CN107846706B - Congestion-avoiding code-aware wireless mesh network multi-path routing method - Google Patents

Congestion-avoiding code-aware wireless mesh network multi-path routing method Download PDF

Info

Publication number
CN107846706B
CN107846706B CN201711339016.5A CN201711339016A CN107846706B CN 107846706 B CN107846706 B CN 107846706B CN 201711339016 A CN201711339016 A CN 201711339016A CN 107846706 B CN107846706 B CN 107846706B
Authority
CN
China
Prior art keywords
node
path
rreq
ack
paths
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201711339016.5A
Other languages
Chinese (zh)
Other versions
CN107846706A (en
Inventor
姚玉坤
李娟�
张毅
李小勇
刘江兵
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chongqing University of Post and Telecommunications
Original Assignee
Chongqing University of Post and Telecommunications
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Chongqing University of Post and Telecommunications filed Critical Chongqing University of Post and Telecommunications
Priority to CN201711339016.5A priority Critical patent/CN107846706B/en
Publication of CN107846706A publication Critical patent/CN107846706A/en
Application granted granted Critical
Publication of CN107846706B publication Critical patent/CN107846706B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/02—Traffic management, e.g. flow control or congestion control
    • H04W28/0289—Congestion control
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/02—Traffic management, e.g. flow control or congestion control
    • H04W28/08—Load balancing or load distribution
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/02—Traffic management, e.g. flow control or congestion control
    • H04W28/10—Flow control between communication endpoints
    • 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
    • H04W40/04—Communication route or path selection, e.g. power-based or shortest path routing based on wireless node resources

Landscapes

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

Abstract

The invention provides a congestion-avoiding code-aware multi-path routing method for a wireless mesh network, and belongs to the technical field of wireless communication. The method comprises the operation in two stages of routing and data transmission during the operation, two new mechanisms of a congestion sensing mechanism and a multi-path dynamic flow distribution mechanism are adopted, in the process of routing discovery, a node with a node cache occupancy rate exceeding a threshold value is used as a high-load node, the high-load node does not need to forward a newly received routing request message and simultaneously reduces the forwarding of the routing request message by a neighbor node of the node, the path with the high-load node in a plurality of paths from a source node to a destination node is eliminated in advance, and the routing overhead is effectively reduced; in the data transmission process, the source node can adjust the flow distribution proportion according to the load condition of the nodes on each path, thereby avoiding network congestion and reducing end-to-end time delay.

