Classifications

• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L45/00—Routing or path finding of packets in data switching networks
  • H04L45/12—Shortest path evaluation
  • H04L45/123—Evaluation of link metrics
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L12/00—Data switching networks
  • H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
  • H04L12/44—Star or tree networks
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L45/00—Routing or path finding of packets in data switching networks
  • H04L45/48—Routing tree calculation
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L45/00—Routing or path finding of packets in data switching networks
  • H04L45/70—Routing based on monitoring results
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L45/00—Routing or path finding of packets in data switching networks
  • H04L45/48—Routing tree calculation
  • H04L45/488—Routing tree calculation using root node determination
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04W—WIRELESS COMMUNICATION NETWORKS
  • H04W40/00—Communication routing or communication path finding
  • H04W40/02—Communication route or path selection, e.g. power-based or shortest path routing
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04W—WIRELESS COMMUNICATION NETWORKS
  • H04W84/00—Network topologies
  • H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks

Definitions

• Network rank is a quantification of the quality of a node's connection to a root node.
• a node's network rank may be advertised to its neighbors and used by the neighbors to evaluate the node as a potential parent.
• a node's rank is based on the number of hops required to reach the root node and other factors, such as the quality of the links between nodes.
• a network may become unstable.
• Brief aberrations in network quality may cause a node to select a parent based on aberrant data or may impact its advertised rank.
• Frequent changes in advertised rank may cause nodes to frequently change parents which may lead to instability in the network.
• the present invention provides a method for calculating and advertising a rank that reflects changes in the network but ignores transitory anomalies.
• the advertised rank is based on the induced rank of the node and the previously advertised rank.
• the induced rank is the node's rank at a point in time and is based on the parent node's rank and the quality of the link between the node and the parent node.
• a weighted running average or rank candidate is used to determine the advertised rank.
• the rank candidate is the weighted sum of the previous rank candidate value and the induced rank and is calculated each time the node advertises its rank. In this manner, only a sustained move to a new induced rank will result in a new advertised rank. This will result in a more stable network.
• a neighboring rank is a rank value that is one step in rank value higher or lower than the previously advertised rank.
• the advertised rank is set to the neighboring rank value.
• the advertised rank is set to the previous advertised rank. The advertised rank is communicated to the other nodes.
• a Link Quality Indicator may reflect the quality of the link between the node and the parent node.
• the value of the LOJ may result in an LQI adjustment to the induced rank.
• the value of the LQI may be calculated using a running average to account for differences in the number of packets received by the node.
• Figure 1 is a block diagram illustrating a portion of an exemplary RF mesh network.
• Figure 2 is a block diagram illustrating the selection of a parent node.
• Figure 3 is a flowchart illustrating an exemplary method of adjusting the advertised rank.
• Figure 4 is a graph showing gradual changes in advertised rank.
• Figure 5 is a flowchart illustrating an exemplary method of adjusting the advertised rank using an upper threshold and a lower threshold.
• Figure 6 is a block diagram illustrating an exemplary node.
• a node's advertised rank may be based on the node's induced rank and a previous rank candidate. Induced rank is the rank of the node at a particular point in time. Rank candidate is based on a running average of the node's past rank. In one implementation, the advertised rank is based on a weighted sum of the induced rank and a previous rank candidate.
• LQI Link Quality Indicator
• the LQI may be based on the average LQI over a time period, as well as a running average LQI.
• the LQI is a weighted sum of the average LQI over a time period and a previous running average LQI.
• Figure 1 is a diagram illustrating a portion of an exemplary RF mesh network.
• the nodes of the network may be configured to transmit and receive communications, as well as relaying communications for other nodes.
• the network may utilize the RPL routing protocol so that each node advertises its network rank to neighboring nodes.
• the network includes a root node 20.
• the network also includes a number of additional nodes, illustrated here as nodes 21 - 31.
• the root node 20 may communicate with additional nodes outside of the network depicted in Figure 1.
• Figure 1 depicts a single root node and eleven additional nodes for illustrative purposes, the RF mesh network can include any number of root nodes and any number of additional nodes.
• Figure 1 illustrates a number of parent-child relationships.
• root node 20 is the parent node for nodes 21, 22, and 23.
• a node may be both a parent and a child.
• node 22 is a child of root node 20 and is also a parent of node 26.
