A method for balancing cluster head node energy in wireless sensor networks

Abstract

The method for balancing the energy of cluster head nodes in a wireless sensor network of the present invention uses the age of information AoI to discard election messages with low freshness when the nodes of the wireless sensor network perform cluster head node preliminary election before clustering, and then selects nodes to form a cluster with a number of hops from the cluster head node as a restriction condition. After K rounds of data transmission / reception, the amount of data processed by each cluster member node is different, resulting in different residual energy of each node, among which the energy consumption of the cluster head node is the most serious. In order to balance the energy of all nodes, the cluster head node is updated at regular intervals, and the current cluster head node directly assigns the cluster member node with the smallest amount of data processed as the new cluster head node through broadcasting by obtaining the amount of data processed by the cluster member nodes. This not only balances the energy of all nodes, but also minimizes the energy consumption of the cluster head node election process, thereby achieving the purpose of extending the life of the wireless sensor network.

Description

Technical Field

[0001] The present invention relates to the field of wireless sensor networks, in particular to a method for balancing the energy of cluster head nodes in a wireless sensor network based on age of information (AoI) and energy consumption balance. Background Art

[0002] Limited by the self-sufficient power supply mode, reducing the energy consumption of wireless sensor nodes plays a very important role in extending the life of nodes, maintaining network topology and extending the life of wireless sensor networks. In wireless sensor networks, nodes are distributed in different clusters. According to the role they play, nodes can be classified into cluster head nodes and cluster member nodes. Cluster member nodes forward the collected data to cluster head nodes, and cluster head nodes collect and fuse the data forwarded by cluster member nodes and then forward it to the base station. Since cluster head nodes need to collect and fuse the data forwarded by cluster member nodes and then forward it to the base station, the energy consumption of cluster head nodes is the most serious in the entire cluster. Therefore, the energy consumption balance of cluster head nodes is crucial to extending the life of wireless sensor networks. Summary of the invention

[0003] The purpose of the present invention is to provide a method for balancing the energy of cluster head nodes in a wireless sensor network based on information age AoI and energy consumption balancing, so as to achieve the purpose of extending the life of the wireless sensor network, and by considering the amount of data processed by the node to evaluate the node energy consumption, the energy consumption caused by using a residual energy evaluation method with too high complexity in the process of estimating the residual energy is avoided.

[0004] The method for balancing the energy of cluster head nodes in a wireless sensor network of the present invention includes the steps of preliminary selection and clustering of cluster head nodes, data processing stage and assigning new cluster head nodes, specifically:

[0005] Step 10: Preliminary selection and clustering of cluster head nodes

[0006] Configure the Election message and the Source_ID_Hop table in each node;

[0007] When the wireless sensor network nodes are preliminarily elected before clustering, the nodes use the age of information AoI to discard the election messages with low freshness, and select the cluster member nodes to form a cluster based on the number of hops from the cluster head node. That is, in the initial state, the preliminaries of the cluster head node follow the principle of minimum ID address priority, and the cluster head node is determined by comparing the "Source Node ID" in the election message that appears periodically in the wireless sensor network.

[0008] The cluster head node broadcasts a SYN message to announce cluster information, and the nodes that receive the message automatically join the cluster;

[0009] After clustering, all nodes in the cluster enter the data processing stage;

[0010] Step 20: Data processing stage

[0011] The cluster head node considers the amount of data that each cluster member node has sent / received each time when assigning time slots, and controls the time when cluster member nodes send / receive data by periodically sending TS_schedule scheduling messages;

[0012] After the cluster member node sends / receives data in the time slot assigned by the cluster head node, it immediately sends a Data_Complete message to the cluster head node, then enters a dormant state and wakes up before the next TS_schedule scheduling message arrives;

[0013] Step 30: Assign a new cluster head node

[0014] After K rounds of data transmission / reception, the current cluster head node obtains the amount of data processed by the cluster member nodes through the Data_complete message reported by the cluster member nodes, and directly assigns the cluster member node with the smallest amount of processed data as the new cluster head node by broadcasting the ClusterHeader_assign message; after the new cluster head node is generated, return to step 2 and immediately broadcast the TS_schedule scheduling message to enter a new round of data processing phase.

