Polling of lighting devices

Abstract

A method for performing a periodic polling procedure collects information on a state of an individual node in a connected network. The method comprises establishing (S10) of a priority policy. The priority policy determines an average polling frequency for at least some of the nodes in the connected network. A request is sent (S20) by a first node to a second node for information on the state of the second node. The request is sent according to the priority policy. The priority policy is based on historical data and / or connection data. The historical data is of previous changes in the state of the respective second node due to commands from other parties than control commands of the first node. The connection data is information about how many control functionality units, different from the first node, that the respective second node is connected through. A connected network is also disclosed.

Description

[0001] 2024PF80455

[0002] 1

[0003] POLLING OF LIGHTING DEVICES

[0004] FIELD OF THE INVENTION

[0005] The present invention generally relates to lighting devices in a network. More specifically, the present invention is related to methods, devices and computer program code for periodic polling to collect information on a state of an individual node in a connected network.

[0006] BACKGROUND OF THE INVENTION

[0007] Connected lighting systems are used frequently today. Such systems are popular since they offer features like scheduling, energy monitoring, sensor-based lighting control and asset management. Such features are also often available remotely. Examples of wireless network protocols that are widely used in today are open standards like Zigbee, Thread, Bluetooth Low Energy (BLE), BLE mesh, Wi-Fi, Wi-Fi direct, and various proprietary network implementations built on top of e.g. the IEEE 802.15.4, IEEE 802.15.1 or IEEE 802.il standards.

[0008] In connected control networks, polling is a process for communication between multiple devices. In a typical configuration, the central controller (e.g., a server) is a node or bridge that sends out requests for information to each device connected to the network. Each device then typically responds to the request by returning the requested information. Polling is commonly used in wireless control networks, where the controller is responsible for gathering data from multiple sensors or actuators. By using polling, the controller can efficiently gather data from each node in the network, without having to constantly maintain a connection with each node. In a lighting context, polling is not only important for individual light states, but also for the scene state in a room, such as to detect the current scene in a room by the states of all lamps in a room.

[0009] In many cases the connected networks can be so large that not all nodes can be reached with a direct link from a central controller, and thus those remote nodes may need the help from one or multiple relaying nodes. Polling may be performed as triggered by an event or as a repeated polling according to some schedule. In bandwidth constrained networks, repeatedly requesting the status of all network nodes (polling) can result in a heavy network2024PF80455

[0010] 2

[0011] load that causes delay for other network traffic. Each poll comprises a request and a response, thus two messages. Each of the message may need to go via multiple hops to between the polling device and the polled device, causing even more network traffic (more air time).

[0012] As light states can change based on powering off / on, e.g., via wall switches, or via e.g. BLE or Matter control, i.e. from other sources than from the bridge itself, the bridge is not guaranteed to have updated information about the light states. Such updated information is relevant for various functions such as automations as well as being able to show the correct state of the lighting devices in a user control device (app). Therefore, the bridge may be configured for regularly polling all lighting devices, e.g. via a Zigbee network or the like. Typically, all lighting devices are polled in serial order with a typical speed of one lighting device per second. Network traffic due to periodic polling is not a big problem for smaller systems, but as the number of lights grows, the total cycle time to get the current state may be slow for larger networks. Further developments allow more connected lighting devices and the upcoming increased direct control of lighting devices, e.g. via BLE connection, may make the current light polling mechanisms either causing network congestion (if polling frequency is increased) and / or long delays for the user to see the correct state (if the polling frequency is maintained and the number of devices and thus the cycle time increases), which is experienced as confusing or annoying.

[0013] Hence, there is a need to improve periodic polling while reducing the risks for network congestion or long delays.

[0014] US20180270137A1 discloses systems and are described for adjusting polling frequency of a premises device, which comprises determining a user interaction with an interface associated with the premises device, and transmitting an instruction to increase a polling frequency of the premises device from a first frequency to a second frequency. The premises device can receive, via a network, one or more commands and a second instruction to decrease the polling frequency from the second frequency to the first frequency. The device can revert the polling frequency to the first frequency in response to the second instruction.