Description

Congestion-avoiding code-aware wireless mesh network multi-path routing method
Technical Field
The invention belongs to the technical field of wireless communication, and relates to a congestion-avoiding code-aware multi-path routing method for a wireless mesh network.
Background
Wireless Mesh Networks (WMNs) are a multipoint-to-multipoint distributed network, and are considered as a new solution to the bottleneck problem of Wireless access in the last mile. The wireless Mesh network is used as an efficient and reliable user access solution, and by virtue of the advantages of low cost, high reliability, self-organization and the like, rapid development and deep research are achieved. The WMN has the advantages of high bandwidth, flexible networking mode, non-line-of-sight transmission, strong robustness and the like. However, in the current practical application of the wireless Mesh network, some problems still remain to be solved. Due to the wireless medium characteristics of WMNs, WMNs are affected by random wireless interference, severe channel fading, etc., which may cause problems such as severe packet loss of the network, increased end-to-end delay, decreased network throughput, and decreased network performance. Ahlswede et al put forward the network coding theory for the first time in 2000, and network coding is an information exchange technology integrating routing and coding, so that the role that nodes in a traditional network only serve as data storage and forwarding is changed, and the traditional data processing mode in the network is broken through essentially. The network coding has the advantages of improving the network throughput, enhancing the network robustness, improving the network security, reducing the transmission delay, saving the node energy consumption and the like. The network coding is applied to the wireless Mesh network, so that the performance of the wireless Mesh network can be effectively improved.
The proposal of the realization scheme of the network coding among the COPE streams is widely researched by a wireless Mesh network coding technology in a unicast communication mode. However, COPE has the following two limitations: whether there is a coding opportunity in the COPE depends on the route that has already been established; in COPE the coding structure is limited to the two-hop region. However, since the number of coding opportunities in a practical network is usually quite limited, a high performance gain cannot be achieved by merely passively utilizing the coding opportunities. Therefore, routing algorithms must be able to discover and create more coding opportunities. BinNi et al first defined the expected number of transmissions by which two nodes successfully exchanged information for encoded transmissions at an intermediate node. And accordingly, an optimal route of coding perception is established. The code perception is to unify the two processes of 'route discovery' and 'code opportunity discovery', namely, to discover the network code opportunity in the process of route discovery. The core idea of code-aware routing is to select a route that is favorable for network coding, and it can actively create coding opportunities by constructing a specific coding structure at the route establishment stage.
However, in order to obtain more coding opportunities, the coding-aware routing algorithm has the problems that the network load is unbalanced and network congestion occurs due to the fact that data streams are gathered in regions. The multipath routing algorithm can fully utilize network resources, improve the fault tolerance of the network and is beneficial to realizing the load balance of the network. Therefore, the combination of coding awareness and multi-path routing can effectively improve the network performance.
The problem of uneven network load can be alleviated by multi-path routing based on coding perception, but the following problems exist in a multi-path routing algorithm based on coding perception through intensive research:
1. there is redundancy of routing messages resulting in a large routing overhead. Since data packet queuing delays are large at high-load nodes and the probability of network congestion occurring is also high, network congestion will severely impact network performance once it occurs. Therefore, the source node should not have a high-load node in the optimal path, and the high-load node still needs to forward the newly received route request and the corresponding route reply message in the route discovery process. The path where there is a high-load node cannot be used as the optimal path, and the high-load node does not need to forward a newly received routing request, which may increase the routing overhead.
2. The flow rate distribution ratio cannot be adjusted according to the load condition of the nodes on the path. The flow distribution proportion of each path is fixed, the problem of constant change of node load is ignored, the flow distribution proportion of the path cannot be reduced when the packet loss rate and the end-to-end delay increase due to congestion of a high-load node in the path in the data transmission process, and the network performance is seriously influenced.
Disclosure of Invention
In order to solve the problems of the existing multi-path routing algorithm based on coding perception, the invention provides a coding perception wireless mesh network multi-path routing method for avoiding congestion.
The coding perception wireless mesh network multi-path routing method for congestion avoidance provided by the invention comprises two new mechanisms of a congestion perception mechanism and a multi-path dynamic flow distribution mechanism. In the process of route discovery, a node with a node cache occupancy rate exceeding a threshold value is used as a high-load node, the high-load node does not need to forward a newly received route request message and simultaneously reduces the forwarding of neighbor nodes of the node to the route request message, and the route with the high-load node in a plurality of routes from a source node to a destination node is eliminated in advance, so that the route overhead is effectively reduced; in the data transmission process, the source node can adjust the flow distribution proportion according to the node load condition of each path, thereby avoiding network congestion and reducing end-to-end time delay.
First, the basic idea and main operation of the new mechanism proposed by the present invention
The basic ideas and main operations of the two new mechanisms of the congestion-aware mechanism and the multi-path dynamic traffic allocation mechanism proposed by the present invention are specifically described below.
