

[0002] It relates to the technical field of Ethernet optical networks with mesh topology. Previous technique

[0003] Mesh topology communication networks have the advantage, compared to star topology networks, of being robust to a certain number of failures. Indeed, if a node or a connection fails, the mesh topology network can reroute data along other paths, which increases the network's resilience to failures. This advantage consequently entails a risk of communication loops occurring within the network.

[0004] When a loop exists in a network, data packets can be sent continuously around the loop, creating unnecessary traffic duplication. This overloads the network, reduces bandwidth availability for other tasks, and leads to a risk of instability and failures in the mesh network.

[0005] To solve the problems related to loops, several loop prevention and management mechanisms have been developed:

[0006] Spanning Tree Protocol (STP): Protocol designed to detect and disable links that create loops, while ensuring that there is always at least one active path between any two nodes in the network.

[0007] Rapid Spanning Tree Protocol (RSTP): Enhancement of STP to accelerate the convergence of network topologies.

[0008] Multiple Spanning Tree Protocol (MSTP): Allows configuration of multiple Spanning Tree instances for VLANs.

[0009] These protocols allow the routing of communications within the mesh topology network to be configured, preventing the occurrence of loops. These protocols rely on disabling certain communication links within the network.

[0010] In the context of an Ethernet optical network, physical communication links are notably provided by electronic equipment called transceivers, which can combine a receiver and a transmitter and include, in particular, a photodiode and a laser. Each network switch, also called a node, includes at least one transceiver per communication link with another network switch. These communication links with other network switches or with Network-connected equipment is accessed via interfaces called ports. The term "port" can refer to a physical interface to which a transceiver is connected, in the case of an optical connection. The term "port" can also refer to a virtual interface corresponding to a data entry or exit point in the firmware of network equipment such as a switch, or in a network virtualization layer. Data transmitted, for example, through different virtual ports can be physically transmitted through a single physical port. In this text, when the term "port" is used to refer to a virtual interface, the term "virtual" is added. Therefore, we will refer to virtual ports.

[0011] To prevent loops, physical and / or virtual ports are disabled during communication routing. However, the transceiver, or a transceiver component, associated with a disabled port remains powered, and in particular its photodiode and laser remain illuminated. This results in suboptimal network power consumption.

[0012] Figure 2 is a representation of a prior art mesh topology Ethernet optical network. In this representation, transceivers are mounted on switches. Each transceiver is associated with either one bidirectional virtual port or two unidirectional virtual ports. Some virtual ports are disabled to prevent switching loops; however, the transceivers to which these virtual ports are associated remain powered. A powered-on transceiver consumes approximately one watt of power. Since an optical Ethernet topology network can include many switches, each switch having many transceivers, the power consumption of such a network is relatively high. Presentation of the invention

[0013] The present invention remedies the aforementioned drawbacks by proposing, according to a first aspect, a method for configuring a mesh topology Ethernet optical network connecting a plurality of devices, the network comprising a plurality of switches, each switch comprising a transceiver, each transceiver comprising a transmitting part and a receiving part, each transmitting or receiving part being associated with a virtual port, the method comprising: - an initialization process involving powering on each switch and, for each switch, powering on each transceiver, - a network map including the identification of possible links allowing communication between two devices, between two switches, or between a device and a switch, a link possible being a succession of virtual ports allowing said communication to be established. - a definition of a network configuration, the network configuration comprising a selection of links from among the possible links, - a network configuration deployment including, for each switch, the disabling of virtual ports not used by the selected links, - a power-off of the transmitting or receiving part of transceivers associated with each disabled virtual port.

[0014] The transmitting part of a transceiver is understood to be the part of an Ethernet optical transceiver dedicated to sending information through an optical fiber. The transmitting part of the transceiver may include, for example, a laser.

[0015] The receiving part of a transceiver is defined as the part of an Ethernet optical transceiver dedicated to receiving information transmitted through an optical fiber. The receiving part of a transceiver may include, for example, a photodiode.

