An Adaptive Multiple Access Method with Load Awareness in Sensor Networks

Abstract

The present invention discloses an adaptive multiple access method for load awareness in a sensor network. By carrying node queue information in data frames, nodes can obtain the resource requirement information of child nodes without using RST / CTS. By introducing a load factor to evaluate and predict the traffic load in the next beacon frame period, nodes can select the most suitable MAC protocol in real time. By introducing two switching thresholds, smooth switching of the node access mechanism is achieved. By grouping nodes according to node load and generating busy tones in the non-transmission phase of time slots, the parent node can quickly sense and timely schedule time-sensitive services. The nodes of the present invention can sense the network load status while predicting the network traffic in the next period of time, adjust the channel access mechanism, perform scheduling in advance, and prevent congestion. When congestion occurs, reasonable scheduling of child nodes is carried out, enabling them to transmit data efficiently and in a timely manner, improving link utilization and network stability.

Description

Technical Field

[0001] The present invention belongs to the field of communication technologies, and particularly relates to an adaptive multiple access method for load awareness in a sensor network. Background Art

[0002] In recent years, communication technologies have been continuously developing, the scale of networks has shown a trend of huge growth, and the functions of nodes are becoming more and more diverse. The compatibility and expandability of networks are getting stronger. For a network, when there is already an existing network, it is not worthwhile to re-establish a new network with the same coverage area. Instead, expanding on the basis of the existing network and adding new nodes will further enlarge the network. At the same time, nodes are becoming more and more intelligent now, and the labor cost is getting higher and higher. Frequent manual replacement of the power supply for nodes is unacceptable. Therefore, there is a greater need for a highly reliable, low-overhead, and low-power consumption channel access mechanism that can support more node communications.

[0003] South China University of Technology proposed a method and device for resource allocation in a wireless body area network based on a hybrid access mechanism in the patent document with the application number 202210275248.3. To reduce the competition conflicts between nodes, improve the utilization rate of channel resources, and maximize the throughput of node transmissions, a channel resource allocation scheme for a wireless body area network based on a hybrid access mechanism was proposed. The scheme includes the following steps: (1) Divide the priorities of nodes according to data characteristics, and divide the types of nodes according to the priorities, where the types of nodes include emergency nodes and ordinary nodes; (2) Determine whether to send new data or old data according to the change of data caching; (3) Perform backoff-differentiated channel competition according to the priorities of nodes; (4) Divide the superframe competition stage adaptively according to the channel state; (5) Determine the number of node time slots in the non-competition stage according to the transmission utility of nodes; (6) Determine the node transmission order in the non-competition stage according to the transmission priorities of nodes.

[0004] However, the defects of this patent are as follows:

[0005] 1. Since this method obtains the channel to send data frames by means of backoff-differentiated channel competition, for large-scale networks, a large number of competition conflicts will occur.

[0006] 2. Although the transmission of time-sensitive services is considered in this method, nodes can only compete for the channel and send data in the initial stage of the competition phase, and since there is no node notifying its neighboring nodes, the transmission efficiency of time-sensitive services will be low, wasting time slot resources.

[0007] 3. Since this method adopts a passive processing method, if the lower-layer nodes make reservations and the upper-layer nodes process them, the nodes cannot actively detect congestion in the network, and a great price has to be paid to alleviate the already existing congestion. Summary of the Invention

[0008] To overcome the deficiencies of the above-mentioned existing technologies, the object of the present invention is to provide an adaptive multiple access method for load awareness in a sensor network. Nodes can sense the network load status and predict the network traffic in the next period of time, adjust the channel access mechanism, perform scheduling in advance to prevent congestion; when congestion occurs, reasonably schedule child nodes to enable them to transmit data efficiently and in a timely manner, and at the same time, when time-sensitive data appears, quickly sense and complete the transmission, improve the link utilization rate, and improve the network stability.

[0009] To achieve the above object, the technical solution adopted by the present invention is as follows:

[0010] By carrying node queue information in the data frame, nodes can obtain the resource requirement information of child nodes without using RST / CTS; by introducing a load factor to evaluate and predict the service load in the next beacon frame period, nodes can select the most suitable MAC protocol in real time; by introducing two switching thresholds, smooth switching of the node access mechanism is achieved; by grouping nodes according to node load and generating busy tones in the non-transmission stage of time slots, the parent node can quickly sense and timely schedule time-sensitive services.