[0015] The number of hops is limited to 2 hops.

[0016] The Election message at least includes:

[0017] The flag indicating whether this Election message is used for the initial election of the cluster head node;

[0018] Hop count that records the number of times the Election message is forwarded;

[0019] Record the Timestamp of when the Election message was sent;

[0020] Record the lifetime of the Election message;

[0021] Record the Destination Node ID, the hardware address of the node that receives the Election message;

[0022] Record the Source Node ID of the node device ID that sends the Election message;

[0023] Record the Forward Node ID of the node that forwards the Election message.

[0024] Record the amount of data that has been processed by the node that sent the Election message.

[0025] The Source ID Hop table in each node includes at least:

[0026] It is used to record the "Source Node ID", "Forward Node ID", the number of times the election message is forwarded, and the node's own ID "Owner's ID" in the election message sent from other nodes whose "Hop count" is the hop count limit. When recording the "Source Node ID", queue them in ascending order according to the "Hop count". If the "Hop count" is the same, queue them in ascending order according to the size of the node ID.

[0027] Before the primary election of the cluster head node is completed, the node continues to send / receive Election messages periodically. If the "Source Node ID" in the newly received Election message is greater than the last item in the Source ID Hop table, the Election message will be discarded; otherwise, the "Source Node ID" in the Election message will be inserted into the corresponding position in the Source ID Hop table, and the last item in the table will be squeezed out.

[0028] The node uses the age of information AoI to discard election messages with low freshness, specifically:

[0029] Before the cluster head node is pre-selected, all nodes first synchronize their clocks and then start sending Election messages for the cluster head node pre-selection;

[0030] When each node receives an Election message, it first determines the freshness of the information age AoI of the Election message based on the sending time "Timestamp" and "Lifetime". If the difference between the local time when the Election message is received and the sending time "Timestamp", that is, the information age AoI is less than the lifetime "Lifetime" in the Election message, it means that the freshness of the Election message is high and proceeds to the next step. Otherwise, the Election message is discarded.

[0031] The SYN message at least includes:

[0032] Cluster ID: Bit, that is, the number of nodes N in the wireless sensor network is rounded up after performing a logarithm operation with base 2;

[0033] Record the ClusterHeader ID of the cluster head node's device ID:

[0034] Member ID1 to Member ID used to record the device IDs of cluster member nodes 1 to i respectively i ;

[0035] Record the amount of data that the cluster head node has processed.

[0036] The TS_schedule scheduling message at least includes:

[0037] Cluster ID: Bit, that is, the number of nodes N in the wireless sensor network is rounded up after performing a logarithm operation with base 2;

[0038] Record the ClusterHeader ID of the cluster head node's device ID;

[0039] Record the Member ID1 to Member ID of the device ID of cluster members 1 to i respectively i ;

[0040] Record the time slot numbers TS1~TS1 assigned by the cluster head node to cluster members 1~i i .

[0041] The Data_complete message at least includes:

[0042] Record the ClusterHeader ID of the cluster head node device ID;

[0043] Record the Member ID of the device ID of cluster member node i i ;

[0044] Data sent / received by cluster member node i in the corresponding time slot i , that is, reporting the amount of processed data to the cluster head node.

[0045] By adopting the technical solution of the present invention, when the nodes of the wireless sensor network perform the initial election of the cluster head node before clustering, the age of information AoI is used to discard the election messages (i.e., Election messages) with low freshness, and then the nodes are selected to form a cluster with a restriction condition of 2 hops from the cluster head node. After K rounds of data transmission / reception, the amount of data processed by each cluster member node is different, resulting in different residual energy of each node. Among them, the energy consumption of the cluster head node is the most serious. In order to balance the energy of all nodes, the cluster head node is updated at regular intervals, and the current cluster head node obtains the amount of data processed by the cluster member nodes through the Data_complete message reported by the cluster member nodes. The cluster member node with the smallest amount of data processed is directly assigned as the new cluster head node by broadcasting the ClusterHeader_assign message, thereby not only balancing the energy of all nodes, but also minimizing the energy consumption of the cluster head node election process, thereby achieving the purpose of extending the life of the wireless sensor network. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 This is a schematic diagram of 2-hop node clustering according to the present invention;