[0016] A virtual port is understood to be a unidirectional virtual interface, for example, of a switch. Information sent from a first switch to a second switch passes through a virtual port of the first switch and a virtual port of the second switch. A transmitting part of a transceiver of the first switch is associated with the virtual port of the first switch, and a receiving part of a transceiver of the second switch is associated with the virtual port of the second switch.

[0017] Powering up each transceiver means powering up, also called supplying electricity, each part, transmitting and receiving, of each transceiver.

[0018] Such steps make it possible to configure a mesh topology Ethernet optical network in a way that avoids switching loops while reducing the power consumption of the network.

[0019] In particular embodiments the invention may further comprise one or more of the following features, taken individually or in all technically possible combinations.

[0020] According to one embodiment, the method according to the first aspect also includes a step of electing a master switch from among the switches, for network mapping, the possible links are identified between each piece of equipment and the master switch and between the master switch and each switch not being elected master switch and, for the definition of a network configuration, an uplink and a downlink are selected from among the possible links for each piece of equipment and for each switch.

[0021] A master switch is defined as one of the switches in a mesh Ethernet optical network that serves, in particular, as a reference point for mapping the network. The switch elected as the master switch could be, for example, the switch with a numerical identifier lower than the numerical identifiers of the other switches in the network. The numerical identifier of each switch in the network could be pre-configured by an operator so as to lead to the election of a master switch located near the predominant equipment connected to the network.

[0022] The uplink is understood to be a data transmission link from the link equipment or switch to the master switch.

[0023] A downlink is understood to be a data transmission link from the master switch to the link equipment or switch.

[0024] Using a master switch as a reference point for mapping the mesh network topology reduces the number of links to be identified. Without a reference point, links can be identified between each piece of equipment and each switch in pairs, resulting in a significantly larger number of links than the number to be identified between the master switch and each individual piece of equipment, and between the master switch and each other switch.

[0025] According to one embodiment, the selection of links among the possible links uses a comparison between a throughput capacity of one of the possible links and a throughput requirement of the link being selected.

[0026] The throughput capacity of a link refers to the data transmission rate available for the communication link. This throughput capacity is constrained by the element in the data transmission chain constituting the link that has the lowest remaining available throughput capacity. The throughput capacity of a link may take into account the throughput requirements of other links already allocated to the elements of the link being selected.

[0027] The bandwidth requirement of the link being selected refers to a projected or pre-configured bandwidth requirement between, for example, one switch and another switch or device for which the link is being selected. The bandwidth requirement of a link between two switches may be relatively low to allow for monitoring and control of the switch. The bandwidth requirement of a link between a switch and a device may be relatively high to meet the device's data transmission needs.

[0028] Such arrangements enable the method according to the invention to configure the network in such a way as to enable the network to meet the data transmission rate requirements of the equipment connected by the network.

[0029] According to one embodiment, the selection of links among the possible links uses a comparison between a response time of one of the possible links and the response time of another possible link.

[0030] Comparing the response times, commonly called "ping", of possible links allows, for example, for preferential use of links allowing a low response time and thus increasing the responsiveness of the network in the transmission of information between equipment.

[0031] According to one embodiment, the selection of links among the possible links uses a comparison between a jitter of one of the possible links and the jitter of another possible link.

[0032] Jitter, also called "jigue," refers to a variation in response time. Comparing the jitter of possible links makes it possible, for example, to favor links with stable response times and thus increase the predictability of data transmission times between devices.

[0033] According to one embodiment, for the definition of a network configuration, at least one virtual port in the succession of virtual ports of the downlink belongs to a switch to which no virtual port in the succession of virtual ports of the corresponding uplink belongs.

[0034] In this way, the downlink and uplink of each piece of equipment and each switch do not take the same path through the mesh topology Ethernet optical network.

