阅读说明：本技术 一种针对温度感知数据的时间相关性去冗余方法 (Time correlation redundancy removing method for temperature sensing data ) 是由 朱容波 李媛丽 王俊 王德军 于 2019-09-30 设计创作，主要内容包括：本发明公开了一种针对温度感知数据的时间相关性去冗余方法,该方法包括：步骤一、获取多个温度传感器采集到的温度感知数据,并对其进行初步处理；步骤二、计算当前采集到的温度感知数据与临时数据集中数据的相似距离；步骤三、计算每个周期内连续冗余温度感知数据的最远时间差,设置相似距离的阈值和最大时间控制的数值,将相似距离与其阈值比较,同时将最远时间差与最大时间控制的数值比较,作为判断冗余的冗余条件；并在计算时加入动态步长控制比较数量；步骤四、输出时间相关性去冗余后的温度感知数据。本发明能保留特殊数据,解决了因过度去冗余而造成的数据过度缺失的问题,使得温度感知数据在去冗余的过程中更合理、更符合用户的接受形式。(The invention discloses a time correlation redundancy removing method for temperature sensing data, which comprises the following steps: acquiring temperature sensing data acquired by a plurality of temperature sensors, and performing primary processing on the temperature sensing data; step two, calculating the similar distance between the currently acquired temperature sensing data and the data in the temporary data set; step three, calculating the farthest time difference of continuous redundant temperature sensing data in each period, setting a threshold value of a similar distance and a numerical value of the maximum time control, comparing the similar distance with the threshold value, and simultaneously comparing the farthest time difference with the numerical value of the maximum time control to serve as a redundant condition for judging redundancy; adding dynamic step length to control the comparison quantity during calculation; and step four, outputting the temperature sensing data after the redundancy removal of the time correlation. The invention can reserve special data, solves the problem of data excessive loss caused by excessive redundancy removal, and ensures that the temperature sensing data is more reasonable and more in line with the acceptance form of a user in the redundancy removal process.)

1. A method for time dependent redundancy elimination for temperature sensing data, the method comprising the steps of:

acquiring temperature sensing data acquired by a plurality of temperature sensors, performing primary processing on the temperature sensing data, constructing a temperature sensing data redundancy removal model based on time series correlation of dynamic change step length, and processing a large amount of redundant information in the data;

judging whether the temperature sensing data currently acquired by the temperature sensor is first sensing data, and if so, marking the first sensing data as non-redundant data; if not, taking the previously acquired temperature sensing data as a temporary data set, and calculating the similar distance between the currently acquired temperature sensing data and the data in the temporary data set;

step three, calculating the farthest time difference of continuous redundant temperature sensing data in each period, setting a threshold value of a similar distance and a numerical value of the maximum time control, comparing the similar distance with the threshold value, and simultaneously comparing the farthest time difference with the numerical value of the maximum time control to serve as a redundant condition for judging redundancy; adding dynamic step length to control the comparison quantity during calculation;

if the redundant condition is met, marking the currently acquired temperature sensing data as redundant data, storing the redundant data into a temporary data set, and continuously acquiring the temperature sensing data; if the condition is not met with the redundancy condition, emptying the previous section of temporary data set, and marking the currently acquired temperature sensing data as non-redundant data for transmission;

and step four, outputting the temperature sensing data after the redundancy removal of the time correlation.

2. The method for removing redundancy of temporal correlation of temperature sensing data according to claim 1, wherein the specific method of step one is:

acquiring temperature sensing data acquired by a plurality of temperature sensors, wherein a large amount of redundant information exists in the data, and performing primary processing on the data; changing T to T₀、t₁、t₂、t₃The corresponding temperature sensing data are respectively expressed as: x ═ X₀、x₁、x₂、x₃、…、x_n}; where T represents a set of times, T₀-t_nRespectively representing time nodes at different moments, X representing t₀-t_nSet of data perceived by a time node, x₀-x_nRepresents t₀-t_nData perceived separately at the moment. Outputting result ═ x using the temperature data sensed for the first time as non-redundant data₀}。

3. The method for removing redundancy of time dependence of temperature sensing data according to claim 2, wherein the specific method of the second step is:

the matrix of similar distances between temperature sensing data is represented as:

wherein, the similar distance matrix is a symmetric matrix, and the element of the diagonal is 0;

the calculation formula of the data similarity distance is as follows:

δ_ij＝|x_i-x_j|，0＜i＜j＜n≤N_max

wherein, delta_ijRepresents t_iAnd t_jData x perceived at time_iAnd x_jN is a range of the number of consecutive redundant data, and N _ max represents the maximum number of consecutive redundant data for limiting the range of consecutive redundant data.

