WO2017030473A1

WO2017030473A1 - Methods and apparatus in a partially meshed network that includes multiple mesh nodes

Info

Publication number: WO2017030473A1
Application number: PCT/SE2015/050872
Authority: WO
Inventors: Jingcheng Zhang; Roma CHIRIKOV; Thomas Rimhagen; Oscar HINRICHSEN
Original assignee: Telefonaktiebolaget Lm Ericsson (Publ)
Priority date: 2015-08-14
Filing date: 2015-08-14
Publication date: 2017-02-23

Abstract

A partially meshed network (10) includes multiple mesh nodes configured to communicate over a set (11) of one or more contention-based channels. A first mesh node (12) receives a time beacon from a second mesh node (14) over one of the channels in the set (11). The beacon indicates a periodic transmission time window within which the second mesh node (14) is configured to transmit over one or more of the channels in the set (11). The first node (12) adjusts the periodic transmission time window within which the first node (12) is configured to transmit over one or more of the channels in the set (11), based on the received beacon, to at least partially overlap the second node's periodic transmission time window. During at least a portion of time outside the first node's periodic transmission time window, the first node (12) operates in a sleep mode in which it turns off one or more receivers at the first node (12).

Description

METHODS AND APPARATUS IN A PARTIALLY MESHED NETWORK THAT INCLUDES

MULTIPLE MESH NODES

TECHNICAL FIELD

The present application generally relates to a mesh network, and particularly relates to communication amongst multiple mesh nodes over a set of one or more contention-based, connectionless channels.

BACKGROUND

A partially meshed network is a communications network made up of multiple mesh nodes organized in a partial mesh topology. The mesh nodes cooperate to propagate transmissions throughout the network by relaying those transmissions from node to node (i.e., hop to hop). Nodes may join and leave the network freely, in an ad-hoc manner.

The mesh nodes of a partially meshed network can communicate using Bluetooth. However, the connection-oriented nature of Bluetooth causes the mesh nodes to consume more power than is acceptable in certain contexts.

Use of Bluetooth Low Energy (BLE) by the mesh nodes reduces the nodes' power consumption as compared to Bluetooth. BLE specifies so-called data channels and

advertisement channels. The advertisement channels are contention-based, whereas the data channels are not. Nodes transmit on an advertisement channel without scheduling the exact times at which those transmissions are performed, meaning that collisions between

simultaneously transmitted messages may occur. A form of collision may also occur in the sense that a node may not be able to receive a transmission at the same time as it is performing a transmission, i.e., if the node is half-duplex.

In traditional BLE approaches, a node sends a connection request to another node over an advertisement channel as part of a predefined connection establishment procedure. Based on this request, the two nodes establish a connection over a data channel and thereafter exchange transmissions over the connection-oriented data channel. BLE mesh networks depart from this traditional BLE approach. Rather than relaying transmissions over the connection- oriented data channels, the nodes relay transmissions over the contention-based,

connectionless advertisement channels.

Despite these improvements, mesh nodes in a mesh network still consume more power than is acceptable in certain contexts.

SUMMARY

According to embodiments herein, mesh nodes in a partially meshed network each transmit over one or more of the contention-based channels in a set (e.g., BLE advertisement channels). Each node does so (e.g., at a random time) within a periodic transmission time window whose length may be for instance autonomously established by that node. Each node broadcasts a time beacon indicating its periodic transmission time window. A node that receives a time beacon from another node advantageously adjusts its periodic transmission time window (e.g., by shifting that window in time) to at least partially overlap that other node's periodic transmission time window. With the two nodes' transmission time windows overlapping in this way, the nodes are configured to operate in a sleep mode during at least a portion of time outside their respective transmission time windows, e.g., based on the assumption that they need not monitor for transmissions from each other outside the overlapping transmission time windows. This proves advantageous, for example, in reducing mesh node power consumption in low throughput BLE mesh networks where collisions between transmissions within overlapping transmission time windows are unlikely.

Some embodiments, for instance, comprise a method in a partially meshed network that includes multiple mesh nodes, including a first mesh node and a second mesh node. The mesh nodes are configured to communicate over a set of one or more contention-based channels. The method comprises, at the first mesh node, receiving a time beacon from the second mesh node in the mesh network over one of the contention-based channels in the set. The time beacon indicates a periodic transmission time window within which the second mesh node is configured to transmit over one or more of the channels in the set. The method comprises adjusting the periodic transmission time window within which the first mesh node is configured to transmit over one or more of the channels in the set, based on the received time beacon, to at least partially overlap the second mesh node's periodic transmission time window. During at least a portion of time outside the first mesh node's periodic transmission time window, the method comprises operating the first mesh node in a sleep mode in which the first mesh node turns off one or more receivers at the first mesh node.

In one or more embodiments, the method further comprises autonomously establishing, by the first mesh node, a length of the first mesh node's periodic transmission time window. The adjusting in the this case comprises adjusting a periodicity of the first mesh node's periodic transmission time window and shifting the first mesh node's periodic transmission time window in time based on the received time beacon.

In one or more embodiments, adjusting comprises calculating, based on the periodic transmission time window indicated by the time beacon, a periodic reception time window within which the first mesh node expects to receive from the second mesh node over one or more of the channels in the set, and adjusting the first mesh node's periodic transmission time window to at least partially overlap the periodic reception time window calculated for the second mesh node.

In one or more embodiments, calculating comprises widening the periodic reception time window based on an accuracy of a sleep timer at the first and/or second mesh node. A sleep timer in this regard is a timer that governs when a mesh node transitions in and out of the sleep mode.

In one or more embodiments, calculating comprises calculating the length of said periodic reception time window as:

Rx _ Win = Tx_ Win + 2 *(SCA_l + SCA₂ ) * CYCLE _ TIME * NUMBER _ OF _ CYCLE

Rx _Win is the length of the periodic reception time window. Tx_ Win is a length of a transmission time window of the second mesh node as indicated in the time beacon received from the second mesh node. SCA_l is an accuracy of a sleep timer at the first mesh node, wherein a sleep timer is a timer based on which a mesh node transitions in and out of said sleep mode. SCA₂ is an accuracy of a sleep timer at the second mesh node. CYCLE _ TIME is a length of a periodic sleep-awake cycle at the second mesh node, as indicated in the time beacon received from the second mesh node. A periodic sleep-awake cycle is a periodically repeating pattern of transitions by the second mesh node between the sleep mode and an awake mode in which the first mesh node turns on said one or more receivers.

