WO2011092612A1

WO2011092612A1 - A method for operating a radio node in a sensor network

Info

Publication number: WO2011092612A1
Application number: PCT/IB2011/050244
Authority: WO
Inventors: Philip Andrew Rudland; Peter Stephen May
Original assignee: Koninklijke Philips Electronics N.V.
Priority date: 2010-01-29
Filing date: 2011-01-19
Publication date: 2011-08-04

Abstract

The present invention relates to a method for operating a network comprising a plurality of radio nodes, each node communicating with at least one of its neighboring nodes, the method comprising the step of linking a first node to a second node, causing the first node to enter in a linked operation mode where the first node transmits relevant information about the first node state during the linked operation mode, wherein the step of linking the first node comprises a link indication of whether the linked operation mode is temporary or permanent.

Description

A METHOD FOR OPERATING A RADIO NODE IN A SENSOR NETWORK

FIELD OF THE INVENTION

The present invention relates to a method for operating a network, such as a wireless sensor network including a plurality of sensor nodes, connected for example wirelessly one to another. In such networks, two nodes may be linked one to another in order to create a special communication mode.

This invention is, for example, relevant for wireless sensor networks like Zigbee networks, or IEEE802.15.4 mesh networks with applications like in Healthcare network for patient monitoring or Lighting solutions, like building automation or home automation.

BACKGROUND OF THE INVENTION

In a wireless sensor network, like in the network 100 shown on Figure 1, a plurality of nodes 101, 102, 103, 104 and 105 are communicating one with each other. The shape of the network can be a star-shaped network like shown on Figure 1 or a mesh network. In a star shaped network, the node 101 is in the centre of the network and communicates directly to its children nodes 102 and 103. In this example, the nodes 104 and 105 are end devices (no further nodes below these nodes) and are children nodes of node 102. This means that they communicate only with node 102. For node 105 to send a message to node 101, node 105 transmits its data to node 102, which relays the message to node 101. When node 101 replies to this message, it transmits its message to node 102 which relays it to node 105.

In such a network, some nodes may be "linked" or "bound" one to another to perform some specific actions or to transmit specific data. For example, the ZigBee Healthcare Profile describes an 802.15.4 and ZigBee mesh network based mechanism for carrying medical data according to the IEEE 1 1073 standard where such a link could become useful, as adopted by the Continua Alliance. Indeed, the profile describes various aspects of network and device operation, including configuration of bindings between nodes. Binding is a form of "linking", as explained above. More specifically, it is a ZigBee mechanism for configuring an embedded device to send relevant messages to a second device.

Neither the ZigBee stack nor the ZigBee Healthcare Profile (HCP) provide any specific support for different sorts of bindings. It is up to the HCP (or their handheld device) to remember to remove the binding before they move on with an unbind command. There could be big problems if they were to forget, such as the sensor continuously attempting to contact a target that is now a long way away, or maybe isn't even in the building any more, or is turned off. This will result in a lot of wasted traffic, and reduced battery lifetimes. However, a handheld device may not know beforehand that it will be moved to a remote location without any possibilities to remove its link with the linked device. SUMMARY OF THE INVENTION

It is an object of the invention to alleviate the above mentioned problems.

It is another object of the invention to enable the linked device to remove the link that is no longer effective so that it can get out of the "linked" operation mode.

To this end, in accordance with a first aspect of the invention, a method is proposed for operating a network comprising a plurality of radio nodes, each node communicating with at least one of its neighboring nodes, the method comprising the step of linking a first node to a second node, causing the first node to enter in a linked operation mode where the first node transmits relevant information about the first node state during the linked operation mode, wherein the step of linking the first node comprises an link indication of whether the linked operation mode is temporary or permanent.

In accordance with a definition of this invention, the sensor itself is provided with a means to distinguish between a binding that is intended to be permanent, and one that is intended to be temporary. It may also be provided with a mechanism to expire temporary bindings if they become unused or unusable, whilst maintaining their ability to support resilient permanent links where required.

