Wireless sensor network charging method for multi-charge nodes

This application is a 371 of international application of PCT application Ser. No. PCT/CN2015/095390, filed on Nov. 24, 2015, which claims the priority benefit of China application no. 201510237380.5, filed on May 12, 2015. The entirety of each of the above-mentioned patent applications is hereby incorporated by reference herein and made a part of this specification.

The invention relates to the field of wireless sensor network energy transmission, in particular, to a wireless sensor network charging method for multi-charge nodes.

Due to developments of magnetic resonance and magnetic coupling wireless energy transmission and research breakthrough of ultra-speed charging battery materials in recent years, much more attention has been put on researches regarding the distribution of moving charging nodes in wireless sensor network (also known as charging trolley) to achieve sensor energy supplement in the network.

Wireless sensor network is typically applied in surveillance sector, hundreds or thousands of sensors distributed in the network are used to collect data and transfer the data to a base station by using its own communication model directly or through jumpers. The base station collects all the information and sends the information to surveillance personnel to achieve the purpose of monitoring. Process of information transmission includes data treatment, integration and node-to-node communication, which are all the main causes of sensor energy consumption.

In order to guarantee the durability of the wireless sensor network (WSNs) during surveillance cycles without consuming all the energy in the network, a practical and effective method must be established to recharge the sensor nodes in the network and to achieve best charging trolleys dispatch which enables the total consumption of the charging trolleys to be the minimum in every charging task.

As put forward by Y. Shi in document, “On renewable sensor network works with wireless energy transfer. Proc. of INFOCOM, IEEE in 2011”, the use of distributing charging trolleys in WSNs, which can visit every sensor in the network periodically, to achieve sensor energy supplement is limited by fixed time intervals.

Researches related to charging methods for wireless sensor network distributed with multiple charging trolleys mainly involve two aspects of problem: one, choosing a shortest path to achieve minimum consumption; and two, scheduling of the charging trolley.

In order to avoid the limitation due to fixed time intervals in the traditional method, the invention provides a wireless sensor network charging method for multi-charge nodes.

The wireless sensor network charging method for multi-charge nodes of the invention is a self-adaptive charging method designed with respect to the independence of each charging trolley.

The self-adaptive charging method is set in the wireless sensor network (WSNs), in which all the charging trolleys are separated and mutually independent, conduct charging tasks within their own ranges, and are capable of automatically selecting a starting time for charging process. The above work is repeated every time after completing the charging task, and it is not limited by time intervals and not affected by other charging trolleys.

Because all the charging trolleys are mutual independent, a degree of complication of the method is only limited within the complication degree of distributing one charging trolley.

Techniques Adopted the Invention:

A WSNs charging method for multi-charge nodes, including the following steps:

(1) establishing a WSNs model: by randomly distributing hundreds or thousands of sensors in a large surveillance area, with q number of charging trolleys and q number of parking lots, wherein a surveillance cycle is T; a set of sensors being V, that is: V={v 1 , v 2 , v 3 . . . } and v 1 , v 2 and v 3 are respectively a first, a second and a third sensors; a battery capacity of each sensor being B, a consumption rate of the j th sensor being ρ j , a low energy alarm threshold being M j , M j =α·B, and 0<α<1, wherein α is a percentage of the low energy alarm threshold M j accounted for the sensor energy B; i being a serial number of the charging trolley which is parked in the parking lot at a position r i , r i =(x i , y i ), 1≤i≤q, x i and y i respectively representing the two dimensional map coordinates of the position r i , q being the quantity of the charging trolleys; a base station being used to collect sensor information and communicate with the charging trolleys; the parking lots being used to recharge the charging trolleys; the battery capacity of every charging trolley being E, a moving speed being a stable value S, and a sensor charging time for every charging trolley being a fixed value C; (2) dividing a field range for every charging trolley; (3) conducting a charging task by each charging trolley.

