CN107295536A - A kind of parallel diagnosis method of testing - Google Patents

Abstract

The present invention relates to a kind of sensor network fault nodal parallel diagnostic test method and system, by setting up based on the diagnostic test structure for treating diagnosis node, treat many wheel sensitive informations of diagnosis node peripheral part node progress to compare, judged to treat whether diagnosis node is malfunctioning node according to sensitive information result of the comparison.Using the parallel diagnosis method of testing and system of the present invention, unnecessary test can be largely reduced, it is to avoid occur signal conflict when signal is transmitted during diagnostic test, can also effectively reduce the time of diagnostic test.

Description

A kind of parallel diagnosis method of testing

Technical field：

The present invention relates to the sensor node in wireless sensor network field, more particularly to a kind of wireless sensor network Fault diagnosis method of testing.

Background technology：

Wireless sensor network is constituted by being deployed in substantial amounts of sensor node in monitored area, passes through side wireless communication One multihop self-organizing network of formula formation.Wireless sensor network can be used for monitoring and protection, medical treatment and nursing, the target of environment The field such as tracking and military detecting.In the most random placement of sensor node external environment out of office, network node quantity is huge, sensing Device type is various, network topology dynamic and the limited characteristic of node resource cause node that various failures occur often. Fault detect in wireless sensor network refers to that, when sensor node breaks down, the data that node is collected are probably It is incorrect, so as to cause wireless sensor network to the false judgment of monitoring information, particularly detection precise requirements compared with High application field, the accuracy of information is vital, so be to the fault detect of wireless sensor network must can not Few, it is also to be worth research.

Research on node failure detection algorithm has had a lot, is broadly divided into five classes：Based on majority vote rule Algorithm, the algorithm based on intermediate value strategy, the algorithm based on decision-making Diffusion Strategy, the algorithm based on weighting, the calculation based on sub-clustering Method.

When we think the node failure of some in awareness network whether state, traditional system diagnostics needs to receive breath network In all test result, then can judge the state of the node, in actual application, the testing time of this method is long And the electricity of consumption great deal of nodes.And based on majority vote rule algorithm, due to needing to treat all neighbours' sections around diagnosis node Point assists the node to be tested, and the use time of diagnostic test is directly proportional to neighbor node number.Meanwhile, the diagnosis of this method is surveyed Examination reliability is easily influenceed by neighbor node failures whether state.

Although the research of wireless sensor network fault diagnostic techniques has had significant progress, develop including certainly Diagnose, based on sub-clustering, based on polytype diagnostic mode including test etc..But as application of higher wireless sensor network is led The expansion in domain, the diversity and complexity of deployed environment bring new challenge, therefore research to the performance of fault diagnosis algorithm The new fault diagnosis algorithm for being applied to wireless sensor network under complex environment is significantly.In addition, in view of sensing The property of finite energy of device node itself, during system diagnostics is carried out on a large scale, it is unnecessary how to efficiently reduce Testing time, it is to avoid meaningless energy loss is also very important.

The content of the invention

This application provides a kind of fault diagnosis method of testing of the sensor node in wireless sensor network, it is only necessary to treats The test result of diagnosis node peripheral part node, largely reduces unnecessary test, and by using designed by the present invention Diagnostic test structure, it is to avoid during diagnostic test signal transmit when occur signal conflict, in addition, the present invention diagnostic test Method with parallel processing, can effectively reduce the time of diagnostic test.

In a first aspect, this application provides a kind of parallel diagnosis method of testing, this method comprises the following steps：

(1) determine to treat diagnosis node；

(2) judge whether to need to set up diagnostic test structure for the band diagnosis node, if it is, performing step (3), such as It is really no, directly perform step (4)；

(3) set up based on the diagnostic test structure for treating diagnosis node；

(4) diagnostic test is performed according to diagnostic test structure；

(5) judged to treat whether diagnosis node is malfunctioning node according to diagnostic test results.