[0035] Such arrangements allow the network to be configured in such a way as to prevent a single failure from systematically resulting in a loss of bidirectional communication between network elements. Furthermore, using different switches for uplink and downlink links makes it possible to define communication links with all network elements using fewer virtual ports. For example, a switch that would otherwise require a dedicated link to maintain its communication with the master switch can benefit from virtual ports being enabled for one of the uplink or downlink links of one of the devices to maintain its communication with the master switch.

[0036] According to one embodiment, the method according to the first aspect of the invention includes a link loss detection, the link loss detection triggering a re-energization of each part of the transceivers previously switched off.

[0037] Link loss detection refers, for example, to the observation of a number of untransmitted data packets that will lead to the conclusion that a communication link is no longer maintained by the network. Link loss is generally caused by a failure of a transceiver or a switch.

[0038] Such arrangements allow, in the event of failure, the network to be reconfigured by establishing, for a lost link, a new link using one or more parts of transceivers previously switched off.

[0039] According to a second aspect, the invention relates to a mesh topology Ethernet optical network comprising a plurality of switches, each switch comprising a transceiver, each transceiver comprising a transmitting part and a receiving part, each transmitting or receiving part being associated with a virtual port, the network is capable of being configured according to a protocol avoiding switching loops by disabling certain virtual ports, the switches are capable of being configured to cut off the power supply to the transmitting or receiving part of the transceivers associated with each disabled virtual port.

[0040] A protocol avoiding switching loops is understood to mean, for example, a method according to the first aspect of the invention, or a protocol such as STP, RSTP, MSTP or SPB “Shortest Path Bridging”.

[0041] Such arrangements allow the mesh topology Ethernet optical network to transmit data between equipment while exhibiting lower energy consumption than prior art Ethernet optical networks and while maintaining reconfiguration possibilities in the event of failure of one or more network elements.

[0042] According to a third aspect, the invention relates to an Ethernet optical switch comprising a transceiver, the transceiver comprising a transmitting part and a receiving part, each transmitting or receiving part being associated with a virtual port, the switch is capable of being configured according to a protocol avoiding switching loops by disabling certain virtual ports of the switch and capable of being configured to cut off the power supply to the transmitting or receiving part of the transceiver associated with each disabled virtual port.

[0043] Such arrangements allow the optical Ethernet switch to cooperate with identical optical Ethernet switches in order to constitute an optical Ethernet network according to the second aspect of the invention.

[0044] According to a fourth aspect, the invention relates to a transceiver comprising a transmitting part and a receiving part, each of the parts of the transceiver being able to be switched on or off independently of the other part of the transceiver.

[0045] Such arrangements allow the transceiver to equip switches according to the third aspect of the invention in order to constitute an Ethernet optical network according to the second aspect of the invention. Presentation of the figures

[0046] The invention will be better understood upon reading the following description, given by way of non-limiting example, and made with reference to the figures:

[0047] [Fig. 1] a schematic representation of an example of the method according to the first aspect of the invention,

[0048] [Fig.2] a schematic representation of an example of an Ethernet optical network configured according to a prior art method,

[0049] [Fig.3] a schematic representation of an example of an Ethernet optical network configured according to an embodiment of the method according to the first aspect of the invention,

[0050] [Fig.4] a schematic representation of an example of an Ethernet optical network configured according to another embodiment of the method according to the first aspect of the invention.

[0051] In these figures, identical reference numerals from one figure to another designate identical or analogous elements. For clarity, the elements shown are not necessarily to the same scale, unless otherwise stated.

[0052] Detailed description of particular embodiments of the invention

[0053] Figure 1 is a schematic representation of an example of method 100 according to the first aspect of the invention. Method 100 for configuring a mesh topology Ethernet optical network connects a plurality of devices 3. The network comprises a plurality of switches 2, each switch comprising a transceiver 4.

