Abstract: Various embodiments disclose a method performed by a first node device in a mesh network that includes identifying a plurality of neighbor node devices that have one or more blocks of a dataset; determining quality scores for respective node devices of the plurality of neighbor node devices; selecting, based on the quality scores, a second node device from the plurality of neighbor node devices; and sending, to the second node device, a first request to receive at least one block of the one or more blocks.

Abstract: Techniques are directed to using communication metric data associated with multiple modulation schemes to achieve a link quality metric that is representative of the link as a whole, across the multiple modulation schemes that may be employed on the link. A calculation of a link quality metric may be triggered by a network layer transmission attempt, with communication metrics accumulated at the link layer of the link. A filter used to calculate the link quality metric may be updated based on network layer transmission attempts, based on successful and/or unsuccessful transmissions at a Media Access Control (MAC) layer of the link. More generally, a calculation of link quality may be triggered by a higher layer transmission attempt while being calculated based on transmission attempts at a lower layer of the link.

Abstract: Various embodiments disclose a method comprising determining, by a first node, a first link latency associated with a second node based on a time period between consecutive slots in a unicast listening schedule for the second node; in response to the first node detecting a failure in transmitting a first frame to the second node, determining, by the first node based on the first link latency, a first backoff time; and in response to determining that the first backoff time has elapsed, retransmitting, by the first node, the first frame to the second node.

Abstract: Techniques for managing multi-mode communications between nodes in a network include initiating, by a first node in a mesh network, a carrier sense multiple access with collision avoidance (CSMA/CA) session while the first node is in a first mode. The CSMA/CA session is associated with a request to transmit a data message to a second node in the mesh network. Initiating the CSMA/CA session includes determining availability of a channel associated with a channel schedule of the second node associated with the first mode.

Abstract: Various embodiments disclose techniques for time-multiplexing multiple listening schedules and physical layer modes that include a first node listening, using a first physical layer mode, for a first network discovery signal; in response to not detecting the first network discovery signal, listening, using a second physical layer mode, for a second network discovery signal; and in response to detecting the second network discovery signal transmitted by a second node using the second physical layer mode, establishing a connection with the second node using the second physical layer mode.

Abstract: Techniques are directed to using communication metric data associated with multiple modulation schemes to achieve a link quality metric that is representative of the link as a whole, across the multiple modulation schemes that may be employed on the link. A calculation of a link quality metric may be triggered by a network layer transmission attempt, with communication metrics accumulated at the link layer of the link. A filter used to calculate the link quality metric may be updated based on network layer transmission attempts, based on successful and/or unsuccessful transmissions at a Media Access Control (MAC) layer of the link. More generally, a calculation of link quality may be triggered by a higher layer transmission attempt while being calculated based on transmission attempts at a lower layer of the link.

Abstract: Techniques for allocating event offsets within a period of transmission are described. A mains-powered device (MPD) may act as a “parent” to one or more battery-powered devices (BPDs). The MPD may assign “event offsets” to each BPD. The event offset is a time by which the BPD's timeslot is “offset” from the start of a periodic cycle of transmissions by the MPD. Thus, each event offset indicates a time that the BPD must be “awake,” i.e., operating its radio receiver and/or performing other functionality. A BPD may spend a substantial fraction of its time in a “sleep” mode, wherein less power is used and fewer functions are performed than during a period of that BPD's event offset. Another BPD may have a different event offset. Communications by the MPD with each child BPD may be substantially uniformly distributed over the period. To increase efficiency, groups of BPDs may receive multicasts.

Abstract: Techniques for establishing communications within a multicast group of nodes included in a mesh network include a method comprising: transmitting, by a node included in a mesh network using unicast, a first request to join a multicast group to a multicast group leader of the multicast group; and in response to determining that no message acknowledging the first request has been received from the multicast group leader, broadcasting, by the node, a second request to join the multicast group.

Abstract: Techniques for allocating event offsets within a period of transmission are described. A mains-powered device (MPD) may act as a “parent” to one or more battery-powered devices (BPDs). The MPD may assign “event offsets” to each BPD. The event offset is a time by which the BPD's timeslot is “offset” from the start of a periodic cycle of transmissions by the MPD. Thus, each event offset indicates a time that the BPD must be “awake,” i.e., operating its radio receiver and/or performing other functionality. A BPD may spend a substantial fraction of its time in a “sleep” mode, wherein less power is used and fewer functions are performed than during a period of that BPD's event offset. Another BPD may have a different event offset. Communications by the MPD with each child BPD may be substantially uniformly distributed over the period. To increase efficiency, groups of BPDs may receive multicasts.

Abstract: Various embodiments disclose a method comprising receiving, at a first node, a listening schedule associated with a second node; determining, by the first node, a link latency associated with the second node based on the listening schedule; in response to the first node detecting a frame transmission failure for a frame being transmitted by the first node to the second node, determining, by the first node based on the link latency, a backoff time; and in response to the first node determining that the backoff time has elapsed, the first node retransmitting the frame from the first node to the second node. In some embodiments, the method also includes determining a frame lifetime value associated with the second node based on the link latency; and in response to determining that a time period corresponding to the frame lifetime value has elapsed, dropping the frame.