1. Congestion awareness mechanism
According to the design method of the encoding perception multi-path routing algorithm in the existing wireless mesh network, when a source node has data to send, a routing request message RREQ is broadcasted, the node receives a new routing request message and forwards the new routing request message until the routing request message reaches a destination node or the time-to-live TTL of the routing request message becomes 0, the destination node receives the routing request message and then unicasts a routing reply message RREP along the original path, and the source node selects a plurality of RREPs and then sends the data from the optimal path. If the source node selects a path with a high load node, the data packet queuing delay is large at the high load node and the probability of network congestion is high, so that the path with the high load node cannot be used as the optimal path for sending data, and the high load node still needs to forward a newly received route request message and a route reply message in the route request process, so that the route overhead is increased.
In order to solve the above problems, the present invention proposes a new mechanism of "congestion awareness mechanism", and the basic idea of the new mechanism is: in the process of route discovery, after receiving a new route request message, the node judges the self load condition through the cache occupancy rate, if the cache occupancy rate of the node exceeds the threshold value, the node load is higher, so that the probability of network congestion of the node is higher, and the newly received route request message is discarded. Because the node does not forward the route request message, the node in the communication range of the node can not receive the route request message and does not need to forward the route request message, and meanwhile, the destination node can not receive the route request message transmitted through the path to which the node belongs, and does not need to send a route reply message to the source node along the path to which the node belongs, the route overhead is obviously reduced. In order to achieve the purpose, the invention provides the following technical scheme:
the new mechanism of the congestion sensing mechanism provided by the invention mainly comprises the following steps:
s1: if a node in the network needs to send a data packet, the node broadcasts and sends a route request message RREQ, and a neighbor node address field needs to be added in the route request message in order to detect coding opportunities existing in all nodes in a path.
The specific information added in the route request message RREQ: and adding a neighbor node address field for storing the neighbor node ID of the node.
S2: if the node receives the route request message RREQ, firstly judging whether the route request message is received, if not, calculating the buffer occupancy rate of the node; otherwise, the route request message is discarded.
The specific calculation formula of the node cache occupancy rate O is as follows:
Figure BDA0001507958400000031
wherein q (c) is the length of the queue in node c, and q (c) is the total length of the queue in node c.
S3: if the cache occupancy is below a threshold TmaxIf yes, updating a route request message RREQ and adding the neighbor node ID of the node to a neighbor node address domain in the RREQ; otherwise, the newly received route request message is abandoned and discarded directly.
The specific method for updating the route request message RREQ comprises the following steps:
firstly, adding 1 to a Hop Count field in an RREQ, and subtracting 1 from a survival time TTL field; the neighbor node ID is then added to the neighbor node address field in the RREQ.
S4: and if the node needs to send the updated route request message RREQ, judging whether a coding opportunity exists according to all the neighbor node IDs added by the RREQ, calculating coding gain, recording the codeable data stream and the coding gain, and sending the updated RREQ.
The specific calculation method of the coding gain of the node comprises the following steps:
firstly, judging whether the node has a coding opportunity, if not, the coding gain G is 0; otherwise, calculating the coding gain G of the coding node, wherein the specific calculation formula is as follows:
Figure BDA0001507958400000042
wherein G iscDenotes the coding gain, p, at node cc,iFor the link l from node c to node ic,iThe two-way delivery rate of (c).
2. Multi-path dynamic traffic allocation mechanism
According to the design method of the coding sensing multi-path routing algorithm in the existing wireless mesh network, after a source node receives a plurality of routing reply messages RREP transmitted along different paths, the routing metric values of all the paths are calculated respectively, m paths with the optimal performance are selected, then the flow distribution proportion of all the paths is calculated, and the flow distribution proportion of all the paths is kept unchanged in the data transmission process. The load of the nodes in the network is constantly changing, so that a congested node may appear in the m paths selected by the source node for transmitting data. In this case, if the flow allocation ratio is not adjusted in time to continue transmitting data packets to the path, the packet loss rate increases, the end-to-end delay increases, and the network throughput decreases.
In order to solve the above problems, the present invention proposes a new mechanism of "multipath dynamic traffic allocation mechanism", and the basic idea of the new mechanism is: after the destination node successfully receives the data packet, an ACK confirmation message is sent to the source node, a flag field is added in the ACK confirmation message, when the node needs to forward the ACK, the load condition is judged according to the cache occupancy rate, if the load is high, the value of the flag field in the ACK is modified, and after the source node receives the ACK, if the value of the flag field in the ACK is modified, the residual flow on the path is distributed to other paths according to the proportion.
In order to achieve the purpose, the invention provides the following technical scheme:
the new mechanism of the multipath dynamic flow distribution mechanism provided by the invention mainly comprises the following steps:
s1: if the destination node receives the route request message, the route reply message RREP is unicast reversely along the transmission path of the route request message, and a Metric field is added in the route reply message RREP in order to enable the source node to successfully calculate the route Metric value of each path.