4. The method for removing redundancy of time dependence of temperature sensing data according to claim 1, wherein the specific method of step three is:

with the increase of the comparison times, the continuous redundant temperature sensing data volume is increased progressively, and the comparison step length is dynamically changed according to the size of different continuous redundant data volumes in the comparison process;

in the process from the 1 st pass to the (n-1) th pass, the comparison times are as follows in sequence:

1、2、3、4、…、n-2；

the comparison step length m of the adjustment is as follows:

here, len (Temp _ Data) is the length of the set for temporarily storing the redundant Data, that is, the length is equal to the number of consecutive redundant Data, and the redundant array changes with the change of the redundant Data amount.Expressed as rounding down on len (Temp _ Data)/2; dynamic step size updating is realized in each redundancy period, and delta is dynamically changed when data comparison is carried out_ij＝|x_i-x_jIn |, x_iAnd x_jJ ═ i, i ═ i + step index positions for both data, reducing computational complexity.

5. The method for removing redundancy of time dependence of temperature sensing data according to claim 1, wherein the specific method of step three is:

the formula for calculating the farthest time difference of the continuous redundant temperature sensing data in each period is as follows:

t＝t_i-t₀

where t denotes the perceived time difference, t_iDenotes the i-th sensing time, t₀Indicating the time at which the data was first perceived.

6. The method for removing redundancy of time dependence of temperature sensing data according to claim 5, wherein the specific method of step three is:

the redundancy judgment method for data redundancy comprises the following steps:

the similarity distance is delta_ijThe farthest time difference is T, the threshold value of the similar distance is th, and the value of the maximum time control is T _ max;

if delta_ijTh and T are less than or equal to T _ max, the marks are used as redundant temperature sensing data, and redundancy removing operation is carried out;

if delta_ijIf > th and T > T _ max, then T is_iData sensed at the moment is promoted to non-redundant temperature sensing data and transmitted as reserved result (x)_n}。

7. The method for removing redundancy of time dependence of temperature sensing data according to claim 1, wherein the specific method of step four is:

and (4) keeping the temperature sensing data generated for the first time and the non-redundant temperature sensing data as a final result data set for outputting.

Technical Field

The invention relates to the technical field of wireless sensor networks, in particular to a time correlation redundancy removing method for temperature sensing data.

Background

With the progress of social science and technology, wireless sensor networks are used in all aspects of life (such as medical treatment, trajectory, earthquake, military, internet of things, agricultural and industrial pollution monitoring and the like), and a large amount of monitoring data is provided for people all the time. Due to the limitations of the sensor network conditions, limited battery energy, memory, capacity, communication range, etc., the network lifetime is limited. The network life is mainly determined by battery energy, and the energy consumption is represented by three aspects of node communication, data processing and data transmission. With the sharp increase of the perception data, a large amount of data transmission is generated, huge battery energy is consumed, and the service life of the network is shortened. Therefore, how to perform the redundancy removal processing on the sensing data becomes a key problem.

For the problem of de-redundancy of sensing data in the WSN, current research mainly focuses on three aspects of temporal correlation, spatial correlation and spatio-temporal correlation. Due to the density of the topological structure of the sensor, a plurality of sensor nodes in a sensing area can jointly record the information of a single event source, and observed values obtained by sensors close in space have high correlation. In addition, because the sensor nodes periodically observe and transmit sensing events, the nature of the physical environment enables high correlation of sampled data of single nodes close in sampling time.

The method mainly aims at analyzing the time correlation of the temperature sensor nodes, carrying out redundancy removal processing on temperature sensing data and improving the performance of the algorithm. The method aims at solving the problem that when data redundancy is removed, only data with a data similarity distance smaller than a threshold th is removed, and after the removal of redundant data is not researched, important information represented by original data can be presented through limited non-redundant data. Meanwhile, in the process of removing redundancy, when data fluctuation is large, the local maximum value and the local minimum value are considered. For the case that the data fluctuation range is extremely small, the data is stable, and the data will face a large amount of situations smaller than the threshold th, which may cause the problem of threshold failure, so that the data in a long period of time is all regarded as redundant data. Finally, the computational power consumption of the sensor is increased when data similarity comparison is performed.

