Low power data transceiver chip and method of operating such a chip

Field of the invention

The present invention relates to a data transceiver comprising a transmit/receive unit for transmitting or receiving a data message (e.g. using half-duplex operation), a processing unit connected to the transmit/receive unit for processing data messages. In a further aspect, the present invention relates to a network controller for transmitting data messages enabling energy saving in data transceivers. In a further aspect, the present invention relates to a method for operating a transceiver chip in a data communication network.

The invention has particular application in wireless infrastructure employing wireless communication devices in which power consumption of the whole infrastructure or power consumption of a battery powered device is an issue. Zigbee, Bluetooth and IEEE802.11 ("WiFi") are examples of wireless protocols that could be used.

Prior art

Such a transceiver chip is known to be used in a low data rate wireless communication network, e.g. operating according to the Zigbee standard (using the

IEEE 802.15.4 Medium Access Control layer protocol). The network may be used as a low power wireless sensor or control network, and e.g. applied in home automation, lighting and security, heating, ventilation and air-conditioning systems or industrial automation. These kinds of applications require sensor and control nodes in the network which can be battery operated.

A low data rate wireless network is typically a distributed sense and control network consisting of a central node, a few relaying nodes and a large number of remote nodes. The network can be arranged in a mesh, tree, star or even peer-to-peer topologies. The remote nodes are usually powered by batteries, by energy recycling or energy scavenging power sources. The remote nodes may communicate directly with a central node or via other relaying or mesh networking nodes. These relaying nodes

may be constantly powered by mains, batteries or by energy recycling/scavenging power sources.

With Zigbee a beaconed wireless network can be configured. In beaconed wireless networks there is a fixed time interval at which the central node collects data and distributes data and commands to remote nodes using beacons. The remote nodes awake from sleep mode just before the beacon transmission from the central node and returns to sleep mode after reception of a beacon frame. A beacon interval may be from a few times per second up to several minutes.

An important functionality of the beacon is to (re-)synchronise the timing of the remote nodes to that of the central node. This synchronisation is required to avoid that the remote nodes loose communication with the central node. The time starting with the remote node wake up, up to the instant the remote node returns to sleep is called a receive window. If the remote nodes wake up too late, a beacon may be missed. If the remote nodes wake up too soon, a battery powered device wastes energy. The drift is caused by timing differences between the two units, mostly due to static and dynamic frequency errors in the reference oscillators. So the longer beacon interval the greater the allowance for drift. This means the receive window must be increased and unnecessary energy usage in this period.

Very often the remote end nodes are powered by a battery or may even use energy recycling or energy scavenging techniques to sustain operation. For battery- powered nodes in extensive wireless networks the cost of battery maintenance/changing is very significant and the extension of battery life is very important, also by sticking to close constraints the use of mass market (low cost) batteries can be enabled. For energy recycling and scavenging techniques the power available particularly over short period is often very limited and it is therefore thus very important that these nodes work as efficiently as possible and stay within close power budgets, particularly energy should not be used on inappropriate or unneeded transactions and that the energy may be preserved for when the energy is needed. Any initiative to reduce energy in the end or relaying nodes can be very rewarding.

Summary of the invention

The present invention seeks to provide a data transceiver and method of operating such a data transceiver, in which the power consumption is reduced which enables long time operation with a restricted power source, e.g. using only a battery as power source. According to the present invention, a data transceiver according to the preamble defined above is provided, in which the processing unit is arranged to interrupt in dependence of at least a received part of the data message the transmit/receive unit to receive a remainder part of the data message.

The invention is based on the following recognition. The energy consumption of a remote node of a low power RF network is dominated by 3 definable modes:

1. Standby power between transactions, when only the timing oscillator is running and device is in sleep mode.

2. Power used in data transmission, at own initiative or when requested by central node. 3. Power used in data reception from the central node during each beacon transmission and in processing these.

The energy consumption in mode 1 (standby mode) is easily defined and limited by the semiconductor process and components used to manufacture the system. The possibilities to reduce the standby power are well documented. The energy consumption in mode 2 (transmission mode) is defined by the radiated RF energy during transmission and the length of message needed by the protocol. Furthermore, transmission from each remote node does not occur on every beacon cycle; only on those cycles where request is made to the node that has to be answered or actions performed. The transmission power for a beacon network may be reduced in order to save energy by measuring the strength of the network integrity and adjusting/reducing transmission power to the most energy efficient level that the link signal quality is not compromised. However, the length of transmission cannot be reduced.