[0015] SUMMARY OF THE INVENTION

[0016] An object of the present technology is to reduce network traffic and to avoid long delays for achieving information of prevailing lighting device states, even for large networks.2024PF80455

[0017] 3

[0018] This and other objects are achieved by providing methods, devices and computer program code having the features in the independent claim. Preferred embodiments are defined in the dependent claims.

[0019] Hence, according to a first aspect of the present invention, there is provided a method for performing a periodic polling procedure to collect information on a state of an individual node in a connected network. The method comprises establishing of a priority policy. The priority policy determines an average polling frequency for at least some of the nodes in the connected network. A request is sent by a first node in the connected network to a second node in the connected network for information on the state of the second node. The first node is often referred to as a hub, central controller, bridge, or gateway. The request is sent according to the priority policy. The priority policy is based on historical data and / or connection data. The historical data is historical data of previous changes in the state of the respective second node due to commands from other parties than control commands of the first node. The connection data is information about how many control functionality units, different from the first node, that the respective second node is connected through.

[0020] Thus, the present invention is based on the idea of sorting out the nodes that statistically are more probable to present changes and prioritize these by increasing the polling frequency for such nodes. Likewise, nodes that statistically are less probable to present changes are instead given a lower priority by decreasing the polling frequency for such nodes.

[0021] The present invention is advantageous in that the network traffic is reduced. A further advantage is that long delays are avoided for a majority of nodes. The present invention is further advantageous in that the information related to the state of the second nodes achieved by the first node is acquired with less (average) latency and is thus more reliable. Information about lighting devices that are changing its state frequently is acquired at a higher rate (and thus latency from change on the node to awareness of such change in the first node is lower on average) than for other lighting devices.

[0022] The present invention is further advantageous in that the method is easily implementable into available network communication systems.

[0023] It will be appreciated that the priority policy is based on historical data of previous changes in the state of the respective second node. These previous changes are due to commands from other parties than control commands of the first node. The historical data comprises preferably one or more of three types of information. A first information is how often the light state changes for each individual second node. The light state changes are light2024PF80455

[0024] 4

[0025] state changes triggered by any other party than the first node. A second information is how large the light state changes are for each individual second node. The light state changes are light state changes triggered by any other party than the first node. A third information is if there is any frequently occurring historical scheduling of power switching on and off each individual second node. The switching may for instance be performed by a user and may be based on routines that occur regularly at the same time of the day or week.

[0026] These three pieces of information, individually or in combination, form a basic information platform on which considerable improvements in state reliability can be achieved.

[0027] According to an embodiment of the present invention, an average polling frequency of a second node is increased if a number of light state changes, triggered by any other party than the first node, during a predetermined period exceeds a threshold. Such other party may e.g. be a controller connecting to a lamp via another network. A lighting device that previously have had its state changed frequently has a higher probability to be changed again. In order to have a reliable state information for such a lighting device, frequent polling is to prefer. Furthermore, the number of light changes may also be very time dependent, e.g. evening, night and daytime may have different patterns, as well as for weekday or weekends. The average polling frequency may therefore also depend on a such a temporal information.

[0028] According to an embodiment of the present invention, an average polling frequency of a second node is increased if the average magnitude of light state changes, triggered by any other party than the first node (10), is larger than a threshold. If e.g. the intensity of a lighting source historically is changed by only small amounts, the overall scene state in a room of the lighting source may not be changed considerably. However, if large changes usually are imposed, the need to have information of the actual state increases.

[0029] Furthermore, the size of light changes may also be very time dependent, e.g. evening, night and daytime may have different patterns, as well as for weekday or weekends. The average polling frequency may therefore also depend on a such a temporal information.