Therefore, the following improvements are made in view of the above three problems:

(1) the data after redundancy removal is analyzed and compared with the original data in a mode of drawing the data into a line graph, and unscientific and unreasonable places for redundancy removal are searched, for example, when redundant data is removed, the most value is lost because the existence of special points such as the maximum value and the minimum value of the data is not considered, and therefore, the control of the maximum time threshold value is added on the basis of an algorithm.

(2) Aiming at the condition that data fluctuation is not obvious and stable, the maximum time threshold value T _ max is increased to solve the problem of threshold value failure, and meanwhile, the accuracy of the maximum value and the minimum value is improved.

(3) In the face of the problem that the energy consumption of the sensor is increased due to the fact that the comparison quantity of similar data calculation is huge in the algorithm, the complexity of data similarity distance calculation is improved through a dynamic step size model, the model mainly proposes to use a piecewise function to change the step size according to the characteristics of redundant similarity distance calculation rules, comparison calculation among data is reduced, the performance of the method is improved, and the loss of calculation energy consumption is reduced.

Disclosure of Invention

The technical problem to be solved by the present invention is to provide a method for removing redundancy of time dependence of temperature sensing data, aiming at the defects in the prior art.

The technical scheme adopted by the invention for solving the technical problems is as follows:

the invention provides a time-dependent redundancy removing method for temperature sensing data, which comprises the following steps:

acquiring temperature sensing data acquired by a plurality of temperature sensors, performing primary processing on the temperature sensing data, constructing a temperature sensing data redundancy removal model based on time series correlation of dynamic change step length, and processing a large amount of redundant information in the data;

judging whether the temperature sensing data currently acquired by the temperature sensor is first sensing data, and if so, marking the first sensing data as non-redundant data; if not, taking the previously acquired temperature sensing data as a temporary data set, and calculating the similar distance between the currently acquired temperature sensing data and the data in the temporary data set;

step three, calculating the farthest time difference of continuous redundant temperature sensing data in each period, setting a threshold value of a similar distance and a numerical value of the maximum time control, comparing the similar distance with the threshold value, and simultaneously comparing the farthest time difference with the numerical value of the maximum time control to serve as a redundant condition for judging redundancy; adding dynamic step length to control the comparison quantity during calculation;

if the redundant condition is met, marking the currently acquired temperature sensing data as redundant data, storing the redundant data into a temporary data set, and continuously acquiring the temperature sensing data; if the condition is not met with the redundancy condition, emptying the previous section of temporary data set, and marking the currently acquired temperature sensing data as non-redundant data for transmission;

and step four, outputting the temperature sensing data after the redundancy removal of the time correlation.

Further, the specific method of the first step of the invention is as follows:

acquiring temperature sensing data acquired by a plurality of temperature sensors, wherein a large amount of redundant information exists in the data, and performing primary processing on the data; set time T ═ T₀、t₁、t₂、t₃、…、t_nThe temperature sensing data corresponding to different time nodes in the } are represented as X sets: x ═ X₀、x₁、x₂、x₃、…、x_n}; where T represents a set of times, T₀-t_nRespectively representing time nodes at different moments, X representing t₀-t_nSet of temperature data, x, sensed by a time node₀-x_nRepresents t₀-t_nData perceived separately at the moment. The temperature sensed for the first timeThe degree data is used as non-redundant data to output result ═ x₀}。

Further, the specific method of the second step of the invention is as follows:

the matrix of similar distances between temperature sensing data is represented as:

wherein, the similar distance matrix is a symmetric matrix, and the element of the diagonal is 0;

the calculation formula of the data similarity distance is as follows:

δ_ij＝|x_i-x_j|，0＜i＜j＜n≤N_max

wherein, delta_ijRepresents t_iAnd t_jData x perceived at time_iAnd x_jN is a range of the number of consecutive redundant data, and N _ max represents the maximum number of consecutive redundant data (for limiting the range of consecutive redundant data).