The energy consumption in mode 3 (receiving mode) is the singular most definable area of energy wastage. Receiver power consumption is not directly related to link signal quality and so cannot be adjusted. A receiver is typically designed to a meet a sensitivity specification.

However, in contrast to transmission mode which occurs occasionally, receiving mode is evoked on each beacon cycle and a typical receiving period is much longer than a transmission period per cycle. Therefore, the energy consumed in receive mode over a number of beacon cycles may exceed the energy consumed in transmit mode by more than an order of magnitude, in spite of the higher instantaneous power consumption in transmission mode than in receiving mode.

In a distributed sense and control network a remote node is addressed occasionally in a beacon frame. According to the invention the data transceiver processes the received part of data message immediately after reception and as soon as the processor unit detects that the current data message does not comprise any information necessarily to be received, the processor interrupts the transmit/receive unit to receive the remainder part of the data message. This shortens the period the data transceiver is in receiving mode and consequently reduces the dissipated power by the transmit/receive unit. In a further embodiment, the processing unit is arranged to power down the transmit/receive unit in dependence of at least a received part of the data message. Using this feature allows to minimize the dissipated power in the transmit/receive mode while the processing unit is still capable to perform some actions while the transmit/receive mode is in power down mode. The reception of a data message may be interrupted when certain conditions are met, and the transmit/receive unit may be powered down instantly.

A data message may comprise a header with at least one destination address (in a Zigbee network the device identification, or alternatively the network identification) and a message payload following the header in time. A data message in the form of a beacon frame may comprise a list of addresses. In a further embodiment of the invention, the data transceiver includes an individual address and the processing unit is arranged to interrupt reception upon deduction that the individual address is not in the list of addresses. Using the list of addresses enables to determine whether the message payload comprises further data for the data transceiver. If not, the power dissipation is reduced by interrupting further reception of the data message.

In the case the list of addresses is a sequentially ordered list (e.g. in numerical order), in a further embodiment of the invention the processing unit is arranged to interrupt the transmit/receive unit receiving the remainder part of the data message

upon detection that the individual address has been passed by. Usage of these features enables to reduce the receiving period further by interrupting reception before the whole address list has been received. This embodiment may be even improved when the message header comprises two ordered lists, a first positioned list with battery powered nodes, and a subsequent list with mains powered nodes or the like. In this case, it is possible to determine whether the receiving node its address is in the message header at an even earlier time, allowing to interrupt the transmit/receive unit even earlier.

In a further embodiment, the individual address is programmable and during installation it has been set to a value from a range at the beginning of the sequentially ordered list. Usage of this feature enables to prioritize the individual remote nodes. By giving a remote node an individual address which is always at the beginning of the sequentially ordered list, the receiving time could be reduced further.

The dissipation of power could be further reduced if the transmitter of a data message could compose messages which take into account whether a transceiver needs minimal power consumption. Therefore, in a further embodiment, the transceiver is arranged to transmit a parameter to enable a controller to take measures to reduce power consumption of the data transceiver.

In most data communication networks, data messages comprise an error correction part, e.g. in the form of a checksum, in order to allow detection and possible correction of transmission errors. In an alternative embodiment, wherein the data message comprises a checksum part, the processing unit is arranged to interrupt reception of a remaining part of the data message including the checksum.

By interrupting the reception of the checksum part, the receiving period is shortened. The handling of errors in the data transmission can sometimes be handled at a higher (application) level. The data transceiver could for example compare the received data with previously received data, for example a measured temperature. If the difference between the two temperatures is too high, the transceiver could decide to skip the latest received data and to stay using the previously received data. This embodiment is especially suitable for applications which transmit non-critical data. In an even further embodiment, the data transceiver is further arranged to determine the signal quality of data messages and to interrupt reception of a remaining part of the data message including the checksum in dependence of the signal quality. Signal quality

can be a combined metric with estimated (receive) signal level, estimated signal to noise ration, estimated signal to interference ratio, detection error signal and/or other signal quality indicator. If the signal quality of previously received data is above a predefined level, this is an indication that data could be received without errors. In that case, the checksum part superfluous end should not necessarily be received. In one embodiment where the data transceiver is arranged to determine the received signal level and/or signal to noise ratio of data messages and processing unit is arranged to obtain the signal quality in dependence of the received signal level and/or signal to noise ratio. In another embodiment, the processing unit is arranged to determine the signal quality by verifying the checksum in data messages.