[0054] The term "include a transceiver" means to include at least one transceiver. The switches in a mesh network topology generally contain more than one transceiver. Transceiver 4 has a transmitting section and a receiving section. The transmitting section of transceiver 4 sends information to another element, such as another switch or piece of equipment, via an optical fiber. The transmitting section of transceiver 4 may include a laser. The receiving section of transceiver 4 receives information from another element, such as another switch or piece of equipment, via an optical fiber. The receiving section may include a photodiode. The process 100 comprises several distinct steps to ensure network configuration. Initialization 101 is the first step of the process 100. This step involves powering on each switch 2.For each switch, each transceiver 4 is also powered on; when one of the transceivers 4 is powered on, each part of the transceiver is powered on. This initial power-up ensures that all network elements are operational and ready for the subsequent steps of the process.

[0055] According to one embodiment, the method 100 includes an election 102 of a master switch from among the switches 2 after initialization 101. This election step 102 determines which switch will serve as the reference point for network mapping. The switch elected as master may be the one with a lower numerical identifier than the other switches in the network. For example, each switch may have a priority number, and the switch with the lowest priority number is elected master. This election establishes a reference point for network mapping. The numerical identifier may be a switch address and may be pre-assigned by a person in charge of network configuration so as to direct the election towards a switch close to the predominant equipment in the network.

[0056] According to one embodiment, the method 100 does not include a step of selecting 102 a master switch.

[0057] A mapping 103 of network 1 constitutes the next step. When the process 100 does not involve the election of a master switch, the mapping 103 involves identifying possible links between switches 2 and / or between devices 3. The possible links to be identified may be all possible pairs of elements, an element being either one of the switches 2 or one of the devices 3. According to another example, the possible links to be identified may be predefined using, for example, configuration files. Some devices 3 may not necessarily need to be connected to each other via network 1.

[0058] When the process 100 involves the election of a master switch, the mapping 103 involves the identification of possible links between each piece of equipment 3 and the master switch, as well as between the master switch and each other switch 2. One of the objectives of the network may be to ensure communications between all the elements, a communication link will therefore be necessary between the master switch and each of the other switches as well as each of the pieces of equipment.

[0059] A possible link is defined as a series of virtual ports connecting two elements. A virtual port is a virtual interface present in a network virtualization layer 1. Each virtual port is associated with one of the switches and ensures unidirectional data transit between that switch and another adjacent element in the network. Two adjacent elements in the network are understood to be two elements physically connected directly, without any other element between them. Each virtual port can be associated with a transceiver segment 4.

[0060] This mapping is essential for understanding the network structure and planning communications. Definition 104 of a network configuration follows the mapping. The network configuration includes a link selection from among the possible links. In order to avoid switching loops, for each pair of elements that will communicate with each other via network 1 and for each communication direction, only one communication link will be selected from among the possible links. This selection may be based on various criteria, such as the throughput capacity, response time, and jitter of the possible links.

[0061] The following substeps are a non-limiting example of selecting a link from among the possible links: 1. Collect, for each link to be selected (for example, for each piece of equipment and each switch), a bandwidth requirement; this bandwidth requirement can be collected, for example, from an architecture file. 2. Sort each link to be selected in descending order of bandwidth requirement in a list of links, 3. Process the first link in the list by selecting the possible link with a throughput capacity closest to the required throughput, the throughput capacity of a possible link being the throughput capacity of the element with the lowest throughput capacity among the elements providing the possible link, 4. Deduce the bandwidth requirement of the processed link from the bandwidth capacity of the elements ensuring the link. 5. Process the next link in the list of links by performing sub-steps 3 and 4 until all links in the list of links have been processed.

[0062] When several possible links have equal throughput capacity, another criterion may be used to choose the link from among the possible links. For example, the response times of the possible links, the jitter of the possible links, or an arbitrary criterion such as a higher or lower identifier number of one of the elements of the possible links. Elements of the possible links are understood to mean, for example, a switch or a virtual port.

[0063] This step of defining a network configuration 104 ensures that the network will be configured to meet the specific data transmission needs presented by the equipment 3.

[0064] The next step is the deployment of the network configuration. During this deployment, for each switch 2, the virtual ports not used by the selected links are disabled. This disabling prevents multiple communication links, also called paths, from being used to transmit data between two devices, thus preventing switching loops from occurring.