The specific information added in the route reply message RREP: and adding a Metric field into the route reply message RREP, and setting the initial value of the Metric field to 0 for storing the route Metric value of the path.
S2: if the intermediate node receives the route reply message RREP, the route Metric value at the node is calculated first, and the route Metric value is added to a Metric domain in the RREP and then the RREP is forwarded.
The specific calculation formula of the routing metric CACM at the node is as follows:
Figure BDA0001507958400000041
wherein G isCIs a coding gain; the ETT is the expected transmission time, which is the time required for successful transmission of one data packet on the link.
S3: if the source node receives M RREPs, firstly, the routing metric values of all paths are respectively calculated, the optimal M paths are selected, secondly, the original flow distribution proportion of all paths is calculated according to a flow distribution formula, and the data grouping starts to be sent.
The specific method for selecting the m optimal paths comprises the following steps:
firstly, respectively calculating the route metric value of the whole CACMp path of the M paths, wherein the calculation formula of the route metric value of the CACMp path is as follows:
Figure BDA0001507958400000051
route metric CACM of pathPIs the sum of the routing metrics CACM at all nodes on the path.
And secondly, selecting the optimal M paths from the M paths. Under the condition that the link quality is the same, the path with a small CACM value has more coding opportunities; in the case where the coding gain is the same, the load and interference on the path having a small CACM value are small. Therefore, the path with the smaller routing metric CACM is preferentially selected. And M paths with small routing metric CACM are selected from the M paths to be used as data transmission paths.
And finally, calculating the original flow distribution proportion of each path, and sending data packets according to the original flow distribution proportion, wherein the original flow distribution proportion of the kth path is as follows:
Figure BDA0001507958400000052
wherein the content of the first and second substances,
Figure BDA0001507958400000053
the route metric value for the k-th path.
S4: and if the destination node receives the data packet sent by the source node, sending ACK to the source node, adding a flag field in the ACK, and feeding back the load condition of the nodes on the path to the source node.
The specific method for feeding back the node load condition in the path to the source node by using the ACK comprises the following steps: and modifying the format of the ACK acknowledgement packet, adding a flag field in the ACK acknowledgement packet, and setting the initial value of the flag field to be 0. When a node with a higher congestion probability appears in the path, the value of the flag field is modified.
S5: if the intermediate node receives the ACK, checking the value of a flag field in the ACK, and if the value of the flag field in the ACK is 1, directly forwarding the ACK; otherwise, judging the load condition according to the cache occupancy rate, if the cache occupancy rate exceeds the threshold value, changing the value of a flag field in the ACK, and continuously forwarding the ACK; otherwise, the ACK is forwarded directly.
The specific method for judging the congestion probability comprises the following steps:
firstly, calculating the cache occupancy rate O of a node:
Figure BDA0001507958400000061
wherein q (c) is the length of the queue in node c, and q (c) is the total length of the queue in node c.
Secondly, comparing the node cache occupancy rate O with a threshold value TmaxThe size of (2). If O is less than threshold value TmaxIf so, the node load is lower and the congestion probability is lower; if O is greater than threshold value TmaxThen the node load is higher and the congestion probability is higher.
And finally, determining whether to modify the value of the flag field in the ACK according to the node load condition. If the node load is higher, the value of the flag field is modified to 1; otherwise, the ACK is sent to the previous-hop node without modification.
S6: if the source node receives the ACK, the value of the flag field is checked. Judging whether the flow distribution proportion needs to be adjusted or not according to the value of the flag field of the ACK, and if the flow distribution proportion needs to be adjusted, redistributing according to an adjustment strategy; otherwise, continuing to send data according to the original proportion.
The specific method of the flow regulation strategy is as follows:
firstly, the source node checks the value of a flag field in the received ACK, if the value of the flag field is 0, the fact that no high-load node exists in the path is indicated, and the flow distribution proportion does not need to be adjusted; if the value of the flag field is 1, it indicates that a high-load node occurs in the path, and the remaining traffic of the path needs to be proportionally allocated to the remaining m-1 paths, where the allocation proportion formula is as follows:
Figure BDA0001507958400000062
wherein m-1 is the rest of the paths after removing the path with the node with higher congestion probability from the m paths for sending data selected by the source node.
The invention has the beneficial effects that: the method uses a congestion sensing mechanism in the route discovery process to remove node paths with higher congestion probability in advance, thereby effectively reducing the route overhead; meanwhile, a multi-path dynamic flow distribution mechanism is adopted, and the flow distribution proportion is adjusted by the source node according to the congestion probability condition of the nodes in the path, so that the occurrence of network congestion can be avoided, the end-to-end time delay and the packet loss rate are reduced, and the network throughput is improved.
Drawings
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described in detail with reference to the accompanying drawings, in which:
FIG. 1 is a schematic diagram of a wireless mesh network model;
in the wireless Mesh network, network nodes can be classified into three categories according to their performance and their role. The first type is a gateway, the gateway is a Mesh router with a protocol conversion function, and the Mesh router allows interconnection and intercommunication between a Mesh network and other external networks (such as the Internet); the second type is a Mesh router, which is a backbone node of a Mesh network and provides access service for a Mesh client; the third type is a Mesh client, which communicates directly with a Mesh router providing access services for it. Compared with the traditional wireless network, the wireless Mesh network has the following four advantages: low deployment cost, good compatibility, self-organization and self-configuration, and high reliability.