NUMBER _ OF _ CYCLE is a number of sleep-awake cycles that have passed since receiving the time beacon from the second mesh node.

In one or more embodiments, calculating a periodic reception time window comprises calculating a time of an edge or midpoint of the periodic reception time window. Adjusting the first mesh node's periodic transmission time window comprises aligning an edge or midpoint of the first mesh node's periodic transmission time window with the calculated edge or midpoint of the periodic reception time window.

Calculating the periodic transmission time window may comprises calculating the start of the periodic reception time window as:

n . ._r ^τ, ,τ, Rx win - Tx win

Rx Win start = DUE TIME - tt = =

^" 2

In this case, Rx _Win _ start s the start of the periodic reception time window. DUE _ TIME is the start of the second mesh node's next transmission time window relative to the time at which the second mesh node broadcasts the received time beacon, tt is an estimate of a

transmission propagation time between the first and second mesh nodes. Rx_Win is a length of the periodic reception time window. Tx _ Win is a length of the second mesh node's periodic transmission time window.

One or more embodiments comprise receiving different time beacons from different mesh nodes in the mesh network, and adjusting comprises independently calculating different periodic reception time windows within which the first mesh node respectively expects to receive from the different mesh nodes over one or more of the channels in the set, and adjusting the first mesh node's periodic transmission time window based on the calculated reception time windows.

In one or more embodiments, the method comprises receiving different time beacons from different mesh nodes in the mesh network. Adjusting comprises calculating a periodic global reception-transmission time window within which the first mesh node expects to receive from the different mesh nodes over one or more of the channels in the set, and within which the first mesh node is configured to transmit over one or more of the channels in the set, and aligning a midpoint of the first mesh node's periodic transmission time window with a midpoint of the periodic global reception-transmission time window.

In one or more embodiments, the method further comprises at the first mesh node, generating a time beacon indicating the first mesh node's periodic transmission time window, as adjusted and broadcasting the generated time beacon in the mesh network over one or more of the channels in the set.

One or more embodiments the first mesh node is a relay node and the method comprises refraining from relaying the received time beacon to any other mesh node in the mesh network, even though the first mesh node is configured to relay at least one other type of communication.

In one or more embodiments, adjusting comprises adjusting the first mesh node's periodic transmission time window based on the periodic transmission time window of mesh nodes within only a defined number of hops from the first mesh node (e.g., a single hop).

In one or more embodiments, the method further comprises, when attempting to join the mesh network, listening by the first mesh node for a time beacon from any mesh node in the mesh network. If no time beacon is detected by listening, the method comprises transmitting by the first mesh node over one or more of the channels in the set within a default periodic transmission time window and adjusting comprises adjusting the first mesh node's periodic transmission time window from the default periodic transmission time window, responsive to detecting the time beacon from the second mesh node during listening.

Embodiments herein also include a method comprising, at the second mesh node, generating a time beacon indicating a periodic transmission time window within which the second mesh node is configured to transmit over one or more of the one or more contention- based channels in the set and broadcasting the time beacon in the mesh network over one or more of the channels in the set. The method comprises during at least a portion of time outside the second mesh node's periodic transmission time window, operating the second mesh node in a sleep mode in which the second mesh node turns off one or more receivers at the second mesh node.

In one or more embodiments, the method further comprises prior to generating the time beacon, receiving a time beacon from the first mesh node in the mesh network over one of the contention-based channels in the set. The time beacon indicates a periodic transmission time window within which the first mesh node is configured to transmit over one or more of the channels in the set. The adjusting comprises adjusting the periodic transmission time window within which the second mesh node is configured to transmit over one or more of the channels in the set, based on the received time beacon, to at least partially overlap the first mesh node's periodic transmission time window. The generating comprises generating the time beacon to indicate the second mesh node's periodic transmission time window, as adjusted.

In one or more embodiments, generating the time beacon to indicate the second mesh node's periodic transmission time window comprises indicating the start of the second mesh node's next transmission time window relative to the time at which the second mesh node broadcasts the time beacon.

In one or more embodiments, broadcasting the time beacon comprises broadcasting the time beacon at a random time within the periodic transmission time window.

In any of these methods, the set of one or more contention-based channels comprises one or more advertising channels configured for advertising the availability of a mesh node to communicate with another mesh node in the mesh network over a connection-oriented channel.

In any of these methods, the time beacon transmitted by a particular mesh node indicates one or more of the following: the start of that mesh node's next transmission time window relative to the time at which the mesh node broadcasts the time beacon; an accuracy of a sleep timer at that mesh node, where a sleep timer is a timer based on which the mesh node transitions in and out of said sleep mode; a length of the mesh node's periodic transmission time window; the next transmission time window in which the next time beacon from that mesh node can be expected; and a length of a periodic sleep-awake cycle at the mesh node. A periodic sleep-awake cycle is a periodically repeating pattern of transitions by the mesh node between the sleep mode and an awake mode in which the mesh node turns on said one or more receivers.

In any of these methods, a transmission time window may be a window of time within which a mesh node performs a transmission at a random time within that window.

In any of these methods, the mesh network may be a Bluetooth Low Energy (BLE) mesh network and the one or more channels comprise one or more advertisement channels.

Embodiments herein also include corresponding apparatus, computer programs, and carriers containing such computer programs.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a block diagram of a partially meshed network according to one or more embodiments.

Figure 2A includes logic flow diagrams of methods respectively implemented by mesh nodes in a partially meshed network according to one or more embodiments. Figure 2B is a block diagram of a transmission time window according to one or more embodiments.

Figure 2C is a block diagram of at least partially overlapping transmission time windows according to one or more embodiments.

Figure 3 is a block diagram of adjustments to a mesh node's periodic transmission time window according to one or more embodiments.

Figure 4A depicts an example sleep-awake cycle of a mesh node according to one or more embodiments.

Figure 4B is a block diagram of information indicated by a time beacon according to one or more embodiments.

Figure 5 is a block diagram of calculations for widening a reception time window according to one or more embodiments.

Figure 6 is a block diagram of mesh nodes isolating their respective transmission time windows to a single hop according to one or more embodiments.

Figure 7 is a block diagram of adjustments to a mesh node's periodic transmission time window according to one or more other embodiments.

Figure 8 is a block diagram of adjustments to a mesh node's periodic transmission time window according to one or more still other embodiments.