[0065] The power-off 106 of the transceiver portion 4 associated with each disabled virtual port is the final step of the method 100. This power-off disconnects the power supply to the transceiver portions 4, including a laser and / or a photodiode that are not required, thus contributing to a significant reduction in network energy consumption. When both the transmitting and receiving portions of one of the transceivers 4 are powered off, one of the transceivers 4 can be completely powered down. The method 100 may also include link loss detection 107. Link loss detection triggers a power-up 108 of each previously powered-off transceiver portion. Link loss is generally caused by a malfunction of a network component.If the elected master switch malfunctions, the steps of election 102, mapping 103, configuration definition 104, configuration deployment 105, and shutdown 106 of unused transceiver parts will bring the network back to a fully functional state. If another network component malfunctions, or if network mapping 103 is performed without a master switch as a reference point, the step of electing a new master switch 102 will not be necessary to reconfigure network 1 after the malfunction. In this way, the network can reconfigure itself in case of failure, using previously shut-down transceiver parts to establish new communication links.

[0066] Figure 3 is a schematic representation of an example of an Ethernet optical network 1 configured according to an embodiment of method 100 according to the first aspect of the invention. Four switches 2 are shown, and for each switch 2, five transceivers 4 are shown. Three devices 3 are also shown. According to one embodiment, in this example, the switch with identifier 1 can be elected as the master switch, and the network 1 is configured according to method 100. The links between devices A, B, and C with the switch with identifier 1 are established. The links between switches 2, 3, and 4 with the switch with identifier 1 are also established. The transceiver components not used to establish these links are switched off, which reduces the power consumption of the network 1 shown.

[0067] In this example, each transceiver provides bidirectional communications, each selected uplink transits through the same transceivers as the corresponding selected downlink.

[0068] Figure 4 is a schematic representation of an example of an Ethernet optical network configured according to another embodiment of the method according to the first aspect of the invention. Similar to the representation in Figure 3, equipment 3, switches 2, and transceivers 4 are shown.

[0069] In this example, each transceiver provides unidirectional communications, but only a portion of each transceiver is used. Four switches 2 are shown, each switch having five transceivers 4. Switch 2 with identifier 1 can be elected as the master switch. The links between the switches and the master switch are unidirectional, and each uplink transits through a different switch than its corresponding downlink. For example, communications from switch with identifier 3 to the master switch transit through switch with identifier 4, while communications from the master switch to switch with identifier 3 transit through switch with identifier 2.

[0070] The network represented by [Fig.3] comprises twenty transceivers 4, therefore forty parts of which twenty-two parts are switched off, in comparison the network represented by [Fig.4] comprises twenty-six switched-off parts.

[0071] The network shown in [Fig.4] therefore shows an additional reduction in energy consumption compared to the network shown in [Fig.3].

[0072] Such provisions therefore make it possible to further reduce the energy consumption of the network 1 configured according to process 100.

[0073] According to a second aspect, the invention relates to a mesh topology Ethernet optical network 1 as shown in Figures [Fig. 3] and [Fig. 4]. The network 1 is configurable by method 100 according to the first aspect of the invention. Each switch 2 has a transceiver 4. A transceiver is understood to mean at least one transceiver; each switch commonly has several transceivers. Each switch 2 may also have information, such as, for example, a numerical identifier, that can be used to elect a master switch from among the switches 2. Each switch 2 also has a circuit breaker mechanism dedicated to each transceiver 4 that it may be equipped with. The circuit breaker mechanism may be digitally controlled to allow each switch 2 to selectively power one or more of its transceiver 4s.

[0074] Such arrangements allow the network 1 to be configured according to the method 100 and thus ensure communications between a plurality of equipment while avoiding the appearance of switching loops, presenting robustness to faults thanks to the possibility of rerouting communications in the event of failure of an element of the network 1 and presenting reduced power consumption compared to prior art mesh topology Ethernet optical networks.