A data message could include a network identification. A network identification indicates for which devices the message is intended. The network identification is normally transmitted in the header of a data message and thus enables to interrupt the receiving of a remainder part of a data message. Therefore, in an alternative embodiment, the processing unit is arranged to retrieve the network identification and to interrupt reception of the remaining part in dependence of the network identification.

A data message could include a destination address. A destination address indicates for which device the message is intended. The destination address is transmitted in the header of a data message and thus enables to interrupt the receiving of a remainder part of a data message. Therefore, in an alternative embodiment, the processing unit is arranged to retrieve the destination address and to interrupt reception of the remaining part in dependence of the destination address.

In a sense and control network some remote nodes have to send occasionally data. For example a temperature sensor has to send the temperature once every five minutes. As described above, the timing of a remote node needs to be synchronized to the central node, to ensure that the remote node wakes up just in time to receive a beacon frame. Therefore, in a further embodiment in which data messages are transmitted at regular time intervals by a transmitter and wherein in a mode the data transceiver is arranged to receive only every n ^th data message, wherein n>l, so as to ensure synchronization with the transmitter. Using this feature enables the data transceiver to be synchronized to the transmitter while the power usage is reduced.

In a second aspect, the present invention relates to a network controller for transmitting data messages to enable energy saving in data transceivers; the network controller comprises:

- a transmit/receive unit for transmitting or receiving data messages from the data transceivers;

- a processing unit connected to the transmit/receive unit for processing data messages; in which the processing unit is arranged to determine from a data message from a data transceiver whether the data transceiver has a restricted power supply, to retrieve from the data message an individual address of the data transceiver, generate data messages to be transmitted to the data transceiver, wherein the transmitted data messages include an ordered list of addresses and the order of the addresses enables the data transceivers to save energy by receiving the data message partially. A network controller with said features enables to optimize the power consumption of the remote nodes. In an embodiment, the processing unit is further arranged to generate a sequentially ordered list of addresses. Using these features enables to optimize the power consumption in data transceivers which are arranged to interrupt receiving a remainder part of the data message upon detection that the individual address has been passed by in the transmitted list of addresses. In another embodiment the processing unit of a network controller is further arranged to generate a list of addresses which comprises first the addresses of data transceivers with a restricted power supply.

In a third aspect, the present invention relates to a data transceiver comprising

- a transmit/receive unit for transmitting or receiving data messages;

- a processing unit connected to the transmit/receive unit for processing data messages; wherein the processing unit is arranged to adapt a wake up interval or timing of a wake up interval in dependence of received messages.

In an embodiment of the data transceiver the processing unit is arranged to determine the timing for a beacon wake up. In another embodiment the processing unit is arranged to determine the timing of the wake up interval in dependence of the actual arrival time of a data message and an estimated arrival time of said data message. In yet another embodiment the processing unit is arranged to determine the wake up interval in dependence of the length of the received message.

A network in which beacons are sent at regular intervals by a central node and wherein the data receiver is arranged to wake up right before the arrival of such

beacons. As mentioned above, beacons allow resynchronization of the timing by remote nodes with regard to the central node. As also mentioned above, the drift differences between the central node and remote node have to be covered in the receive window (time between remote node wake up and remote node returns to sleep). More precisely, the remote node will wake up a certain time before it expects the arrival of the beacon to cover drift. The present invention will apply adaptation of this certain time by means of timer margin tracking to shorten the wake up time and/or receive window while maintaining margin for the dynamic drift.

In a further aspect, the present invention relates to a method for operating a data transceiver in a data communication network, comprising:

- receiving data messages;

- interrupting in dependence of at least a received part of a data message reception of a remainder part of said data message.

In an embodiment the method further comprises the action power down a transmit/receive unit of the transceiver in dependence of at least said received part of the data message.

Short description of drawings

The present invention will be discussed in more detail below, using a number of exemplary embodiments, with reference to the attached drawings, in which

Fig. 1 shows a simplified block diagram of an embodiment of a data transceiver chip according to the present invention;

Fig. 2 shows a schematic representation of a data message as used by the transceiver chip of Fig. 1;

Detailed description of exemplary embodiments

In Fig. 1, a simplified block diagram of a wireless data transceiver 1 is shown of an embodiment of the present invention. The data transceiver 1 is a low power data device, of which the power supply is provided by an external battery or energy accumulator 2. Furthermore, the data transceiver 1 is connectable to an external antenna 3 for transmitting or receiving RF signals.

