Abstract: A technique for managing transfer of data frames having a pre-determined mutual order is presented. Each data frame contains an order indicator, for example a sequence number, indicative of the position of the data frame in the pre-determined mutual order. A transmitting end provides each data frame with a satisfaction indicator that indicates whether acknowledgement about reception of earlier transmitted data frames is expected from a receiving end. When the satisfaction indicator of a data frame indicates that no acknowledgement is expected, the receiving end knows that the data frame can be treated as a beginning of a received data flow when checking the integrity of the received data flow on the basis of the order indicators. Thus, there is no need to establish a logical connection between the transmitting and receiving ends prior to the transfer of the data frames.

Abstract: A network element includes control entities (102, 103) and controllable entities (104-105) such as line interface units. The control entities configure the controllable entities to enable the network element to operate in a data transfer network. Each control entity maintains a configuration data-set in accordance with configuration information based on data received from the data transfer network and transmits the configuration data-set to one or more controllable entities. The control entity applies a configuration rule on the configuration information and updates the configuration data-set on the basis of the result of the configuration rule. The configuration rule is stateless so that its result is independent of earlier results of the configuration rule and of earlier configuration information. The control entities are enabled to keep their configuration data-sets nearly congruent with each other because they use the same configuration information and the same stateless configuration rule.

Abstract: A network element of a software-defined network includes a processing system for maintaining a control system for managing data. The control system is maintained in accordance with configuration data received from a controller of the software-defined network. In a case where the control system has no entry corresponding to frame information related to a received data frame, the processing system checks whether status information indicates availability of the frame information to the controller. If the status information does not indicate the availability, the processor system arranges the frame information to be available to the controller and updates the status information to indicate the availability. Thus, data traffic related to data-driven learning and occurring between the network element and the controller can be reduced with the aid of the above-mentioned status information.

Abstract: A network element of a software-defined network comprises a data transfer interface (210) for receiving and transmitting data and a processing system (215) for constructing a software-defined data path and a fixed-functionality data path such as for example an IP/MPLS or L2 switching path. The software-defined data path comprises look-up tables for selecting software-defined actions and the fixed-functionality data path defines fixed-functionality actions. The processing system is adapted to convert metadata associated with data managed by the software-defined data path to a data format suitable for the fixed-functionality data path when one or more of the fixed-functionality actions are needed in conjunction with forwarding the data. One of the fixed-functionality actions is selected at least partly on the basis of which one of the look-up tables was most recently accessed, and the determined fixed-functionality action is carried out so as to initiate the one or more needed fixed-functionality actions.

Abstract: A method for optimizing entity-specific configuration systems for functional entities of a network element of a software-defined network is presented. The entity-specific configuration systems are based on a network element-specific configuration system related to the whole network element and comprising configuration entries for determining operations to be carried out by the network element. The method comprises forming (401) relation data that indicates, for each of the functional entities, those of the configuration entries which are irrelevant to the functional entity under consideration so that operation of the functional entity is independent of these configuration entries. The relation data enables the entity-specific configuration systems to be constructed (402) so that the entity-specific configuration system of each functional entity is free from those configuration entries which, according to the relation data, are irrelevant to the functional entity under consideration.

Abstract: A configuration system for configuring the control-plane of a software-defined network is presented. The configuration system determines an entity-specific group of control-plane functionalities for each of controller entities located in a distributed way in the software-defined network. The control-plane functionalities are suitable for configuring data forwarding network elements, such as routers, to support services to be provided by the software-defined network. The configuration system composes, on the basis of the entity-specific groups, entity-specific portions of control data and transmits the entity-specific portions of the control data to the controller entities so as to configure each controller entity to support the corresponding entity-specific group of the control-plane functionalities.

Abstract: A method and a device for defining implementation of a look-up table for a network element of a software-defined network “SDN” is presented. The network element includes hardware for implementing the look-up table in two or more mutually alternative ways. The method includes selecting (301) an optimal, or at least a suitable, one of the mutually alternative ways to implement the look-up table on the basis of a) the hardware available for implementing the look-up table, b) information about one or more look-up keys, and c) at least one of the following: the maximum number of entries of the look-up table, an average time between successive look-ups, an average time between successive modifications of the look-up table. The method enables the network element to be configured in a way that the hardware resources of the network element are utilized in an optimal or at least a suitable way.