• a parent node is a node that is between the root node and the child node and that communicates directly with the child node. The child node will periodically re-evaluate its parent node and may select a new parent node if the new parent node provides a better quality connection to the root node.
• root node 20 is a collector for an AM I network and nodes 21-31 include metering devices for measuring resource consumption at different premises.
• Figure 2 illustrates the selection of a parent node by node 250.
• the portion of the network shown in Figure 2 includes root node 200 and nodes 210, 220, 240, 250, 260.
• Node 250 selects a parent node to minimize its own rank.
• rank is expressed in the form of binary numbers, or multiples of 256.
• Rank may also be expressed in the form of hexadecimal numbers, or multiples of 100. Other numbering systems may also be used.
• node 250 To select a parent node, node 250 considers the advertised rank of its neighboring nodes and the link quality between it and each neighboring node. Node 250 maintains LOJ information for its neighboring nodes based on its communications with the nodes. The LOJ may result in one or more rank values being added to the rank calculation as an LQI adjustment. In one implementation, when the LQI is below a threshold, indicating that the link quality is good, then the LQI adjustment is zero, when the LQI is between a first value and the threshold, indicating that the link quality is fair, one rank value may be added, and when the LQI is above the first value, indicating that the link quality is poor, two rank values may be added.
• the advertised rank of node 210 is 256
• the advertised rank of node 220 is 256
• the advertised rank of node 240 is 512
• the advertised rank of node 260 is 512.
• the LQI between nodes 210 and 250 is above the first value
• the LQI between nodes 240 and 250 is between the first value and the threshold
• the LQI between nodes 220 and 250 is above the first value
• the LQI between nodes 250 and 260 is below the threshold.
• Node 250 may consider nodes 210, 220, 240, and 260 as potential parent nodes and select the one that provides node 250 with the lowest rank.
• node 250 may determine the impact of the selection of the different potential parent nodes on its rank. If node 250 selects node 210 as its parent, then its rank is 1024 (advertised rank of node 210 + 2 rank values for LQI adjustment + 1 rank value for one hop). If node 250 selects node 240 as its parent, then its rank is 1024 (advertised rank of node 240 + 1 rank value for LQI adjustment + 1 rank value for one hop). If node 250 selects node 220 as its parent, then its rank is 1024 (advertised rank of node 220 + 2 rank values for LQI adjustment + 1 rank value for one hop).
• node 250 selects node 260 as its parent, then its rank is 768 (advertised rank of Node 260 + no LQI adjustment + 1 rank value for one hop). Since node 250 is seeking to minimize its own rank, it selects node 260 as its parent.
• the advertised rank of node 210 is 256
• the advertised rank of node 220 is 512
• the advertised rank of node 240 is 512
• the advertised rank of node 260 is 768.
• the LQI between nodes 210 and 250 is above the first value
• the LQI between nodes 240 and 250 is between the first value and the threshold
• the LQI between nodes 220 and 250 is below the threshold
• the LQI between nodes 250 and 260 is below the threshold.
• node 250 selects node 240 as its parent, then its rank is 1024 (advertised rank of node 240
• node 250 selects node 220 as its parent, then its rank is 768 (advertised rank of node 220 + no LQI adjustment + 1 rank value for one hop). If node 250 selects node 260 as its parent, then its rank is 1024
• a node selects only one parent node. If there are multiple potential parent nodes that result in the same rank for the node, then the node may give one of the factors used to determine its rank more weight. One option is to select the potential parent having the fewest number of hops to the root. Another option is to select the potential parent with the best LQI. Other options may use multiple factors or different factors to select one of the nodes as its parent when there is no difference in rank for the node.
• the advertised rank of node 210 is 256
• the advertised rank of node 220 is 256
• the advertised rank of node 240 is 512
• the advertised rank of node 260 is 512.
• the LQI between nodes 210 and 250 is between the threshold and the first value
• the LQI between nodes 240 and 250 is below the threshold
• the LQI between nodes 220 and 250 is above the first value
• the LQI between nodes 250 and 260 is between the threshold and the first value.
• node 250 selects node 210 as its parent, then its rank is 768 (advertised rank of node 210 + 1 rank values for LQI adjustment + 1 rank value for one hop).
• node 250 selects node 240 as its parent, then its rank is 768 (advertised rank of node 240 + no LQI adjustment + 1 rank value for one hop). If node 250 selects node 220 as its parent, then its rank is 1024 (advertised rank of node 220 + 2 rank values for LQI adjustment + 1 rank value for one hop). If node 250 selects node 260 as its parent, then its rank is 1024 (advertised rank of node 260 + 1 rank value for LQI adjustment + 1 rank value for one hop). In this example, nodes 210 and 240 result in the same rank for Node 250.
• node 250 selects node 210 as its parent.
• the node may determine the number of hops to the root using a routing metric, such as the Hop Count Object defined by the RPL specification. If the criteria for selecting a parent node includes the best LQI between the parent and child, then node 250 selects node 240 as its parent.
