CN104320835A - Improvement method based on LDS node sleeping strategy - Google Patents

Abstract

The invention belongs to the energy conservation field of wireless sensor networks, and provides an improvement method based on an LDS node sleeping strategy. A conception of data activeness h is introduced, the data activeness is used as a factor to influence the node sleeping probability, and the greater the data activeness is, the smaller the sleeping probability of the node is. The invention designs a method capable of realizing the requirement of energy conservation and prolonging network lifetime. Under the circumstance of guaranteeing the energy conservation of the wireless sensor network, the effictiveness of data measurement in the sensor network is improved while an LDS algorithm is improved.

Description

Based on improving one's methods of LDS node dormancy strategy

Technical field

The present invention relates to radio sensing network energy-saving field, particularly relate to improving one's methods based on LDS node dormancy strategy.

Background technology

Due to wireless sense network low-power consumption, low rate, low cost characteristic, so design wireless sense network time must consider energy-conservation factor.In communication energy-saving: for network system, energy-conservationly mainly comprise two kinds of modes, a kind of is under the condition of node dense distribution, redundancy, by periodic scheduling, makes the node dormancy of redundancy to extend the life span of whole network; Another kind is the efficient routing mechanism of design energy and improves topology of networks, is reached the object extending network lifetime and improve energy efficiency by the traffic load in balancing network; For individual equipment, main energy saving means is dormancy, by closing wireless transceiver in the period without communication requirement, avoids the energy dissipation that idle listening causes.

For individual node, we will consider the life-span of battery and the accuracy of measurement.The life-span that we wish battery, the longer the better, accuracy is more high better simultaneously, find under the condition that traffic rate is identical through correlative study, the sampling period of transducer and the battery life of equipment are proportional, but the sampling period of prolongation transducer can affect control precision.Formulating that relevant dormancy strategy can average out in accuracy and energy saving thus will be the direction that we study.Wherein have two kinds of dormancy algorithms, the first is RS algorithm, this algorithm: cluster head node enters resting state with a certain probability selection node, and node only needs the random number reading random number generator generation, and then judges whether dormancy.Whether the sleep of program interior joint is random distribution, for all nodes, their dormancy probability is all the same, but the program is not considered in single-hop networks, the energy that the node far away apart from the node of cluster head consumes can be larger, can affect the life cycle of network like this.In addition this network is that the angle of statistical probability selects whether dormancy, and in the face of accident, its efficiency is lower comparatively speaking.The second is LDS algorithm, this algorithm is to the improvement of RS algorithm on space factor, but in the face of accident, there will be the unreasonable of node measurement Resourse Distribute, its efficiency is still lower, and therefore we also will consider the impact of measurement result on dormancy probability.

Have studied common dormancy algorithm in the Central China University of Science and Technology in 2008 one section of Master's thesis " the node dormancy algorithm research of industrial wireless sensor network " and have RS algorithm, but RS algorithm does not embody the three-dimensional effect of node, then also been proposed the innovatory algorithm LDS algorithm of RS, the network life cycle of this algorithm improves a lot, but its efficiency is not very high, and author proposes the node dormancy dispatching algorithm based on AHP simultaneously, this algorithm adopts and is divided into model, but algorithm is complicated, not too practical for the computing capability that wireless senser is lower.

And they are from macroscopic perspective to improve network lifetime, but in the face of a certain specific event, above-mentioned algorithm just can not embody, thus causes the efficiency of measurement on the low side.So the Coverage-preserving density control algorithm designing a lightweight will be research emphasis realizing energy-conservation and balance that is efficiency.