[0047] Figure 2 This is a schematic diagram of the Election message format in the present invention;

[0048] Figure 3 It is the Source_ID_Hop representation in the present invention;

[0049] Figure 4 It is a schematic diagram of a SYN message in the present invention;

[0050] Figure 5 This is a schematic diagram of a 1-hop forwarding node of the present invention;

[0051] Figure 6 This is a schematic diagram of the TS_schedule scheduling message in the present invention;

[0052] Figure 7 This is a schematic diagram of the Data_complete message in the present invention;

[0053] Figure 8 This is a schematic diagram of the ClusterHeader_assign message in the present invention;

[0054] Fig. 9 The present invention is a flow chart of the cluster head node determination method. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Generally, the components of the embodiments of the present invention described and shown in the drawings here can be arranged and designed in various different configurations.

[0056] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0057] When the wireless sensor network nodes of the present invention perform the initial election of the cluster head node before clustering, the nodes use the age of information AoI to discard the election messages with low freshness, and then select the nodes to cluster based on the number of hops from the cluster head node as the restriction condition. In order to avoid excessive end-to-end transmission delay, the number of hops is set to 2 in the present invention, that is, the node whose data can reach the cluster head node after at most one forwarding can become a member of the cluster.

[0058] like Figure 1 As shown in the figure, assuming that node A is the cluster head node, the nodes around cluster head node A (white background) will form a cluster with cluster head node A according to the 2-hop criterion. Assuming that the number of nodes in the wireless sensor network is N, the number of nodes in each cluster is set to n, and the value of n can be customized by the network administrator according to the actual situation. In order to maximize the network life, if the amount of data to be collected each time is larger or (and) the collection frequency is higher, the average energy consumption of the node is higher, and the value of n should be small; if the amount of data to be collected each time is smaller or (and) the collection frequency is lower, the average energy consumption of the node is lower, and the value of n should be large.

[0059] like Fig. 9 As shown, the method for balancing the energy of cluster head nodes in a wireless sensor network of the present invention includes the steps of preliminary selection and clustering of cluster head nodes, data processing stage and assignment of new cluster head nodes, specifically:

[0060] Step 10: Preliminary selection and clustering of cluster head nodes

[0061] Step 11: Configure the Election message

[0062] In the initial state, the cluster head node is selected based on the principle of minimum ID address priority, that is, the nodes in the wireless sensor network are based on the principle of 2 hops. The cluster head node is determined by comparing the "Source Node ID" in the election message that appears periodically in the wireless sensor network, such as Figure 2The Election message shown includes:

[0063] ①F: 1-bit byte, indicating the function of the Election message. The default value is '0'. If it is '1', it indicates that the Election message is used for the initial election of the cluster head node.

[0064] ②Hop count: 2 bytes, recording the number of times the Election message is forwarded. The default value is '00'. The value increases by 1 each time it is forwarded.

[0065] ③Timestamp: 6-bit byte, recording the time when the Election message was sent;

[0066] ④Lifetime: 6 bytes, recording the lifetime of the Election message;

[0067] ⑤R: 1 byte, reserved for future use, the default value is '0';

[0068] ⑥Destination Node ID: 16-bit byte, records the hardware address of the node that receives the Election message. When used for cluster head node election, all 16 bits are '1';

[0069] ⑦Source Node ID: 16 bits, recording the device ID of the node that sends the Election message;

[0070] ⑧Forward Node ID: 16 bits, recording the device ID of the node forwarding the Election message. For an Election message that can be directly delivered to the cluster head node in one hop, all the values ​​here are '0'.

[0071] ⑨Data: variable length, records the amount of data that has been processed by the node that sent the Election message, in bytes;

[0072] ⑩PADDING: Variable length. When the length of "Data" is not an integer multiple of 16 bits, it is used to pad it to an integer multiple of 16 bits.