Preferably, the foundation specifically includes following steps based on the diagnostic test structure for treating diagnosis node：

By treating diagnosis node S_uOr the maximum number of plies L of diagnostic test structure described in default_maxWith maximum child node number C_max, wherein 1≤C_max≤|N(S_u) |, N (S_u) it is to treat diagnosis node S_uNeighbor node set.Generate S_uCandidate's child node Collection, selects C_maxThe most node of individual neighbor node numberAs the first layer set of node of deagnostic structure；

For i-th layer of node S_{I, j}, wherein 2≤i≤L_maxAnd 1≤j≤C_max, determine node S_{I, j}Candidate's child node collection A_{I, j}.If | A_{I, j}| >=1 and i ＜ L_max- 1, then in A_{I, j}In select that the candidate's child node conduct for possessing most neighbor nodes S_{I+1, j}Or one candidate's child node of random selection is used as S_{I+1, j}；If | A_{I, j}| >=1 and i=L_max- 1, then randomly choose one Candidate's child node is used as S_{I+1, j}；If | A_{I, j}|=0 or i=L_max, then just stop choosing S_{I, j}Child node.

Preferably, by S_uWhole neighbor nodes candidate's child node collection is added to generate S_uCandidate's child node collection.

Preferably, node S is determined_{I, j}Candidate's child node collection A_{I, j}Specially directly use S_{I, j}Neighborhood generation A_{I, j}。

Preferably, node S is determined_{I, j}Candidate's child node collection A_{I, j}Specific method is：

To all and node S_{I, j}Adjacent node y, node is put into if node y meets following two condition by node y S_{I, j}Candidate's child node collection A_{I, j}In：

(1) node y and node S_uDistance be equal to node S_{I, j}With node S_uDistance add 1；

(2) node y and node S_uI-th layer of deagnostic structure in except node S_{I, j}Outside all nodes it is all non-conterminous.

Preferably, it is described to be specially according to the execution diagnostic test of diagnostic test structure：

First round diagnostic test：To 1 all≤j≤C_maxAndBy S_{3r-1, j}To S_{3r-2, j}Sensed Information compares, wherein, l_jRepresent S_uDeagnostic structure in jth arrange the actual maximum number of plies；

Second wheel diagnostic test：To 1 all≤j≤C_maxAndBy S_{3r, j}To S_{3r-1, j}Carry out sensing letter Breath compares；

Third round diagnostic test：To 1 all≤j≤C_maxAndBy S_{3r+1, j}To S_{3r, j}Carry out sensing letter Breath compares；

The sensitive information is more specific to be：

Node x is sent to node y compares request and sensing moment t；

Node x and y senses environmental information to obtain the sensing value v of oneself in moment t simultaneously_xWith v_y；

Node y is by v_yIt is sent to node x；

Node x is received after y measured value, by v_xWith v_yCompare, if | v_x-v_y|≤θ, wherein θ are the thresholds being previously set It is worth, then sensitive information comparative result C (x, y)=0；If | v_x-v_y| ＞ θ, then sensitive information comparative result C (x, y)=1.

It is described to be judged to treat whether diagnosis node is that malfunctioning node is specially according to diagnostic test results：

To 1 all≤j≤C_max, by S_{1, j}Collect the sensitive information comparative result that three-wheel is completed before jth is listed in Then the comparative result is sent to S_u；S_uSelect sensing Since the first continuous 0 most that first node layer arranged of number of starting to S in information comparative result_uCarry out sensitive information ratio Compared with；If sensitive information result of the comparison is 0, S is judged_uIt is malfunctioning node；Otherwise, it is determined that S_uIt is fault-free node.

Preferably, the environmental information includes temperature, humidity.

Second aspect, this application provides a kind of parallel diagnosis test system that 1-8 either method is required for perform claim System, the system includes：

Determining module, for determining to treat diagnosis node；

Judge module, diagnostic test structure is set up for judging whether to need for the band diagnosis node, if it is, logical Know that diagnostic test structure sets up module to set up based on the diagnostic test structure for treating diagnosis node, if it is not, then notifying diagnosis to survey Die trial block performs diagnostic test；

Diagnostic test structure sets up module, for setting up based on the diagnostic test structure for treating diagnosis node；

Diagnostic test module, for performing diagnostic test according to diagnostic test structure；

Diagnostic result judge module, for being judged to treat whether diagnosis node is malfunctioning node according to diagnostic test results.