The data transceiver 1 comprises a transmit/receive unit 11 (Tx/Rx), a processing unit 10 connected to the transmit/receive unit 10 arranged to process base band signals

to be transmitted or arranged to process received base band signals. In this embodiment, the data transceiver 1 is arranged to operate according to the Zigbee protocol (IEEE 802.15.4) in order to be used in a Personal Area Network (PAN), which is an example of a low power wireless sensor and control network. Thus, the processing unit 10 is operational to provide the data frame (de-)composition in accordance with the Zigbee protocol. In a Zigbee PAN, the nodes operate in a half duplex manner, i.e. the transmit/receive unit 11 is either in transmitting mode or in receiving mode. Exchange of data between the processing unit 10 and transmit/receive unit 11 is obtained using a first data bus 18, connecting the processing unit 10; transmit/receive unit 11 and a first register unit 12. The first register unit 12 may be implemented in the form of memory cells in semiconductor technology. The size of this first register unit 12 can be adapted to account for the projected use of the first register unit 12, and may e.g. be 20-30 bytes in size.

The data transceiver 1 may further comprise an application unit 14 for the execution of an application program associated with the use of data transceiver 1 in the personal area network. The application unit 14 may e.g. comprise a standard type of microprocessor with associated memory comprised in the integrated circuit design of the data transceiver 1. Although not indicated in the simplified diagram of Fig. 1, this application unit 14 may be equipped with interface circuitry, e.g. for connecting with an external sensor (not shown) or an external device (not shown). Again, in this embodiment, data exchange between the application unit 14 and processing unit 10 is accomplished using a second data bus 17 and a second register unit 13, e.g. in the form of semiconductor technology memory cells.

The data transceiver 1 further comprises a power supply unit 15, which is arranged to provide operating power to the various (functional) components of the data transceiver 1, i.e. the transmit/receive unit 11, processing unit 10, application unit 14 (and the first and second register units 12, 13 if arranged as volatile memory). This may be accomplished using a power supply bus 16 as indicated in Fig. 1. When the data transceiver 1 is used as a typical Zigbee node, the power is usually provided by a battery 2. This battery 2 may be chosen to be able to support the operation of the total transceiver chip, i.e. transmit/receive unit 11 in either transmit or receive mode (due to the half duplex operation), the processing unit 10, and the optional application unit 14.

For long time operation on one battery charge, it would be desirable to be able to use batteries with a long life time, e.g. a lithium button cell.

Further features may be provided in other embodiments of the present invention in order to lower the instantaneous power consumption of the transceiver chip 1. One group of embodiments relates to minimizing the power consumption of the transmit/receive unit 11 , more in particular by minimizing the power consumption during the receive mode of the transmit/receive unit 11. In battery operated nodes using a transceiver chip 1 in a low power data transmission network, one of the largest waste areas in terms of power consumption is that nodes are listening for messages that are intended for other nodes. When there are many nodes in a network, or several networks operating in the same space, which may happen in domestic and commercial property applications, data traffic can be high. Further, neighbor networks at larger distances can be received at rather low receive levels. Several possibilities exist to reduce the duration of unnecessary reception of data. It should be noted that the scope of the invention is not limited to battery operated nodes. Power saving is anyway important. Even when all nodes in a network have a mains supply, the enormous number of nodes in a building with the features according to the invention will result in a significant reduction of energy costs. Therefore, the invention is suitable for nodes with any kind of power supply.

Typically in a low power beacon network, the receive period is much longer than transmit period and therefore total energy consumption in receive mode may be higher. A full beacon cycle will be based on the following basic steps:

1. The remote node wakes up in time to receive incoming beacon frame from a central node.

2. The remote node receives the beacon frame transmitted from the central node.

3. The remote node interprets the beacon.

4. If necessary, the remote node switches into transmit mode, if not then proceed to step 8. 5. The remote node transmits data to the central node

6. The remote node switches back to receive mode

7. The remote node awaits the acknowledgement from the central node. If no acknowledgement is received, repeat from step 4.

8. The remote node switches back into sleep mode and a beacon interval timer is started to enable to wake up in time to receive the next beacon.