Figure 9 is a block diagram of a mesh node according to one or more embodiments. Figure 10 is a block diagram of a mesh node according to one or more other

embodiments.

DETAILED DESCRIPTION

Figure 1 illustrates a partially meshed network 10, such as a Bluetooth Low Energy (BLE) mesh network. The network 10 includes multiple mesh nodes (MN), including mesh node 12 and mesh node 14, organized in a partial mesh topology. With the nodes only partially meshed in this way, at least two nodes in the network 10 communicate only indirectly via one or more intermediate nodes, such that communication occurs over multiple hops. Each

intermediate node in this regard operates as a relay node that relays transmissions from one node to another.

Each mesh node is configured to communicate with one or more other mesh nodes over a set 1 1 of one or more contention-based channels. As shown, for example, the set 1 1 of one or more contention-based channels between mesh node 12 and mesh node 14 includes multiple contention-based channels 13, 15 interspersed in frequency amongst non-contention- based channels 17. In embodiments where the network 10 is a BLE mesh network, for instance, channels 13, 15 may be advertising channels whereas channels 17 may be data channels. In this case, the BLE advertising channels do not even include a carrier sense media access mechanism to avoid collisions. Regardless, the contention-based nature of the channel(s) in the set 1 1 means that mesh nodes transmit over the channel(s) without scheduling the exact times at which those transmissions are performed. Transmissions over the channel(s) are therefore susceptible to collision. For example, if multiple mesh nodes simultaneously perform separate transmissions to a target node over the same contention-based channel, such jeopardizes the target node's ability to recover those separate transmissions.

The contention-based nature of the channel(s) in the set 1 1 also threatens the ability of the mesh nodes to conserve power. Without any scheduling as to exactly when a mesh node performs a transmission, a mesh node intended to receive that transmission (e.g., for relaying) would otherwise have to continually keep its receiver(s) on so as not to miss the transmission. This would consume a significant amount of power at the mesh node, e.g., approximately 18mA for a BLE mesh node. A form of collision may also occur if a mesh node is not able to receive a transmission at the same time as it is performing a transmission, i.e., if the node is half-duplex.

Figure 2A in this regard shows methods respectively implemented by mesh node 12 (as an example of a first mesh node) and mesh node 14 (as an example of a second mesh node) in the partially meshed network 10.

According to the respective methods shown in Figure 2A, mesh node 14 generates a time beacon (Block 200). The time beacon indicates a periodic transmission time window within which mesh node 14 is configured to transmit over the one or more contention-based channels in the set 1 1 . In at least some embodiments, as shown in Figure 2B, a transmission time window in this regard is a time window W that repeats periodically and that spans multiple possible time instances Tl - TN at which mesh node 14 may (i.e., is permitted to) transmit over the channel(s) in the set 1 1 (Figure 2B shows N as being 13). Such transmissions may be of the time beacon or of any other type of data. In this case, the time beacon generally indicates the time window W , without more particularly indicating at which specific time instance Tn within that window W a transmission by mesh node 14 will actually occur. Regardless, having generated this time beacon, mesh node 14 broadcasts the time beacon in the mesh network 10 over one or more of the channels in the set 1 1 (Block 210). Mesh node 14 may do so

periodically or occasionally.

Mesh node 12 correspondingly receives the time beacon from mesh node 14 over one of the contention-based channels in the set 1 1 (Block 220). Mesh node 12 then adjusts the periodic transmission time window within which mesh node 12 is itself configured to transmit over the channel(s) in the set 1 1 , based on the received time beacon, to at least partially overlap mesh node 14's periodic transmission time window (Block 230). As shown briefly in Figure 2C, for instance, mesh node 12's transmission time windows W1 -W3 before adjustment do not overlap mesh node 14's transmission time windows W1 -W2 at all. After adjustment, though, mesh node 12's transmission time windows W1 -W3 overlaps mesh node 14's transmission time window W1 -W3 for at least a portion of time. In other examples not shown, though, partial overlap may already occur for one or more transmission time windows, and the adjustment herein achieves a greater degree of overlap.

Regardless, for this interval of overlap between the mesh nodes' transmission time windows, transmissions by the mesh nodes 12, 14 are susceptible to collision on the channel(s) in the set 1 1 . But with the mesh nodes' periodic transmission time windows at least partially overlapping in this way, mesh node 12 operates in a sleep mode during at least a portion of time outside mesh node 12's periodic transmission time window (Block 240). In the sleep mode, mesh node 12 turns off one or more receivers at mesh node 12. Alternatively or additionally, mesh node 12 turns off one or more transmitters at mesh node 12. Where mesh node 12 simply has a single radio for both reception and transmission, such may entail turning off the radio in full or part.

Indeed, armed with information about when mesh node 14 is configured to transmit, and having adjusted when mesh node 12 is itself configured to transmit, mesh node 12 exploits the overlap in transmission time windows as certainty that it will not miss receiving a transmission from node 14 if it turns off its receiver(s) or radio during at least a certain portion of time outside its transmission time window. With its receiver(s)/radio turned off, mesh node 12

advantageously conserves power in sleep mode without adversely affecting its ability to receive transmissions, e.g., as needed to serve as a relay node in the partially meshed network 10. As shown in Figure 2A, mesh node 14 is likewise configured to operate in the sleep mode during at least a portion of time outside mesh node 14's periodic transmission time window (Block 250).

In at least some embodiments, overlap between the mesh nodes' transmission time windows advantageously conserves mesh node power while proving inconsequential or acceptable in terms of collisions. In one embodiment, for example, latency requirements in the partially meshed network are low enough (e.g., 100ms per hop), transmission durations are low enough, and/or data transmission rates are high enough (e.g., 1 Mbps) to ensure that the likelihood of transmission collision is low, despite transmission time window overlap. Such may be the case for instance in a BLE mesh network.

To nonetheless reduce the likelihood of collision further, a transmission time window in at least some embodiments is a window of time within which a mesh node performs a transmission at a random time within that window (e.g., as opposed to at a predetermined time within the window, as contemplated by other embodiments). That is, in at least some embodiments, mesh node 14 randomly selects a time within its periodic transmission time window (e.g., time T5 in Figure 2B) and transmits at that randomly selected time. The mesh node 14 in one or more embodiments performs this random selection on a window-by-window basis so as to transmit at a potentially different time in different successive windows.