• a node's rank is dynamic and may change as the rank of its parent or the LQI between it and its parent changes.
• the term induced rank as used herein, refers to the node's rank at a particular time. It is based on the advertised rank of the parent node, plus one rank value for the hop from the parent node to the child, plus any LQI adjustment for the link between the parent node and the child node.
• a node advertises its rank to its neighboring nodes.
• a node may advertise a rank referred to herein as its advertised rank that may be different than its induced rank.
• the advertised rank is calculated and sent out in a
• DIO DODAG Information Object
• the frequency of the transmission of the DIO message is variable. In one example, the DIO is sent approximately once per hour, but it may be sent more frequently after a network change. In a mesh network, the DIO may be broadcast to the other nodes in the network. In one exemplary system, the frequency of the calculation of the induced rank is
• the advertised rank is set to be equal to the induced rank. Subsequently, the advertised rank is determined as described below.
• Rank Candidate (3 ⁇ 4 * previous Rank Candidate) + (1 ⁇ 4 * current Induced Rank).
• the rank candidate is determined each time the node advertises its rank.
• a neighboring rank is a rank value that is one step in rank value higher or lower than the previously advertised rank. For example, if the previously advertised rank is 768, the neighboring rank values are 512 and 1024.
• the difference between the rank candidate and the closest neighboring rank is compared to a threshold, referred to herein as a rank threshold, in 308.
• the rank threshold is configurable, and may vary according to the installation. The rank threshold may be the same for all nodes throughout the network. In one example, the rank threshold is set to approximately 10% of one rank value.
• the process continues to 310 and the advertised rank is set to the neighboring rank value.
• the rank candidate value is also set to the advertised rank value, for the purpose of calculating the rank candidate in the next iteration.
• the process continues to 312 and the advertised rank is set to the previous advertised rank.
• Figure 4 illustrates how the advertised rank may change over time and how the advertised rank relates to the induced rank.
• the y-axis represents the rank value.
• the x-axis represents sequential time periods at which the advertised rank is determined. In one implementation the advertised rank is determined whenever the node transmits a DIO message. Although the divisions along the x-axis are shown as being equally spaced, the time between communications may vary. For the purposes of illustration, one rank value is 256 and the rank threshold is 20.
• the advertised rank, the induced rank, and the rank candidate are all 512.
• the induced rank is 768.
• the rank candidate may be calculated using the equation provided above. If so, then the rank candidate is 576 (.75(previous rank candidate of 512) + .25(induced rank of 768)).
• the rank candidate is compared to the neighboring rank, 768. Since the difference between the rank candidate (576) and the neighboring rank (768) is greater than the rank threshold (192 > 20), the advertised rank is set to be the previous advertised rank or 512.
• the advertised rank is 512
• the rank candidate is 744
• the induced rank is 768.
• the rank candidate may be calculated using the equation provided above. If so, then the rank candidate is 750 (.75(previous rank candidate of 744) + .25(induced rank of 768)). Since the difference between the rank candidate (750) and the neighboring rank (768) is within the rank threshold (18 ⁇ 20), the advertised rank is set to a new rank value of 768.
• the advertised rank, the rank candidate, and the induced rank are all 768.
• the induced rank is 512.
• the rank candidate may be calculated using the equation provided above. If so, then the rank candidate is 704 (.75(previous rank candidate of 768) + .25(induced rank of 512)). Since the difference between the rank candidate (704) and the neighboring rank (512) is greater than the rank threshold (192 > 20), the advertised rank is set to the previous advertised rank of 768.
• the advertised rank is 768
• the rank candidate is 520
• the induced rank is 512.
• the rank candidate may be calculated using the equation provided above. If so, then the rank candidate is 518 (.75(previous rank candidate of 520) + .25(induced rank of 512)). Since the difference between the rank candidate (518) and the neighboring rank (512) is within the rank threshold (6 ⁇ 20), the advertised rank is set to a new rank value of 512.
• Figure 4 illustrates that the advertised rank changes more slowly than the induced rank.
• Figure 4 used the same rank threshold for a neighboring rank above the previous advertised rank or below the previous advertised rank, some implementations may use different rank thresholds. This is illustrated in Figure 5.
• an upper threshold may be used to determine when to increase the advertised rank and a lower threshold may be used to determine when to decrease the advertised rank.
• this method starts with the rank candidate determined in step 304 in Figure 3.
• the node evaluates whether the rank candidate is above or below the previous advertised rank. When the rank candidate is above the previous advertised rank, the node proceeds to 506 and determines the difference between the rank candidate and the upper neighboring rank. At 508, the node compares the difference calculated in 506 to the upper threshold. When the difference is within the upper threshold, the node proceeds to step 310 in Figure 3. When the difference is outside of the upper threshold, the node proceeds to step 312 in Figure 3. In this example, the advertised rank is only changed when the rank candidate is within the upper threshold.