[0011] An adaptive multiple access method for load awareness in a sensor network includes the following steps;

[0012] Step 1: The node counts the types and numbers of child nodes (routing nodes and sensing nodes) included in the node itself, and analyzes and calculates the switching threshold 1 for switching to the scheduling access mechanism according to the relevant theory of the Markov process, and then executes Step 2;

[0013] Step 2: Count the service load of the child nodes included in the node itself, determine the number of groups for time-sensitive service grouping and the maximum number of members in each group according to the node clock resolution, and then execute Step 3;

[0014] Step 3: At each node, sort its child nodes from largest to smallest according to the service load, and select a group that can meet the maximum delay allowed for time-sensitive services for each child node according to the sorting. Broadcast the grouping result to the child nodes through the beacon frame, and then execute Step 4;

[0015] Step 4: When the node generates time-sensitive services, if the node channel access mechanism is contention access, select an idle time slot and upload the time-sensitive data from the child node to the parent node in a timely manner; if the node channel access mechanism is parent node scheduling access, judge whether the next transmission time slot of the node can meet the delay requirement of the time-sensitive service. If it can be met, the node uploads the time-sensitive data from the child node to the parent node in the next transmission time slot. If it cannot be met, execute Step 5;

[0016] Among them, step 3 is the preliminary preparation work, and step 4 is how the node should handle events when a certain situation occurs;

[0017] Step 5: When the node generates time-sensitive data, a busy tone is generated at the corresponding moment of the group where the node is located. The parent node determines the node number that generates the time-sensitive data according to the generated time slot of the busy tone and the specific position where the busy tone moment is located, and constructs a command frame in the downlink time slot or carries scheduling information in the ACK frame, so that the node that generates the time-sensitive service uploads data in the next time slot;

[0018] Among them, the parent node groups all child nodes. When a child node generates time-sensitive data and the parent node requires the child node to schedule and upload data, when the upload time slot allocated by the parent node cannot meet the delay requirement of the time-sensitive service, the child node finds the corresponding position in the time slot corresponding to the group allocated by the parent node to generate a channel busy tone, and the parent node determines which node generates the time-sensitive data according to the generated time slot and position of the busy tone;

[0019] Step 6: The node counts the service load of each node, and selects the average service load with non-zero load when the node competes for access as the handover threshold 2;

[0020] Step 7: The node collects the queue information of each child node according to the queue information field carried in the received data frame, and calculates the service load factor that measures the current service load; the data frame is the data frame carried by the parent node according to the queue information field in the data frame uploaded by the child node;

[0021] If the node currently uses the contention access mechanism, then:

[0022] If the service load factor is less than the handover threshold 1, use contention access and execute step 8;

[0023] If the service load factor is greater than the handover threshold 1, use scheduled access and execute step 9;

[0024] If the node currently uses the scheduled access mechanism, then:

[0025] If the service load factor is less than the handover threshold 2, use contention access and execute step 8;

[0026] If the service load factor is greater than the handover threshold 2, use scheduled access and execute step 9;

[0027] Step 8: The node allocates time slots for the nodes known to have data to be sent in the beacon frame, and the remaining idle time slots are contended for access by each node;

[0028] Step 9: The node calculates the number of time slots assigned to each routing node (the parent node assigns time slots to the child nodes), and uses the weighted round-robin method to assign time slots to the sub-routing nodes. If the time slots can meet the time slot requirements of each routing node, the sensing nodes are scheduled in the remaining time slots; if the time slots cannot meet the time slot requirements of each routing node, the number of time slots assigned to each node is reduced according to the load ratio of the child nodes, and the node broadcasts the final time slot allocation result in the beacon frame to inform the child nodes;

[0029] Step 10: When the node detects data transmission in the channel but fails to parse the valid data, the node considers that a collision occurs in the channel. When the node detects multiple collisions within a beacon frame period, it directly changes the access mechanism to scheduled access and schedules all routing nodes in the beacon frame.

[0030] In step 1, the switching threshold 1 for switching to the scheduled access mechanism is calculated based on the local topology of the network and the type of child nodes. Assuming the network load of each node, according to the Markov-related theory, the packet channel access probability and the throughput of the node under different loads are calculated, and compared with the number of packets generated by the child nodes. The maximum load when the throughput is equal to the number of packets generated by the child nodes is selected as the switching threshold 1.

[0031] In step 2, for the node to determine the maximum number of members in each group, the node determines the non-transmission time length of each time slot based on the time slot length, the transmission time length of the data packet, and the clock accuracy of the node, thereby determining the number of members in each group. According to the delay requirements of the time-sensitive service, the number of groups is determined. The specific calculation is as follows:

[0032] N = t time-s / t slot

[0033] n = t slot-free / t min

[0034] Where N represents the number of groups, t time-s represents the maximum delay allowed for the time-sensitive service, t slot represents the length of the time slot, n represents the maximum number of members allowed in each group, t slot-free represents the non-transmission time length of the time slot, t min represents the minimum time length for the node to distinguish the channel busy tone.

[0035] In step 3, for each child node, a group that can meet the maximum delay allowed for the time-sensitive service is selected. The specific steps are as follows:

[0036] The parent node calculates the maximum delay for each of its sub - routing nodes to transmit time - sensitive services based on the allocation result of the time slots allocated to the child nodes. In the order of the service load size of each sub - routing node, packets are selected for each sub - routing node. The criterion for selecting packets is to calculate the probability that the delay of transmitting time - sensitive services by the computing node exceeds the time limit when allocating each packet. The packet with the smallest probability of exceeding the time limit for transmitting time - sensitive services is selected as the result of packet allocation for that node.

[0037] The node in step 5 determines which nodes have generated time - sensitive services based on the time slot where the busy tone is generated and the specific position of the time when the busy tone is generated. It is the child nodes that have generated time - sensitive services. The analysis node calculates that it cannot meet the time - delay requirements of time - sensitive services when uploading time - sensitive services in the next transmission time slot. A busy tone is generated at the corresponding position in the nearest time slot corresponding to the group where the child node is located. The parent node determines the nodes that have generated time - sensitive services based on the time slot and the specific time of the busy tone.

[0038] The node in step 6 counts the service load of each child node and calculates the switching threshold 2 for switching to contention access. When counting the load, it is the number of packets to be sent in the known child node queue of the counting node and the load of the node in the past several beacon frame periods. The average service load when the load of the node is not zero during contention access is selected as the switching threshold 2.

[0039] The node in step 7 calculates the load factor. Based on the queue information in the data frame received by the node, the service load factor value in the previous beacon frame period, and the change in service load from the previous beacon frame period to the period before the previous one, the service load factor of the current beacon frame is weighted and calculated as follows:

[0040] D(t)=a*m(Num / num)+(1 - a)*[b*D(t - 1)+(1 - b)*d(D)]

[0041] Where D(t) represents the service load factor at the current time; Num is the total number of child nodes; num is the number of known nodes to be sent; m is the number of known packets to be sent; D(t - 1) is the service load factor in the previous statistical period; d(D) is the change in service load from the period before the previous one to the previous statistical period; a and b represent the weight values and 0 ≤ a, b ≤ 1.

[0042] When the load factor calculated by the node in step 7 during scheduled access is less than the switching threshold 2, the node switches from scheduled access to contention access; otherwise, it maintains scheduled access to reduce the number of access mechanism switches.

[0043] The step where the node in step 7 collects the sub-node queue information according to the queue information field in the received data requires the sub-nodes to carry the queue information when sending data, without using additional data packets to notify the parent node of the resource requirement information. The parent node reads this field while receiving the data packet to obtain the resource requirement information of the sub-node sending the message.

[0044] The step where the node in step 8 allocates time slots for the sub-nodes with non-empty transmission queues is that the node directly allocates time slot resources for transmitting the data in the transmission queue according to the queue information carried in the received data frame, reducing the competition pressure.

[0045] Advantages of the present invention:

[0046] First, compared with the existing methods, the present invention uses a dynamic allocation method to allocate time slots on demand according to the sub-node type and node load. For the unallocated time slots, the remaining nodes can access the channel through competition; during scheduling, time slot resources are allocated according to the node load, improving the overall network transmission efficiency.

[0047] Second, compared with the existing methods, the present invention enables the sub-nodes to send busy tone notifications to the parent node at different times during the non-transmission phase of the time slot to indicate the generation of time-sensitive services, and reduces the waiting time of other nodes through the parent node scheduling and transmission method. At the same time, the parent node groups all sub-nodes to support the time-sensitive services of all nodes to be transmitted on time as much as possible.

[0048] Third, compared with the existing methods, the present invention dynamically switches the channel access mode by calculating the load factor and switching threshold. Using the competition method at low load can ensure low data delay and low node overhead. When a node has multiple data frames to send, simple scheduling reduces competition collisions. At the same time, at high load, full scheduling by the parent node improves channel utilization, reduces conflicts, and increases network capacity.

[0049] Fourth, compared with the existing methods, the present invention reduces the overhead by carrying the node queue information in the data frame and centralized scheduling in the parent node beacon frame, without using the RTS / CTS handshake, and at the same time enables the protocol to be applied to devices with lower transmission rates. Description of the Drawings

[0050] Figure 1 It is a schematic diagram of the network time slot structure of the present invention.

[0051] Figure 2 It is a schematic diagram of the time slot grouping and the position where the busy tone is generated in the present invention.

[0052] Figure 3 It is a schematic diagram of the node access mechanism switching process of the present invention.

[0053] Figure 4 This is a schematic diagram of the node scheduling process of the present invention.

[0054] Figure 5 This is a schematic diagram of the scheduling and transmission process of time-sensitive services of nodes in the present invention. Detailed implementation manners

[0055] The present invention will be further described in detail below with reference to the accompanying drawings.

[0056] (1) Node model: Each routing node is set to use two channels. Channel 1 is used to communicate with the parent node of the node, and Channel 2 is used to communicate with the child nodes. The channels are isolated from each other, and the Channel 2 of adjacent routing nodes will not conflict. The sensing node only uses one channel, which is the Channel 2 of the adjacent routing node, and is used for data interaction with the connected routing node.

[0057] (2) Time frame structure: As shown in the attached... Figure 1 As shown, a beacon frame contains multiple time slots. The first time slot within each beacon frame period is the beacon frame time slot, which is used to broadcast the beacon frame and the broadcast time slot allocation result. Each time slot is further divided into an uplink time slot and a downlink time slot. Data is uploaded in the uplink time slot and downloaded in the downlink time slot. Among them, the uplink time slot has several backoff units and one transmission unit.

[0058] (3) Adaptive switching of the channel accounting mechanism that combines competition and scheduling. The node counts the service load of the child nodes and adaptively selects the channel access method according to the service load. When the service load is low, the parent node designates the child node to use CSMA-CA (Carrier Sense Multiple Access with Collision Avoidance) to compete for access to the channel; when the service load is high, the parent node uses a scheduling method to control the child node to access the channel. Whether it is competitive access or scheduled by the parent node, the child node should carry the node queue information when uploading the data frame, so that the parent node can know the resource requirements of each child node. If the child node has multiple data packets to transmit, the child node can carry the number of data packets to be sent cached in the node in the transmitted data packet. The parent node can count the number of packets in the pending transmission queue of the child node after receiving the data packet of the child node, so as to evaluate the number of time slots required by each child node, and allocate transmission time slots for each child node according to this demand situation, and put the allocation result in the beacon frame or the acknowledgment frame broadcast to inform the child node.

[0059] As shown in the attached... Figure 3As shown, after the routing node accesses the network, it first notifies the child nodes to access the channel using CSMA-CA, and calculates the packet channel access probability and throughput of the nodes under different traffic loads according to the local network topology and child node types. It selects the maximum traffic load that the contention access mechanism can provide stable service, that is, the traffic load when the throughput per unit time is equal to the generation amount per unit time as the switching threshold 1 for the scheduling access. The node calculates the traffic load factor that measures the current traffic load by weighted calculation of the number of data packets to be sent by the known child nodes and the load factor of the previous period. When the load factor is greater than the switching threshold 1, the parent node scheduling access mechanism is used. To reduce the number of times of switching the channel access mechanism, the node selects the average traffic load when there is traffic load in the contention access as the switching threshold 2 for the contention access. When the load factor is less than the threshold 2, the parent node broadcasts to notify the child nodes to switch to the contention access to reduce the service transmission delay.

[0060] The calculation method of the load factor is as follows:

[0061] D(t) = a * m * (N / n) + (1 - a) * [b * D(t - 1) + (1 - b) * d(D)] (0 < a, b < 1)

[0062] Where D(t) represents the traffic load factor at the current time; N is the total number of child nodes; n is the number of known nodes to be sent, m is the number of known data packets to be sent, D(t - 1) is the traffic load factor of the previous statistical period, and d(D) is the change in traffic load from the previous statistical period to the period before the previous one.

[0063] (4) The time slot scheduling scheme is as shown in the appendix Figure 4 As shown. When the parent node switches to the scheduling access mechanism, the parent node should count the traffic loads of each routing child node, and use the loads of each routing node as weights to allocate time slots to each routing node in a weighted round-robin manner until the time slots allocated to all sub-routing nodes can transmit the loads of each routing node; if the time slots are not enough to meet the needs of all routing nodes, calculate the difference, and distribute the difference to each sub-routing node according to the ratio of the loads of each sub-routing node; if there are unallocated time slots, then schedule the sub-sensing nodes according to the serial numbers of the sensing nodes. When allocating time slots, the time slots allocated to each child node should be as scattered as possible. Calculate the time slots that each routing node can get in a cycle, and allocate time slots to each routing node in the order of load size. The nodes with larger loads are allocated time slots first.

[0064] (5) Transmission of time-sensitive services. There are different services in the network. For time-sensitive data, under the low-load contention access mechanism, the node can immediately contend for time slots to transmit time-sensitive services; when the parent node performs scheduling for data transmission, as shown in the appendix Figure 5As shown, the node groups all its child nodes. The number of groups is related to the duration of the non - transmission part in the uplink time slot and the node clock accuracy. Map the remainder of each time slot number divided by the number of groups to each group, as shown in the appendix Figure 2 As shown. The node sorts the child nodes according to the load size of the child nodes, and distributes the child nodes to the group that minimizes the transmission delay of time - sensitive services. When the node has time - sensitive services to transmit, compare the time when the node reaches the next transmission time slot with the maximum delay allowed for the time - sensitive service. When the allowed delay of the time - sensitive service is small, the node sends a busy tone in the time slot where the group is located to inform the parent node of the generation of the time - sensitive service. The parent node differentiates different nodes by sensing the time difference of the busy tone, and adopts a high response ratio first scheduling algorithm in the downlink time slot to allocate time slots for the node to transmit time - sensitive services and sense the services generated by the node. When the parent node allocates a time slot for the child node to transmit time - sensitive services, it selects the time slot occupied by the node with the smallest load that meets the time - delay requirements of the time - sensitive service.

[0065] Such as Figures 3 - 5 As shown: The specific adaptive multiple access method of the present invention is as follows:

[0066] Step 1: The node counts the types and numbers of its child nodes, and analyzes and calculates the switching threshold 1 for switching to the scheduling access mechanism according to the relevant theory of the Markov process, and then executes Step 2;

[0067] Step 2: The node counts the service load of each child node, determines the number of groups of time - sensitive service groups and the maximum number of members in each group according to the node clock resolution, and then executes Step 3;

[0068] Step 3: The node sorts the child nodes according to the service load size, selects a group that can meet the maximum delay allowed for the time - sensitive service for each child node according to the sorting, broadcasts the grouping result to the child nodes through a beacon frame, and then executes Step 4;

[0069] Step 4: When the node generates a time - sensitive service, if the node's channel access mechanism is contention access, select an idle time slot to upload the time - sensitive data in a timely manner; if the node's channel access mechanism is parent - node scheduling access, judge whether the node's next transmission time slot can meet the time - delay requirements of the time - sensitive service. If it can be met, the node uploads the time - sensitive data in the next transmission time slot. If it cannot be met, execute Step 5;

[0070] Step 5: The node generates time - sensitive data, generates a busy tone at the corresponding moment in the group where the node is located. The parent node determines the node number that generates the time - sensitive data according to the time slot when the busy tone is generated and the specific position where the busy tone moment is located, and constructs a command frame in the downlink time slot or carries scheduling information in the ACK frame, so that the node that generates the time - sensitive service uploads data in the next time slot;

[0071] Step 6: Each node calculates its own service load, and selects the average service load of the nodes with non-zero load during contention access as the handover threshold 2;

[0072] Step 7: The node collects the queue information of each child node according to the queue information field carried in the received data frame, and calculates the service load factor that measures the current service load;

[0073] If the node currently uses a contention access mechanism, then:

[0074] If the service load factor is less than the handover threshold 1, use contention access and execute Step 8;

[0075] If the service load factor is greater than the handover threshold 1, use scheduled access and execute Step 9;

[0076] If the node currently uses a scheduled access mechanism, then:

[0077] If the service load factor is less than the handover threshold 2, use contention access and execute Step 8;

[0078] If the service load factor is greater than the handover threshold 2, use scheduled access and execute Step 9;

[0079] Step 8: The node allocates time slots in the beacon frame for the nodes known to have data to be sent, and the remaining idle time slots are contended for access by each node;

[0080] Step 9: The node calculates the number of time slots allocated to each routing node, and uses the weighted round-robin method to allocate time slots to the sub-routing nodes. If the time slots can meet the time slot requirements of each routing node, the sensing nodes are scheduled in the remaining time slots; if the time slots cannot meet the time slot requirements of each routing node, the number of time slots allocated to each node is reduced according to the load ratio of the child nodes, and the node broadcasts the final time slot allocation result in the beacon frame to inform the child nodes;

[0081] Step 10: When a node detects data transmission in the channel but fails to parse valid data, the node considers that a collision has occurred in the channel. When a node detects multiple collisions within a beacon frame period, it directly changes the access mechanism to scheduled access and schedules all routing nodes in the beacon frame. In Step 1, the handover threshold 1 for switching to scheduled access is calculated based on the local topology of the network and the type of child nodes. Assuming the network load of each node, according to the Markov-related theory, the packet channel access probability and the throughput of the node under different loads are calculated, and compared with the number of packets generated by the child nodes. The maximum load at which the throughput is equal to the number of packets generated is selected as threshold 1.

[0082] In step 2, the node determines the maximum number of members in each group. The node determines the non - transmission time length of each time slot based on the time slot length, the transmission time of the data packet, and the clock accuracy of the node, thereby determining the number of members in each group. According to the delay requirement of the time - sensitive service, the number of groups is determined. The specific calculation is as follows:

[0083] N = t time-s / t slot

[0084] n = t slot-free / t min

[0085] Where N represents the number of groups, t time-s represents the maximum delay allowed for the time - sensitive service, t slot represents the length of the time slot, n represents the maximum number of members allowed in each group, t slot-free represents the non - transmission time length of the time slot, t min represents the minimum time length for the node to distinguish the channel busy tone.

[0086] In step 5, the node determines which nodes generate the time - sensitive service according to the time slot where the busy tone is generated and the specific position where the busy tone is generated. The child nodes that generate the time - sensitive service calculate and analyze that the time - sensitive service cannot meet the delay requirement when uploaded in the next transmission time slot, and generate a busy tone at the corresponding position in the nearest time slot corresponding to the group where the child node is located. The parent node determines the nodes that generate the time - sensitive service according to the time slot and the specific time where the busy tone is located.

[0087] In step 6, the node statistics the traffic load of each child node and calculates the switching threshold 2 for switching to contention access. When statistics the load, it statistics the number of packets to be sent in the known child node queue of the node and the load of the node in the past several beacon frame periods, and selects the average traffic load when the load is not zero during contention access as the switching threshold 2.

[0088] In step 7, the node calculates the load factor. Based on the queue information in the data frame received by the node, the traffic load factor value in the previous beacon frame period, and the change in traffic load in the previous beacon frame period, the traffic load factor of the current beacon frame is weighted and calculated. The specific calculation is as follows:

[0089] D(t)=a*m(Num / num)+(1 - a)*[b*D(t - 1)+(1 - b)*d(D)]

[0090] Among them, D(t) represents the traffic load factor at the current time; Num is the total number of child nodes; num is the number of known nodes to be sent; m is the number of known data packets to be sent; D(t - 1) is the traffic load factor in the previous statistical period; d(D) is the change in traffic load in the previous statistical period compared to the period before the previous one; a and b represent the weight values and 0 ≤ a, b ≤ 1.

[0091] When the load factor calculated by the node in step 7 during scheduling access is less than the handover threshold 2, the node switches from scheduling access to contention access; otherwise, it maintains scheduling access, reducing the number of handovers of the access mechanism.

[0092] In step 7, the node collects the queue information of child nodes according to the queue information field in the received data, which requires the child nodes to carry the queue information when sending data, without using additional data packets to notify the parent node of the resource requirement information. The parent node reads this field while receiving the data packet to obtain the resource requirement information of the child node sending the message.

[0093] In step 8, the node allocates time slots for child nodes with non-empty transmission queues. The node directly allocates time slot resources for the data in the transmission queue according to the queue information carried in the received data frame, reducing the contention pressure.

Claims

1. An adaptive multiple access method for load awareness in a sensor network, characterized in that, It includes the following steps; Step 1: The node counts the types and numbers of the child nodes it contains, analyzes and calculates the switching threshold 1 for switching to the scheduling access mechanism according to the relevant theories of the Markov process, and executes Step 2; Step 2: The node counts the traffic loads of the child nodes it contains, determines the number of packets of the time-sensitive service packets and the maximum number of members in each packet according to the node clock resolution, and executes Step 3; Step 3: At each node, the child nodes are sorted in descending order according to the traffic load, and a packet that can meet the maximum delay allowed by the time-sensitive service is selected for each child node according to the sorting. The packet result is broadcast to the child nodes through a beacon frame, and Step 4 is executed; Step 4: When a node generates time-sensitive services, if the node channel access mechanism is contention access, it selects an idle time slot and uploads the time-sensitive data from the child node to the parent node in time; if the node channel access mechanism is parent node scheduling access, it judges whether the next transmission time slot of the node can meet the delay requirement of the time-sensitive service. If it can be met, the node uploads the time-sensitive data from the child node to the parent node in the next transmission time slot. If it cannot be met, Step 5 is executed; Step 5: When the node generates time-sensitive data, a busy tone is generated at the corresponding moment of the packet where the node is located. The parent node determines the node number that generates the time-sensitive data according to the generated time slot of the busy tone and the specific position where the busy tone moment is located, and constructs a command frame in the downlink time slot or carries scheduling information in the ACK frame, so that the node that generates the time-sensitive service uploads data in the next time slot; Step 6: The node counts the traffic loads of each node, and selects the average traffic load when the load is not zero during contention access of the node as the switching threshold 2; Step 7: The node collects the queue information of each child node according to the queue information field carried in the received data frame, and calculates the traffic load factor for measuring the current traffic load; If the node currently uses the contention access mechanism, then: If the traffic load factor is less than the switching threshold 1, use contention access and execute Step 8; If the traffic load factor is greater than the switching threshold 1, use scheduling access and execute Step 9; If the node currently uses the scheduling access mechanism, then: If the traffic load factor is less than the switching threshold 2, use contention access and execute Step 8; If the traffic load factor is greater than the switching threshold 2, use scheduling access and execute Step 9; Step 14: The node allocates time slots for the nodes known to have data to be sent in the beacon frame, and the remaining idle time slots are contended for access by each node; Step 15: The node calculates the number of time slots allocated to each routing node, and uses the weighted round-robin method to allocate time slots to the sub-routing nodes. If the time slots can meet the time slot requirements of each routing node, the sensing nodes are scheduled in the remaining time slots; if the time slots cannot meet the time slot requirements of each routing node, the number of time slots allocated to each node is reduced according to the child node load ratio. The node broadcasts the final time slot allocation result in the beacon frame to inform the child nodes; Step 10: When a node detects that there is data transmission in the channel but fails to parse valid data, the node considers that a collision has occurred in the channel. When a node detects multiple collisions within a beacon frame period, it directly changes the access mechanism to scheduled access and schedules all routing nodes in the beacon frame.

2. The adaptive multiple access method for load awareness in a sensor network according to claim 1, wherein In Step 1, the switching threshold 1 for switching to the scheduled access mechanism is calculated based on the local topology of the network and the types of child nodes. Assuming the network load of each node, according to the relevant Markov theory, the packet channel access probability and the throughput of the node under different loads are calculated and compared with the number of packets generated by the child nodes. The maximum load at which the throughput is equal to the number of packets generated by the child nodes is selected as the switching threshold 1.

3. An adaptive multiple access method for load awareness in a sensor network according to claim 1, characterized in that In Step 2, for the node to determine the maximum number of members in each packet, the node determines the non-transmission time length of each time slot based on the time slot length, the transmission time length of the data packet, and the clock accuracy of the node, thereby determining the number of members in each packet. According to the delay requirements of the time-sensitive service, the number of packets is determined. The specific calculation is as follows: N = t time-s / t slot n = t slot-free / t min Where N represents the number of groups, and t time-s represents the maximum latency allowed for time-sensitive services, and t slot represents the length of a time slot, n represents the maximum number of members allowed for each group, and t slot-free represents the non-transmission time length of a time slot, and t min represents the minimum time length for a node to distinguish a channel busy tone.

4. An adaptive multiple access method for load awareness in a sensor network according to claim 1, characterized in that Step 3 is to select a packet for each child node that can meet the maximum delay allowed by the time-sensitive service. The specific steps are as follows: The parent node calculates the maximum delay for each sub-routing node to transmit the time-sensitive service according to the allocation result of the time slots allocated to the child nodes. In the order of the service load size of each sub-routing node, a packet is selected for each sub-routing node. The criterion for selecting a packet is to calculate the probability that the transmission of the time-sensitive service by the node allocated to each packet exceeds the time limit. The packet with the smallest probability that the transmission of the time-sensitive service exceeds the time limit is selected as the result of allocating the packet for the node.

5. The adaptive multiple access method for load awareness in a sensor network according to claim 1, characterized in that, In Step 5, the node determines which nodes have generated the time-sensitive service based on the time slot where the busy tone is generated and the specific position where the busy tone is generated. The child node that generates the time-sensitive service calculates and analyzes that the transmission of the time-sensitive service by the analysis node in the next transmission time slot cannot meet the delay requirements of the time-sensitive service, and generates a busy tone at the corresponding position in the nearest time slot corresponding to the packet where the child node is located. The parent node determines the node that generates the time-sensitive service based on the time slot where the busy tone is located and the specific time.

6. The adaptive multiple access method for load awareness in a sensor network according to claim 1, wherein In Step 6, the node counts the service load of each child node and calculates the switching threshold 2 for switching to contention access. When counting the load, the number of packets to be sent in the known child node queue of the node and the load of the node in the past several beacon frame periods are counted. The average service load when the load of the node is not zero during contention access is selected as the switching threshold 2.

7. An adaptive multiple access method for load awareness in a sensor network according to claim 1, characterized in that In Step 7, the node calculates the load factor. Based on the queue information in the data frame received by the node, the service load factor value in the previous beacon frame period, and the change in the service load in the previous beacon frame period, the service load factor of the current beacon frame is weighted and calculated. The specific calculation is as follows: D(t) = a * m(Num / num) + (1 - a) * [b * D(t - 1) + (1 - b) * d(D)] Among them, D(t) represents the traffic load factor at the current time; Num is the total number of child nodes; num is the number of known nodes to be sent; m is the number of known data packets to be sent; D(t - 1) is the traffic load factor in the previous statistical period; d(D) is the change in traffic load in the previous statistical period compared to the period before the previous one; a and b represent the weights and 0 ≤ a, b ≤ 1.

8. The adaptive multiple access method for load perception in a sensor network according to claim 1, wherein When calculating that the load factor at the node in step 7 is less than the handover threshold 2 during scheduling access, the node switches from scheduling access to contention access; otherwise, it maintains scheduling access, reducing the number of handovers of the access mechanism.

9. The adaptive multiple access method for load awareness in a sensor network according to claim 1, wherein In step 7, the node collects the queue information of child nodes according to the queue information field in the received data, which requires the child nodes to carry the queue information when sending data, without using additional data packets to notify the parent node of the resource requirement information. The parent node reads this field while receiving the data packet to obtain the resource requirement information of the child node sending the message.

10. The adaptive multiple access method for load awareness in a sensor network according to claim 1, wherein, In step 8, the node allocates time slots for child nodes with non-empty send queues, which means that the node directly allocates time slot resources for the data in the transmission queue according to the queue information carried in the received data frame.

Images