In some embodiments, the approaches described above amount to only coarsely synchronizing different nodes' transmissions, so as to lump their transmissions into confined time windows while also intentionally promoting variability in the exact timing of those contention-based transmissions. Both of these conserve mesh node power. Lumping transmissions into confined time windows creates deterministic opportunities for transmission silence and thereby receiver/transmitter/radio power down. And the contention-based nature of the transmissions avoids power-consuming signaling that would otherwise be required overhead for exchanging scheduling information amongst the nodes. Moreover, the approaches may realize a fully distributed, self-adaptive technique for coarse synchronization, without requiring a central or special node to manage synchronization.

Furthermore, in one or more embodiments, the above approaches provide each mesh node 12, 14 with the freedom to autonomously establish the length of its own periodic transmission time window. That is, mesh node 12 retains independent control over its periodic transmission time window's length, even when adjusting that window based on the time beacon received from mesh node 14. For instance, mesh node 12 adjusts its periodic transmission time window for at least partial overlap by adjusting the window's periodicity and/or shifting the window in time (without necessarily adjusting the window's length), based on the time beacon received from node 14. In this way, mesh node 12 maintains control over the duty cycle of its sleep mode.

In at least some embodiments, mesh node 12 adjusts its periodic transmission time window by calculating a periodic reception time window within which mesh node 12 expects to receive from mesh node 14 over the channel(s) in the set 1 1 . Mesh node 12 calculates this periodic reception time window based on mesh node 14's periodic transmission time window, as indicated by the time beacon received from mesh node 14. Mesh node 12 then adjusts its periodic transmission time window to at least partially overlap the periodic reception time window calculated for mesh node 14.

In one such embodiment, mesh node 12 calculates mesh node 14's periodic reception time window to account for one or more components of uncertainty about that window. For instance, mesh node 12 may widen mesh node 14's periodic reception time window based on an accuracy of a sleep timer at mesh node 12 and/or 14. With their respective sleep timers governing when mesh node 12 and 14 transition in and out of the sleep mode, the accuracy of those timers (e.g., due to clock drift) affects when their respective transmissions actually occur. In a BLE mesh network, for instance, a 50 parts per million (PPM) crystal may introduce a maximum of +50/AS clock drifting per second. Widening the calculated reception time window to absorb transmissions from mesh node 14 that are premature or late due to sleep timer inaccuracy ensures mesh node 12 does not miss receiving a transmission from mesh node 14. Figure 3 illustrates this window widening in a context where mesh node 12 joins the mesh network 10 for the first time and mesh nodes 12, 14 transmit at a randomly selected time within their respective periodic transmission time windows.

As shown in Figure 3, mesh node 12 attempts to join the mesh network 10. When doing so, mesh node 12 listens during interval 300 for a time beacon from any mesh node in the network 10. If mesh node 12 does not detect a time beacon during interval 300, the node transmits over one or more of the channels in the set 1 1 within a default periodic transmission time window (not shown). Mesh node 12 may do so based on an assumption that it is the first node in the network 10. As shown, though, mesh node 14 has already joined the network 10 and has a periodic transmission time window TX _WinlA (established based on an existing reception time window RX _Win _ E ) within which it performs transmissions at a random time. Mesh node 12's initial listening interval 300 happens to overlap mesh node 14's periodic transmission time window TX _WinlA in this example. As such, mesh node 12 receives a time beacon 310 that mesh node 14 transmits at a random time within the window TX _Win\A . Responsive to detecting this beacon 310 and based on the periodic transmission time window TX _Win\A indicated by the beacon 310, mesh node 14 adjusts its own periodic transmission time window from the default window (not shown) to window TX _Win\2. Mesh node 12's transmission time window TX _Win\2 in this regard at least partially overlaps mesh node 14's transmission time window TX _Win\A .

More specifically, mesh node 12 calculates a periodic reception time window

RX _Win\A within which mesh node 12 expects to receive from mesh node 14 over one or more of the channels in the set 1 1 . In doing so, mesh node 12 widens that reception time window RX _Win\A based on the accuracy of a sleep timer at mesh node 12 and/or 14. As shown, for instance, mesh node 14's reception window RX _Win\A is wider than mesh node 14's transmission time window TX _Win\A to account for sleep timer inaccuracy. Having calculated mesh node 14's reception window RX _Win\ , mesh node 12 adjusts its

transmission time window TX _Winl2 to at least partially overlap that reception window

RX _Win\A .

Consider a specific example where mesh node 12 performs this adjustment based on certain information indicated by the time beacon 310 received from mesh node 14. As shown in Figure 4A, mesh node 14's periodic transmission time window is characterized by the length TX _Win\A of any given window. The window is also characterized by the start of mesh node 14's next transmission time window relative to the time at which mesh node 14 broadcasts the time beacon 310, here shown as the difference DUE _TIME . For example, before sending the beacon 310, the value of DUE _TIME may be calculated and assigned into beacon. In at least some embodiments, time beacon 310 indicates both DUE _TIME and TX _Win\A .

In a simple embodiment, mesh node 12 calculates the next transmission and/or reception time window for mesh node 14 as occurring DUE _ TIME after it receives time beacon 310 from mesh node 14, and for a length of TX _Win\A . Mesh node 12 adjusts its periodic transmission time window TX _Win\2 to at least partially overlap that calculated window.

In other embodiments, though, time beacon 310 includes additional information about the accuracy of mesh node's sleep timer. As shown in Figure 4B, for instance, the time beacon 310 includes not only DUE _TIME 440 and TX _Win 430, but also fields SCA 400,

CYCLE _TIME 410, and TB _ EXP 420. The field SCA (sleep clock accuracy) 400 describes the accuracy of mesh node 14's sleep timer, e.g., in terms of parts per million (PPM). The field CYCLE _TIME 410 indicates a length of mesh node 14's periodic sleep-awake cycle (shown in

Figure 4A). This cycle is a periodically repeating pattern of transitions by mesh node 14 between the sleep mode and an awake mode in which mesh node 14 turns on one or more receivers at mesh node 14 (i.e.., the ones turns off in sleep mode). Finally, the field TB _ EXP

420 indicates the rate at which mesh node 14 broadcasts a time beacon. Such a time beacon 310 may be in various informational formats including a protocol description unit (PDU) and a text format.