Abstract: A network element includes control entities (102, 103) and controllable entities (104-105) such as line interface units. The control entities configure the controllable entities to enable the network element to operate in a data transfer network. Each control entity maintains a configuration data-set in accordance with configuration information based on data received from the data transfer network and transmits the configuration data-set to one or more controllable entities. The control entity applies a configuration rule on the configuration information and updates the configuration data-set on the basis of the result of the configuration rule. The configuration rule is stateless so that its result is independent of earlier results of the configuration rule and of earlier configuration information. The control entities are enabled to keep their configuration data-sets nearly congruent with each other because they use the same configuration information and the same stateless configuration rule.

Abstract: A lookup system includes a ternary content-addressable memory TCAM configured to carry out a lookup on the basis of a lookup key so as to produce a preliminary lookup result. The lookup system further includes a subsystem configured to produce a final lookup result on the basis of the preliminary lookup result and auxiliary data including at least a part of bit values of do-not-care bit positions of the lookup key such that the preliminary lookup result is independent of the bit values of the do-not-care bit positions. The required number of TCAM-lines can be reduced because different alternatives of the auxiliary data corresponding to different possible bit values of the do-not-care bit positions of the lookup key are handled with the subsystem instead of the TCAM.

Abstract: A configuration system for configuring the control-plane of a software-defined network is presented. The configuration system determines an entity-specific group of control-plane functionalities for each of controller entities located in a distributed way in the software-defined network. The control-plane functionalities are suitable for configuring data forwarding network elements, such as routers, to support services to be provided by the software-defined network. The configuration system composes, on the basis of the entity-specific groups, entity-specific portions of control data and transmits the entity-specific portions of the control data to the controller entities so as to configure each controller entity to support the corresponding entity-specific group of the control-plane functionalities.

Abstract: A technique for managing transfer of data frames having a pre-determined mutual order is presented. Each data frame contains an order indicator, for example a sequence number, indicative of the position of the data frame in the pre-determined mutual order. A transmitting end provides each data frame with a satisfaction indicator that indicates whether acknowledgement about reception of earlier transmitted data frames is expected from a receiving end. When the satisfaction indicator of a data frame indicates that no acknowledgement is expected, the receiving end knows that the data frame can be treated (303) as a beginning of a received data flow when checking the integrity of the received data flow on the basis of the order indicators. Thus, there is no need to establish a logical connection between the transmitting and receiving ends prior to the transfer of the data frames.

Abstract: A method for configuring network elements of a software-defined network includes receiving (401) configuration data at one of the network elements, constructing (402) a configuration system of the network element on the basis of the configuration data, addressing (403) by the network element the configuration data or an appropriate portion of it to at least one other of the network elements, and transmitting (404) in accordance with the configuration system at least the appropriate portion of the configuration data to the other network element. The other network element, in turn, constructs its configuration system on the basis of the configuration data received by it and possibly transmits at least a portion of the configuration data to still one or more other network elements. Thus, the network elements represent a single entity from the viewpoint of a controller of the software-defined network. Hence, the scalability of the software-defined network is improved.

Abstract: A network element of a software-defined network includes a data transfer interface (211) for receiving and transmitting data and a processing system (212) for constructing, in accordance with configuration data received from a controller system of the software-defined network, a configuration system for managing data to be forwarded. The processing system is adapted to configure a first portion of the configuration system to associate, with data received at the network element, a configuring entity-specific identifier (CEID) that identifies a configuring entity, for example an application, related to the received data. The configuring entity-specific identifier constitutes at least a part of metadata associated with the received data. Appropriate other portions of the configuration system are configured to use this metadata when determining actions to be carried out in conjunction with the received data.