[0073] Step 12: Configure Source_ID_Hop table

[0074] Each node has a Source_ID_Hop table (see Figure 3) It is used to record the "Source Node ID", "Forward Node ID" in the Election messages sent by other nodes with a maximum "Hop count" of 2 hops, the number of times the Election message is forwarded (obtained through "Hop count"), and the node's own ID "Owner’s ID". When recording the "Source Node ID", it is queued according to the ascending order of "Hop count"; in the case of the same "Hop count", it is queued according to the ascending order of the node ID size. The Source ID Hop table can record data of n nodes, where n is the number of nodes in each cluster and can be customized by the network administrator according to the actual situation.

[0075] Before the primary election of the cluster head node ends, the node will continuously send / receive Election messages periodically. If the "Source Node ID" in the newly received Election message is greater than the last item in the Source ID Hop table, the Election message will be discarded; otherwise, the "Source Node ID" in the Election message will be inserted into the corresponding position in the Source ID Hop table, and the last item in the table will be squeezed out. For example: Before the primary election of the cluster head node ends, the Source ID Hop table of a certain node is full. At this time, an Election message from node X is received, with a "Hop count" of '01', and Owner’s ID < X’ID < A’ID. The information of node X will be inserted Figure 3 above "A’ID" in the table, and the information of node Y in the table will be squeezed out of the Source ID Hop table.

[0076] Step 13. Primary election and clustering of the cluster head node

[0077] Before conducting the primary election of the cluster head node, all nodes first perform clock synchronization, and the time required for synchronization can be customized by the network administrator according to the actual situation. After all nodes are clock synchronized, they start sending Election messages to conduct the primary election of the cluster head node;

[0078] Step 13.2: When each node receives an Election message, it first determines the freshness of the information age AoI of the Election message based on the sending time "Timestamp" and "Lifetime". If the difference between the local time when the Election message is received and the sending time "Timestamp", that is, the information age AoI is less than the "Lifetime" in the Election message, it means that the freshness of the Election message is high, and it goes to step 13.3. Otherwise, the Election message is discarded.

[0079] Step 13.3: After all nodes have exchanged multiple rounds of Election messages, the node with the smallest node ID will be able to obtain the information of n nodes that meet the maximum hop count of 2 hops, thus becoming the cluster head node;

[0080] Step 13.4, the cluster head node broadcasts Figure 4 The SYN message shown is used to announce cluster information, and the node that receives the message automatically joins the cluster;

[0081] Step 13.5: After clustering, all nodes in the cluster enter the data processing stage.

[0082] like Figure 4 The SYN message shown includes:

[0083] ①Cluster ID: bits, that is, the number of nodes N in the wireless sensor network is rounded up after the logarithm operation with base 2 is performed, and the length is variable.

[0084] ②PADDING: Variable length. When the length of "Cluster ID" is not an integer multiple of 16 bits, it is used to pad it to an integer multiple of 16 bits.

[0085] ③ClusterHeader ID: records the device ID of the cluster head node;

[0086] ④Member ID1~Member ID i : Record the device ID of cluster member nodes 1 to i;

[0087] ⑤Data: variable length, records the amount of data that the cluster head node has processed, in bytes, and the default value is all '0';

[0088] ⑥PADDING: variable length. When the length of "Data" is not an integer multiple of 16 bits, it is used to pad it to an integer multiple of 16 bits.

[0089] Step 20: Data processing stage

[0090] Each time the cluster head node assigns a time slot, it must avoid the situation where the 2-hop nodes near the node cannot send / receive data due to the sleep of the 1-hop node. Therefore, for each election message sent by the node, the cluster head node must also record the corresponding previous hop node ID "Forward Node ID" to ensure that there is always a 1-hop forwarding node (i.e., the previous hop node "Forward Node") that is not in sleep near the 2-hop node.

[0091] like Figure 5 As shown in the figure, assuming that after the initial selection of the cluster head node, node A becomes the cluster head node, for the 2-hop nodes C and D, the 1-hop nodes B and E are its forwarding nodes (this information can be obtained through the Source_ID_Hop table). Therefore, when performing time slot scheduling, if node C or D does not sleep, nodes B and E cannot sleep at the same time.