In at least some embodiments, mesh node 12 calculates the length of the periodic reception time window for mesh node 14, with window widening based on this additional information in time beacon 310. For example, mesh node 12 in one embodiment calculates the length Rx_Win of the periodic reception time window as:

Rx _ Win = Tx _Win + 2 * ( SCA_l + SCA₂ ) * CYCLE _ TIME * CYCLE _ NUM , where Tx _Win is the length of mesh node 14's transmission time window as indicated by field

430 in the time beacon 310, SCA_l is the accuracy of mesh node 12's own sleep timer, SCA₂ is the accuracy of mesh node 14's sleep timer as indicated by field 400 in time beacon 310, and where CYCLE _TIME is the length of mesh node 14's periodic sleep-awake cycle as indicated by field 410 in time beacon 310. This defines mesh node 14's duty cycle in at least some embodiments. Furthermore, mesh node 12 calculates CYCLE _ NUM as being the number of sleep-awake cycles that have passed since receiving the time beacon 310. In one embodiment, mesh node 12 tracks this number of cycles that have passed, with that number having a ceiling of TB _ EXP indicated by field 420 in time beacon 310.

Calculating the length of the periodic reception time window in this way advantageously accounts for increasing uncertainty that occurs with increasing time beacon 310 staleness. Specifically, as the number of sleep-awake cycles CYCLE _ NUM since receiving a time beacon 310 increases, the calculated Rx Win will lengthen until the next time beacon 310 is received, i.e., TB _ EXP cycles since the last beacon. Thus, in certain embodiments the expected Rx_Win may not be the same in each cycle. In the same or other embodiments, mesh node 12 calculates the start time Rx_ Win_ start of mesh node 14's reception time window as:

Rx win— Tx win

Rx Win start = DUE TIME - tt = =

^" 2

where Rx_Win is the calculated length of the reception time window, Tx Win is the length of mesh node 14's transmission time window as indicated by field 430 in the time beacon 310, DUE _TIME s the start of the mesh node 14's next transmission time window relative to the time at which the mesh node 14 broadcasts the received time beacon 310, as indicated by field 440, and tt is an estimate of the transmission propagation time between mesh node 12 and mesh node 14.

Figure 5 depicts this calculation graphically. As shown, mesh node 14 transmits time beacon 310 at time Tt , but mesh node 12 does not receive the beacon 310 until time Tr , resulting in a propagation delay tt . Mesh node 12 inspects the time beacon 310 to obtain

DUE _TIME (as shown DUE _ TIME = Tx _ s - Tt ), as well as Tx_Win (as shown

Tx _Win = Tx _ e - Tx _ s ) . After having calculated Rx Win as described above, to obtain a length of Rx_Win = Rx_ e - Rx_ s , mesh node 12 calculates Rx_Win _ start using

DUE _TIME , Tx _Win , and Rx Win .

Of course, although the example above assumed a non-zero propagation delay, in some embodiments, tt can be assumed to be zero.

Note that, in view of window widening, the periodic nature of a mesh node's transmission and/or reception time windows may be coarse, at least with respect to the windows' start time.

In another sense, though, the periodic nature of the windows is preserved when viewed from the midpoint of the windows.

Irrespective of the particular manner in which mesh node 12 calculates mesh node 14's reception time window, mesh node 12 in at least some embodiments itself generates a time beacon indicating mesh node 12's periodic transmission time window, as adjusted. Mesh node

12 then broadcasts the generated time beacon in the mesh network 10 over one or more of the channels in the set 1 1 .

Notably, mesh node 12 in this case signals its adjusted periodic transmission time window. In one or more embodiments, mesh node 12 does so instead of propagating signaling indicating mesh node 14's periodic transmission time window. That is, mesh node 12 refrains from relaying the time beacon it received from mesh node 14. In at least some embodiments, this is despite mesh node 12 being configured to relay at least one other type of communication

(e.g., as a relay node in the network 10). This effectively isolates mesh node 12's periodic transmission time window (and its effects) to other mesh nodes that are within only a single hop from mesh node 12 (i.e., its direct neighbors), rather than propagating that window throughout the whole mesh network 10. In other words, the time beacon is transmitted as a broadcast with hop 1 , meaning that all mesh nodes within propagation range receive it, but the beacon is never forwarded by any mesh node.

Limiting the impact of each mesh node's (autonomously selected) transmission time window to only a subset of mesh nodes in the network 10 in this way advantageously conserves mesh node power, especially if some mesh nodes select to operate in the sleep mode for shorter periods of time than others (e.g., for increased performance requirements). Of course, although described as limiting transmission time window propagation to only a single hop, propagation could also be limited to any defined number of hops so as to exclude a subset of the network 10.

Correspondingly, a mesh node in such embodiments may be configured to only adjust its periodic transmission time windows based on the periodic transmission time windows of mesh nodes within a defined number hops (e.g., one). Figure 6 shows an example where a mesh node 12 is isolated by its neighbors, including mesh node 14 and mesh node 16. As shown in Figure 6, mesh node 14 and mesh node 16 may communicate with each other if mesh node 12 relays transmissions between them, since mesh node 12 is within the radio

propagation range of mesh node 14 and mesh node 16. Mesh node 14 and mesh node 16 are not neighbors of each other, while mesh node 12 is the neighbor of both mesh node 14 and mesh node 16. Mesh node 14, mesh node 12 and mesh node 16 have transmission time windows Tx_Win_14, Tx_Win_12, and Tx_Win_16, respectively. The effect of mesh node 14's window Tx_Win_14 is only directly felt by mesh node 12, and the effect of mesh node 16's window Tx_Win_16 is only directly felt by mesh node 12 as well. For example, if mesh node 14 declares a long Tx_Win_14 in its time beacon, only mesh node 12 needs to adapt to that long window; mesh node 16 is not affected since it only need to adapt to mesh node 12's

Tx_Win_12.

Of course, if mesh node 12 is within radio propagation range of multiple other mesh nodes, mesh node 12 may receive transmit beacons broadcasted from multiple mesh nodes, not just node 14. In this case, mesh node 12 may adjust its periodic transmission time window, based on the received beacons, to at least partially overlap one or more of the other mesh nodes' transmission time windows.