[0030] According to an embodiment of the present invention, an average polling frequency of a second node is decreased during periods corresponding to frequently occurring historical periods of the second node being switched off as scheduled by any other party than the first node. In periods having a high probability of inactivity, polling may be decreased considerably. This may e.g. be lighting devices that regularly are switched off during nights. Alternatively, lighting devices may instead be switched on when a surrounding light level2024PF80455

[0031] 5

[0032] becomes too low. Repeated such periodic behavior may therefore be used to adapt a polling frequency.

[0033] According to an embodiment of the present invention, an average polling frequency of a second node is increased if the second node is connected to a control functionality unit different from the first node. The connection to another unit having control capabilities indicates that there is an increase probability that a change will occur. In a further embodiment, the average polling frequency of the second node is further increased if the second node is connected to more than one control functionality unit different from the first node. The more possibilities there are for controlling the second node, the higher is the likelihood that a change will appear. The average polling frequency can therefore be increased based on the number of connected control functionality units.

[0034] According to an embodiment of the present invention, the historical data and the number of connected control functionality units may be complemented by further information. An average polling frequency of a second node may be decreased if the second node is not momentarily reachable. The time to a next change of the state is in average lower during periods when the second node is not reachable.

[0035] According to an embodiment of the present invention, the method for performing a periodic polling procedure further comprises receiving of information on the state of the respective second node, and updating the priority policy based on the received information on the state of the respective second node. The behavior of the second nodes may vary in time, and it is therefore preferred to keep the priority policy well updated.

[0036] According to a second aspect of the present invention, there is provided a connected network comprising a plurality of nodes. The connected network is configured to perform a periodic polling procedure for collecting information on a state of an individual node in the connected network. The connected network thereby comprises a first node out of the plurality of nodes configured to establish a priority policy. The priority policy determines an average polling frequency for at least some of the nodes in the connected network. The first node is further configured to send requests to respective second nodes out of the plurality of node in the connected network, for information on the state of the respective second node. The requests are sent according to the priority policy. The priority policy is based on historical data of previous changes in the state of the respective second node due to commands from other parties than control commands of the first node and / or how many2024PF80455

[0037] 6

[0038] control functionality units, different from the first node, the second node is connected through.

[0039] According to a third aspect of the present invention, there is provided a computer program code which, when the computer program code is executed by processing means in or connected to a first node according to the second aspect, cause the processing means to instruct the first node to perform or at least initiate the method according to the first aspect.

[0040] The second and third aspects leads to the same advantages as for the first aspect.

[0041] Further objectives of, features of, and advantages with, the present invention will become apparent when studying the following detailed disclosure, the drawings and the appended claims. Those skilled in the art will realize that different features of the present invention can be combined to create embodiments other than those described in the following.

[0042] BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The aspects of the present invention will now be described in more detail, with reference to the appended drawings showing embodiment(s) of the invention.

[0044] In the drawings, like reference characters generally refer to the same parts throughout the different figures. Also, the drawings are not necessarily to scale, emphasis instead generally being placed upon illustrating the principles of the invention.

[0045] FIG. 1 demonstrates an embodiment of a connected network comprising a plurality of nodes configured to perform a polling procedure;

[0046] FIG. 2 shows schematically a flow chart of a method for performing a periodic polling procedure according to an exemplifying embodiment of the present invention; and FIG. 3 is a time diagram of a possible polling scheme.

[0047] DETAILED DESCRIPTION

[0048] Figure. 1 demonstrates a connected network 1 comprising a plurality of nodes 10, 20 configured to perform a periodic polling procedure. The periodic polling procedure is intended for collecting information on a state of an individual node 20 in the connected network 1. The connected network 1 can be a local network under the control of a local coordinator to serve a certain control purpose. The connected network 1 can also be connected to the cloud 2 or a backbone network 3, via a first node 10, also denoted as a hub, central2024PF80455

[0049] 7

[0050] controller, bridge, or gateway. The connected network 1 may also be connected to a user interface, where a user may expect to see the correct state of each light.