• the node proceeds to 510 and determines the difference between the rank candidate and the lower neighboring rank.
• the node compares the difference from 510 to the lower threshold. If the difference is within the lower threshold, the node proceeds to step 310 in Figure 3. If the difference is outside of the lower threshold, the node proceeds to step 312 in Figure 3.
• the advertised rank is only changed when the rank candidate is within the lower threshold.
• LQI Link Quality Indicator
• SN R signal to noise ratio
• the node includes a translation table that is based upon the RF characteristics of the node's transceiver hardware. The translation table is used to convert the SN R to LQI so that a "good" SN R value is translated to a "good” LQI value for RPL.
• the LQI scale is consistent in RPL across hardware platforms, whereas the translation tables are adapted for the different hardware platforms used by the nodes.
• packet success rate may be used is to adjust the LQI based on packet success rate. For example, if the SNR value is "good”, but the packet success rate is bad, then the translation table provides a "good” LQI value, but the packet success rate adjusts the LQI value so that it is not quite as “good”. In this manner, the LQI reflects both SN R and other factors that affect packet success rate.
• a node Since a node is likely to receive more packets from certain nodes, e.g., its parent, there may be significant variations in the number of packets used to calculate LQI values for different nodes.
• a running average calculation may be used. An average LQI is determined over a predetermined time period by adding the LQI values for the packets received from a node during the time period and dividing the total by the number of received packets.
• a running average LQI is determined by a weighted sum of the average LQI and the previous running average LQI. In one example, the following is used:
• LQI(running average) (3 ⁇ 4 * LQI(previous running average)) + (1 ⁇ 4 * LQI(average)).
• the average LQI for the previous time period may be used.
• the time period used to calculate the average LQI may be independent of when the node communicates its advertised rank.
• Figure 6 illustrates an exemplary node that may be used in the network of Figure 1.
• a node can include a meter located at a premises, which measures the consumption of a resource such as gas, water, or electric power. Such a meter can be part of an RF network used for AMI.
• Other examples of nodes include a router, collector or collection point, host computer, hub, or other electronic device that is attached to a network and is capable of sending, receiving, or forwarding information over a
• a node 600 may include a network interface 604 and an antenna 602 so that it may communicate with like nodes and/or other devices in the mesh network.
• a network interface 604 may communicate with the microcontroller 612 using the bus 606, which enables the node to function like a computer, carrying out computer and command functions to provide implementations of the present invention described herein.
• the microcontroller 612 may include a processor 608 and a storage medium, such as memory 610.
• the processor 608 can be any suitable processing device or group of devices configured to execute instructions stored in the memory.
• the memory 610 can be any suitable non-transitory computer-readable medium for providing computer-executable program instructions, such as firmware or other operating instructions for controlling one or more hardware components of a node, to the processor 608.
• Examples of a non-transitory computer-readable medium may include, but are not limited to, a memory chip, ROM, RAM, an ASIC, a configured processor, electronic storage, optical storage, magnetic storage, or other storage device from which a computer processor can read instructions.
• the instructions may include processor-specific instructions generated by a compiler and/or an interpreter from code written in any suitable computer- programming language.
• the memory may be located internal to the node or accessible by the node via a network, for example.
• a node may also include a crystal oscillator (i.e.
• a clock to provide timekeeping and an energy storage device (i.e. a battery) to provide backup power.
• Some nodes may be powered only by an energy storage device.
• Some nodes include additional components, such as a utility meter that measures the consumption of a resource or service.

Landscapes

• Engineering & Computer Science (AREA)
• Computer Networks & Wireless Communication (AREA)
• Signal Processing (AREA)
• Mobile Radio Communication Systems (AREA)
• Data Exchanges In Wide-Area Networks (AREA)

Priority Applications (2)

Applications Claiming Priority (2)

Publications (1)

Family

ID=57068214

Family Applications (1)

Country Status (4)

Cited By (1)

Families Citing this family (12)

Family Cites Families (7)

• 2016
  • 2016-08-31 US US15/252,514 patent/US10230630B2/en active Active
  • 2016-09-19 JP JP2018517221A patent/JP6670930B2/ja active Active
  • 2016-09-19 WO PCT/US2016/052444 patent/WO2017058547A1/en not_active Ceased
  • 2016-09-19 CA CA2997424A patent/CA2997424C/en active Active

Non-Patent Citations (4)

Also Published As

Similar Documents

Legal Events