In at least some embodiments where mesh node 12 receives different time beacons from different mesh nodes, mesh node 12 in this regard independently calculates different periodic reception time windows within which mesh node 12 respectively expects to receive from the different mesh nodes over one or more of the channels in the set 1 1 . That is, the different reception time windows are dedicated to individual mesh nodes and may be for instance updated when time beacons are respectively received from those nodes. Regardless, in this case, mesh node 12 adjust its periodic transmission time window based on the calculated reception time windows, e.g., so that its transmission time window at least partially overlaps one or more of the independently calculated reception time windows. In one or more embodiments, the independently calculated reception time windows could end up being distributed randomly in time due to lack of coordination. In an effort to combat this, and confine the reception time windows to defined periods (and thereby create opportunities for sleep mode), mesh node 12 may perform a sort of window translation procedure. Figure 7 shows an example of this procedure.

As shown in Figure 7, mesh node 12 receives different time beacons from three different mesh nodes, resulting in three different independently calculated reception time windows Rx_Win , Rx_Win\ , and Rx _Win3 in cycle 320A . In view of this, mesh node 12 calculates a so-called periodic global reception-transmission time window TxRx_ Win shown in cycle 320B. This is a window TxRx_Win within which mesh node 12 expects to receive from the three different mesh nodes over one or more of the channels in the set 1 1 , and within which the mesh node 12 is configured to transmit over one or more of the channels in the set 1 1 . In

embodiments where mesh node 12 listens throughout its transmission time window as well, at least while not transmitting, the global window may define the total interval that mesh node 12 listens for transmissions from other mesh nodes. As shown, for instance, this window

TxRx _Win spans the three different independently calculated reception time windows

Rx_Win , Rx_Win\ , and Rx _Win3 , as well as mesh node 12's own transmission time window Tx JWin . Having calculated the global window TxRx JWin , mesh node 12 aligns a midpoint of its periodic transmission time window Tx Win with a midpoint of the periodic global reception-transmission time window TxRxJWin , as shown in cycle 320C.

In at least some embodiments, mesh node 12 ultimately aligns the windows in this way by recursively correcting skewness between the windows TxJWin and TxRx JWin . Skewness in this regard is defined as the difference between the midpoint of window Tx JWin and the midpoint of window TxRx Win . Mesh node 12 translates its transmission time window

TxJWin in time by this skewness in order to offset and minimize the skewness. Mesh node 12 may do so in conjunction with broadcasting its time beacon, so that the time beacon can indicate its skew-corrected transmission time window.

In one embodiment, for example, mesh node 12 calculates the skewness every time before sending its own time beacon. Mesh node 12 compensates for the skewness in the DUE JTIME field of the beacon. Eventually, the center of the global window TxRxJWin becomes the center of mesh node 12's transmission time period TxJWin . It may take several cycles to complete the alignment and such alignment may need to be updated if the

transmission times of other mesh nodes change (e.g. as they update their alignment).

Alternatively, other alignments may occur such as aligning the start of a transmission window with the start of a reception time window. Illustrated yet other embodiments, Figure 8 shows an example mesh node 12 adjusts its transmission time window 620A to partially overlap with only one of its independently calculated reception time windows 61 OA, 630A. As shown in cycle 600A, aligning the middle of mesh node 12's transmission time window 620A with the middle of the previously calculated global window TxRx Win would cause mesh node 12 to miss opportunities to enter sleep mode. Indeed, if mesh node 12 did enter sleep mode between reception windows 61 OA and 630A, it would be interrupted by its own transmission time window 620A. Accordingly, in some embodiments, mesh node 12 aligns the midpoint of its transmission time window 620A with the midpoint of a selected one of the reception time windows 630C, as shown in cycles 600B and 600C.

Note that the partially meshed network 10 herein may be a wired or wireless network. It may be a personal area network, a Bluetooth network, or a Bluetooth Low Energy network. The mesh nodes (e.g. mesh node 12) can be any device configured to communicate in a partially meshed topology.

Example embodiments are described herein with reference to mesh nodes depicted in Figure 1 (e.g. mesh node 12 and mesh node 14). Those skilled in the art will appreciate that such embodiments are merely an example and these embodiments herein are applicable to other mesh nodes in other multi-hop mesh networks.

Embodiments herein also include corresponding apparatus. Embodiments for instance include a mesh node (e.g. mesh node 12 as a "first" mesh node). Figure 9 illustrates additional details of a mesh node 12 in accordance with one or more embodiments. The mesh node 12 is configured, e.g., via any functional means or units, to implement the processing described above. The mesh node 12 may include for instance a receiving means or unit 920 for receiving a time beacon; an adjusting means or unit 950 for adjusting a mesh node's periodic

transmission time window; and a control means or unit 910 for operating the mesh node in a sleep mode. Such control means may comprise turning off all or part of a radio unit of a mesh node. It may also comprise using a lower frequency clock.

In at least some embodiments, a mesh node 12 comprises one or more processing circuit(s) 980 configured to implement the above processing, such as by implementing functional means or units. In one embodiment, for example, the processing circuit(s) 980 implement functional means or units as respective circuits. The circuits in this regard may comprise circuits dedicated to performing certain functional processing and/or one or more microprocessors in conjunction with memory 990. In embodiments that employ memory 990, which may comprise one or several types of memory such as read-only memory (ROM), random-access memory, cache memory, flash memory devices, optical storage devices, etc., the memory stores program code that, when executed by the one or more for carrying out one or more microprocessors, carries out the techniques described herein.

In one or more embodiments, the mesh node 12 also comprises one or more communication interfaces 970. The one or more communication interfaces include various components for sending and receiving data and control signals. More particularly, the interface(s) include one or more transmitters that is each configured to use known signal processing techniques, typically according to one or more standards, and is configured to condition a signal for transmission (e.g., over the air via one or more antennas). Similarly, the interface(s) include one or more receivers that is each configured to convert signals received (e.g., via the antenna(s)) into digital samples for processing by the one or more processing circuits.

Embodiments may also include other types of mesh nodes (e.g. mesh node 14 as a "second" mesh node). Figure 10 illustrates additional details of a mesh node 14 in accordance with one or more embodiments. The mesh node 14 is configured, e.g., via any functional means or units, to implement the processing described above. The mesh node 14 may include for instance a generating means or unit 1020 for generating a time beacon, a transmitting means or unit 1030 for broadcasting the time beacon; and control means or unit 1010 for operating the mesh node in a sleep mode. Such control means may comprise turning off all or part of the radio unit of a mesh node. It may also comprise using a lower frequency clock.