[0051] The connected network 1 may operate according to a wired communication protocol, such as KNX, Modular Digital Communication System (MODBUS), Building Automation and Control Network (B ACnet), or Digital Addressable Lighting Interface (DALI).

[0052] The connected network 1 may operate according to a wireless communication protocol, such as to implement large scale information distribution and collection in a wireless control system with a plurality of nodes 10, 20. The connected network 1 can be used for lighting control and / or building automation. It is preferred that the wireless communication protocol supports multi-hop routing, which e.g. can be Zigbee, Thread, Bluetooth Mesh, WiFi mesh, WirelessHART, SmartRF, CityTouch, IP500, Z-Wave, or any other mesh or treebased technology. Considering that the connected network 1 may comprise a large number of nodes 10, 20, transmission between nodes may rely on multi -hop routing.

[0053] As mentioned above, in bandwidth constrained connected networks 1, periodic polling repeatedly requesting the status of all network nodes 10, 20 can be a heavy network load that may cause delays in the status information gathering itself as well as delays for other network traffic.

[0054] To use a Zigbee network as an example. A Zigbee network has limited bandwidth, but it may comprise a large number of nodes. In order to show the actual state of all lamps in a lighting control App correctly, the first node 10 has to collect state information from all second nodes 20, typically by polling. The polling serves for gathering information about any changes of the states of the nodes, also for out-of-band light changes, such as e.g. by local buttons, BLE, powerup / down, Zigbee remote controls etc. Preferably, a correct set of information should be available within a few seconds.

[0055] Polling is not only important for individual light states, but also for a scene state in a room. A current scene state in a room may be detected by the collective information of states of e.g. all lighting equipment in the room.

[0056] Typically, each poll comprises a request and a response, thus two messages. Each message may go via multiple hops to the polled device, i.e. a second node 20, and back to the polling device, i.e. the first node 10.2024PF80455

[0057] 8

[0058] The first node 10 may be a coordinator, a central controller, a bridge, or a gateway of the connected network 1. The other nodes, i.e. the second nodes 20, to be polled by the first node 10 may be routers or end nodes, in a mesh network context.

[0059] In a lighting context, a node 10, 20 may be comprised in a lighting device, a luminaire, a sensor, an actuator, such as a switch to serve for the communication function of the lighting device, the luminaire, the sensor, or the switch. A node 10, 20 may also be comprised in a HVAC system, a smart refrigerator, a smart oven, other smart white goods, or a remote controller in a broader building / home automation context.

[0060] The connected network 1 can, as illustrated in Fig. 1, be a mesh network. In a mesh network, devices are connected to each other directly and indirectly to allow for multiple communication paths between nodes. This means that data can be sent from one node to another through multiple routes, creating a redundant and resilient network. Mesh networks are commonly used in wireless sensor networks, home automation, and industrial applications, where reliability and flexibility are critical. As shown in Fig. 1, the mesh network may comprise a first node 10, which serves as a coordinator or a central controller. The first node 10 is used to manage the communication between the other nodes, the second nodes 20, in the connected network 1, to distribute and / or collect information to the second nodes 20 in the connected network 1, or act as a central point of the mesh network for interfacing with another network, e.g. a backbone network 3. There are a plurality of router nodes that are responsible for forwarding data packets between nodes in the network to enable multi-hop routing. In addition to router nodes, there is another type of nodes called end nodes, which are connected to the network but does not participate in the routing of data packets. End nodes are devices that are primarily responsible for sensing, actuating, or controlling a specific function or application. Sometimes, end nodes may have limited processing power, memory, and communication capabilities compared to router nodes or coordinator nodes. They are designed to be simple and low-cost, enabling them to be deployed in large numbers throughout a network. End nodes in a mesh network usually communicate with other nodes in the network through router nodes or a parent node. They send data packets to the network and receive data packets from the network, but they do not participate in the routing of data packets.