In a traditional beacon network configuration each and every node will wake up on every beacon interval and listen to that beacon for relevant content then either react or wait in standby mode until the next cycle.

Significant energy can be preserved if the remote node may establish that any data message has no relevant content for a specific remote node. At the moment of determination that the data message has no relevant content the end node will shut down until the next beacon frame and the procedure will be repeated. The earlier in reception of a data message the determination is performed, the more energy may be preserved in said beacon cycle.

In this group of embodiments, the transceiver chip 1 is arranged to analyze the data received instantaneously. This can e.g. be implemented in the transmit/receive unit 11, or in the power supply unit 15 using e.g. logic circuitry or an embedded software program. In Fig. 2 a simplified diagram is shown of a data packet or message 20 as transmitted by a data source in the data network (e.g. a Full Function device (FFD) in a Zigbee network), and which may be intended for a specific node using a transceiver chip 1. The message 20 comprises a data part 23, which comprises the data payload of the message, e.g. instruction codes, commands, actual data, etc. This data part 23 is preceded by a destination address 22, e.g. a MAC-address of the node for which the data packet 20 is intended. A MAC-address is a unique individual address of a device in a network. Depending on the specific application, multiple destination addresses 22 may be present. Furthermore, the destination address 22 may be preceded by an identification of the network associated with the destination address (e.g. DstPANId, Destination Personal Area Network Identification in Zigbee protocol terms). At the end of the message 20, a checksum 24 (CRC Cyclic Redundancy Checksum) or another type of error detection and/or correcting code may be provided. The structure of the actual message 20 used in an actual network, e.g. the earlier mentioned Zigbee network, may be more complex, but in general, the destination address (es) 22 and network identification 21 are included in a header part of the message 20, and the other items are included in a part of the message 20 following the header part (the checksum 24 usually being the last part of the message 20). The

checksum 24 could be calculated over the header part 21, 22 and the data payload part 23 of the data message.

The data transceiver 1 may be arranged to interrupt receiving further data and consequently further processing of data which reduces the power consumption. In an advantageous embodiment the transmit/receive unit is power down and consequently will not consume any power. The interruption could be performed in reaction to a number of predetermined circumstances:

In a first embodiment the processing unit of a remote node is arranged to deduce that the individual address of the remote node is not in the list of destination addresses in a data message, e.g. a beacon frame. This feature enables to interrupt the reception of the data payload part and the CRC part. Upon deduction that the individual address is not in the list of addresses, the remote node could immediately return back to standby mode until the next beacon frame, without receiving the whole beacon frame. It should be noted that the list of addresses could comprise only one address. In a second embodiment is assumed that the list of addresses in de data message is a sequentially ordered list. The addresses could be alphabetical or numerical or a combination. The addresses could be in increasing or decreasing order. The processing unit is arranged to interrupt reception upon detection that the individual address has been passed by. In case of an increasing ordered list, as soon as the processing unit detects that an address is received which is higher than its own individual address and its own individual address was not in the list, the processing unit 10 interrupts the transmit/receive unit 11 to receive the remaining addresses, the data part 23 and the CRC part 24 of the data message.

In a third embodiment a network controller allocates network/MAC addresses to those specific nodes with restricted power supply, e.g. battery powered or energy scavenging nodes or the like. The allocated addresses are preferably in a range that corresponds to the beginning of a sequentially ordered list. To enable this a data transceiver is arranged to transmit a data message including a parameter indicating that the network controller take appropriate measures to reduce power consumption in the specific remote node. The parameter could be a flag or bit in a data message to be transmitted by the remote node. Furthermore, the remote node is arranged to program or reassign its individual address. The programming of the individual address of such a remote node is preferably performed during installation of set node in the network and

a network controller may then accommodate this as a part of the installation and take appropriate measures, such as the composition of the order of addresses in the address list.

Separate destination address lists or fields may be provided in the format of the message 20 for battery powered nodes and otherwise powered nodes, so that a receiving node can determine that its own address is not comprised in the list of battery powered node addresses. In an alternative embodiment, the destination addresses are also ordered numerically in each of the fields, such that the receiving node can determine that the list of destination addresses for battery powered nodes has ended by detecting a jump back in the destination addresses.

Comparable measures can be taken by detecting the network identification 21 in the message header. When this network identifier 21 differs from the network identification of which a particular node is part, the rest of the message 20 may be ignored, and the transmit/receive unit 11 in receive mode may be powered down for the remainder of the message 20 duration.