[0092] In addition, in order to ensure the energy balance of nodes to the greatest extent, the cluster head node also considers the amount of data that the node has sent / received each time it assigns a time slot. The cluster head node periodically sends TS_schedule scheduling messages (see Figure 6 ) Controls the time when a node sends / receives data. The first TS_schedule message should be sent slightly later than the SYN message broadcast so that all members in the cluster have enough time to complete the parsing of the SYN broadcast message. The sending interval of TS_schedule messages can be defined by the network administrator according to actual conditions.

[0093] like Figure 6 The TS_schedule scheduling message shown includes:

[0094] ①Cluster ID: consistent with the "Cluster ID" in the SYN message;

[0095] ②PADDING: Variable length. When the length of "Cluster ID" is not an integer multiple of 16 bits, it is used to pad it to an integer multiple of 16 bits.

[0096] ③ClusterHeader ID: records the device ID of the cluster head node;

[0097] ④Member ID1~Member ID i : 16 bits, recording the device IDs of cluster members 1 to i respectively;

[0098] ⑤TS1~TS i: The length is variable, recording the time slot number assigned by the cluster head node to cluster members 1~i. The total duration of the allocated time slot is less than the sending interval of the TS_schedule scheduling message, ensuring that the node using the last time slot has finished sending / receiving data before the next TS_schedule scheduling message arrives;

[0099] ⑥PADDING: variable length, when "TS1~TS i "When the length is not an integer multiple of 16 bits, it is used to pad it to an integer multiple of 16 bits;

[0100] After the node sends / receives data in the time slot assigned by the cluster head node, it immediately sends a Data_Complete message to the cluster head node (see Figure 7 ), then enters a dormant state and wakes up before the next TS_schedule scheduling message arrives; the Data_complete message includes:

[0101] Record the ClusterHeader ID of the cluster head node device ID;

[0102] Record the Member ID of the device ID of cluster member node i i ;

[0103] Cluster member node i sends / receives data "Data i ", that is, reporting the amount of processed data to the cluster head node, in bytes;

[0104] PADDING: When the length of "Data" is not an integer multiple of 16 bits, it is used to pad it to an integer multiple of 16 bits;

[0105] Step 30: Assign a new cluster head node

[0106] After K rounds of data transmission / reception (the value of K can be customized by the network administrator according to the actual situation), the amount of data processed by each cluster member node is different, resulting in different residual energy for each node. Among them, the energy consumption of the cluster head node is the most serious. In order to balance the energy of all nodes, the cluster head node must be updated every once in a while. In order to minimize the energy consumption during the cluster head node election process, the current cluster head node obtains the amount of data processed by the cluster member nodes through the Data_complete message reported by the cluster member nodes, and broadcasts the ClusterHeader_assign message (see Figure 8) directly assigns the cluster member node with the smallest amount of processed data as the new cluster head node. The ClusterHeader_assign message is generated periodically in the wireless sensor network, and the length of the period is the above-mentioned K value. After the new cluster head node is generated, return to step 2 and immediately broadcast the TS_schedule scheduling message to enter a new round of data processing phase.

