Research on topology control of wireless sensor network based on energy level

Research on topology control of wireless sensor network based on energy level

absrtact: in the planning and design of wireless sensor network, it is the first important problem to reduce the energy consumption of the nodes, prolong their working time and maximize the life cycle of the network. This paper designs a topology control strategy based on node energy level, which is designed for excessive energy consumption of nodes near the aggregation node, and avoids the network failure due to the premature depletion of these nodes, which makes the node energy consumption more balanced and prolongs the life of the network. Finally, the validity of the method is verified by the program simulation.

Keywords: wireless sensor network, energy level, network topology

A Ask questions

The minimum energy consumption route is a path that selects the minimum energy consumption of the node from the data source to all the paths of the aggregation node; the least hop routing is the path that selects the fewest hops from the data source to the aggregation node. However, in wireless sensor networks, if the data is transmitted frequently using the same path, the nodes on the path will be prematurely invalidated by the energy consumption, so that the whole network can be divided into isolated parts and reduce the lifetime of the whole network. In order to balance the network load and the energy consumption of the nodes, when some nodes consume more energy, while others have little energy consumption, the corresponding measures should be taken to make the node energy consumption relatively average, so as to avoid the failure of some key energy-consuming nodes (such as sink node or cluster head node), which can cause the whole network to be paralyzed. Therefore, many researchers have given a lot of topological control algorithms from the point of view of energy consumption of equilibrium nodes.

In this paper, a topology control strategy based on node energy level is designed to avoid the network failure of these nodes due to the premature exhaustion of energy and prolong the life of the network.

Two System Model 2.1 Network model in this paper, N sensor nodes are randomly distributed within a square region A, and the sensor network has the following properties: (1) The only base station is deployed in a distant location outside of region A;

(2) Each node has a unique identity;

(3) The initial energy of each node is basically the same;

(4) Sensor node is not moved after deployment;

(5) The location information of the node is known;

(6) Transmission power can be adjusted when the transmitting node communicates with the receiving node at different distances;

(7) The energy consumption of the nodes in each round is not uniform.

2.2 Wireless Communication model

In this paper, we use the first-order wireless energy model which is more commonly used now: fusing n kbit information and transmitting it to d distance, the energy consumed is:

(1) The energy consumed by receiving the data is:

(2) Sending the data consumes energy:

In addition, most protocols and algorithms employ data aggregation techniques to reduce the amount of data sent and received,

So as to achieve the purpose of saving energy. The strategy of this paper also uses data aggregation technology to reduce energy loss.

In addition, the paper assumes that the wireless channel is symmetric, that is, the energy consumed by transmitting messages from the node to Node B equals the energy consumed by transmitting the message m to the node port from Node B.

Three Topology Control Policy Description 3. 1 node Set layer ID

This policy layers all nodes in the sensor network at a distance from the base station, when the system is initialized

Set a layer ID for each node. Nodes with the same layer ID are at the same level, and their distances to the base station are essentially the same, and the energy overhead of communication with the base station is similar. The closer the base station is, the smaller the layer ID of the node, in the policy, the node with the small layer ID is called the upper node, the node with the layer ID is called the lower node, and the node with the same layer ID is called the same layer node. The system divides the sensor nodes in the whole network into multiple levels according to the distance from the base station at the time of initialization. During the layering process, the base station sends a layer message to all nodes in the network first. Each node determines its distance from the base station based on the signal strength of the layer message. This paper uses the failure of free space

Reduction Model:

wherein, for the base station transmit power, to receive the signal power, D is the transmission distance, is the decay

The coefficient of reduction, usually set to constant, so the distance from the sensor node to the base station can be calculated in the formula:

Where, for Constant, D is the size of the layer radius (usually taken), if

，

The node will record its own ID layer as K.