Further, the specific method of the third step of the invention is as follows:

with the increase of the comparison times, the continuous redundant temperature sensing data volume is increased progressively, and the comparison step length is dynamically changed according to the size of different continuous redundant data volumes in the comparison process;

in the process from the 1 st pass to the (n-1) th pass, the comparison times are as follows in sequence:

1、2、3、4、…、n-2；

the comparison step length m of the adjustment is as follows:

here, len (Temp _ Data) is the length of the set for temporarily storing the redundant Data, that is, the length is equal to the number of consecutive redundant Data, and the redundant array changes with the change of the redundant Data amount.

This is shown as rounding down to len (Temp _ Data)/2. Dynamic step size updating is realized in each redundancy period, and delta is dynamically changed when data comparison is carried out_ij＝|x_i-x_jIn |, x_iAnd x_jJ ═ i, i ═ i + step index positions for both data, reducing computational complexity.

Further, the specific method of the third step of the invention is as follows:

the formula for calculating the farthest time difference of the continuous redundant temperature sensing data in each period is as follows:

t＝t_i-t₀

where t denotes the perceived time difference, t_iDenotes the i-th sensing time, t₀Indicating the time at which the data was first perceived.

Further, the specific method of the third step of the invention is as follows:

the redundancy judgment method for data redundancy comprises the following steps:

the similarity distance is delta_ijThe farthest time difference is T, the threshold value of the similar distance is th, and the value of the maximum time control is T _ max;

if delta_ijTh and T are less than or equal to T _ max, the marks are used as redundant temperature sensing data, and redundancy removing operation is carried out;

if delta_ijIf > th and T > T _ max, then T is_iData sensed at the moment is promoted to non-redundant temperature sensing data and transmitted as reserved result (x)_n}。

Further, the specific method of the fourth step of the invention is as follows:

and (4) keeping the temperature sensing data generated for the first time and the non-redundant temperature sensing data as a final result data set for outputting.

The invention has the following beneficial effects: the invention relates to a time correlation redundancy removing method for temperature sensing data, which comprises (1) carrying out initial analysis on the temperature sensing data

And analyzing special values (maximum value, minimum value and the like) existing in the experimental data, and comparing whether nodes specially existing in the data after redundancy removal are influenced or not according to the special values, so that the analyzability of the data is lost.

(2) The problem of information loss caused by the fact that a user cannot receive any information for a long time due to the fact that a large amount of data is continuously removed due to the fact that a threshold value fails is solved

Aiming at the existing redundancy removing algorithm, the problem that the temperature sensing data change cannot be received by a user for a long time due to excessive redundancy removing caused by the fact that the data are extremely gentle is not considered, so that a time control variable between continuous redundant data is introduced, and an optimal maximum time threshold T _ max of sensing time is found, so that the special data can be kept, the redundant data can be removed in an acceptable maximum time interval, the analysis of the data by the user is not influenced, the analysis of the data is not influenced under the condition that the redundant data are removed to the maximum, and the effect of reducing the transmission energy consumption of the data is achieved.

(3) Solving the problem of computing energy consumption caused by large computation amount of the data redundancy removing similar distance

Aiming at some existing redundancy removing algorithms, a problem of computing energy consumption also exists. Because the similarity calculation is carried out on the data, the calculation energy consumption is increased, and because the data volume is large, the accumulation of all the calculation energy consumption is very large consumption, therefore, the time correlation data redundancy removal method (TDS) based on the maximum time threshold and the dynamic step size model is provided for the defects and shortcomings of the existing algorithm, and the loss of the energy consumption is effectively reduced. Compared with the DAWF (data acquisition window function) method, the energy consumption increasing rate is increased by 5%.

The experimental result shows that in the process of removing redundancy of the temperature sensing data, the temperature sensing data is maintained in the error range acceptable to the user, so that the retention of special data is guaranteed, the problem of data lack in the past caused by excessive redundancy removal is solved, and the temperature sensing data is more reasonable and more in accordance with the form acceptable to the user in the process of removing redundancy. However, since the calculation of the data similarity distance is increased while removing redundancy, a large amount of calculation power consumption is increased.

Furthermore, the efficiency of the algorithm is improved, and a data redundancy removing model of the time series correlation of the dynamic step length is provided. By removing redundancy and reducing computational power consumption, the overall power consumption is reduced by 93.41% compared to a data transmission method that does not do anything.