In at least some embodiments, a mesh node 14 comprises one or more processing circuit(s) 1080 configured to implement the above processing, such as by implementing functional means or units. In one embodiment, for example, the processing circuit(s) 1080 implement functional means or units as respective circuits. The circuits in this regard may comprise circuits dedicated to performing certain functional processing and/or one or more microprocessors in conjunction with memory 1090. In embodiments that employ memory 1090, which may comprise one or several types of memory such as read-only memory (ROM), random-access memory, cache memory, flash memory devices, optical storage devices, etc., the memory stores program code that, when executed by the one or more for carrying out one or more microprocessors, carries out the techniques described herein.

In one or more embodiments, the mesh node 14 also comprises one or more communication interfaces 1070. The one or more communication interfaces include various components for sending and receiving data and control signals. More particularly, the interface(s) include one or more transmitters that is each configured to use known signal processing techniques, typically according to one or more standards, and is configured to condition a signal for transmission (e.g., over the air via one or more antennas). Similarly, the interface(s) include one or more receivers that is each configured to convert signals received (e.g., via the antenna(s)) into digital samples for processing by the one or more processing circuits.

Those skilled in the art will appreciate that a single mesh node may comprise one or more functional means or units, processing circuits, memory, or communication interfaces of mesh node 12 and mesh node 14 in accordance with one or more embodiments. For example, a single mesh node may comprise a generating means or unit, a transmitting means or unit, a receiving means or unit and a control means or unit(s). That is, a mesh node herein may be capable of performing as both mesh node 12 and mesh node 14, e.g., at different times.

Those skilled in the art will also appreciate that embodiments herein further include corresponding computer programs.

A computer program comprises instructions which, when executed on at least one processor of a mesh node, cause the node to carry out any of the respective processing described above. A computer program in this regard may comprise one or more code modules corresponding to the means or units described above.

Embodiments further include a carrier containing such a computer program. This carrier may comprise one of an electronic signal, optical signal, radio signal, or computer readable storage medium.

Those skilled in the art will recognize that the present invention may be carried out in other ways than those specifically set forth herein without departing from essential

characteristics of the invention. The present embodiments are thus to be considered in all respects as illustrative and not restrictive, and all changes coming within the meaning and equivalency range of the appended claims are intended to be embraced therein.

Claims

CLAIMS What is claimed is:

1 . A method in a partially meshed network (10) that includes multiple mesh nodes, including a first mesh node (12) and a second mesh node (12), configured to communicate over a set (1 1 ) of one or more contention-based channels, the method comprising:

at the first mesh node (12), receiving (220) a time beacon from the second mesh node (14) in the mesh network (10) over one of the contention-based channels in the set (1 1 ), wherein the time beacon indicates a periodic transmission time window within which the second mesh node (14) is configured to transmit over one or more of the channels in the set (1 1 );

adjusting (230) the periodic transmission time window within which the first mesh node (12) is configured to transmit over one or more of the channels in the set (1 1 ), based on the received time beacon, to at least partially overlap the second mesh node's periodic transmission time window; and

during at least a portion of time outside the first mesh node's periodic transmission time window, operating (240) the first mesh node (12) in a sleep mode in which the first mesh node (12) turns off one or more receivers at the first mesh node (12).

2. The method of claim 1 , further comprising autonomously establishing, by the first mesh node, a length of the first mesh node's periodic transmission time window, and wherein said adjusting comprises adjusting a periodicity of the first mesh node's periodic transmission time window and/or shifting the first mesh node's periodic transmission time window in time based on the received time beacon.

3. The method of any of claims 1 -2, wherein said adjusting comprises calculating, based on the periodic transmission time window indicated by the time beacon, a periodic reception time window within which the first mesh node (12) expects to receive from the second mesh node (14) over one or more of the channels in the set (1 1 ), and adjusting the first mesh node's periodic transmission time window to at least partially overlap the periodic reception time window calculated for the second mesh node (14).

4. The method of claim 3, wherein said calculating comprises widening said periodic reception time window based on an accuracy of a sleep timer at the first and/or second mesh node (14), wherein a sleep timer is a timer that governs when a mesh node transitions in and out of said sleep mode.

5. The method of any of claims 3-4, wherein said calculating comprises calculating the length of said periodic reception time window as:

Rx _ Win = Tx_Win + 2* ( SCA_l + SCA₂ ) * CYCLE _ TIME * CYCLE _ NUM

wherein:

Rx_ Win is the length of said periodic reception time window;

Tx_Win is a length of a transmission time window of the second mesh node

(14) as indicated in the time beacon received from the second mesh node (14);

SCA₁ is an accuracy of a sleep timer at the first mesh node (12), wherein a sleep timer is a timer based on which a mesh node transitions in and out of said sleep mode;

SCA₂ is an accuracy of a sleep timer at the second mesh node (14);

CYCLE _TIME is a length of a periodic sleep-awake cycle at the second mesh node (14), as indicated in the time beacon received from the second mesh node (14), wherein a periodic sleep-awake cycle is a periodically repeating pattern of transitions by the second mesh node (14) between the sleep mode and an awake mode in which the first mesh node (12) turns on said one or more receivers; and

CYCLE _ NUM is a number of sleep-awake cycles that have passed since receiving the time beacon from the second mesh node (14).

6. The method of any of claims 3-5, wherein calculating the periodic reception time window comprises calculating the start of the periodic reception time window as:

Rx win— Tx win

Rx Win start = DUE TIME - tt = =

^" 2

wherein:

Rx_Win _ start is the start of the periodic reception time window;

DUE _TIME s the start of the second mesh node's next transmission time window relative to the time at which the second mesh node (14) broadcasts the received time beacon;

tt is an estimate of a transmission propagation time between the first and second mesh nodes;

Rx_Win is a length of the periodic reception time window; and

Tx_Win is a length of the second mesh node's periodic transmission time window.

7. The method of any of claims 1 -6, comprising receiving different time beacons from different mesh nodes in the mesh network (10), and wherein said adjusting comprises independently calculating different periodic reception time windows within which the first mesh node (12) respectively expects to receive from the different mesh nodes over one or more of the channels in the set (1 1 ), and adjusting the first mesh node's periodic transmission time window based on the calculated reception time windows.

8. The method of any of claims 1 -7, comprising receiving different time beacons from different mesh nodes in the mesh network (10), and wherein said adjusting comprises calculating a periodic global reception-transmission time window within which the first mesh node (12) expects to receive from the different mesh nodes over one or more of the channels in the set (1 1 ), and within which the first mesh node (12) is configured to transmit over one or more of the channels in the set (1 1 ), and aligning a midpoint of the first mesh node's periodic transmission time window with a midpoint of the periodic global reception-transmission time window.