In another embodiment the processing unit 10 is arranged to interrupt reception immediately after reception of the data part 23 of a message. In this embodiment the CRC part 24 will not be received by the remote node. There are several situations wherein the reception of the CRC part could be omitted. Although in this case the checksum part 24 is not being received, the error detection and correction of transmitted data may be handled at a higher level, e.g. in the application being executed in the application unit 14. This could be done for non- critical data, such as temperature in a heating system. Non-critical data could be defined to be data which varies smoothly in time. An application unit could verify the integrity of the non-critical data by comparing the data with previously received data. If the difference between both data is within a predefined range the data is regarded to be valid, if not the data is not used and the remote node will wait for the next data message with data.

Alternatively, the data transceiver 1 may be arranged to detect the quality of reception of a message 20. The signal to noise and/or signal to interference ratio of the received signal could be a criterion for determining the signal quality or quality of reception. When the quality is above a predetermined level (e.g. related to the received signal strength), it is likely that the message 20 will be received without errors. In that

case, the reception of the message 20 can be limited to the data part 23 without the checksum part 24. The receive mode of the transmit/receive unit 11 may thus be powered down at an earlier stage, thus reducing power consumption. The signal quality could be determined over the receive part of a message or extended to the average signal quality over the last n data messages, where n is an integer greater than 1. The signal quality could also be determined by verifying the CRC in subsequent data messages. By evaluating the relative number of erroneous frames, the quality of reception could be measured. When the number of failures is falling within acceptable bounds, the remote node will allow interrupting the reception before the end of the data message by interrupting reception at the end of the data part 23 and consequently skipping the CRC-part 24 of a data message.

In the case where a part of a control system is temporarily not functioning or is not used for parts of the day or other periods, then the network controller may instruct specific sensors nodes to skip up and coming beacons cycles as the information is not needed for a relatively long period with respect to the period of a beacon cycle. The node will only need to receive that minimum beacon rate to ensure synchronisation is retained between the remote node and the central controller. In this case the data transceiver is arranged to receive only every n ^th data message, wherein n>l, so as to ensure synchronization with the transmitter. A significant reduction of power consumption can be realized with this feature.

This procedure may also happen when a system is running in a stabile condition and less frequent sampling is needed. Often in the start-up phase systems need more intense monitoring than in stable running. Consequently, the end node will wake up at a multiple of the beacon interval. Example is again a domestic heating system. It is often not used, so in these periods the room thermostats do not have to report temperature and have only to idle until the system is again active, the central controller has only to ensure that the synchronisation is not lost and so the beacon period at the remote node may be extended in effect to save energy and used only to retain network binding. From the embodiments given above, is clear that the power consumption can be further reduced in remote nodes when appropriate measures are taken in the network controller. A network controller has a similar composition as a remote node. A network controller is normally powered by a mains supply. A network controller

comprises a transmit/receive unit for transmitting or receiving data messages to/from data transceivers and a processing unit connected to the transmit/receive unit for processing data messages. To reduce power in remote nodes the processing unit is arranged to determine from a data message from a data transceiver whether the data transceiver has a restricted power supply. Furthermore the processing unit is arranged to retrieve from the data message an individual address of the data transceiver. To enable power reduction the network controller generates data messages to be transmitted to data transceivers, wherein the transmitted data messages include an ordered list of addresses and the order of the addresses enables the data transceivers to save energy by receiving the data message partially. This could be done by generating a sequentially ordered list. In another embodiment the list comprises two ordered lists. The first list comprises only addresses of limited power nodes and the second list the other nodes. By detecting a discontinuity in the order of the list, the remote node could determine whether the first list has been transmitted. The network controller could also be arranged to submit commands to program the individual address of a remote node. This enables to assign an individual address to a remote node which enables the remote node to reduce power by deducing very efficiently that its individual address is not in the list of addresses.