Drawings

The invention will be further described with reference to the accompanying drawings and examples, in which:

FIG. 1 is a time series data based redundancy analysis model according to an embodiment of the present invention;

FIG. 2 is a time series correlation data based redundancy elimination decision process analysis according to an embodiment of the present invention;

FIG. 3 is a flow chart of a temporal correlation de-redundancy algorithm according to an embodiment of the present invention;

FIG. 4 is a raw data presentation of a node of an embodiment of the present invention;

FIG. 5 is a diagram of an SDT algorithm analysis of redundant results at different thresholds according to an embodiment of the present invention;

FIG. 6 is a data presentation of the DAWF method redundancy elimination of an embodiment of the present invention;

FIG. 7 is an energy consumption analysis of a TDS method of an embodiment of the invention;

FIG. 8 is an energy consumption analysis of the DAWF algorithm according to an embodiment of the invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The time correlation redundancy removing method for the temperature sensing data comprises the following steps:

acquiring temperature sensing data acquired by a plurality of temperature sensors, performing primary processing on the temperature sensing data, constructing a temperature sensing data redundancy removal model based on time series correlation of dynamic change step length, and processing a large amount of redundant information in the data;

judging whether the temperature sensing data currently acquired by the temperature sensor is first sensing data, and if so, marking the first sensing data as non-redundant data; if not, taking the previously acquired temperature sensing data as a temporary data set, and calculating the similar distance between the currently acquired temperature sensing data and the data in the temporary data set;

step three, calculating the farthest time difference of continuous redundant temperature sensing data in each period, setting a threshold value of a similar distance and a numerical value of the maximum time control, comparing the similar distance with the threshold value, and simultaneously comparing the farthest time difference with the numerical value of the maximum time control to serve as a redundant condition for judging redundancy; adding dynamic step length to control the comparison quantity during calculation;

if the redundant condition is met, marking the currently acquired temperature sensing data as redundant data, storing the redundant data into a temporary data set, and continuously acquiring the temperature sensing data; if the condition is not met with the redundancy condition, emptying the previous section of temporary data set, and marking the currently acquired temperature sensing data as non-redundant data for transmission;

and step four, outputting the temperature sensing data after the redundancy removal of the time correlation.

In order to more reasonably remove a large amount of redundant data generated in the sensor, the temperature sensing data is reasonably analyzed, the following scenes are comprehensively considered, and the algorithm is improved, so that the data expression is more reasonable, and the energy consumption is effectively reduced. We have found that:

(1) under the condition that the sensing event is not changed, the temperature sensing data has a small change and floating range in a short time;

(2) under the condition that the temperature sensing data has extremely small fluctuation, data redundancy can be removed within an error range acceptable by a user, so that the energy consumption of data transmission is reduced.

To verify the feasibility of this invention, an analysis was performed using temperature sensing data from the intel berkeley laboratory. This data, perceived collection for temperature, collected about forty thousand pieces of data per sensor node, for a total of 54 sensor nodes, with a data volume of about two million. We will carry out the relevant work around the temperature sensing data of this laboratory, the overall flow chart of the algorithm, as shown in fig. 3. The redundancy problem can be represented visually as shown in fig. 1.

The method comprises the following steps: acquiring data acquired by a plurality of temperature sensors, wherein a large amount of redundant information exists in the data, and then carrying out primary processing on the data, and carrying out the following analysis:

changing T to T₀、t₁、t₂、t₃、…、t_nThe corresponding temperature sensing data are respectively X ═ X₀、x₁、x₂、x₃、…、x_nCalculation and analysis by data similarity distance equation (1) { x }₀、…、x_n-1The similar distance between them, first, the first sensed temperature data is used as non-redundant data to transmit result ═ x₀}。

Representation of the similarity distance matrix between data:

wherein the similarity distance matrix is a symmetric matrix, and the diagonal is 0.

The calculation formula of the data similarity distance is as follows:

δ_ij＝|x_i-x_j|，0＜i＜j＜n≤N_max (2)

wherein, delta_ijRepresents t_iAnd t_jData x perceived at time_iAnd x_jN is a range of the number of consecutive redundant data, and N _ max represents the maximum number of consecutive redundant data (for limiting the range of consecutive redundant data).