Further, in the aforesaid step (2), detail steps of dividing the field range for every charging trolley are:

(2.1) generating an expansion node: by using a closed circuit η to encircle q number of charging trolleys and their parking lots in the sensor, wherein the circuit η does not include other sensor node therein, thus the circuit η is the expansion node;

(2.2) by using the expansion node as a root node, the sensor node being the smallest spanning tree ψ generated by a tree branch node;

(2.3) decomposing the smallest spanning tree y: by taking q number of parking lots as the root nodes, decomposing the smallest spanning tree into q number of un-intersected root trees, and the upper limit for the total number of nodes in every root tree being:

K = ⌈ A q ⌉ ,

wherein A is the total number of sensor nodes in the sensor network, q is the quantity of charging trolleys, and ┌·┐ indicates to round up; A being the total number of sensor nodes in the sensor network;

(2.4) connecting outer surrounding nodes of every root tree to form q number of circuits ζ 1 ζ 2 . . . ζ q , wherein the circuit ζ i (i=1, 2, . . . , q) represents the field range of the i th charging trolley.

Further, in the aforesaid step (3), detail steps of conducting the charging task by each charging trolley are:

(3.1) the charging trolleys receiving a total number of alarm sensor nodes and a serial number of an alarmed sensor node, and initializing: l=0, j=0, wherein l is the total number of alarm sensor nodes being received, and j is the serial number of the alarmed sensor node;

(3.2) the charging trolleys starting to receive signals emitted from a low energy sensor, the charging trolleys updating the values of l,j after receiving the alarm signals and generating a shortest charging path s l for every l value, and the charging trolleys calculating an energy discriminant vector Q, wherein l is the total number of alarm sensor node being received, and j is the serial number of the alarmed sensor node;

(3.3) if an element Q ij of the energy discriminant vector Q is ≤5% B and j=1, 2, . . . , l, wherein l is the total number of alarm sensor nodes being received, j is serial number of the alarmed sensor node, and B represents the battery capacity of each sensor, then conducting the step (3.4) based on the shortest charging path s l , otherwise, returning back to the step (3.2);

(3.4) conducting the charging task: that is, the charging trolleys start to charge the sensors;

(3.5) the charging trolleys determining whether returning back to the parking lots for recharging is required, and if it is required, then the charging trolleys return back to parking lots for recharging; otherwise, returning back to the step (3.2).

Further, in the aforesaid step (3.2), the method of generating the shortest charging path s l is: a set of all the alarmed sensors being V c =(v 1 c , v 2 c , v 3 c . . . v l c ); by using the charging trolleys as the start-point, selecting a path e m with the smallest Euclidean distance in l number of paths that connect with the charging trolleys, wherein e m connects another alarmed sensor node v i c with the charging trolleys; next, by using the sensor node v i c as the start-point, selecting a path e n with the smallest Euclidean distance from the remaining l−1 paths which excludes the path e m , wherein e n connects the alarmed sensor node v i c with the alarmed sensor node v i c ; and in this analogy, obtaining the shortest charging path s l in which the charging trolleys are being used as the start-point and l number of to-be-charged sensor nodes are being passed by.

Further, in the aforesaid step (3.2), the meaning of the energy discriminant vector Q is: a vector consisting of the minimum residual energy under conditions of assuming that the charging task starts when l number of alarm signals are received, and the j th alarmed sensor node is arranged to be charged at the last.

Further, in the aforesaid step (3.2), the form of the energy discriminant vector Q is: └Q l1 Q l2 . . . Q lj . . . Q ll ┘, j=1, 2, . . . , 1; and a method of calculating the element Q lj of the energy discriminant vector is:

Q lj = M j - ρ j · [ D S + l · C + ( t - t j ) ] ,

wherein M j is a low energy alarm threshold, the threshold is set as M j =20% B, B represents the battery capacity of each sensor, ρ j is an energy consumption rate of the j th alarmed sensor node, l is the total number of alarm sensor nodes being received, D is a total length of the shortest charging path corresponded by l, S is the moving speed of the charging trolleys, C is a charging time required by each sensor, C is a constant value, t is the current time that the charging trolleys calculate the vector element, and t j is an alarming time of the j th alarm sensor node recorded by the charging trolleys.

Further, in the aforesaid step (3.5), bases for the charging trolleys to determine whether returning back to parking lots for recharging is required are: a remaining energy E μ of the charging trolleys being ≤5% E, wherein E μ is assumed to be the energy remained in the charging trolleys after the charging trolleys returned back to the parking lots to conduct the charging task again, and E is a battery capacity of the charging trolleys;

E μ is calculated according to: E μ =E′−(Q r→r i +Q c +ΔQ), wherein E′ is a remaining energy at the current moment when the charging trolleys made the determination, Q r→r′ i , is an energy consumed while the charging trolleys completed the charging task and returned back to parking lots, ΔQ is an extra energy loss by the charging trolleys under the influence of surrounding environment, and Q c is an energy expected to be consumed when conducting the charging task again.