3. 2-layer nodes into chains

The basic information for any node includes the Node ID, layer ID, remaining energy resenergy, energy threshold Thresenegry, and the next node NextNode, which is five tuples:

Broadcast (Nodeid,layerid,resenergy,thresenegry,nextnode)

The node ID is uniformly encoded. The layer ID indicates the distance of the node from the base station and is set when the system is initialized. The remaining energy records the current remaining energy of the node. The energy threshold is used to determine if the node has sufficient energy to be used as the leader node of the layer. The node ID, layer ID, and nextnode do not change in the sensor node life cycle. The remaining energy will change dynamically with the energy consumption of the node, and the energy threshold is broadcast by the base station at irregular intervals.

After the layer ID is determined, all nodes in each layer broadcast broadcast to this layer node, thus determining the highest energy node in the layer as the first leader node of the layer and initializing the node information, the node sends the message broadcast (Nodeid,layerid, Resenergy, Thresenegry, nextnode), this layer receives this message from the nearest node, and feeds back its node identity Nodeid information, when the node initializes its NextNode node to Nodeid. Similarly, the node also sends broadcast to the node closest to itself, the node that receives the message also feeds back its nodeid as NextNode, and so on, until all nodes of this layer are linked, it is important to note that the last node of this layer does not have the next node. So its nextnode is empty.

3. Data transfer between 3-tier nodes

The network connection information Records basic information about the link to the upper Leader sensor node, including the node ID, layer ID, and remaining energy resenergy. The connection information is represented by 3 queues: Parent node Queue (parents), sibling node queue (siblings), and child node queue (children). These 3 queues are built at system initialization, the parent queue holds the layer ID smaller than its own upper neighbor node, the sibling queue holds the same layer ID as its own neighbor node, and the sub-queue storage layer ID is larger than its own lower neighbor node.

All nodes on each layer send the information collected by the sensor to the leader node of the layer. At the same time, the data of the leader node of the lower node is collected, and then the data is fused and sent to the upper node. And so on, until the data is sent to the sink node.

It is conceivable that because leader nodes receive and send large amounts of data, the energy will consume more quickly.

For the balance of energy consumption, when the remaining energy Resenergy of the leader node is less than Thresenegry, the leader node of this layer will be changed to NextNode node.

With the energy consumption of the nodes, when the remaining energy of all the nodes in this layer is less than thresenegry, the layer leader nodes will be converted logically to the lower leader nodes, that is, the sensor signals collected by the nodes in this layer are sent to the lower nodes. The lower leader node of this layer sends data to the leader node on the upper level of the layer.

3. 4 Top-level node migration

Nodes that communicate directly with sink nodes are called top-level nodes. The top node satisfies the following two conditions: 1) The parent node queue is empty, 2) the remaining energy of the node itself is greater than the pre-set threshold value.

Because the top node communicates directly with the base station, its energy consumption is faster than that of the non-top-level nodes. In the When top layer node

When the energy of the leader node is less than the pre-set threshold Thresenegry, it no longer qualifies for communication with the base station. In this case, the strategy transfers the task of communicating with the base station to the leader node, which is a little farther from the base station but with sufficient energy, a mechanism called top-level node migration. Through the top node migration, the system can evenly distribute the energy loss to all nodes in the network. In the case of a top-level node migration, the top-level node swaps itself with the parent-child of the underlying neighbor node. The node first receives data from the child nodes, and then aggregates the received data with its own data.

If the node receives multiple header messages, the message forwarding with the highest energy level is selected, and the message with low energy level is discarded. When the energy of a layer node is insufficient and the energy of all neighboring nodes in its sibling queue is less than the set threshold, the node needs to swap the parent-child relationship with the adjacent underlying node:

(1) Transfer the adjacent downlevel nodes in the node's sub-queue to the parent queue;

(2) Transfer the node from the parent queue of the corresponding downlevel node to the child queue.

When the node itself or the same node in the sibling queue has sufficient energy, the data is transferred parallel to the nodes in this layer, and when the node itself and the same node in the sibling queue have insufficient energy, the transfer of this node to the lower node

Parent-child relationships.