Abstract: A network element of a software-defined network includes a processing system for maintaining a control system for managing data. The control system is maintained in accordance with configuration data received from a controller of the software-defined network. In a case where the control system has no entry corresponding to frame information related to a received data frame, the processing system checks whether status information indicates availability of the frame information to the controller. If the status information does not indicate the availability, the processor system arranges the frame information to be available to the controller and updates the status information to indicate the availability. Thus, data traffic related to data-driven learning and occurring between the network element and the controller can be reduced with the aid of the above-mentioned status information.

Abstract: A network element (101-104) of a software-defined network is adapted to construct a configuration system on the basis of configuration data received from a controller (105, 106) of the software-defined network. The network element is arranged to transmit, in response to an event indicating a need to verify the configuration system, status information to the controller. The event may include for example a loss and a subsequent reestablishment of a connection between the network element and the controller. The status information indicates a first portion of the configuration data which has been implemented in the configuration system prior to the event. After receiving the status information, the controller is able transmit an appropriate second portion of the configuration data to the network element so as to enable verification of an appropriate portion of the configuration system. Thus, there is no need to verify the whole configuration system after the event.

Abstract: In a data transfer network, a network device (102) receives at its first data interface an identifier message transferred via a first data transfer path from another network device (106). The network device compares the identifier message to earlier received identifier messages. If the identifier message matches an earlier received identifier message received at another data interface and transferred via a second data transfer path from the other network device, the network device updates its forwarding database to express that the other network device is accessible also via the first data interface, and transmits address information related to the other network device via the first data interface so as to determine the first data transfer path to be an alternative for the second data transfer path. The first and second data transfer paths can be used for protecting each other and/or for load sharing.

Abstract: A network element of a software-defined network includes functional entities (321-325) capable of transferring data between each other and providing connections to other network elements. One or more of the functional entities are configurable with configuration data received from a controller of the software-defined network. The network element is adapted to transmit, to the controller, entity-specific capability descriptors related to the functional entities configurable with the configuration data. Each entity-specific capability descriptor expresses operations capable of being carried out by the respective functional entity. As the controller is made aware of the capabilities of the functional entities, the goal of the software-defined networking to provide controlled management of the data-forwarding functionality of a network element as a whole can be extended to the functional entities, e.g.

Abstract: A method for defining a look-up system having functionality suitable for a network element of a software-defined network “SDN” includes maintaining (401) data specifying a set of look-up tables so that the look-up system is capable of being constructed in alternative ways using one or more look-up tables of the set of look-up tables and each of the look-up tables has a look-up table specific resource consumption estimate indicating a load factor of at least one component of the network element when running operations related to the look-up table. The method includes selecting (402), on the basis of the look-up table specific resource consumption estimates, such one of the alternative ways of constructing the look-up system which provides a look-up system that is optimal, or at least suitable, for the network element.

Abstract: A method for configuring a network element of a software-defined network “SDN” includes retrieving (401), for different look-up tables, a look-up table specific resource consumption estimate indicating a load factor of at least one component of the network element when running operations related to the look-up table, determining (402), on the basis of the look-up table specific resource consumption estimates, a resource consumption estimate for at least one look-up system including one or more of the look-up tables and having functionality suitable for the network element, and sending (404) to the network element, if the network element is capable (403) of running the look-up system, configuration data defining the look-up system. The method provides a possibility to avoid configuring network elements to overload and, on the other hand, the resource consumption estimates related to different network elements can be used for optimizing the software-defined network as a whole.

Abstract: A network element of a software-defined network includes a data transfer interface (211) for receiving and transmitting data and a processing system (212) for constructing, in accordance with configuration data received from a controller system of the software-defined network, a configuration system for managing data to be forwarded. The processing system is adapted to configure a first portion of the configuration system to associate, with data received at the network element, a configuring entity-specific identifier (CEID) that identifies a configuring entity, for example an application, related to the received data. The configuring entity-specific identifier constitutes at least a part of metadata associated with the received data. Appropriate other portions of the configuration system are configured to use this metadata when determining actions to be carried out in conjunction with the received data.