Q c is calculated according to: Q c =λ·D′+(l−l′)·B−(B′ l′+1 +B′ l′+2 + . . . B′ l ), wherein λ is an energy consumption of the charging trolleys per unit distance, l is the total number of alarm sensor nodes being received, l′ represents a total number of alarm released sensor nodes at the moment that the charging trolleys made the determination, D′ is a total length of the shortest path generated corresponding to l at the moment that the charging trolleys made the determination, B′ l′+1 , is a remaining energy of the l′+1 th alarm sensor node, B′ l′+2 is a remaining energy of the l′+2 th alarm sensor node, B′ l′ is a remaining energy of the l′ th alarm sensor node, and B is the battery capacity of the sensors.

The invention, as compared to the convention technology, has the following advantages:

(1) The sensor energy consumption rate and the remaining energy can be considered as together as a whole and do not need to be analyzed solely, thereby avoiding time constraints set forth by charging at fixed time intervals, and being capable of self-adaptively changing a charging strategy based on the variation of the sensor energy information in the network.

(2) By utilizing the invention, all the charging trolleys are mutually independent, and the method complexity is equivalent to the complexity of deploying one charging trolley, and thus is simple and reliable.

FIG. 1 is a flow chart illustrating a wireless sensor network charging method according to the invention:

FIG. 2 is a block diagram illustrating the steps of dividing the field range for every charging trolley according to the invention.

The invention will now be described in detailed with reference to the preferred embodiments and accompanying drawings.

A wireless sensor network charging method for multi-charge nodes, including the following steps:

(1) establishing a WSNs model: by randomly distributing hundreds or thousands of sensors in a large surveillance area, with q number of charging trolleys and q number of parking lots q, wherein a surveillance cycle is T;

a set of sensors is V, that is: v={v 1 , v 2 , v 3 . . . } and v 1 , v 2 and v 3 are respectively a first, a second and a third sensors; a battery capacity of each sensor being B, a consumption rate of the j th sensor being ρ j , a low energy alarm threshold being M j , M j =α·B, and 0<α<1, wherein α is a percentage of the low energy alarm threshold M j accounted for the sensor energy B;

i being a serial number of the charging trolley which is parked in the parking lot at a position r i , r i =(x i , y i ), 1≤i≤q, x i and y i respectively representing the two dimensional map coordinates of the position r i , q being the quantity of the charging trolleys; a base station being used to collect sensor information and communicate with the charging trolleys; the parking lots being used to recharge the charging trolleys; the battery capacity of every charging trolleys being E, a moving speed being a stable value S, and a sensor charging time for every charging trolley being a fixed value C;

(2) dividing a field range for every charging trolley;

(3) conducting a charging task by each charging trolley.

Further, in the aforesaid step (2), detail steps of dividing the field range for every charging trolley are:

(2.1) generating an expansion node: by using a closed circuit q to encircle q number of charging trolleys and their parking lots in the sensor network, wherein the circuit η does not include other sensor node therein, thus the circuit η is the expansion node;

(2.2) by using the expansion node as a root node, the sensor node being the smallest spanning tree ψ generated by a tree branch node;

(2.3) decomposing the smallest spanning tree ψ: by taking q number of parking lots as the root nodes, decomposing the smallest spanning tree into q number of un-intersected root trees, and the upper limit for the total number of nodes in every root tree being:

K = ⌈ A q ⌉ ,

wherein A is the total number of sensor nodes in the sensor network, q is the quantity of charging trolleys, and ┌·┐ indicates to round up; A is the total number of sensor nodes in the sensor network;

(2.4) connecting outer surrounding nodes of every root tree to form q number of circuits ζ 1 , ζ 2 . . . ζ q , wherein the circuit ζ i (i=1, 2, . . . , q) represents the field range of the i th charging trolley.

Further, in the aforesaid step (3), detail steps of conducting the charging task by each charging trolley are:

(3.1) the charging trolleys receiving a total number of alarm sensor nodes and a serial number of an alarmed sensor node, and initializing: l=0, j=0, wherein l is the total number of alarm sensor nodes being received, and j is the serial number of the alarmed sensor node.