Sensor network obvious defect is exactly finite energy, so in WSN design, be first also most important challenge be exactly energy efficiency, namely realize the perfect adaptation of energy and efficiency.Generally we are by making part of nodes dormancy, and part of nodes enlivens to realize energy-conservation, gives an example below to illustrate:

A certain plat as shown in Figure 1, this divide into three parts, coastal, forest, desert, in a period of time of coastal area, variations in temperature is very little, and desert area is relatively very large in this period variations in temperature.Now in desert be coastally uniformly distributed the temperature that temperature sensor node measures desert and coastal area respectively, the distribution density of their sensor node is all the same.We suppose that desert area is the same with the dormancy probability of coastal area sensor node now, the result that can cause like this is exactly, the unreasonable of Resourse Distribute is measured in desert area and coastal area, corresponding is, the larger ratio of desert area variations in temperature is very fast, the dormancy probability of temperature nodes should be smaller, allow more node be in measuring state as far as possible, such data measured just more accurately can reflect the fluctuation situation of this area's temperature, and the variations in temperature of coastal area is less slow, the dormancy probability of temperature nodes should be larger, a little less node is allowed to be in active state, thus realize energy-conservation.Traditional RS dormancy mechanism is in tufted net, in realizing bunch, each node dormancy probability is impartial, but it does not consider from the distant node of cluster head the consumed energy that to communicate with cluster head large, can cause like this fringe node can very fast energy ezpenditure totally, and the LDS dormancy mechanism introduced afterwards introduce node and cluster head distance this because usually affecting dormancy probability, particularly, distance cluster head is far away, dormancy probability is larger, but, still the irrational problem that there is node measurement Resourse Distribute can be there is in above-mentioned Fig. 1 in LDS, thus, need to improve LDS dormancy mechanism.

Summary of the invention

The present invention is directed to above-mentioned prior art Problems existing to make improvements, namely the technical problem to be solved in the present invention is to provide a kind of improving one's methods based on LDS node dormancy strategy, thisly devise a kind of algorithm can realize energy-conservation requirement based on improving one's methods of LDS node dormancy strategy, extend network lifetime.When ensureing that wireless sense network is energy-conservation, LDS algorithm is improved to the validity that simultaneously improve Sensor Network DATA REASONING.In order to solve the problems of the technologies described above, the invention provides following technical scheme:

A kind of improving one's methods based on LDS node dormancy strategy, introduces the concept of a data liveness h, and because usually affecting node dormancy probability, data activity degree is larger as one for this data activity degree, and the dormancy probability of node is less.

In LDS dormancy mechanism, the Probability p (x) that node enters dormancy is linear with the distance x to cluster head, hypothetical probabilities is p (x), and cluster head and bunch interior nodes thereof are all distributed in the border circular areas that centered by cluster head, radius is R, shown in following formula:

$p\left(x\right)=\left\{\begin{array}{c}\mathrm{Cf}\left(x\right),0\&le;x\&le;x_s\\ 1,x_s<x\&le;R\end{array}\right.,$

If these area measure data are comparatively active for cluster node, its corresponding dormancy probability is less, and we introduce liveness to affect dormancy probability, then the new probability formula of the LDS dormancy strategy improved becomes as follows:

$p(x,h)=\left\{\begin{array}{c}\mathrm{Cf}(x,h),0\&le;x\&le;x_s,0<h<t\\ 1,x_s<x\&le;R\end{array}\right.,$

Wherein wherein ε is this bunch of group's dormancy node dormancy probability, and wherein h is larger and p (x, h) is less.There is the kind balancing method of several measurement data liveness (h) below:

Select gradiometry method, wherein β is coefficient, x _nfor a certain node n-th measures numerical value, T _nfor n-th measuring period of a certain node.

Also can select variance mensuration, wherein wherein β is coefficient, x _nfor a certain node n-th measures numerical value, T _nfor n-th measuring period of a certain node.

And select measurement relative entropy method, if the absolute value measuring numerical difference in the measurement data unit interval is ▽ x, this its measurement entropy can represent: $H_C\left(X\right)=-\underset{-\&infin;}{\overset{\&infin;}{\&Integral;}}{p}_X\left(x\right)\log p_X\left(x\right)\mathrm{dx}$

Then liveness h=β H _c(X), wherein β, κ are all coefficients, and ▽ x is the absolute value measuring numerical difference in the unit time.