There is a need for at least two sorts of linking in such networks. A first sort of linkings is effectively permanent linkings, being used for long term fixed connectivity, such as between a sensor and a fixed monitoring system. The second sort of linkings are effectively temporary linkings, in that they are set up on demand, for example when a Health Care Professional (HCP) needs to access the data, and then removed when they are done.

Moreover, in some embodiments of the invention, it is proposed to give the nodes the ability to remove the bindings between them automatically, by carrying out some unbinding tests. This permits to avoid the necessity to send an unbinding request when a station will move out of range of the network, and to avoid considerable waste of energy and battery life for nodes continuing to transmit a message the other linked station can no longer listen to the transmissions.

In accordance with a second aspect of the invention, it is proposed a radio node comprising means for communicating with at least one neighboring node, the node comprising means for linking to a second node, means for operating in a linked operation mode with the second node where the radio node is arranged for transmitting relevant information about its state during the linked operation mode, wherein the means for linking to the first node are arranged for storing a link indication on whether the linked operation mode is temporary or permanent.

These and other aspects of the invention will be apparent from and will be elucidated with reference to the embodiments described hereinafter.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will now be described in more detail, by way of example, with reference to the accompanying drawings, wherein:

Fig. 1 is a block diagram of a wireless sensor network in which the invention is implemented.

Fig. 2 is the representation of a Binding Request in accordance with the invention.

Fig 3 is a flow chart representing a method in accordance with a method used in one aspect of the invention.

DETAILED DESCRIPTION OF THE INVENTION

The present invention relates to the binding of nodes in a network. When two nodes are bound one to another, they enter in a specific operation mode in which the state of a first node may infer some actions on the second node. This implies a specific communication between the two nodes in order to indicate the state a first node to the second node.

In an exemplary embodiment of the invention, the network is a wireless network comprising radio nodes, like a wireless sensor network or a Zigbee network. For example, such a network can be a mesh network or a star shaped network like network 100 illustrated on Figure 1.

In such a network, as explained in the preamble of the description, a plurality of nodes 101, 102, 103, 104 and 105 are communicating one with each other. In this example, the nodes 104 and 105 are end devices (no further nodes below these nodes) and are children nodes of node 102. This means that nodes 104, 105 communicate only with node 102. For node 105 to send a message to node 101 , node 105 transmits its data to node 102, which relays the message to node 101. When node 101 replies to this message, it transmits its message to node 102 which relays it to node 105. This exemplary shape of network is used here for illustrative purpose only. Indeed, the invention can be implemented in different types of networks. In this example, the node 105 is linked to the node 101. This means for instance that if the state of node 105, this may cause an action at the node 101. For example, if node 105 comprises a switch, once the switch of this node 105 is actuated and e.g. switched to the ON position, a lamp connected to the node 101 turned on. In another application, the node 105 may be a wireless sensor on a patient which is linked to the node 101 which could be a handheld terminal for a doctor. Then, at regular intervals, the wireless sensor of node 105 may measure a parameter like blood pressure and the node 105 transmits data representative of the sensed blood pressure value to the node 101. It is to be noted that in this latter example, node 101 being a handheld device is mobile and is likely to move during operation in the network. As a consequence, the star shaped network may not be the optimal configuration and a mesh network may be preferred.

From these two examples, it is clear that a binding may last various durations of time (months, years or forever in the first example, and few minutes in the second example). In the second example, it is likely that the node 101 moves out of range of node 105. In case the two nodes are still operating in binding mode when this occurs, the node 105 is unaware that the node 101 is no more able to receive data. In order to avoid the node 105 wasting resources and energy by transmitting at regular intervals although node 101 cannot listen anymore, in accordance with the first embodiment of the invention, it is proposed to specify to the nodes whether the link between them is permanent or temporary. By doing so, the node 105 could stop the binding operation mode when required.