(3.2) the charging trolleys starting to receive signals emitted from a low energy sensor, the charging trolleys updating the values of l,j after receiving the alarm signals and generating a charging path s l for every l value, and the charging trolleys calculating an energy discriminant vector Q, wherein l is the total number of alarm sensor nodes being received, and j is the serial number of the alarmed sensor node;

(3.3) if an element Q lj of the vector quantity is ≤5% B and j=1, 2, . . . , l, wherein l is the total number of alarm sensor nodes being received, j is the serial number of the alarmed sensor node, and B represents the battery capacity of each sensor, then conducting the step (3.4) based on the shortest charging path s l , otherwise, returning back to the step (3.2);

(3.4) conducting the charging tasks: that is, the charging trolleys start to charge the sensors;

(3.5) the charging trolleys determining whether returning back to the parking lots for recharging is required, and if it is required, then the charging trolleys return back to the parking lots for recharging; otherwise, returning back to the step (3.2);

In the aforesaid step (3.2), the method of generating the shortest charging path s l is: a set of all the alarmed sensors being V c =(v 1 c , v 2 c , v 3 c . . . v l c ); by using the charging trolleys as the start-point, selecting a path e m with the smallest Euclidean distance in l number of paths that connect with the charging trolley, wherein e m connects another alarmed sensor node v i c with the charging trolleys; next, by using the sensor node v i c as the start-point, selecting a path e n with the smallest Euclidean distance from the remaining l−1 paths which excludes the path e m , wherein e n connects the alarmed sensor node v i c with the alarmed sensor node v i c ; and in this analogy, obtaining the shortest charging path s l in which the charging trolleys are being used as the start-point and l number of to-be-charged sensor nodes are being passed by.

In the aforesaid step (3.2), the meaning of the energy discriminant vector Q is: a vector consisting of the minimum residual energy under conditions of assuming that the charging task starts when l number of alarm signals are received, and the j th alarmed sensor node is arranged to be charged at the last.

In the aforesaid step (3.2), the form of the energy discriminant vector Q is: └Q l1 Q l2 . . . Q lj . . . Q ll ┘, j=1, 2, . . . , l; and a method of calculating the element of Q lj of the energy discriminant vector is:

Q lj = M j - ρ j · [ D S + l · C + ( t - t j ) ] ,

wherein M j is a low energy alarm threshold, the threshold is set as M j =20% B, B represents the battery capacity of each sensor, ρ j is an energy consumption rate of the j th alarmed sensor node, l is the total number of alarm sensor nodes being received, D is a total length of the shortest charging path corresponded by l, S is the moving speed of the charging trolleys, C is a charging time required each sensor, C is a constant value, t is the current time that the charging trolleys calculate the vector element, and t j is an alarming time of the j th alarm sensor node recorded by the charging trolleys.

In the aforesaid step (3.5), bases for the charging trolleys to determine whether returning back to parking lots for recharging, is required are: a remaining energy E μ of the charging trolleys being ≤5% E, wherein E μ is assumed to be the energy remained in the charging trolleys after the charging trolleys returned back to parking lots to conduct the charging task again, and E is a battery capacity of the charging trolleys;

E μ is calculated according to: E μ =E′−(Q r→r i +Q c +ΔQ), wherein E′ is a remaining energy at the current moment when the charging trolleys made the determination, Q r→r i is an energy consumed while the charging trolleys completed the charging task and returned back to the parking lots, ΔQ is an extra energy loss by the charging trolleys under the influence of surrounding environment, and Q c is an energy expected to be consumed when conducting the charging task again.

Q c is calculated according to: Q c =λ·D′+(l−l′)·B−(B′ l′+1 +B′ l′+2 + . . . B′ l ), wherein λ is an energy consumption of the charging trolleys per unit distance, l is the total number of alarm sensor nodes being received, l′ represents a total number of alarm released sensor node at the moment that the charging trolleys made the determination, D′ is a total length of the shortest path generated corresponding to l at the moment that the charging trolleys made the determination, B′ l′+1 is a remaining energy of the l′+1 th alarm sensor node, B′ l′+2 is a remaining energy of the l′+2 th alarm sensor node, B′ l′ is a remaining energy of the l′ th alarm sensor node, B is the battery capacity of sensors.