Improving one's methods based on LDS node dormancy strategy provided by the invention, its beneficial effect is: devise a kind of algorithm and can realize energy-conservation requirement, extend network lifetime.When ensureing that wireless sense network is energy-conservation, LDS algorithm is improved to the validity that simultaneously improve Sensor Network DATA REASONING.

Accompanying drawing explanation

Accompanying drawing is used to provide a further understanding of the present invention, and forms a part for specification, together with embodiments of the present invention for explaining the present invention, is not construed as limiting the invention.In the accompanying drawings:

Fig. 1 is regional feature schematic diagram;

Fig. 2 is the temperature-time data and curves figure in a certain place of the node measurement utilizing LDS dormancy strategy;

Fig. 3 is the temperature-time data and curves figure in a certain place of the present invention one node measurement.

Embodiment

The present invention discloses a kind of improving one's methods based on LDS node dormancy strategy, introduces the concept of a data liveness h, and because usually affecting node dormancy probability, data activity degree is larger as one for this data activity degree, and the dormancy probability of node is less.

In LDS dormancy mechanism, the Probability p (x) that node enters dormancy is linear with the distance x to cluster head, hypothetical probabilities is p (x), and cluster head and bunch interior nodes thereof are all distributed in the border circular areas that centered by cluster head, radius is R, shown in following formula:

$p\left(x\right)=\left\{\begin{array}{c}\mathrm{Cf}\left(x\right),0\&le;x\&le;x_s\\ 1,x_s<x\&le;R\end{array}\right.,$

If these area measure data are comparatively active for cluster node, its corresponding dormancy probability is less, and we introduce liveness to affect dormancy probability, then the new probability formula of the LDS dormancy strategy improved becomes as follows:

$p(x,h)=\left\{\begin{array}{c}\mathrm{Cf}(x,h),0\&le;x\&le;x_s,0<h<t\\ 1,x_s<x\&le;R\end{array}\right.,$

Wherein wherein ε is this bunch of group's dormancy node dormancy probability, and wherein h is larger and p (x, h) is less.

Measurement data liveness h, can select gradiometry method,

Wherein β is coefficient, x _nfor a certain node n-th measures numerical value, T _nfor n-th measuring period of a certain node.The method can the change speed of very clear and definite directly perceived reflected measurement data, but but existing defects, if data fluctuations is very little, node will be caused to be in resting state for a long time, thus make node can not catch bursty data in time, if data variation is very fast, node will be caused in running order for a long time, thus consume battery very soon.

Measurement data liveness h, can also select variance mensuration, wherein wherein β is coefficient, x _nfor a certain node n-th measures numerical value, T _nfor n-th measuring period of a certain node.The method is the improvement to gradiometry method, during in order to avoid h=0, causes node to be in resting state for a long time.

Measurement data liveness h, and measurement relative entropy method can be selected, if the absolute value measuring numerical difference in the measurement data unit interval is ▽ x, this its measurement entropy can represent: $H_C\left(X\right)=-\underset{-\&infin;}{\overset{\&infin;}{\&Integral;}}{p}_X\left(x\right)\log p_X\left(x\right)\mathrm{dx}$

Then liveness h=β H _c(X), wherein β, κ are all coefficients, and ▽ x is the absolute value measuring numerical difference in the unit time.The method very truly can reflect the impact of data variation entropy on dormancy period.

Below in conjunction with embodiment, the present invention is further illustrated:

Embodiment 1, as table 1: the temperature-time data being a certain place of the node measurement utilizing LDS dormancy strategy,

Table 1

Curve chart corresponding to table 1 is as Fig. 2.

Can find out that from table 1 and Fig. 2 the change of data is smaller between 0-6T, and the change of data is larger between 6-10T.When a certain period, variations in temperature was violent, when accurately will describe temperature-time figure, need to measure more data in this period, so should shorten measuring period thus obtain more data in 6-10T during this period of time interval, the image of description also can be more accurate.