[0061] Polling is typically governed by the first node 10. The first node 10 is responsible for initiating and controlling communication between the devices in the network. In order to cause a device to retrieve data or perform an, the first node 10 sends requests to2024PF80455

[0062] 9

[0063] one or more devices at the second nodes 20. The second nodes 20 in the connected network respond to the requests with the requested data or an acknowledgement of the action performed. Polling can be done in different ways, such as periodic polling or demand polling. In periodic polling, the coordinator polls the nodes on a regular basis.

[0064] The first node 10 comprises typically processing means 11. The processing means is configured to process computer program code 12. The computer program code 12 may be stored in a memory of the first node 10 or may be retrievable from an external source. When the computer program code 12 is executed by the processing means 11 of the first node 10, the processing means 11 of the first node 10 instructs the first node 10 to perform or at least initiate processes according to the present invention.

[0065] Alternatively, or in combination, the processing of at least a part of the computer program code 12 may be performed in an external processing means, e.g. in the cloud. The external processing means is then connected to the first node 10 to instruct the first node 10 to perform or at least initiate the processes according to the present invention.

[0066] As mentioned above, polling is not only important for individual light states, but also for the scene state in a room 40, such as to detect the current scene in a room 40 by the states of all lamps in a room.

[0067] Any change of state of a second node 20 that is the result of an instruction or other communication with the first node 10 is consequently known by the first node 10.

[0068] However, there are typically also other control functionality units 30 that may influence the state of the second nodes 20. Non-exclusive examples could be via wall switches, or via e.g. BLE control arrangements. State changes executed by such other control functionality units 30 are not necessarily known at once by the first node 10 and have to be searched for by the polling procedure. The main focus of the polling procedure thus concerns state changes initiated by such other control functionality units 30.

[0069] Figure 2 shows schematically a flow chart of an embodiment of a method for performing a periodic polling procedure to collect information on a state of an individual node in a connected network. In step S10, a priority policy is established. The priority policy determines an average polling frequency for at least some of the nodes in the connected network. In step S20, a request is sent according to the priority policy. The request is sent by a first node in the connected network to a second node in the connected network, and is intended for information on the state of the second node. The priority policy is based on historical data and / or on how many control functionality units, different from the first node,2024PF80455

[0070] 10

[0071] that the second node is connected through. The historical data is historical data of previous changes in the state of the respective second node due to commands from other parties than control commands of the first node.

[0072] The above use of the priority policy reduces the network traffic. At the same time, long delays for achieving information of prevailing lighting device states are avoided, even for large networks. By adapting the average polling frequency based on conditions that may be influenced by the probability of occurrence of changes, it is possible to reduce the total amount of polling traffic within the connected network. It also may lead to a quicker average update after that a certain state change has occurred. The minor drawback is that some state changes, postulated to have a low probability of occurrence, may be detected and updated with a longer detection time. However, since these occasions are relatively rare, the average time for updates is reduced.

[0073] The historical data may change with time. In a preferred embodiment, indicated by dotted lines in Figure 2, further steps are provided. In step S30, information on the state of the respective second node is received. In step S40, the priority policy is updated based on the received information on the state of the respective second node. In this way, the historical data may gradually change to follow changes in the pattern of use of the different second nodes.

[0074] Returning to Figure 1, a connected network 1 is illustrated. The connected network 1 comprises a plurality of nodes 10, 20. The connected network 1 is configured to perform a periodic polling procedure for collecting information on a state of an individual node in the connected network 1. The connected network 1 therefore comprises a first node 10, out of the plurality of nodes, configured to establish a priority policy. The priority policy determines an average polling frequency for nodes in the connected network. The first node 10 in the connected network 1 is further configured to send requests to second nodes 20 out of the plurality of node 10, 20. The first node is often referred to as a bridge. The request is sent according to the priority policy for information on the state of the respective second node 20. The priority policy is based on historical data of previous changes in the state of the respective second node due to other sources than control commands of the first node and / or on how many control functionality units 30, different from the first node 10, that the respective second node is connected through.