9. The method of any of claims 1 -8, further comprising:

at the first mesh node (12), generating a time beacon indicating the first mesh node's periodic transmission time window, as adjusted; and

broadcasting the generated time beacon in the mesh network (10) over one or more of the channels in the set (1 1 ).

10. The method of any of claims 1 -9, wherein the first mesh node (12) is a relay node, and wherein the method comprises refraining from relaying the received time beacon to any other mesh node in the mesh network (10), even though the first mesh node (12) is configured to relay at least one other type of communication.

1 1 . The method of any of claims 1 -10, wherein said adjusting comprises adjusting the first mesh node's periodic transmission time window based on the periodic transmission time windows of mesh nodes within only a defined number of hops from the first mesh node (12).

12. The method of any of claims 1 -1 1 , further comprising:

when attempting to join the mesh network (10), listening by the first mesh node (12) for a time beacon from any mesh node in the mesh network (10);

if no time beacon is detected by said listening, transmitting by the first mesh node (12) over one or more of the channels in the set (1 1 ) within a default periodic transmission time window; wherein said adjusting comprises adjusting the first mesh node's periodic transmission time window from said default periodic transmission time window, responsive to detecting the time beacon from the second mesh node (14) during said listening.

13. A method in a partially meshed network (10) that includes multiple mesh nodes, including a first mesh node (12) and a second mesh node (14), configured to communicate over a set (1 1 ) of one or more contention-based channels, the method comprising:

at the second mesh node (14), generating (200) a time beacon indicating a periodic transmission time window within which the second mesh node (14) is configured to transmit over one or more of the one or more contention-based channels in the set (1 1 );

broadcasting (210) the time beacon in the mesh network (10) over one or more of the channels in the set (1 1 ); and

during at least a portion of time outside the second mesh node's periodic transmission time window, operating (250) the second mesh node (14) in a sleep mode in which the second mesh node (14) turns off one or more receivers at the second mesh node (14).

14. The method of claim 13, further comprising, prior to generating the time beacon:

receiving a time beacon from the first mesh node (12) in the mesh network (10) over one of the contention-based channels in the set (1 1 ), wherein the time beacon indicates a periodic transmission time window within which the first mesh node (12) is configured to transmit over one or more of the channels in the set (1 1 ); and

adjusting the periodic transmission time window within which the second mesh node (14) is configured to transmit over one or more of the channels in the set (1 1 ), based on the received time beacon, to at least partially overlap the first mesh node's periodic transmission time window;

wherein said generating comprises generating the time beacon to indicate the second mesh node's periodic transmission time window, as adjusted.

15. The method of any of claims 13-14, wherein generating the time beacon to indicate the second mesh node's periodic transmission time window comprises indicating the start of the second mesh node's next transmission time window relative to the time at which the second mesh node (14) broadcasts the time beacon.

16. The method of any of claims 13-15, wherein broadcasting the time beacon comprises broadcasting the time beacon at a random time within the periodic transmission time window.

17. The method of any of claims 1 -16, wherein the set (1 1 ) of one or more contention-based channels comprises one or more advertising channels configured for advertising the availability of a mesh node to communicate with another mesh node in the mesh network (10) over a connection-oriented channel.

18. The method of claim 1 -17, wherein the time beacon transmitted by a particular mesh node indicates one or more of:

the start of that mesh node's next transmission time window relative to the time at which the mesh node broadcasts the time beacon;

an accuracy of a sleep timer at that mesh node, wherein a sleep timer is a timer based on which the mesh node transitions in and out of said sleep mode; a length of the mesh node's periodic transmission time window;

the next transmission time window in which the next time beacon from that mesh node can be expected; and

a length of a periodic sleep-awake cycle at the mesh node, wherein a periodic sleep- awake cycle is a periodically repeating pattern of transitions by the mesh node between the sleep mode and an awake mode in which the mesh node turns on said one or more receivers.

19. The method of any of claims 1 -18, wherein a transmission time window is a window of time within which a mesh node performs a transmission at a random time within that window.

20. The method of any of claims 1 -19, wherein the mesh network (10) is a Bluetooth Low Energy (BLE) mesh network and the one or more channels comprise one or more

advertisement channels.

21 . A first mesh node (12) in a partially meshed network (10), wherein the first mesh node (12) is a first one of multiple mesh nodes in the mesh network (10) that communicate over a set (1 1 ) of one or more contention-based channels, the mesh node configured to:

receive a time beacon from a second mesh node (14) in the mesh network (10) over one of the contention-based channels in the set (1 1 ), wherein the time beacon indicates a periodic transmission time window within which the second mesh node (14) is configured to transmit over one or more of the channels in the set (1 1 );

adjust a periodic transmission time window within which the first mesh node (12) is configured to transmit over one or more of the channels in the set (1 1 ), based on the received time beacon, to at least partially overlap the second mesh node's periodic transmission time window; and

during at least a portion of time outside the first mesh node's periodic transmission time window, operate in a sleep mode in which the first mesh node (12) turns off one or more receivers at the first mesh node (12).

22. The first mesh node of claim 21 , configured to implement the method of any of claims 2-12 or 17-20.

23. A second mesh node (14) in a partially meshed network (10) that includes multiple mesh nodes, including a first mesh node (12) and the second mesh node (14), wherein the mesh nodes are configured to communicate over a set (1 1 ) of one or more contention-based channels, the second mesh node (14) configured to:

generate a time beacon indicating a periodic transmission time window within which the second mesh node (14) is configured to transmit over one or more of the one or more contention-based channels in the set (1 1 );

broadcast the time beacon in the mesh network (10) over one or more of the channels in the set (1 1 ); and

during at least a portion of time outside the second mesh node's periodic transmission time window, operate in a sleep mode in which the second mesh node (14) turns off one or more receivers at the second mesh node (14).

24. The second mesh node of claim 23, configured to implement the method of any of claims 14-20.

25. A computer program comprising instructions which, when executed by at least one processor of a mesh node, causes the mesh node to carry out the method of any of claims 1 -20.

26. A carrier containing the computer program of claim 25, wherein the carrier is one of an electronic signal, optical signal, radio signal, or computer readable storage medium.