Abstract: Techniques for allocating event offsets within a period of transmission are described. A mains-powered device (MPD) may act as a “parent” to one or more battery-powered devices (BPDs). The MPD may assign “event offsets” to each BPD. The event offset is a time by which the BPD's timeslot is “offset” from the start of a periodic cycle of transmissions by the MPD. Thus, each event offset indicates a time that the BPD must be “awake,” i.e., operating its radio receiver and/or performing other functionality. A BPD may spend a substantial fraction of its time in a “sleep” mode, wherein less power is used and fewer functions are performed than during a period of that BPD's event offset. Another BPD may have a different event offset. Communications by the MPD with each child BPD may be substantially uniformly distributed over the period. To increase efficiency, groups of BPDs may receive multicasts.

Abstract: Various embodiments disclose a method that includes: attempting to detect, with a first transceiver associated with a first node, a network discovery signal, wherein the attempting is performed according to (a) a first listening schedule associated with a first physical layer mode and (b) a second listening schedule associated with a second physical layer mode; detecting, with the first transceiver, the network discovery signal during a slot associated with the first listening schedule; and in response to detecting the network discovery signal, establishing, with the first node, a connection between the first node and the second node using the first physical layer mode.

Abstract: Techniques are directed to using communication metric data associated with multiple modulation schemes to achieve a link quality metric that is representative of the link as a whole, across the multiple modulation schemes that may be employed on the link. A calculation of a link quality metric may be triggered by a network layer transmission attempt, with communication metrics accumulated at the link layer of the link. A filter used to calculate the link quality metric may be updated based on network layer transmission attempts, based on successful and/or unsuccessful transmissions at a Media Access Control (MAC) layer of the link. More generally, a calculation of link quality may be triggered by a higher layer transmission attempt while being calculated based on transmission attempts at a lower layer of the link.

Abstract: One embodiment of the present invention sets forth a technique for establishing communications within a multicast group of nodes included in a mesh network. The technique includes detecting that a first message related to the member node joining the multicast group has not been received from a multicast group leader included in the multicast group. The technique also includes in response, generating a first broadcast message for the member node that includes a multicast join request. The technique further includes forwarding the first broadcast message to one or more nodes included in the mesh network that are direct neighbors of the member node, wherein at least one node included in the one or more nodes further forwards the first broadcast message based on a first maximum hop limit.

Abstract: Various embodiments disclose a method comprising receiving, at a first node, a listening schedule associated with a second node; determining, by the first node, a link latency associated with the second node based on the listening schedule; in response to the first node detecting a frame transmission failure for a frame being transmitted by the first node to the second node, determining, by the first node based on the link latency, a backoff time; and in response to the first node determining that the backoff time has elapsed, the first node retransmitting the frame from the first node to the second node. In some embodiments, the method also includes determining a frame lifetime value associated with the second node based on the link latency; and in response to determining that a time period corresponding to the frame lifetime value has elapsed, dropping the frame.

Abstract: Various embodiments disclose a method performed by a first node device in a mesh network that includes identifying a plurality of neighbor node devices that have one or more blocks of a dataset; determining quality scores for respective node devices of the plurality of neighbor node devices; selecting, based on the quality scores, a second node device from the plurality of neighbor node devices; and sending, to the second node device, a first request to receive at least one block of the one or more blocks.

Abstract: Various embodiments disclose a method that includes: attempting to detect, with a first transceiver associated with a first node, a network discovery signal, wherein the attempting is performed according to (a) a first listening schedule associated with a first physical layer mode and (b) a second listening schedule associated with a second physical layer mode; detecting, with the first transceiver, the network discovery signal during a slot associated with the first listening schedule; and in response to detecting the network discovery signal, establishing, with the first node, a connection between the first node and the second node using the first physical layer mode.

Abstract: Various embodiments disclose a method that includes sending, by a first node to a plurality of neighbor nodes, a query associated with a dataset, wherein the dataset includes multiple blocks; receiving, at the first node from a set of the plurality of neighbor nodes, one or more responses to the query; based on a first quality ranking of individual nodes in the set, selecting, by the first node, a second node from the set; and sending, by the first node, a first request to the second node for at least one block of the multiple blocks.

Abstract: Techniques for allocating event offsets within a period of transmission are described. A mains-powered device (MPD) may act as a “parent” to one or more battery-powered devices (BPDs). The MPD may assign “event offsets” to each BPD. The event offset is a time by which the BPD's timeslot is “offset” from the start of a periodic cycle of transmissions by the MPD. Thus, each event offset indicates a time that the BPD must be “awake,” i.e., operating its radio receiver and/or performing other functionality. A BPD may spend a substantial fraction of its time in a “sleep” mode, wherein less power is used and fewer functions are performed than during a period of that BPD's event offset. Another BPD may have a different event offset. Communications by the MPD with each child BPD may be substantially uniformly distributed over the period. To increase efficiency, groups of BPDs may receive multicasts.

Abstract: Source routing techniques include sending data across several networks, while limiting source routing overhead. For example, the source routing techniques may use a first address format to route data to nodes along a routing path that are within a first network where a source node is located, and use a second address format to route the data to a node along the routing path that is within a second, different network. The node in the second network may similarly route the data through the second network using the first address format for nodes within the second network and, if needed, route the data to a node within a third network using the second address format. This may be repeated for any number of networks to reach a destination.