In the particular example of ZigBee, the linking explained above is a "binding". This biding is a standard mechanism defined by which an application on one device knows which device(s) to send its application messages to. Central to it is the "bind request" command, which is a standard message format by which a 3rd device can instruct a 1st device to send its application messages to a 2nd device. ZigBee also provides a mechanism by which two simple devices can arrange to be bound together - both send an "end device bind request" to a control point (the ZigBee Coordinator) which matches them up, determines which is the source, and sends that node a bind request specifying the other matched nodes as the target.

On receipt of a bind request an entry is added into the binding table of the recipient, instructing it to send application messages to the specified target. All messages which are "sent via the binding table" then go to all recipients in the binding table.

Unbind request is the reverse of the bind request. In a conventional network, this request is necessary to remove the binding of two nodes. Note that an application may choose to send to a specified target, ie: not via the binding table. This mechanism is generally used for replying to requests (eg: RX:get value request, TX: get value response to the sender of the request). It might also be used, for example, to build an application layer equivalent of the binding mechanism, in which the application itself stores the target addresses, rather than having them stored by the ZigBee layers.

ZigBee binding is actually based on endpoints, ie: sub-device IDs. For example if a dual switch housing has two switches on the front, one might be on endpoint 1 (EP1) and one on endpoint 2 (EP2). Binding from EP1 to lamp 1 and EP2 to lamp 2 would cause an On command to be sent from EP 1 to lamp 1 when the first switch is pressed, and an On command to be sent from EP2 to lamp 2 when the 2nd switch is pressed. There can be multiple bindings to and from endpoints, for example we could additionally add a binding from switch 2 to lamp 1, such that when switch 2 was pressed lamp 1 would be sent an On message as well. The targets have endpoints too, so you could have a single node control 2 lamps and have the different lamps on different endpoints.

Thus, in accordance with an example of the first embodiment, the node 101 when binding may send a Binding Command 200 as represented on Figure 2. This Binding Command may comprise a destination node field 201 indicative of the target node 105, a source field 202 indicative of the transmitting node 101 and a Command Field 203 indicative that this command is a Binding Command. In this example, each binding might have a flag or a field 204 associated with it indicating whether it is permanent or temporary. This flag may be one bit (e.g. 0 for permanent, 1 for temporary). Along with this flag, it is possible to have a field indicative an unbinding test. For example, if, as explained after, the unbinding test is a number of failed transmission attempts, this field may be indicative of the number of consecutive unsuccessful attempts to reach before unbinding. This may also be a time duration since the last successful transmission to the other node.

In order to save some data resource, the flag may be omitted and one specific value of the field may mean that the binding is permanent. For example, the value 0 may be reserved for the case where the binding is permanent. Then, the value indicated in the unbinding test field may be done by an index (e.g. 1 for 3 attempts, 2 for 5 attempts, 3 for 10 attempts). The value indicated in the unbinding test may be chosen in dependence on some characteristics of the binding. For example, if the communication channel characteristics are so bad that it is likely that a message is lost when transmitted, the number chosen for the unbinding test may be higher. Similarly, if the binding is critical (in case of patient monitoring), the value chosen for the unbinding test may be higher. On the contrary, if one of the linked node is highly mobile, it is likely that it will go away during operation and the value chosen for the unbinding test may be lower.

If the linking is temporary then, if a fixed threshold number of data messages (eg: 10) fail to be delivered in a row, then the binding should be considered to have failed, and the record of the binding should be deleted. Delivery failure might be detected through lack of an acknowledgement from the destination.

These 10 messages may be counted per target device, such that if any 10 messages to the same target fail in a row then all temporary bindings to that target are removed. The counter could be reset if any acknowledgement or other message is received from this target. There may be a combination of permanent and temporary bindings to a target device, and only the temporary bindings would be deleted. The system would keep trying to send messages to targets indicated by a permanent binding.