FIG. 2 is a schematic diagram of a network topology for inter-stream network coding;
the coding perception refers to detecting coding opportunities existing in the network in the process of route discovery, fully utilizing the potential coding opportunities in the network and further improving the network throughput. In the network model exemplified in fig. 2, it is assumed that the source node a sends the packet p along the path a → b → c → h1To destination node h, source node d sends packet p along path d → c → i → j2To the destination node j. In the conventional data transmission mode, a data packet p is completed1Successful delivery, p, along b → c → h2Successful delivery along d → c → i requires a total of 4 transmissions. Transmitting a data packet p at a source node a according to the broadcast characteristics of the wireless medium1Nodes b and j then receive and cache p1Node b continues forwarding packet p1When node c receives and caches p1(ii) a Sending a data packet p at a source node d2Nodes c and h then receive and cache p2When node c receives and caches p1. Successful receipt of packet p at node c1And p2Judging whether the node c has coding opportunity according to coding conditions, and broadcasting and sending coded packets by the node c after coding
Figure BDA0001507958400000071
Node h receives
Figure BDA0001507958400000072
Then store it with the cached p2To carry out
Figure BDA0001507958400000073
Successfully obtain the original data packet p1Similarly, when the node i will encode the packet
Figure BDA0001507958400000074
To j, node j will forward
Figure BDA0001507958400000075
With its cached p1To carry out
Figure BDA0001507958400000076
Can successfully obtain the original data packet p2. And, only 3 transmissions are needed to complete packet p in this process1Successful delivery, p, along b → c → h2Successful delivery along d → c → i results in a significant improvement in the transmission efficiency of the data packet.
FIG. 3 is a flow diagram of a congestion awareness mechanism;
the congestion awareness mechanism is mainly used in the route discovery process and is used for eliminating paths with high load nodes in advance. After the intermediate node receives the route request message RREQ broadcast by the source node, firstly calculating the self cache occupancy rate, judging the node load condition according to the cache occupancy rate, and if the cache occupancy rate is lower than a threshold value TmaxIf yes, updating a route request message RREQ and adding the neighbor node ID of the node into the RREQ; otherwise, the newly received route request message is abandoned and discarded directly. Before the node sends the updated route request message RREQ, judging whether a coding opportunity exists according to all neighbor node IDs added by the RREQ, and if the coding opportunity exists, recording the codeable data stream and sending the updated RREQ; otherwise, the updated RREQ is directly sent.
Fig. 4 is a flow chart of a dynamic traffic allocation mechanism.
In order to avoid the occurrence of network congestion during data transmission, a dynamic traffic allocation mechanism is used. When the node receives the route reply message RREP, the route metric is calculated first, and then the route metric value is added into the RREP to be forwarded to the previous hop node. When a source node receives M RREPs from different paths, firstly, the routing metric value of each path is calculated, M optimal paths with small routing metric values are selected, secondly, the original flow distribution proportion of each path is calculated according to a flow distribution formula, and data grouping starts to be sent. And when the destination node successfully receives the data packet sent by the source node, sending ACK to the source node, and adding a flag field with an initial value of 0 in the ACK. When the intermediate node receives the ACK, judging the congestion probability according to the cache occupancy rate; if the congestion probability is higher, the value of a flag field in the ACK is modified to 1, and the ACK is continuously forwarded; otherwise, the ACK is forwarded directly. After the source node receives the ACK, the value of the flag field is checked. Judging whether the flow distribution proportion needs to be adjusted or not according to the value of the flag field of the ACK, if the value of the flag field is 1, the flow distribution proportion needs to be adjusted, and reallocating according to an adjustment strategy; otherwise, continuing to send data according to the original proportion.
Detailed Description
For more detailed description, the coding-aware wireless mesh network multipath routing method for congestion avoidance according to the present invention is described as follows, with reference to the accompanying drawings:
a congestion-avoiding coding-aware multi-path routing method for a wireless mesh network is disclosed, which uses a congestion-aware mechanism in the route discovery process to remove node paths with higher congestion probability in advance, thereby effectively reducing the routing overhead; meanwhile, a multi-path dynamic flow distribution mechanism is adopted, and the source node adjusts the flow distribution proportion according to the congestion probability condition of the nodes in the path, so that the network congestion can be avoided. The specific method comprises the following steps:
s1: if a node in the network needs to send a data packet, the node broadcasts and sends a route request message RREQ, and a neighbor node address field needs to be added in the route request message in order to detect coding opportunities existing in all nodes in a path.
The specific information added in the route request message RREQ: and adding a neighbor node address field for storing the neighbor node ID of the node.
S2: if the node receives the route request message RREQ, firstly judging whether the route request message is received, if not, calculating the buffer occupancy rate of the node; otherwise, the route request message is discarded.
The specific calculation formula of the node cache occupancy rate O is as follows:
Figure BDA0001507958400000081
wherein q (c) is the length of the queue in node c, and q (c) is the total length of the queue in node c.
S3: if the cache occupancy is below a threshold TmaxIf yes, updating a route request message RREQ and adding the neighbor node ID of the node to a neighbor node address domain in the RREQ; otherwise, the newly received route request message is abandoned and discarded directly.