Step two: establishing similar distance for calculating temperature sensing data based on dynamic step comparison

Aiming at the performance and energy consumption problems caused by data similarity distance calculation, a comparison mode of dynamic change step length is provided for carrying out an optimization algorithm. In the case of incremental continuous redundant temperature data, the data similarity distance is still solved by one-to-one comparison in the original model, which increases a large amount of unnecessary calculation. Due to the time correlation relationship of the data, the smaller the sensing time difference between the temperature data is, the higher the possibility of similarity between the data is; meanwhile, in order to avoid the situation that the data similarity distance is very small or very large and cannot be identified under special conditions, the comparison step size is dynamically changed according to the sizes of different continuous redundant data quantities, as shown in fig. 2.

The comparison times from the 1 st pass to the n-1 st pass are as follows: 1. 2, 3, 4, …, n-2. Thus, without improvement, the total number of times is: (n-1) (n-2)/2.

Therefore, the invention proposes that the dynamic step size becomes larger as the continuous redundant data increases, so as to reduce the step size of comparison. Representation of the step size:

where len (Temp _ Data) is the length of the temporary redundancy array, i.e., equal to the amount of consecutive redundancy Data. .

This is shown as rounding down to len (Temp _ Data)/2. Since the redundant array will change with the change of the redundant data amount, the dynamic step update can be realized in each redundant period through the formula in (4), so that when data comparison is performed, the index positions of two data can be dynamically changed, thereby reducing the comparison amount, greatly reducing the complexity of the calculation of the similarity distance (1) (2), and dynamically changing the delta_ij＝|x_i-x_jIn |, x_iAnd x_jJ is i, i is i + step index position of two data, thereby promoting the computational efficiency of similar distance and reducing the computational energy consumption.

Step three: and improving the redundancy condition for judging the redundancy, and taking the condition (2) in the step one and the condition (4) in the step three as the factors for judging the redundancy.

And calculating the farthest time difference of the continuous redundant temperature sensing data in each period:

t＝t_i-t₀(4)

another variable is introduced, the maximum time control T _ max:

if delta_ijTh and T are less than or equal to T _ max, the marks are used as redundant temperature sensing data, and redundancy removing operation is carried out;

if delta_ijIf > th and T > T _ max, then T is_iData sensed at the moment is promoted to non-redundant temperature sensing data and transmitted as reserved result (x)_n}。

Step four: and outputting a non-redundant result.

Each other is less than threshold th, therefore, it belongs to redundant data, redundant removal is performed, the first generated sensing data and non-redundant temperature sensing data are retained, and then the final result set, result, is { x ═₀、x_n}。

The invention mainly adopts the technical scheme that the model is improved as follows:

1) constructing a temperature sensing data redundancy removing model based on time series correlation;

through continuous acquisition of temperature sensing data and similarity calculation between the temperature sensing data and the temperature sensing data, T ═ T { (T) } is calculated₀、t₁、t₂、t₃、…、t_nThe corresponding temperature sensing data are respectively X ═ X₀、x₁、x₂、x₃、…、x_n}; by analysis of data similarity distances, { x₀、…、x_n-1The similarity distance between the two data is less than threshold th, so that the data belongs to redundant data, redundancy removal is carried out, the first generated sensing data and non-redundant data are reserved, and the final result set is that result is equal to { x }₀、x_n}; time range for analyzing redundant data, t ═ t_i-t₀There are two cases, one within and one greater than the user acceptable range. Therefore, to solve this problem, another variable, the maximum time control T _ max, is introduced. By calculating the maximum time range of the redundant data, the redundant data can be further reduced to non-redundant data when the maximum time is greater than T _ maxTherefore, the problem that the data cannot be received for a long time due to excessive removal of redundant data of the sensing data is solved.

2) Constructing a data redundancy removal model based on the time series correlation of the dynamic change step length;

because the similar distance calculation of the data is introduced in the data redundancy removing model based on the time series correlation, a large amount of energy consumption for calculation is increased, and therefore, the concept of dynamic step size is introduced to reduce the calculation amount of the similar distance of the data. The variation of the dynamic step size m is improved mainly according to the root cause causing the increase of the computing energy consumption. Since the more consecutive redundant data, the larger the calculation amount, the more the comparison step is dynamically changed according to the amount of consecutive redundant data.