There are also other ways to detect communication failures, such as receiving an error message from the system indicating that the route to a specified target has failed at some point, or that a new route cannot be discovered.

Another simple approach would be to make a rule that following the failure of an existing route, route discovery is not undertaken for devices which only have temporary bindings. As such, when the underlying network finds that the target device has moved out of range, any temporary bindings would be removed. The device might still carry out its normal route discovery mechanism if there are one or more permanent bindings still in place.

Other metrics might be used to identify a failed link, for example a binding to a target device might be designated as having expired after 10 minutes. This expiry time could be reset whenever data is received. Expiry could additionally/alternatively be automatic after 10 minutes, and a manager device wanting to avoid this expiry [possibly in the case that there is little traffic but the link needs to remain] might have to issue a new binding (or renew the existing binding somehow) to avoid it being automatically removed. For example, after 9.5 minutes it could send an unbind message, quickly followed by a new bind message. Alternatively we could create an application rebind message, or simply send another bind message by itself.

A variant of this invention is to have a discrete (or continuous) scale of metadata associated with each binding (or with all bindings to or from a given device or endpoint, etc), and modify the use of these bindings according to statistics relating to that binding. Essentially this is an analogue equivalent of the invention described above, where there is scope for more fine grained detail than simply which of two categories each binding falls into, and more scope for varying the usage of that binding than just whether or not to delete it on failure of a certain condition. As a fairly simple but concrete example, each endpoint could be given a trouble threshold and an expunge threshold, and the sensor software could be arranged to drastically reduce its efforts to use a certain binding if the number of failures exceeds the trouble threshold, but to only delete it if it exceeds its specified expunge threshold.

Further it is possible to allocate a different hop count limit for each target device or binding entry. Then if the HCP's device is still in the building, but a long way away, it could still be made to expire. For example, a binding might be set to a maximum of only 2 hops, so that messages no longer get as far as the target device, and acknowledgements are not received. Probably this would expire quickly when the HCP leaves the room, even though they are still connected to the same ZigBee network.

A method in accordance with an embodiment of the invention is illustrated on Figure 3. This method might have the following steps:

S301 : Indicate which of multiple bindings are temporary and which are permanent. This may be done, as seen before, by the use of the Binding Request Command illustrated on Figure 2. Once the binding of the devices is done, the linked devices operate in the Binding Mode (Step S302).

- From step S303 to S306, in Binding mode, each node Detect when a binding has failed. This is done by detecting (S303) the unsuccessful attempts. If an attempt is successful, the counter is reset (S304). On the contrary, if an unsuccessful attempt of transmission is detected, the counter is incremented (S305) and then compared to the unbinding test value (S306).

- Once the counter reaches the unbinding test value, and if the failed binding is designated as a temporary binding, then the binding is deleted (S 308). If the binding is permanent, it is not deleted.

In a variant of the above embodiments, it is proposed a solution to indicate easily to all stations which of multiple bindings are temporary and which are permanent:

- A solution can be to have all bindings which are to destination endpoints numbered in the range 0x01-0x7F be designated as permanent bindings, and those to endpoints in the range 0x80-0xff be designated as temporary bindings. The node requesting the binding ensures that a suitable endpoint number is used as the destination of the binding for the behaviour required. Alternatively a new table could be created with one entry per binding, and a boolean variable indicating permanent/temporary status for that binding.

Alternatively a new table could be created with one entry per target device indicating permanent/temporary, and all bindings associated with that target device using the value from the table.

In variants of the above embodiments, it is proposed alternative solutions to detect when a binding has failed:

A solution can be a counter keeping track of the number of failed attempts to send data to a target (ie: attempts to send for which no acknowledgement was received)

- If an acknowledgement is received ok then the counter is reset to 0.

If a message is sent for which no acknowledgement is received then the counter is incremented.

If the counter exceeds a certain threshold (eg: 10) then the binding is considered to have failed.