The specific method for updating the route request message RREQ comprises the following steps:
firstly, adding 1 to a Hop Count field in an RREQ, and subtracting 1 from a survival time TTL field; the neighbor node ID is then added to the neighbor node address field in the RREQ.
S4: and if the node needs to send the updated route request message RREQ, judging whether a coding opportunity exists according to all the neighbor node IDs added by the RREQ, calculating coding gain, recording the codeable data stream and the coding gain, and sending the updated RREQ.
The specific calculation method of the coding gain of the node comprises the following steps:
firstly, judging whether the node has a coding opportunity, if not, the coding gain G is 0; otherwise, calculating the coding gain G of the coding node, wherein the specific calculation formula is as follows:
Figure BDA0001507958400000091
wherein G iscDenotes the coding gain, p, at node cc,iFor the link l from node c to node ic,iThe two-way delivery rate of (c).
S5: if the destination node receives the route request message, the route reply message RREP is unicast reversely along the transmission path of the route request message, and a Metric field is added in the route reply message RREP in order to enable the source node to successfully calculate the route Metric value of each path.
The specific information added in the route reply message RREP: and adding a Metric field into the route reply message RREP, and setting the initial value of the Metric field to 0 for storing the route Metric value of the path.
S6: if the node receives the route reply message RREP, the route Metric value at the node is calculated first, and the route Metric value is accumulated to a Metric domain in the RREP and then forwarded.
The specific calculation formula of the routing metric CACM at the node is as follows:
Figure BDA0001507958400000092
wherein G isCIs a coding gain; the ETT is the expected transmission time, which is the time required for successful transmission of one data packet on the link.
The routing metric CACM of said pathPThe specific calculation formula of (A) is as follows:
Figure BDA0001507958400000093
route metric CACM of pathPIs the sum of the routing metrics CACM at all nodes on the path.
S7: if the source node receives M RREPs, firstly, the routing metric values of all paths are compared, the optimal M paths are selected, secondly, the original flow distribution proportion of all paths is calculated according to a flow distribution formula, and data grouping starts to be sent.
The specific method for selecting the m optimal paths comprises the following steps:
first, the magnitude of the route metric values of the M paths are compared. And secondly, selecting the optimal M paths from the M paths. Under the condition that the link quality is the same, the path with a small CACM value has more coding opportunities; in the case where the coding gain is the same, the load and interference on the path having a small CACM value are small. Therefore, the path with the smaller routing metric CACM is preferentially selected. And M paths with small routing metric CACM are selected from the M paths to be used as data transmission paths. And finally, calculating the original flow distribution proportion of each path, and sending data packets according to the original flow distribution proportion, wherein the original flow distribution proportion of the kth path is as follows:
Figure BDA0001507958400000101
wherein the content of the first and second substances,
Figure BDA0001507958400000102
the route metric value for the k-th path.
S8: and if the destination node receives the data packet sent by the source node, sending ACK to the source node, adding a flag field in the ACK, and feeding back the load condition of the nodes on the path to the source node.
The specific method for feeding back the node load condition in the path to the source node by using the ACK comprises the following steps: and modifying the format of the ACK acknowledgement packet, adding a flag field in the ACK acknowledgement packet, and setting the initial value of the flag field to be 0. When a node with a higher congestion probability appears in the path, the value of the flag field is modified.
S5: if the intermediate node receives the ACK, checking the value of a flag field in the ACK, and if the value of the flag field in the ACK is 1, directly forwarding the ACK; otherwise, judging the load condition according to the cache occupancy rate, if the cache occupancy rate exceeds the threshold value, changing the value of a flag field in the ACK, and continuously forwarding the ACK; otherwise, the ACK is forwarded directly.
The specific method for judging the congestion probability comprises the following steps:
firstly, calculating the cache occupancy rate O of a node:
Figure BDA0001507958400000103
wherein q (c) is the length of the queue in node c, and q (c) is the total length of the queue in node c.
Secondly, comparing the node cache occupancy rate O with a threshold value TmaxThe size of (2). If O is less than threshold value TmaxIf so, the node load is lower and the congestion probability is lower; if O is greater than threshold value TmaxThen the node load is higher and the congestion probability is higher.
And finally, determining whether to modify the value of the flag field in the ACK according to the node load condition. If the node load is higher, the value of the flag field is modified to 1; otherwise, the ACK is sent to the previous-hop node without modification.
S10: if the source node receives the ACK, the value of the flag field is checked. Judging whether the flow distribution proportion needs to be adjusted or not according to the value of the flag field of the ACK, and if the flow distribution proportion needs to be adjusted, redistributing according to an adjustment strategy; otherwise, continuing to send data according to the original proportion.
The specific method of the flow regulation strategy is as follows:
firstly, the source node checks the value of a flag field in the received ACK, if the value of the flag field is 0, the fact that no high-load node exists in the path is indicated, and the flow distribution proportion does not need to be adjusted; if the value of the flag field is 1, it indicates that a high-load node occurs in the path, and the remaining traffic of the path needs to be proportionally allocated to the remaining m-1 paths, where the allocation proportion formula is as follows:
Figure BDA0001507958400000111
wherein m-1 is the rest of the paths after removing the path with the node with higher congestion probability from the m paths for sending data selected by the source node.