[0107] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of various changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for balancing the energy of cluster head nodes in a wireless sensor network, characterized in that ,It includes three steps of the initial selection and clustering of cluster head nodes, the data processing phase and the assignment of new cluster head nodes, specifically: Step 10: Preliminary selection and clustering of cluster head nodes Configure the Election message and the Source_ID_Hop table in each node: The Election message at least includes: The flag indicating whether this Election message is used for the initial election of the cluster head node; Hop count that records the number of times the Election message is forwarded; Record the Timestamp of when the Election message was sent; Record the lifetime of the Election message; Record the Destination Node ID, the hardware address of the node that receives the Election message; Record the Source Node ID of the node device ID that sends the Election message; Record the Forward Node ID of the node that forwards the Election message. Record the amount of data that has been processed by the node that sent the Election message; The Source_ID_Hop table in each node at least includes: "Source Node ID", "Forward Node ID", and the number of times the Election message is forwarded, which are sent from other nodes with a "Hop count" of at most 2 hops. When recording the "Source Node ID", the nodes are queued in ascending order according to the "Hop count". If the "Hop count" is the same, the nodes are queued in ascending order according to the size of the node ID. Before the primary election of the cluster head node is completed, the node continues to send or receive election messages periodically. If the "Source Node ID" in the newly received election message is greater than the "Source NodeID" in the last row of the Source_ID_Hop table, the election message is discarded; otherwise, the "Source Node ID" in the election message is inserted into the corresponding position in the Source_ID_Hop table, and the last row of node information in the table will be squeezed out; When the wireless sensor network nodes are preliminarily elected before clustering, the nodes use the age of information AoI to discard the election messages with low freshness, and select the cluster member nodes to form a cluster based on the maximum number of hops from the cluster head node. That is, in the initial state, the preliminaries of the cluster head node follow the principle of minimum ID address priority, and the cluster head node is determined by comparing the "Source Node ID" in the election message that appears periodically in the wireless sensor network. The cluster head node broadcasts a SYN message to announce cluster information, and the nodes that receive the message automatically join the cluster; After clustering, all nodes in the cluster enter the data processing stage; Step 20: Data processing stage The cluster head node considers the amount of data that each cluster member node has sent / received each time when assigning a time slot, and controls the time when the cluster member nodes send / receive data by periodically sending TS_schedule scheduling messages; After the cluster member node sends / receives data in the time slot assigned by the cluster head node, it immediately sends a Data_Complete message to the cluster head node, then enters a dormant state and wakes up before the next TS_schedule scheduling message arrives; Step 30: Assign a new cluster head node After K rounds of data transmission / reception, the current cluster head node obtains the amount of data processed by the cluster member nodes through the Data_complete message reported by the cluster member nodes, and directly assigns the cluster member node with the smallest amount of processed data as the new cluster head node by broadcasting the ClusterHeader_assign message; after the new cluster head node is generated, return to step 20 and immediately broadcast the TS_schedule scheduling message to enter a new round of data processing phase.

2. The method for balancing cluster head node energy in a wireless sensor network according to claim 1, characterized in that: The maximum number of hops is 2 hops.

3. The method for balancing cluster head node energy in a wireless sensor network according to claim 1, characterized in that The node uses the age of information AoI to discard election messages with low freshness, specifically: Before the cluster head node is pre-selected, all nodes first synchronize their clocks and then start sending Election messages for the cluster head node pre-selection; When each node receives an Election message, it first determines the freshness of the information age AoI of the Election message based on the sending time "Timestamp" and "Lifetime". The difference between the local time when the Election message is received and the sending time "Timestamp" is set as the information age AoI. If the information age AoI is less than the lifetime "Lifetime" in the Election message, it means that the freshness of the Election message is high and proceeds to the next step. Otherwise, the Election message is discarded. The next step is that after all nodes have exchanged multiple rounds of Election messages, the node with the smallest node ID will be able to obtain information of n nodes that meet the maximum hop number of 2 hops, thereby becoming the cluster head node.

4. The method for balancing cluster head node energy in a wireless sensor network according to claim 1, characterized in that The SYN message at least includes: Cluster ID: Bit, that is, the number of nodes N in the wireless sensor network is rounded up after performing a logarithm operation with base 2; Record the ClusterHeader ID of the cluster head node's device ID; Member ID1 to Member ID used to record the device IDs of cluster member nodes 1 to i respectively i ; Record the amount of data that the cluster head node has processed.

5. The method for balancing cluster head node energy in a wireless sensor network according to claim 1, characterized in that The TS_schedule scheduling message at least includes: Cluster ID: Bit, that is, the number of nodes N in the wireless sensor network is rounded up after performing a logarithm operation with base 2; Record the ClusterHeader ID of the cluster head node's device ID; Record the Member ID1 to Member ID of the device ID of cluster members 1 to i respectively i ; Record the time slot numbers TS1~TS1 assigned by the cluster head node to cluster members 1~i i .

6. The method for balancing cluster head node energy in a wireless sensor network according to claim 1, characterized in that The Data_complete message at least includes: Record the ClusterHeader ID of the cluster head node device ID; Record the Member ID of the device ID of cluster member node i i ; Data sent / received by cluster member node i in the corresponding time slot i , that is, reporting the amount of processed data to the cluster head node.

Images