[0075] Figure 3 is a time diagram of a possible polling scheme in a connected network. The connected network comprises in this illustrative example a first node and at2024PF80455

[0076] 11

[0077] least four second nodes, denoted A, B, C and D. The second node A is a node connected to one control functionality unit besides any control functionality of the first node. This may e.g. be a wall switch. This wall switch is used with a medium frequency. Requests of the polling process, represented as bars in the diagram, are sent with an average time interval of AtA. The second node B is a node that in previous time periods has changed its state relatively frequently. It is likely that this use pattern is continued and therefore requests of the polling process are sent with an average time interval of Atn, where Atn is smaller than AtA. The second node C is a node that is state-changed relatively frequently and furthermore is connected to more than one control functionality unit besides the first unit. Requests of the polling process are sent even more often, with an average time interval of Ate. The second node D is a node that seldom changes its state and that is not connected to any outer control functionality unit. The average time interval Ato is therefore long. Note that since the available bandwidth for the polling process may change in time, the time intervals illustrated in Figure 3 are average times. The actual nominal time may therefore differ. However, the relation between the sizes of the different times intervals of the different second nodes is qualitatively the same. The time delay before the first node becomes aware of a state change, illustrated as stars in the figure, is, if the signaling time is neglected, represented by the distance between each star and the closest following bar. Most of these distances are short, which means a low average delay.

[0078] The average polling frequency and the criteria therefore is thus a central feature of the presently proposed polling process. Different criteria may be used for different kinds of networks. In a preferred embodiment, the priority policy is based at least on historical data of previous changes in the state of the respective second node due to commands from other parties than control commands of the first node. The historical data comprises at least one of three quantities. A first quantity is how often the light state changes for each individual second node. These light state changes are triggered by any other party than the first node. A second quantity is how large the light state changes are for each individual second node. Also here, these light state changes are triggered by any other party than the first node. A third quantity is if there is any frequently occurring historical scheduling of power switching on and off each individual second node.

[0079] In one preferred embodiment, an average polling frequency of a second node is increased if a number of light state changes, triggered by any other party than the first node, exceeds a threshold. A second node that historically has changed its state frequently, above2024PF80455

[0080] 12

[0081] the threshold, has a high probability of experiencing a new change of state relatively soon, and by reducing the average polling frequency such probably state change may be found within very short delays.

[0082] In one preferred embodiment, an average polling frequency of a second node is increased if the magnitude of light state changes, triggered by any other party than the first node, are larger than a threshold. Since a large change of state, above a definable threshold, may considerably influence e.g. a room lighting, it is of interest to find such large changes quick. This is made by shortening the average time for second nodes that usually provide large state changes.

[0083] In one preferred embodiment, an average polling frequency of a second node is decreased during periods corresponding to a frequently occurring historical periods of the second node being switched off. This may for instance be a second node that is only active daytime on working days. The probability of a change occurring at nights or during weekends is then very low.

[0084] A second node that is connected to one external control functionality unit, i.e. different from the first node, is probably a second node that is intended to change its state in different ways and probably also relatively frequent. Therefore, in one embodiment, an average polling frequency of a second node is increased if the second node is connected to a control functionality unit different from the first node. If the second is further connected to more external control functionality unit, this typically indicates that it is requested to be able to change its state in different ways, which typically is the case for second nodes being changed often. In one embodiment, the average polling frequency of the second node is further increased if the second node is connected to more than one control functionality unit different from the first node.