Claims (1)

1. A coding-aware wireless mesh network multi-path routing method for congestion avoidance is characterized in that:
during the operation, the method comprises the operation of two stages of routing and data transmission, and two new mechanisms of a congestion perception mechanism and a multipath dynamic flow distribution mechanism are respectively adopted;
in the process of route discovery, a node of which the node cache occupancy rate exceeds a threshold value is used as a high-load node, the high-load node does not need to forward a newly received route request message RREQ, meanwhile, the forwarding of a neighbor node of the high-load node to the RREQ is reduced, and paths of the high-load node existing in a plurality of paths from a source node to a destination node are eliminated in advance;
in the data transmission process, the source node can adjust the flow distribution proportion according to the load condition of the nodes on each path;
the congestion awareness mechanism comprises the following steps:
s1: if the source node in the network needs to send a data packet, the source node broadcasts a RREQ, the purpose of the RREQ is to detect coding opportunities existing in all nodes in a path, the RREQ comprises a neighbor node address domain, and the neighbor node address domain is used for storing a neighbor node ID of the source node;
s2: if the neighbor node receives the RREQ, firstly judging whether the RREQ is received or not, and if not, calculating the buffer occupancy rate of the neighbor node; otherwise, discarding the RREQ;
s3: if the cache occupancy is below a threshold TmaxIf yes, the RREQ is updated, and the neighbor node ID of the neighbor node is added to the neighbor node address field in the RREQ(ii) a Otherwise, the newly received RREQ is abandoned and forwarded, and the RREQ is directly discarded;
s4: if the neighbor node needs to send the updated RREQ, judging whether a coding opportunity exists according to all neighbor node IDs added by the RREQ, calculating coding gain, recording the codeable data stream and the coding gain, and sending the updated RREQ; wherein the coding gain of the neighbor node is specifically calculated as:
judging whether the neighbor node has a coding opportunity, if not, the coding gain G is 0; otherwise, calculating the coding gain G of the neighbor node, wherein the specific calculation formula is as follows:
Figure FDA0002759462600000011
wherein G iscDenotes the coding gain, p, at node cc,iFor the link l from node c to node ic,iThe two-way delivery rate of (d);
the multi-path dynamic flow distribution mechanism comprises the following steps:
s1: if the destination node receives the RREQ, starting to reversely unicast a route reply message RREP along the transmission path of the RREQ, wherein the RREP comprises a Metric domain, and the initial value of the Metric domain is set to be 0 and used for storing the route Metric value of the path, so that the source node can successfully calculate the route Metric value of each path; wherein, the calculation formula of the route metric value is as follows:
Figure FDA0002759462600000012
wherein G isCIs a coding gain; ETT is the expected transmission time, which is the time required to successfully transmit one data packet on the link;
s3: if the source node receives M RREPs, firstly, the routing metric values of all paths are respectively calculated, and the optimal M paths are selected; secondly, performing a first step; calculating the original flow distribution proportion of each path according to a flow distribution formula, and starting to send data packets;
the specific method for selecting the m optimal paths comprises the following steps: firstly, respectively calculating a routing metric value CACMp of each path in M paths, wherein the calculation formula of the CACMp is as follows:
Figure FDA0002759462600000021
wherein, the CACMPThe sum of the CACM at all nodes on each path;
secondly, selecting the optimal M paths from the M paths, and preferentially selecting the path with a smaller routing metric value CACM;
and finally, calculating the original flow distribution proportion of each path, and sending data packets according to the original flow distribution proportion, wherein the original flow distribution proportion of the kth path is as follows:
Figure FDA0002759462600000022
wherein the content of the first and second substances,
Figure FDA0002759462600000023
a route metric value for the kth path;
s4: if the destination node receives the data packet sent by the source node, sending ACK to the source node, adding a flag field of a flag bit in the ACK, if a high-load node appears in a path, changing the value of the flag field to 1 by the high-load node, and feeding back the load condition of the node on the path to the source node;
s5: if the intermediate node receives the ACK, checking the value of a flag field in the ACK, and if the value of the flag field in the ACK is 1, directly forwarding the ACK; otherwise, judging the load condition according to the cache occupancy rate; if the cache occupancy rate exceeds a threshold value, changing the value of a flag field in the ACK, and continuously forwarding the ACK; otherwise, directly forwarding the ACK;
s6: if the source node receives the ACK, checking a value of a flag field; judging whether the flow distribution proportion needs to be adjusted or not according to the value of the flag field of the ACK, and if the flow distribution proportion needs to be adjusted, redistributing according to an adjustment strategy; otherwise, continuing to send data according to the original proportion;
the adjustment strategy specifically comprises the following steps:
and proportionally distributing the residual flow of the path to the rest m-1 paths, wherein the distribution proportion has the following calculation formula:
Figure FDA0002759462600000024
wherein m-1 is the rest paths of the m paths for sending data selected by the source node, except the path of the node with the highest congestion probability.
CN201711339016.5A 2017-12-14 2017-12-14 Congestion-avoiding code-aware wireless mesh network multi-path routing method Active CN107846706B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711339016.5A CN107846706B (en) 2017-12-14 2017-12-14 Congestion-avoiding code-aware wireless mesh network multi-path routing method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711339016.5A CN107846706B (en) 2017-12-14 2017-12-14 Congestion-avoiding code-aware wireless mesh network multi-path routing method