The data transceiver 1 of an end node will wake up to receive beacons transmitted by a central node. In start up phase the end node will wake up at every beacon interval or at fixed multiple of beacon intervals, while keeping time margin to cover the static and dynamic offset. Beacons allow resynchronization of the timing of remote nodes with regards to the timing of the central node. The drift differences between the timing of the central node and a remote node have to be covered in the receive window. The receive window is the time between the wake up instant and return to sleep instant of a remote node. It should be noted that the receive window could also be defined to be the period in time the transmit/receive unit 11 is active. The end node must wake up a certain time before the arrival of a data message to be able to receive the data message. When a transmit/receive unit 11 of data receiver 1 is receiving the preamble of a data packet 20, it will start up its internal synchronization and detect a start-of- frame delimiter to allow detection of the bits of data packet 20. (The preamble and start-of- frame delimiter preceding data packet 20 are not depicted in Figure 2.) Transmit/receive unit 11 forwards the actual time stamp value with regard to the

beacon start-of- frame instant to processing unit 10. In a previous wake up period the processing unit 10 had estimated an expected time stamp value with regard to arrival time of the next beacon start-of frame instant. After reception the actual time stamp value from the transmit/receive unit 11, the processing unit 10 determines the differences between the expected time stamp value and the actual time stamp value to adapt the timing for the wake up to receive the next beacon and to generate an expected time stamp value for the beacon in the next wake up period. In this way the average time between the instant of wake up of the data transceiver and the instant of the reception of a beacon is reduced. Consequently, the wake up period or receive window of the data transceiver will be shorter. The adaptation of the timing for the next beacon wake up uses a processing algorithm with slow feedback of difference between the actual and expected time stamp value and a bias value that gives sufficient margin to cover the dynamic offset.

In an other embodiment the processing unit is arranged to determine the wake up interval in dependence of an indicator in a received message indicating the length of the received message. In IEEE 802.15.4 the header of a message comprises a field containing the frame length information. The frame length information specifies the number of octets contained in the payload of the data message. In the event the reception of a data message is interrupted in dependence of at least a received part of the data message as described above, the processing unit knows the number of octets received and can calculate the remaining number of octets not yet received. The remaining number of octets defines the minimum period in which the receiver should not necessarily be able to receive data. In some cases, a data message has to be acknowledged by a receiver. In this case, the period to transmit to acknowledge could be added to this minimum period, as a new data message will not be transmitted before the acknowledge is received by the transmitter. By means of an indicator indicating the length of a data message and expected responses on said data message, a period in which the transmitter should not be able to receive data could be determined. When taking in to account this period, this will ensure that the receiver will not wake up too early or is not started up before the last octet of the corresponding data message or acknowledge is transmitted. In this way the power consumption can be further reduced.

The power consumption could be further improved, by taking into account the startup time of the respective units of a wireless data transceiver 1. For example, the startup time of a transmit/receive unit 11 is be 4 ms, whereas the startup time of an processing unit 10 and application unit is 2 ms. By means of the message length indicator, the period up to the next message can be determined. By taking into account the startup time of the respective units in the transceiver, the maximum sleep time for each of the unit to ensure that each of the respective units is started up in time, can be calculated. For example, if by means of the message length indicator is determined that the next data message will not be send within 5 ms, the transmit/receive unit 11 must be start-up after approximately 1 ms ( = 5 - 4 ms) and the processing unit 10 and application unit 14 after approximately 3 ms ( = 5 - 2 ms). In this example the transmit/receive unit 11 could be switched off for 1 ms and the processing unit 10 and application unit 14 could be switched off for 3 ms. If the estimated time to receive the next data message is for example 3 ms, the transmit/receive unit 11 should not be switched off whereas the processing unit 10 and application unit 14 could be switched off for 1 ms. Therefore, an unit could only be switched off when the expected time up to the transmission of the next data message is longer than the start-up time of the corresponding unit.

The power reduction could be further reduced by taking into the power needed to startup an unit and the power consumption when the unit is waiting to receive data. It could happen that the average power P _s tartu _P needed to startup an unit in a startup period Tstartup is larger than the average power consumption when said unit is waiting. In this case the minimum time needed to decide that the unit can be switched off depends on the equation T _mm = (P _s tartup ^χ T _st artup) / Pidie- Using this parameters enables to reduce the power consumption further, without reducing the receiving performance of the transceiver.

It should be noted that in some wireless communication standards a data message has a fixed length. In this case the time until the next data message can still be determined and the methods to minimize the power consumption described above can be applied. The possibility to switch each of the units off and on independently or in dependence of the availability of data supplied from one unit to another unit enables to reduce the power consumption further, which lengthen the battery lifetime.

Several embodiment of the present invention have been described above by way of exemplary embodiments. Various modifications and variations for the elements described with respect to these embodiments may be made by the skilled person without departing from the scope of the present invention, which is defined by the appended claims.