[0085] If a second node is not reachable for the moment, e.g. if the second node has been turned off or the communication therewith is closed down or broken, it is difficult to achieve any state information therefore. Efforts can be made to re-establish a contact with that second node, but this may involve different start-up procedures, which may be very time consuming. Furthermore, if the second node is completely turned off, there can be no state change before it is turned on again. The probability that a state change is to occur for such a non-reachable second node is very low. Therefore, in one embodiment, an average polling frequency of a second node is decreased if the second node is not momentarily reachable.2024PF80455

[0086] 13

[0087] The person skilled in the art realizes that the present invention by no means is limited to the preferred embodiments described above. On the contrary, many modifications and variations are possible within the scope of the appended claims. For example, additional criteria for selecting average polling frequencies may be used, cooperating with the above described ones.

Claims

2024PF8045514CLAIMS1. A method for performing a periodic polling procedure to collect information on a state of an individual node in a connected network (1); the method comprising:- establishing (S10) a priority policy determining an average polling frequency for at least some of the nodes in the connected network;- sending a request (S20) according to the priority policy, by a first node (10) in the connected network (1) to a second node (20) in the connected network (1), for information on the state of the second node (20);wherein the priority policy is based on historical data of previous changes in the state of the respective second node (20) due to commands from other parties than control commands of the first node (10); wherein the historical data comprises at least one of:how often a light state changes, triggered by any other party than the first node (10), for each individual second node (20);how large are the light state changes, triggered by any other party than the first node (10), for each individual second node (20); andif there is any frequently occurring historical scheduling of power switching on and off each individual second node (20)2. The method according to claim 1, wherein an average polling frequency of a second node (20) is increased if a number of light state changes, triggered by any other party than the first node (10), during a predetermined period exceeds a threshold.

3. The method according to claim 1 or 2, wherein an average polling frequency of a second node (20) is increased if the average magnitude of light state changes, triggered by any other party than the first node (10), is larger than a threshold.

4. The method according to any of the claims 1 to 3, wherein an average polling frequency of a second node (20) is decreased during periods corresponding to frequently occurring historical periods of the second node (20) being switched off by any other party than the first node (10).2024PF80455155. The method according to any of claims 1 to 4, wherein the priority policy is further based on how many control functionality units (30), different from the first node (10), the second node is connected to.

6. The method according to claim 5, wherein an average polling frequency of a second node (20) is increased if the second node (20) is connected to a control functionality unit (30) different from the first node (10).

7. The method according to claim 6, wherein the average polling frequency of the second node (20) is further increased if the second node (20) is connected to more than one control functionality unit (30) different from the first node (10).

8. The method according to any of the claims 1 to 7, wherein an average polling frequency of a second node (20) is decreased if the second node (20) is not momentarily reachable.

9. The method according to any of the claims 1 to 8, further comprising the steps of:- receiving (S30) information on the state of the respective second node (20), and- updating (S40) the priority policy based on the received information on the state of the respective second node (20).

10. A connected network (1) comprising a plurality of nodes (10, 20) configured to perform a periodic polling procedure for collecting information on a state of an individual node in the connected network (1); the connected network (1) comprising:- a first node (10) out of the plurality of nodes configured to establish a priority policy determining an average polling frequency for at least some of the nodes in the connected network;wherein the first node (10) is further configured to send requests, according to the priority policy to respective second nodes (20) out of the plurality of node (10, 20) in the connected network (1), for information on the state of the respective second node (20);2024PF8045516wherein the priority policy is based on historical data of previous changes in the state of the respective second node (20) due to commands from other parties than control commands of the first node (10); wherein the historical data comprises at least one of:how often a light state changes, triggered by any other party than the first node (10), for each individual second node (20);how large are the light state changes, triggered by any other party than the first node (10), for each individual second node (20); andif there is any frequently occurring historical scheduling of power switching on and off each individual second node (20).

11. A computer program code which, when the computer program code is executed by processing means (11) in or connected to a first node (10) according to claim 10, cause the processing means (11) to instruct the first node (10) to perform or at least initiate the method (SI) of claims 1-9.

Images