Publications (2)

Publication Number Publication Date
CN107846706A CN107846706A (en) 2018-03-27
CN107846706B true CN107846706B (en) 2021-02-05

Family

ID=61665056

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711339016.5A Active CN107846706B (en) 2017-12-14 2017-12-14 Congestion-avoiding code-aware wireless mesh network multi-path routing method

Country Status (1)

Country Link
CN (1) CN107846706B (en)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111263419B (en) * 2020-01-17 2021-12-28 西安交通大学 Unmanned aerial vehicle-based dynamic routing method for stereo heterogeneous network in emergency scene
CN112003787B (en) * 2020-08-14 2022-09-09 北京灵汐科技有限公司 Routing path determining method, device, control equipment and storage medium
CN111988827B (en) * 2020-08-28 2022-07-29 苏州永联芯电子有限公司 Route optimization method and system of Bluetooth mesh network with mobile node
CN112566208B (en) * 2020-11-30 2022-05-03 浙江大学 Intelligent power grid wireless data communication path searching method
CN112822725B (en) * 2020-12-30 2023-03-31 国网甘肃省电力公司信息通信公司 Wireless heterogeneous network multilink data distribution method based on service priority
CN112954719B (en) * 2021-03-25 2023-10-13 盐城工学院 Traffic matching method for network coding perception route in wireless multi-hop network
CN114079969B (en) * 2021-08-16 2023-11-28 珠海市杰理科技股份有限公司 Data transmission method and device, readable storage medium and node equipment

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101415248A (en) * 2008-11-21 2009-04-22 中山大学 Method for establishing layer-striding dynamic source route protocol based on load balance
CN104320811A (en) * 2014-09-18 2015-01-28 南京邮电大学 Routing method for traffic distribution by using node information transmission capacity
CN105897605A (en) * 2016-04-08 2016-08-24 重庆邮电大学 IPv6-based power line carrier communication network congestion control method

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110225312A1 (en) * 2010-03-10 2011-09-15 Thomson Licensing Unified cache and peer-to-peer method and apparatus for streaming media in wireless mesh networks
CN102223671B (en) * 2010-04-15 2014-02-19 华为技术有限公司 Method and communication equipment for data transmission in wireless multihop network

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101415248A (en) * 2008-11-21 2009-04-22 中山大学 Method for establishing layer-striding dynamic source route protocol based on load balance
CN104320811A (en) * 2014-09-18 2015-01-28 南京邮电大学 Routing method for traffic distribution by using node information transmission capacity
CN105897605A (en) * 2016-04-08 2016-08-24 重庆邮电大学 IPv6-based power line carrier communication network congestion control method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
移动自组网负载自适应的多路径传输方法研究;周通;《中国学术文献网络出版总库》;20170110;第4.3.3节,第4.4.4节第2段 *

Also Published As

Publication number Publication date
CN107846706A (en) 2018-03-27

Similar Documents

Publication Publication Date Title
CN107846706B (en) Congestion-avoiding code-aware wireless mesh network multi-path routing method
CN110139319B (en) Routing method for minimizing transmission delay of high dynamic delay network
CN101674630B (en) Implementation method of cross-layer routing capable of perceiving congestion
CN111479306B (en) Q-learning-based flight ad hoc network QoS routing method
Chen et al. Congestion-aware routing protocol for mobile ad hoc networks
Guo et al. Analysis of general network coding conditions and design of a free-ride-oriented routing metric
Tekaya et al. Multipath routing mechanism with load balancing in ad hoc network
CN108449271B (en) Routing method for monitoring path node energy and queue length
CN104010343A (en) Wireless network optimization method for intelligent meter reading system
CN106850436B (en) Mine hybrid wireless mesh network routing method based on virtual potential energy field
Darabkh et al. An improved reactive routing protocol over mobile Ad-hoc networks
CN111132236B (en) Multi-unmanned aerial vehicle self-organizing network MPR node selection method based on improved OLSR protocol
Jhang et al. C2AR: coding and capacity aware routing for wireless ad hoc networks
CN103037436B (en) Network coding perception wireless mesh routing protocol based on hybrid wireless mesh protocol (HWMP)
Nikseresht et al. Interference-aware multipath routing for video delivery in wireless multimedia sensor networks
CN103108372A (en) Interference sensing cross-layer routing method based on node sending and receiving capacity
Wannawilai et al. AOMDV with sufficient bandwidth aware
Singal et al. LSMRP: Link stability based multicast routing protocol in MANETs
Jingrong et al. Opportunistic cooperative routing in multi-radio multi-channel wireless sensor networks
Nage et al. TCP-aware network coding with opportunistic scheduling in wireless mobile ad hoc networks
Su et al. Single phase admission control for QoS-routing protocol in ad hoc networks
Gu et al. Network coding-aware routing protocol in wireless mesh networks
Lu et al. Improved multi-path AODV protocols for real-time video transport over Mobile Ad Hoc Networks
Argyriou Network coding in IEEE 802.11 wireless LANs with an enhanced channel access scheme
EP2141865A1 (en) Method for determining a route for the transmission of a sequence of data packets and ad-hoc mesh network

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant