A layered and clustered underwater acoustic communication network-oriented demand-aware hybrid MAC protocol design method

Abstract

The application belongs to the technical field of MAC protocol of underwater acoustic communication network, and relates to a demand-aware hybrid MAC protocol design method for layered and clustered underwater acoustic communication network, comprising: a surface sink node distributes network initialization information to each cluster head node by multicasting a beacon frame; each cluster head node allocates bandwidth and rate based on the location information and channel conditions of the member nodes in the cluster; the member nodes perform parallel intra-cluster uplink transmission of data packets in the allocated sub-band according to the bandwidth and rate allocation results; the cluster head node aggregates the received data and transmits to the surface sink node in the allocated time slot; the surface sink node dynamically adjusts the time slot allocation of the next period according to the business load and transmission state of each cluster, and feeds back the reception results and time slot information to the cluster head node through an acknowledgement frame; and the member nodes adjust the transmission behavior according to the listening and sensing collision avoidance mechanism. The method is suitable for efficient access control and reliable data transmission in layered and clustered underwater acoustic communication network.

Description

Technical Field

[0001] This invention relates to the field of underwater acoustic communication network MAC protocol technology, and to a demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks. Background Technology

[0002] Underwater acoustic communication networks are widely used in scenarios such as seismic source zone observation, tsunami warning, and marine environmental monitoring. In these applications, multiple underwater nodes are typically deployed in the sea area to collect seabed seismic signals, pressure fluctuations, and environmental parameters, and then upload the monitoring data to a surface aggregation node. To reduce network management complexity and improve network scalability, existing underwater acoustic communication networks typically adopt a hierarchical and clustered organization. Member nodes within a cluster are responsible for data acquisition and reporting, while the cluster head node is responsible for data aggregation within the cluster and further forwarding the aggregated data to the surface aggregation node.

[0003] However, underwater acoustic channels are characterized by narrow bandwidth, long latency, and significant differences in link quality, posing considerable challenges to the design of MAC protocols in hierarchical and clustered underwater acoustic communication networks. In such networks, member nodes within a cluster share the same channel for uplink access, while the cluster head node needs to both receive data uploaded by members within the cluster and send aggregated data to the surface aggregation node within its allocated time slots. Without an access and scheduling mechanism that matches the hierarchical and clustered two-level link topology, timing and resource coupling can easily occur between concurrent reporting within the cluster, aggregation reception by the cluster head, and uplink forwarding by the cluster head. This can lead to problems such as mutual constraints on intra-cluster and inter-cluster scheduling, uplink window conflicts, queue backlogs, decreased throughput, and increased end-to-end latency.

[0004] Existing clustered underwater acoustic media access control technologies mainly fall into the following categories: one category suppresses collisions and improves transmission controllability through reservation or scheduling mechanisms; another category improves the protocol's adaptability to service load fluctuations by introducing queue-driven or adaptive scheduling mechanisms; in addition, there are hybrid access designs that combine different multiple access methods. While these technologies can reduce collision probability or improve scheduling flexibility to some extent, they still have the following shortcomings in hierarchical clustered aggregation scenarios: (1) Inter-cluster scheduling cannot simultaneously address both the queuing backlog problem under high load and the time slot waste problem under low load; (2) The quality of links within the cluster varies, and the communication rates that can be achieved by each member node are different, which can easily lead to local congestion and uneven bandwidth utilization, thereby weakening the performance gains brought by parallel transmission within the cluster. (3) Under the half-duplex constraint of the cluster head node, there is a risk of conflict between the uplink aggregation and transmission process of the cluster head and the data reception process of the member nodes in the cluster. This can easily lead to data loss at the cluster head, which in turn affects the network transmission efficiency and reliability.

[0005] Therefore, existing technologies still lack a demand-aware hybrid MAC protocol suitable for hierarchical and clustered underwater acoustic communication networks, which can simultaneously solve problems such as inter-cluster time slot scheduling, intra-cluster resource allocation, and cluster head transmit / receive conflict avoidance. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks.

[0007] In view of this, the present invention proposes a demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks, comprising the following steps: Step 1: The surface convergence node sends network-wide initialization information to each cluster head node via multicast beacon frames; Step 2: Each cluster head node allocates bandwidth and rate to its cluster members based on the location information of the member nodes and channel conditions. Step 3: Based on the bandwidth allocation results and rate allocation results, each member node performs intra-cluster parallel uplink transmission to the corresponding cluster head node within the allocated sub-frequency band; Step 4: Each cluster head node aggregates the received intra-cluster data and sends the aggregated data to the surface aggregation node within the uplink time slot allocated by the surface aggregation node. Step 5: The surface aggregation node dynamically adjusts the time slot allocation for the next cycle based on the service load and transmission status of each cluster, and sends the reception result and time slot allocation information back to the cluster head node through an acknowledgment frame; Step 3 also includes: each member node adjusts its transmission behavior according to the listening-aware conflict backoff mechanism to avoid conflicts between the uplink transmission of member nodes and the uplink transmission window of the cluster head node sending aggregated data to the surface convergence node.

[0008] As an improvement to the above method, the network-wide initialization information in step 1 includes: network node location information, network-wide time synchronization information, and an initial cluster head time slot allocation scheme; wherein, The network node location information refers to the spatial coordinates of each cluster head node and member node within the cluster in the network. The network-wide time synchronization information is used to provide a unified time reference for each cluster head node; The initial cluster head time slot allocation scheme is used to provide each cluster head node with the time slot start information and time slot length information required for the first round of inter-cluster uplink scheduling.

[0009] As an improvement to the above method, step 2 specifically includes: Each cluster head node allocates bandwidth according to the channel matching principle, giving priority to lower frequency bands to member nodes that are farther away from the cluster head node, and priority to higher frequency bands to member nodes that are closer to the cluster head node. Under the conditions of satisfying the total bandwidth constraint within the cluster and the minimum bandwidth constraint of the member nodes, the goal is to achieve a balanced communication rate within the cluster by minimizing the difference between the maximum and minimum communication rates of each member node within the cluster. The cluster head node writes the optimized bandwidth allocation results and communication rate allocation results of each member node into the bandwidth allocation frame and multicasts them to the member nodes within the cluster, so that each member node within the cluster obtains the corresponding uplink transmission frequency band and communication rate parameters.

[0010] As an improvement to the above method, the parallel uplink transmission in step 3 adopts frequency division multiple access.

[0011] As an improvement to the above method, in step 5, the surface aggregation node dynamically adjusts the time slot allocation for the next cycle based on the service load and transmission status of each cluster, including the calculation process of the aggregation cycle; specifically including: Let the first Wheel, cluster head have There are 1 member node, and the uplink transmission of each member node follows a Poisson process with a rate of 1. If the nodes are independent of each other, then the cluster head The receiving process follows a Poisson distribution, corresponding to the load. for: ; Aggregation cycle Refers to cluster head No. Round upload time slot start to the first The time interval between the start of the upload time slot and the cluster head during this period. Receive and cache data from member nodes; the specific calculation formula is as follows: in, For the first wheel hub head The time slot length, For the first wheel hub head The allocated time slot length, The response time for the ACK frame. It is the total number of cluster heads. i Indicates the cluster head number. k This indicates the sequence number of the current cluster head.

[0012] As an improvement to the above method, in step 5, the surface aggregation node dynamically adjusts the time slot allocation for the next cycle based on the service load and transmission status of each cluster, including a time slot demand calculation process, specifically including: The time required for newly generated data packets to be uploaded within an aggregation cycle is calculated using the following formula, which is the basic upload requirement. : in, For cluster head upload speed, The original data packet length. Cluster head fusion rate; Calculate the compensation requirement according to the following formula. : in, For the first The number of data bits that need to be sent in the next round if the previous round was unsuccessful. The volatility margin is calculated using the following formula. :

[0013] in, It is an adjustable margin coefficient; The cluster head is estimated using the following formula. No. Upload slots of the wheel : .

[0014] As an improvement to the above method, in step 5, the surface aggregation node dynamically adjusts the time slot allocation for the next cycle based on the service load and transmission status of each cluster, including a time slot optimization process, specifically including: The cluster head time slot allocation problem is modeled as a demand-matching optimization problem, which is solved by minimizing the allocated time slot length. With actual transmission needs The deviation between them achieves unified suppression of delayed transmission costs and idle losses, and its optimization objective can be expressed as: in, and These represent the minimum and maximum allowed time slot lengths, respectively.

[0015] As an improvement to the above method, each member node adjusts its sending behavior according to the listening-aware conflict backoff mechanism, specifically including: Before sending data, each member node listens to the channel occupancy status on its allocated sub-band to determine whether the corresponding cluster head node is in the uplink transmission phase of sending aggregated data to the surface aggregation node. When the listening result indicates that the cluster head node is in the uplink transmission phase, the member node does not immediately send the current data, but performs backoff processing on the data to be sent.

[0016] As an improvement to the above method, the backoff process specifically includes: The data to be sent is stored in the local backoff queue. After the monitoring results indicate that the uplink transmission of the cluster head node has ended and the channel has returned to idle, the data to be sent is rescheduled according to the order in the backoff queue.

[0017] Compared with the prior art, the advantages of the present invention are: 1. This invention provides a demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks, which solves the problem of mismatch between the two-level link access and scheduling mechanism in hierarchical and clustered networks; 2. Through the demand-aware adaptive time slot allocation mechanism, the uplink time slot length of the next cycle can be dynamically adjusted according to the service load and transmission status of each cluster head node, which can reduce queue backlog under high load, reduce time slot idle waste under low load, and improve the channel utilization of inter-cluster uplink scheduling. 3. Through the rate-balanced bandwidth allocation mechanism, bandwidth resources can be adaptively allocated according to the location information of member nodes and channel conditions within the cluster, reducing the communication rate differences between member nodes within the cluster, alleviating local congestion and bandwidth utilization imbalance, and improving the efficiency of parallel transmission within the cluster. 4. Through the listening-aware conflict backoff mechanism, member nodes within the cluster can proactively avoid the uplink transmission window of the cluster head node, reducing transmit and receive conflicts under the half-duplex constraint of the cluster head node, reducing data loss within the cluster, and improving network transmission reliability. Attached Figure Description

[0018] Figure 1 This is a flowchart of a demand-aware hybrid MAC protocol for a hierarchical and clustered underwater acoustic communication network in one embodiment. Figure 2 This is a diagram of the underwater acoustic communication network structure in the embodiment; Figure 3 This refers to the frame structure design in the embodiment; Figure 4 This is a schematic diagram of conflict awareness and backoff for cluster member nodes in the embodiment; Figure 5 This is a schematic diagram showing the comparison of the average end-to-end latency of the protocol proposed in this invention with the TDMA and AMA-MAC protocols under different node traffic loads. Figure 6This diagram illustrates the comparison of packet delivery rates between the protocol proposed in this invention and the TDMA and AMA-MAC protocols under different node traffic loads. Detailed Implementation

[0019] A demand-aware hybrid MAC protocol for hierarchical and clustered underwater acoustic communication networks is proposed to achieve efficient access control and reliable data transmission in such networks.

[0020] According to the present invention, a demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks includes the following steps: 1. During the network initialization phase, the surface convergence node multicasts beacon frames to each cluster head node, disseminating network-wide initialization information. This network-wide initialization information includes network node location information, network-wide time synchronization information, and the initial cluster head uplink time slot allocation scheme.

[0021] 2. After receiving the network-wide initialization information, each cluster head node allocates intra-cluster communication resources according to the location information of its member nodes and the channel conditions of the corresponding links, obtains the bandwidth allocation results and communication rate allocation results for each member node, and sends the bandwidth allocation results and communication rate allocation results to its member nodes.

[0022] 3. Based on the received bandwidth allocation results and communication rate allocation results, each member node performs intra-cluster parallel uplink transmission to the corresponding cluster head node within the allocated sub-frequency band.

[0023] 4. Each cluster head node aggregates the received intra-cluster data and sends the aggregated data to the surface aggregation node within the uplink time slot allocated by the surface aggregation node.

[0024] 5. The surface convergence node dynamically adjusts the uplink time slot allocation results for each cluster head node in the next cycle based on the data transmission completion status, service load, and buffered data to be sent of each cluster head node in the current cycle, and feeds back the data reception results and time slot allocation information for the next cycle to each cluster head node through acknowledgment frames.

[0025] During the process, each member node adjusts its transmission behavior according to the monitoring-aware conflict backoff mechanism to avoid conflicts between the uplink transmission of member nodes and the uplink transmission window of the cluster head node sending aggregated data to the surface convergence node. Before transmission, member nodes monitor the channel occupancy status on the allocated sub-band. When they detect that the corresponding cluster head node is in the uplink transmission phase, they perform backoff processing on the data to be transmitted and reschedule the transmission after the channel becomes idle, thereby reducing the transmit and receive conflicts under the half-duplex constraint of the cluster head node.

[0026] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0027] Example refer to Figure 1 This is a flowchart of a demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks. The method of this invention includes the following steps: 1. The surface convergence node multicasts beacon frames to each cluster head node and sends network initialization information to each cluster head node; 2. Each cluster head node allocates intra-cluster communication resources based on the location information and channel conditions of its member nodes, obtains the bandwidth allocation results and communication rate allocation results for each member node, and sends the bandwidth allocation results and communication rate allocation results to its member nodes. 3. Based on the received bandwidth allocation results and communication rate allocation results, each member node performs intra-cluster parallel uplink transmission to the corresponding cluster head node within the allocated sub-frequency band. 4. Each cluster head node aggregates the received intra-cluster data and sends the aggregated data to the surface aggregation node within the uplink time slot allocated by the surface aggregation node. 5. The surface convergence node dynamically adjusts the uplink time slot allocation results for each cluster head node in the next cycle based on the data transmission completion status, service load, and buffered data to be sent of each cluster head node in the current cycle, and feeds back the data reception results and time slot allocation information for the next cycle to each cluster head node through acknowledgment frames. During the process, each member node adjusts its transmission behavior according to the listening and sensing conflict backoff mechanism to avoid conflicts between the uplink transmission of member nodes and the uplink transmission window of the cluster head node sending aggregated data to the surface convergence node.

[0028] Specifically: 1. Downlink initialization The underwater acoustic communication network model used in this embodiment is as follows: Figure 2 As shown. In this embodiment, the network first enters the downlink initialization phase, where the surface convergence node multicasts beacon frames to each cluster head node to complete the distribution of network-wide initialization information. The frame structure design is as follows. Figure 3 As shown, the beacon frame includes network node location information, network-wide time synchronization information, and the initial cluster head uplink time slot allocation scheme.

[0029] The network node location information refers to the spatial coordinates of each cluster head node and its member nodes within the network; the network-wide time synchronization information is used to provide a unified time reference for each cluster head node to support the timing coordination during subsequent uplink transmission of cluster head aggregated data and the confirmation frame feedback process; the initial cluster head uplink time slot allocation scheme is used to provide each cluster head node with the time slot start information and time slot length information required for the first round of inter-cluster uplink scheduling.

[0030] 2. The cluster head performs rate-balanced bandwidth allocation among cluster members. Each cluster head node allocates intra-cluster communication resources based on the location information of its member nodes and channel conditions, thereby obtaining the bandwidth allocation results and communication rate allocation results for each member node.

[0031] First, the cluster head node allocates frequency bands according to channel matching principles, prioritizing low-frequency bands for cluster member nodes farther from the cluster head and high-frequency bands for cluster member nodes closer to the cluster head to reduce path loss. Then, while satisfying the total cluster bandwidth constraint and the minimum bandwidth constraint for each member node, the bandwidth of each member node is further optimized. Optimize to minimize intra-cluster communication rate The maximum-minimum difference is optimized as follows:

[0032] The constraints are:

[0033] in, The minimum bandwidth allowed for member nodes. The number of member nodes within the cluster. This represents the total bandwidth within the cluster.

[0034] The cluster head node writes the optimized bandwidth allocation results and communication rate allocation results into the bandwidth allocation frame (BA) and multicasts them to the member nodes within the cluster. This enables each member node within the cluster to obtain the corresponding FDMA uplink transmission frequency band and communication rate parameters, providing a basis for subsequent parallel uplink transmission within the cluster.

[0035] 3. Cluster members upload data in parallel using FDMA. In this embodiment, after receiving the bandwidth allocation frame (BA) from the cluster head node, each member node transmits data according to the sub-band allocation result and communication rate allocation result carried therein. Different member nodes occupy different sub-bands and transmit data to their respective cluster head nodes at corresponding communication rates, thereby achieving intra-cluster parallel uplink transmission using frequency division multiple access.

[0036] 4. The cluster head aggregates the received data and uploads it in a dedicated time slot. In this embodiment, after receiving data uploaded by member nodes within their cluster, each cluster head node aggregates the data to form aggregated data uploaded between clusters. The surface aggregation node allocates dedicated uplink time slots to each cluster head node, and each cluster head node sends aggregated data to the surface aggregation node within its corresponding uplink time slot. When the amount of aggregated data exceeds the transmission capacity of the current time slot, the unsent data is retained in a local cache, awaiting transmission in subsequent cycles.

[0037] 5. Adaptive time slot allocation based on demand awareness at the water surface convergence node. In this embodiment, demand-aware adaptive time slot allocation is a dynamic time slot allocation mechanism with periodic feedback. After each upload cycle, the surface aggregation node dynamically adjusts the time slot length of each cluster head for the next cycle based on network load information and data transmission completion status.

[0038] (5.1) Calculation of polymerization cycle Assume the first Cycle, cluster head have There are 8 members, and each member node sends packets using a Poisson process with a rate of 1. If the nodes are independent of each other, then the cluster head The receiving process still follows a Poisson distribution, and the total load is: Aggregation cycle Refers to cluster head No. Round upload time slot start to the first The time interval between the start of the upload time slot and the cluster head during this period. Receive and cache data from member nodes; the specific calculation formula is as follows: in, For the first wheel hub head The time slot length, For the first wheel hub head The allocated time slot length, The response time for the ACK frame. It represents the total number of cluster heads.

[0039] (5.2) Calculation of time slot requirements Based on the statistical information from this round, estimate the upload time slots required for the next round. Each time slot needs to include three parts: 1) Basic upload requirements Based on the traffic load of member nodes and data fusion efficiency, calculate the time required for newly generated data packets to be uploaded within an aggregation cycle. This requirement reflects the data transmission overhead under normal circumstances.

[0040] in, For cluster head upload speed, The original data packet length. This represents the cluster head fusion rate.

[0041] 2) Compensation requirements In the current round, if the cluster head node fails to complete the transmission of all data due to insufficient time slots, the pruned data will be transmitted in the next round. Furthermore, if the corresponding data is not correctly received by the surface sink node after the cluster head completes transmission in the first round, the surface sink node will mark the cluster head's position as 0 in the returned ACK frame, thus triggering the cluster head to retransmit in the next round. To avoid excessive overhead and transmission delay, the protocol stipulates that each data packet can only be retransmitted once at most. The compensation requirement calculation formula is as follows: in, This represents the number of data bits that were not successfully transmitted in this round and need to be transmitted in the next round.

[0042] 3) Fluctuation margin To account for potential instantaneous load unevenness caused by traffic fluctuations and node location changes, the protocol allocates an additional time slot fluctuation margin to each cluster head. The specific calculation formula is as follows: in, It is an adjustable margin coefficient.

[0043] Therefore, cluster head No. The estimated time slot required for the wheel is: (5.3) Time slot optimization In this embodiment, two key costs in the cluster head time slot allocation process are jointly optimized: delayed transmission cost and idle time slot loss.

[0044] Cluster head nodes can only send aggregated data to surface sink nodes within their allocated uplink time slots in each TDMA cycle. When the actual transmission demand for aggregated data exceeds the current time slot length, the excess portion will be delayed to the next communication cycle, introducing additional queuing delay and energy consumption. Conversely, when the allocated time slot length is significantly greater than the actual demand, the remaining time slots will be idle, resulting in wasted channel resources. Therefore, properly matching the time slot length with the actual transmission demand is key to improving the overall performance of the protocol.

[0045] Let the first Wheel cluster head The allocated time slot length is The corresponding actual transmission requirements are .when Data that cannot be transmitted within the current time slot will be transmitted later, and the delay duration can be expressed as: Conversely, when the allocated time slot length exceeds the actual transmission demand, i.e. When there are excess time slots, they cannot be effectively utilized, forming idle time slots. The loss can be expressed as: Taking into account both types of costs, this protocol models the cluster head time slot allocation problem as a demand-matching optimization problem. By minimizing the deviation between the allocated time slot length and the actual transmission demand, it achieves unified suppression of delayed transmission costs and idle losses. Its optimization objective can be expressed as: in, and These represent the minimum and maximum allowed timeslot lengths, respectively, used to constrain the feasible range of timeslot allocation and avoid delayed transmission caused by excessively short timeslots or serious resource waste caused by excessively long timeslots.

[0046] 6. Member node listening and awareness backoff mechanism When cluster member nodes upload data to the cluster head using FDMA, the cluster head needs to handle two types of transceiver tasks simultaneously: receiving uplink data from cluster member nodes and sending aggregated data to the surface aggregation node in its dedicated TDMA time slot. Due to the limitations of the half-duplex underwater acoustic transceiver system, the cluster head cannot reliably receive data from cluster members simultaneously during uplink transmission. If the data arrival window of a member node overlaps with the uplink transmission window of the cluster head, the data packet upload fails, leading to packet loss within the cluster. Assume the cluster... Members At any moment Generate packets Let the transmission rate of this member on the allocated sub-band be... The data packet length is If (bit), then the reception time window of the data packet at the cluster head can be represented as: On the other hand, setting cluster heads In the The uplink transmission window within each TDMA cycle is: in This is the actual start time of transmission for the dedicated timeslot of this cluster head. Let this be the end time of the aggregated data transmission. Then, the "cluster head transmission – member arrival" conflict can be described by the overlap of the two time intervals: When the above conditions are met, the arrival process of the data packet of the member node overlaps with the uplink transmission process of the cluster head, which causes the data packet to fail to be received at the cluster head, thus triggering packet loss within the cluster.

[0047] To reduce this "cluster head send - member arrive" conflict, this paper designs a listener-aware conflict avoidance mechanism in member nodes, such as... Figure 4 As shown.

[0048] Before transmission, member nodes enter a listening state to determine channel occupancy on their allocated sub-bands, identifying whether the cluster head is in the uplink transmission phase. Since the cluster head transmits aggregated data uplink, the transmission duration is relatively long, and it occupies all available bandwidth resources during the uplink phase. Therefore, during its transmission, the sub-band allocated to the member node will continuously exhibit a stable channel busy characteristic, allowing the member node to effectively determine whether the cluster head is transmitting. If the listening result indicates the presence of a cluster head uplink signal, the member node does not transmit directly but marks the data packet to be transmitted as a backoff packet and stores it in a local backoff queue. The backoff queue uses first-in-first-out management to maintain the service arrival order.

[0049] When the monitoring results indicate that the cluster head uplink transmission has ended and the channel has returned to idle, the member node triggers the backoff queue release process: the backoff packets are rescheduled according to the queue order and transmitted sequentially within the available window.

[0050] Figure 5 This is a schematic diagram showing the comparison of the average end-to-end latency of the protocol proposed in this invention with the TDMA protocol and the AMA-MAC protocol under different node traffic loads in the embodiment. The results show that the average end-to-end latency of this protocol is always at the lowest level under different traffic loads.

[0051] Figure 6 This is a schematic diagram showing the comparison of packet delivery rates of the TDMA protocol and AMA-MAC protocol proposed in this invention under different node traffic loads in the embodiments. It can be seen from the results that the delivery rate of this protocol is the highest when the traffic load is greater than 0.26 packets / s.

[0052] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent substitutions to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A requirement-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks, comprising the following steps: Step 1: The surface convergence node sends network-wide initialization information to each cluster head node via multicast beacon frames; Step 2: Each cluster head node allocates bandwidth and rate to its cluster members based on the location information of the member nodes and channel conditions. Step 3: Based on the bandwidth allocation results and rate allocation results, each member node performs intra-cluster parallel uplink transmission to the corresponding cluster head node within the allocated sub-frequency band; Step 4: Each cluster head node aggregates the received intra-cluster data and sends the aggregated data to the surface aggregation node within the uplink time slot allocated by the surface aggregation node. Step 5: The surface aggregation node dynamically adjusts the time slot allocation for the next cycle based on the service load and transmission status of each cluster, and sends the reception result and time slot allocation information back to the cluster head node through an acknowledgment frame; Step 3 also includes: Each member node adjusts its transmission behavior based on the listening and sensing conflict backoff mechanism to avoid conflicts between the uplink transmission of member nodes and the uplink transmission window of the cluster head node sending aggregated data to the surface convergence node.

2. The demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks according to claim 1, characterized in that, The network-wide initialization information in step 1 includes: network node location information, network-wide time synchronization information, and the initial cluster head time slot allocation scheme; wherein, The network node location information refers to the spatial coordinates of each cluster head node and member node within the cluster in the network. The network-wide time synchronization information is used to provide a unified time reference for each cluster head node; The initial cluster head time slot allocation scheme is used to provide each cluster head node with the time slot start information and time slot length information required for the first round of inter-cluster uplink scheduling.

3. The demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks according to claim 1, characterized in that, Step 2 specifically includes: Each cluster head node allocates bandwidth according to the channel matching principle, giving priority to lower frequency bands to member nodes that are farther away from the cluster head node, and priority to higher frequency bands to member nodes that are closer to the cluster head node. Under the conditions of satisfying the total bandwidth constraint within the cluster and the minimum bandwidth constraint of the member nodes, the goal is to achieve a balanced communication rate within the cluster by minimizing the difference between the maximum and minimum communication rates of each member node within the cluster. The cluster head node writes the optimized bandwidth allocation results and communication rate allocation results of each member node into the bandwidth allocation frame and multicasts them to the member nodes within the cluster, so that each member node within the cluster obtains the corresponding uplink transmission frequency band and communication rate parameters.

4. The demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks according to claim 1, characterized in that, The parallel uplink transmission in step 3 adopts frequency division multiple access.

5. The demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks according to claim 1, characterized in that, In step 5, the surface aggregation node dynamically adjusts the time slot allocation for the next cycle based on the service load and transmission status of each cluster, including the calculation process of the aggregation cycle; specifically, it includes: Let the first Wheel, cluster head have There are 1 member node, and the uplink transmission of each member node follows a Poisson process with a rate of 1. If the nodes are independent of each other, then the cluster head The receiving process follows a Poisson distribution, corresponding to the load. for: ; Aggregation cycle Refers to cluster head No. Round upload time slot start to the first The time interval between the start of the upload time slot and the cluster head during this period. Receive and cache data from member nodes; the specific calculation formula is as follows: in, For the first wheel hub head The time slot length, For the first wheel hub head The allocated time slot length, The response time for the ACK frame. It is the total number of cluster heads. i Indicates the cluster head number. k This indicates the sequence number of the current cluster head.

6. The demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks according to claim 5, characterized in that, In step 5, the surface aggregation node dynamically adjusts the time slot allocation for the next cycle based on the service load and transmission status of each cluster, including a time slot demand calculation process, specifically including: The time required for newly generated data packets to be uploaded within an aggregation cycle is calculated using the following formula, which is the basic upload requirement. : in, For cluster head upload speed, The original data packet length. Cluster head fusion rate; Calculate the compensation requirement according to the following formula. : in, For the first The number of data bits that need to be sent in the next round if the previous round was unsuccessful. The volatility margin is calculated using the following formula. : in, It is an adjustable margin coefficient; The cluster head is estimated using the following formula. No. Upload slots of the wheel : 。 7. The demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks according to claim 6, characterized in that, In step 5, the surface aggregation node dynamically adjusts the time slot allocation for the next cycle based on the service load and transmission status of each cluster, including a time slot optimization process, specifically including: The cluster head time slot allocation problem is modeled as a demand-matching optimization problem, which is solved by minimizing the allocated time slot length. With actual transmission needs The deviation between them achieves unified suppression of delayed transmission costs and idle losses, and its optimization objective can be expressed as: in, and These represent the minimum and maximum allowed time slot lengths, respectively.

8. The demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks according to claim 1, characterized in that, Each member node adjusts its sending behavior according to the listening-aware conflict backoff mechanism, specifically including: Before sending data, each member node listens to the channel occupancy status on its allocated sub-band to determine whether the corresponding cluster head node is in the uplink transmission phase of sending aggregated data to the surface aggregation node. When the listening result indicates that the cluster head node is in the uplink transmission phase, the member node does not immediately send the current data, but performs backoff processing on the data to be sent.

9. The demand-aware hybrid MAC protocol design method for hierarchical and clustered underwater acoustic communication networks according to claim 1, characterized in that, The retreat process specifically includes: The data to be sent is stored in the local backoff queue. After the monitoring results indicate that the uplink transmission of the cluster head node has ended and the channel has returned to idle, the data to be sent is rescheduled according to the order in the backoff queue.

Images