As table 2: be utilize the temperature-time data that the present invention is based on a certain place of a node measurement of improving one's methods of LDS node dormancy strategy,

Table 2

Curve chart corresponding to table 2 is as Fig. 3.

As can be seen from table 2 and Fig. 3, result of the present invention improves in this time period measuring frequency larger to data fluctuation ratio, in the time period that data fluctuations is smaller, measuring frequency reduces relatively, contrast adopts the node measurement of LDS dormancy measurement strategies to find, although pendulous frequency does not become many, the accuracy of rendering image improves, and measures efficiency and is thus significantly improved.

To sum up, the energy consumption of the balanced single-hop networks interior joint of the present invention's energy, extends network lifetime.For the region that measurement environment is different, the reasonable distribution measuring resource can be realized.In the region that measurement data is active, can relatively accurate measurement data, in the region that measurement data is more blunt, relatively can avoid the waste of energy, thus realize energy-conservation.Realize energy-conservation while ensure the validity measured, and between energy saving and validity, seek a kind of balance.

The foregoing is only the preferred embodiments of the present invention, be not limited to the present invention, although with reference to previous embodiment to invention has been detailed description, for a person skilled in the art, it still can be modified to the technical scheme described in foregoing embodiments, or carries out equivalent replacement to wherein portion of techniques feature.Within the spirit and principles in the present invention all, any amendment done, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (5)

1. improving one's methods based on LDS node dormancy strategy, is characterized in that: the concept introducing a data liveness h, and because usually affecting node dormancy probability, data activity degree is larger as one for this data activity degree, and the dormancy probability of node is less.

2. improving one's methods based on LDS node dormancy strategy according to claim 1, it is characterized in that: in LDS dormancy mechanism, the Probability p (x) that node enters dormancy is linear with the distance x to cluster head, hypothetical probabilities is p (x), cluster head and bunch interior nodes thereof are all distributed in the border circular areas that centered by cluster head, radius is R, shown in following formula:

$p\left(x\right)=\left\{\begin{array}{c}\mathrm{Cf}\left(x\right),0\&le;x\&le;x_s\\ 1,x_s<x\&le;R\end{array}\right.$

If these area measure data are comparatively active for cluster node, its corresponding dormancy probability is less, and we introduce liveness to affect dormancy probability, then the new probability formula of the LDS dormancy strategy improved becomes as follows:

$p(x,h)=\left\{\begin{array}{c}\mathrm{Cf}(x,h),0\&le;x\&le;x_s,0<h<t\\ 1,x_s<x\&le;R\end{array}\right.$

Wherein wherein ε is this bunch of group's dormancy node dormancy probability, and wherein h is larger and p (x, h) is less.

3. improving one's methods based on LDS node dormancy strategy according to claim 1, is characterized in that: measurement data liveness h, selects gradiometry method, wherein β is coefficient, x _nfor a certain node n-th measures numerical value, T _nfor n-th measuring period of a certain node.

4. improving one's methods based on LDS node dormancy strategy according to claim 1, is characterized in that: measurement data liveness h, selects variance mensuration, wherein wherein β is coefficient, x _nfor a certain node n-th measures numerical value, T _nfor n-th measuring period of a certain node.

5. improving one's methods based on LDS node dormancy strategy according to claim 1, it is characterized in that: measurement data liveness h, select and measure relative entropy method, if the absolute value measuring numerical difference in the measurement data unit interval is ▽ x, this its measurement entropy can represent: $H_C\left(X\right)=-\underset{-\&infin;}{\overset{\&infin;}{\&Integral;}}{p}_X\left(x\right)\log p_X\left(x\right)\mathrm{dx}$

Then liveness h=β H _c(X), wherein β, κ are all coefficients, and ▽ x is the absolute value measuring numerical difference in the unit time.