[0075] According to a third aspect, the invention relates to an Ethernet optical switch as shown in Figures [Fig. 3] and [Fig. 4]. The switch 2 is suitable for configuration according to method 100 in order to constitute an element of the network 1. The switch 2 has transceivers 4 and mechanisms for supplying power to each transceiver component is powered off, and the power supply to each component is cut off. These mechanisms can be digitally controlled, for example, by the "firmware" of switch 2. "Firmware" refers to low-level software embedded in the switch's computer system, enabling control and configuration of the switch.

[0076] According to a fourth aspect, the invention relates to a transceiver 4 comprising a transmitting part and a receiving part. The transceiver 4 is configured to allow each of its parts to be powered independently of the other part. In this way, the transceiver 4 according to the fourth aspect of the invention can be used to equip switches 2 in order to form a network 1 according to the second aspect of the invention.

Claims

Demands

1. A method (100) for configuring a mesh topology Ethernet optical network (1) connecting a plurality of devices (3), the network comprising a plurality of switches (2), each switch comprising a transceiver (4), each transceiver (4) comprising a transmitting part and a receiving part, each transmitting or receiving part being associated with a virtual port, the method (100) comprising: - an initialization (101) comprising powering up each switch (2) and, for each switch, powering up each transceiver (4), - a mapping (103) of the network (1) comprising the identification of possible links enabling communication between two devices (3) or between two switches (2) or between a device (3) and a switch (2), a possible link being a succession of virtual ports enabling said communication, - a definition (104) of a network configuration,the network configuration including a selection of links from among the possible links, - a deployment (105) of the network configuration including, for each switch (2), a deactivation of the virtual ports not used by the selected links, - a power-off (106) of the transmitting or receiving part of the transceivers associated with each deactivated virtual port.

2. A method (100) according to claim 1 further comprising a step of electing (102) a master switch from among the switches (2), wherein, for network mapping (103), possible links are identified between each piece of equipment (3) and the master switch and between the master switch and each switch (2) not being elected master switch and wherein, for defining (104) a network configuration, an uplink and a downlink are selected from among the possible links for each piece of equipment (3) and for each switch (2).

3. A method (100) according to any one of the preceding claims wherein the selection of links among the possible links uses a comparison between a throughput capacity of one of the possible links and a throughput requirement of the link being selected.

4. A method (100) according to any one of the preceding claims wherein the selection of links among the possible links uses a comparison between a response time of one of the possible links and the response time of another possible link.

5. A method (100) according to any one of the preceding claims wherein the selection of links among the possible links uses a comparison between a jitter of one of the possible links and the jitter of another possible link.

6. Method (100) according to claim 2, wherein, for the definition (104) of a network configuration, at least one virtual port of the downlink virtual port sequence belongs to a switch (2) to which no virtual port of the corresponding uplink virtual port sequence belongs.

7. Method (100) according to any one of the preceding claims comprising a link loss detection (107), the link loss detection triggering a re-energization (108) of each previously switched-off transceiver part.

8. Ethernet optical network with mesh topology (1) comprising a plurality of switches (2), each switch (2) comprising a transceiver (4), each transceiver (4) comprising a transmitting part and a receiving part, each transmitting or receiving part being associated with a virtual port, the network is capable of being configured according to a protocol avoiding switching loops by disabling certain virtual ports, the switches are capable of being configured to cut off the power supply to the transmitting or receiving part of the transceivers (4) associated with each disabled virtual port.

9. An Ethernet optical switch (2) comprising a transceiver (4), the transceiver (4) comprising a transmitting part and a receiving part, each transmitting or receiving part being associated with a virtual port, the switch (2) being capable of being configured according to a protocol avoiding switching loops by disabling certain virtual ports of the switch (2) and capable of being configured to cut off 15 the power supply of the transmitting or receiving part of the transceiver (4) associated with each disabled virtual port.

10. Transceiver (4) comprising a transmitting part and a receiving part, each part of the transceiver (4) being able to be turned on or off independently of the other part of the transceiver (4).