Note that other algorithms might be more effective in certain cases, such as for more quickly expiring a node of the edge of reception, for example rather than setting the counter to 0 on receipt of an acknowledgement, we could instead decrement the counter (if it's nonzero!).

If the failed binding is a designated as temporary binding, then delete it. Otherwise do not delete it:

The binding table can be interrogated using the APSME-GET.request primitive, requesting to get the binding table entry.

A binding can be deleted by issuing the APSME-UNBIND.request primitive, indicating the parameters of the binding that you wish to remove.

- Bindings designated as permanent would not be deleted.

In case a node is linked to a plurality of other nodes at the same time, the links may be estimated independently. It means that if a failed temporary link is removed, the other temporary links will not be removed (unless they are also detected as failed).

It is to be noted that, although in the previous examples, all the links between the nodes are wireless, the invention could be implemented in a network where a part of the connections between the nodes are wired. Indeed, some nodes may be connected one to another in a wired way. Moreover, some devices may have a gateway role and be connected to another subnetwork through an Ethernet wired connection. It is to be noted that the method of the invention may implemented by software solutions loaded on the radio nodes. The radio nodes may thus comprise a processor coupled to a transceiver and a memory, like flash memory or hard drive or any other data storage means.

In the present specification and claims the word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. Further, the word "comprising" does not exclude the presence of other elements or steps than those listed.

The inclusion of reference signs in parentheses in the claims is intended to aid understanding and is not intended to be limiting.

From reading the present disclosure, other modifications will be apparent to persons skilled in the art. Such modifications may involve other features which are already known in the art of radio communication.

Claims

1. A method for operating a network comprising a plurality of radio nodes, each node communicating with at least one of its neighboring nodes, the method comprising the step of linking a first node to a second node, causing the first node to enter in a linked operation mode where the first node transmits relevant information about the first node state during the linked operation mode, wherein the step of linking the first node comprises a link indication of whether the linked operation mode is temporary or permanent.

The method of claim 1, wherein the linked operation mode comprises the first node transmitting a message indicative of the first node state each time the first node state changes.

The method of claim 1 or 2, wherein the linked operation mode comprises the first node transmitting a message indicative of the first node state at regular time intervals.

The method of any of the preceding claims, wherein the first node is a sensor node having a sensing device for measuring a parameter value and wherein the first node state comprises the parameter value.

The method of any of the preceding claims, wherein the link indication is indicative of a first period of time.

The method of claim 5, further comprising the first node removing the link with the second node if the time elapsed since the last successful communication with the second node is greater than the first period of time.

The method of claim 5 or 6, wherein the link is considered as permanent if the indicated first period of time is a predetermined value.

8. The method of any of the preceding claims, wherein the link indication is indicative of a maximal number of unsuccessful attempts.

9. The method of claim 8, further comprising the first node removing the link with the second node if the number of unsuccessful attempts reaches the maximal number of unsuccessful attempts.

10. The method of claim 8 or 9, wherein the first node considers an attempt of communication unsuccessful if it does not receive an acknowledgement message of the transmitted message.

11. The method of claims 8, 9 or 10, wherein the network comprises at least a third node, wherein the first node and the third node are operating in a temporary linked operation mode with the second node, and wherein the first node counts one unsuccessful attempt when the third node communication attempt to the second node is unsuccessful.

12. The method of any of the preceding claims, wherein the link indication is embedded in a sub-address used in the linked operation mode of the first node.

13. The method of claim 12, wherein the linked operation mode is temporary if the sub- address value is in a first range and permanent otherwise.

14. A radio node comprising means for communicating with at least one neighboring node, the node comprising means for linking to a second node, means for operating in a linked operation mode with the second node where the radio node is arranged for transmitting relevant information about its state during the linked operation mode, wherein the means for linking to the first node are arranged for storing a link indication on whether the linked operation mode is temporary or permanent.