As the network energy continues to drain, the top node will be migrated from the node closest to the base station to the nearest base station

The node. Finally, the energy of all the nodes in the system is less than the energy threshold value thresenegry. Because the packet sent to the base station contains the remaining energy value of the sending node, when the base station discovers that the energy value of the sending node is less than Thresenegry, the thresenegry is multiplied:

The base station then broadcasts the new threshold to all nodes, and all nodes thresenegry their own. Back to the top-level node from the base station 0-tier node. The system begins the next round of top-level node migrations until the network energy is exhausted.

3. The 5 policy steps are as follows:

1): Each node initializes the layer ID, remaining energy resenergy. Get each layer Layerid according to the distance.

2): Each node broadcasts a broadcast (NODEID,LAYERID,RESENERGY,NULL,NULL) message at this level, obtaining the highest energy node of this layer as the leader node of this layer.

3): The leader node as the chain head, the shortest distance as the scale of all the nodes of this layer are connected to a chain.

4): Each leader node receives the data of the lower leader node and merges it with the data obtained at this layer and then passes up a layer of leader node.

5): When the remaining energy Resenergy of the top-level leader node communicating directly with the sink node is lower than Thresenegry, the parent-child relationship between the top leader node and the lower leader node is exchanged through the top-level migration node. If the remaining energy Resenergy of all leader nodes is lower than thresenegry, go to 4).

Four. Program Emulation 4.1 Simulation Environment

This emulator platform: window7+vs2012+opencv2.4.4. OpenCV2.4.4 specific configuration method, please refer to the official website: http://wiki.opencv.org.cn/index.php.

4.2 Concrete Steps

1) randomly generate nodenum=100 nodes in the 600*600 pixel image, where the point in the upper-left corner is the sink node with coordinates (0,0). 1:

Figure 1

2) obtain each layer Layerid according to the distance from the sink node.

Figure 2

3) to obtain each layer of the leader node, the program to each layer of the leftmost node as a layer leader node, it should be assumed to start so the node energy is basically the same. 3 (the 5 green nodes on the left represent the leader nodes of each layer):

Figure 3

4) with each layer of the leader node as the chain head, the shortest distance as the scale of all the nodes of this layer connected to the chain, 4:

Figure 4

5) leader nodes are migrated in this layer. Assume that the energy consumed by the signal sent in each layer is squared away from the distance.

Fig. 5 Fig. 6

Fig. 7 Fig. 8

Fig. 9 Fig. 10

Where the color of the node represents the remaining energy of the node Resenergy:

Blue: resenergy > 80%

Yellow: 80% > Resenergy > 60%

Orange: 60% > Resenergy > 40%

Red: 40% > Resenergy > 20%

Grey: 20% > Resenergy > 10%

Black: 10% > Resenergy > 0

A large black dot represents the first sensor node that runs out of power.

4.3 Results comparison

A comparison of the variable power transmit signal (Figure 10) and the constant power transmit signal (Figure 11):

Fig. 10 Fig. 11

You can see: Figure 10 The top layer of energy is basically depleted, and figure 11 energy consumption is uneven, there are many orange nodes and yellow nodes.

It is shown from Figure 12 and figure 13 that the transfer of the layer leader node of the variable power transmit signal can save energy.

Fig. 12 Fig. 13

4.4 Source Code

See appendix.

Five. Summary

This chapter draws on the advantages of several classical algorithms, such as Leach algorithm, teen algorithm, pegasis algorithm, PEDAP algorithm, and gives a topology control strategy based on node energy level, which divides the nodes in the network into different levels according to the distance sink nodes. The nodes with high energy levels in each layer are selected to communicate with each other, and then the nodes closest to the sink node and the remaining energy greater than the given threshold are selected to communicate directly with the sink node, thus avoiding premature death of the nodes near the sink node and causing the network to die. This strategy makes the energy consumption of nodes in the network more balanced, thus prolonging the life of the network.

Research on topology control of wireless sensor network based on energy level