And (3) redundancy removal and energy consumption analysis through experimental data:

in this experimental part, the data from the intel berkeley research laboratory were used mainly for research analysis. The laboratory has totally arranged 54 sensor nodes, monitors the temperature change condition of different positions of the whole laboratory respectively, and the sensor nodes collect data once every 0.5 minute, collects data of about one month, and the data volume of each node is about forty thousand, totally 54 nodes, therefore, the total data volume reaches two million pieces, and the data volume is huge. In the preliminary test, four ten thousand pieces of data of one node are mainly used for analysis and improvement.

(1) The raw data of the node shows, as shown in fig. 4:

in fig. 4, the x-axis represents time and the y-axis represents temperature. The temperature varied dramatically with time, especially near time 430 minutes, reaching a minimum, then increasing with time to reach a maximum peak around 750 minutes, then decreasing again with time to begin at a minimum at 1800 minutes, then increasing again, and reaching a maximum at 2250 minutes and beginning to decrease again. Since the variability of the data is very large, it is very intuitive that the data is the most specific maximum node in the data at around 430, 750, 1800 and 2250 minutes. Therefore, in the data redundancy removing process, the data redundancy removing situation of the four positions needs to be particularly focused.

(2) The effect graph of the node after redundancy removal is shown in fig. 4:

in fig. 5, the abscissa is time in minutes and the ordinate is temperature in degrees celsius. Under the condition that the threshold th under the SDT (based on the maximum time threshold and the dynamic step size model) method is respectively 0, 1, 0.5, 0.4 and 0.25, the drawn line graphs can visually reflect the special value in the data, the maximum value at the peak and the minimum value at the trough. As can be seen from fig. 5, the amount of data is reduced for different threshold cases, but the characteristics closest to the original data are reflected with the least data. The maximum value at the peak and the maximum value at the trough are infinitely close to the original data. Analysis of data volume and accuracy at different thresholds, as shown in table 1:

TABLE 1 non-redundant data case at SDT method threshold

In fig. 6, the abscissa is time in minutes and the ordinate is temperature in degrees celsius. When the threshold th under the DAWF method is 0, 0.01, 0.02, 0.03, 0.04, 0.05, respectively, the line graphs drawn respectively can visually show that the data after redundancy removal is distributed above and below the original data, which is a necessary phenomenon caused by the fact that the DAWF method performs mean solution on the redundant data. The average value becomes smaller than the original data during the data straight-line descending process, and becomes larger than the original data during the data straight-line ascending process. Thus, the de-redundant data will appear to float above or below the original data. The original data is only close to but cannot be equal to the original data, so that an error exists between the original data and the redundancy-removed data.

Analysis of data volume and accuracy at different thresholds, as shown in table 2:

TABLE 2 DAWF method non-redundant data case under different thresholds

(3) The energy consumption analysis situation for the node redundancy removal is shown in fig. 7;

as shown in fig. 7, the abscissa represents the threshold value and the ordinate represents the energy consumption increase rate. In fig. 7, the blue curve is the trend before the performance of the TDS method is improved, and therefore, it can be clearly seen that the energy consumption increases first and decreases sharply as the threshold value increases, which is caused by the increase of the calculation energy consumption as the number of comparison times of the similar distance between the data increases. Therefore, analysis of algorithm performance is particularly important. After the performance improvement is added, as shown in the figure, the red curve (after the performance improvement of the TDS method) has a trend that the energy consumption increasing rate is obviously increased along with the increase of the threshold value. And when the threshold value is 0.4, the energy consumption increasing rate reaches 93.41%.

As shown in fig. 7, the abscissa represents the threshold value and the ordinate represents the energy consumption increase rate. In the figure, the curve is the performance improvement rate of the DAWF method, and it can be clearly seen that as the threshold value increases, the energy consumption improvement rate increases rapidly within 0.01 of the threshold value, and then is stable. At the threshold of 0.05, the energy consumption improvement rate was 88.09%.

From the results of fig. 7 and 8, it is shown that the SDT method is about 5% more energy saving than the DAWF method.

It will be understood that modifications and variations can be made by persons skilled in the art in light of the above teachings and all such modifications and variations are intended to be included within the scope of the invention as defined in the appended claims.