Brief description of the drawings

Fig. 1 is parallel diagnosis method of testing flow chart of the present invention

Fig. 2 is based on the deagnostic structure schematic diagram for treating diagnosis node

Fig. 3 is the flow chart of the diagnostic test structure in structural map 2

Fig. 4 is the first round test schematic diagram of the embodiment of the present invention

Fig. 5 is the second wheel test schematic diagram of the embodiment of the present invention

Fig. 6 is the third round test schematic diagram of the embodiment of the present invention

Fig. 7 is the fourth round test schematic diagram of the embodiment of the present invention

Embodiment

Describe the specific embodiment of present inventive concept in detail now with reference to accompanying drawing.

First, the implication of the term for being used in the present invention is defined as follows：In wireless sensor network, if two Sensor node p and node q distance is within mutual communication radius, then it is adjacent just to claim the two nodes, that is, is saved Point p is the hop node apart from node q.We claim node p and node q distance to be k-hop, represent the fewest number of hops between p and q For k.Dis (p, q) is made to represent the distance between node p and node q.

As shown in figure 1, after the parallel diagnosis method of testing of the present invention starts, it is necessary first to determine whether to treat diagnosis section Point sets up diagnostic test structure, if it is not required, then directly starting diagnostic test, if it is desired, then construction is based on treating diagnosis section The diagnostic test structure of point；After the diagnostic test structure is constructed, diagnostic test is carried out using the diagnostic test structure.

Fig. 2 show it is exemplary based on the diagnostic test structural representation for treating diagnosis node, wherein, S_uTo treat diagnosis section Point, the exemplary diagnostic test structure is one five layers of structure, the nodes that each layer includes can with it is identical can not also Together, S is made_{I, j}Represent the node of i-th layer of jth row in the diagnostic test structure, n_iAll node numbers of i-th layer of expression.

The method for building up of above-mentioned diagnostic test structure is explained referring to Fig. 3.

We define the candidate's child node collection set up and used in the diagnostic test configuration process first, i.e.,：If node x is Treat diagnosis node S_u, then x candidate's child node integrate all neighbor node collection as node x；If node x is to treat diagnosis node S_u's I-th node layer (2≤i≤L in deagnostic structure_max- 1), then the selection mode of node x candidate's child node collection is as follows：

To all node ys adjacent with node x, node y is put into node x's if node y meets following two condition Candidate's child node is concentrated：

(1) node y and node S_uDistance be equal to node x and node S_uDistance add 1.

(2) node y and node S_uI-th layer of deagnostic structure in all nodes in addition to node x it is all non-conterminous.

It is node S to make A_uDeagnostic structure in x same layer all sets of node, and N (x) for node x neighbours save Point set, the generation code of node x candidate's child node collection is described as follows：

Input：Node x, treats diagnosis node S_u

Output：Node x candidate's child node collection B

Referring to Fig. 3, setting up the method for the diagnostic test structure includes：

1. by treating diagnosis node S_uOr the maximum number of plies L of diagnostic test structure described in default_maxWith maximum child node Number C_max, wherein 1≤C_max≤|N(S_u) |, generate S_uCandidate's child node collection, select C_maxThe individual most nodes of neighbor node number As the first layer set of node of deagnostic structure；Wherein, S_uCandidate's child node integrate as node S_uIt is all Neighbor node collection.

2. since the second layer, i-th layer of node S is selected as steps described below_{I, j}, wherein 2≤i≤L_maxAnd 1≤j≤ C_max.We use set A_{I, j}Represent node S_{I, j}Candidate's child node collection, as node S_{I, j}In the presence of, with above-mentioned candidate's child node collection Production method produce or directly use S_{I, j}Neighborhood generation A_{I, j}；If | A_{I, j}| >=1 and i ＜ L_max- 1, then in A_{I, j} In select that the candidate's child node for possessing most neighbor nodes as S_{I+1, j}Or one candidate's child node conduct of random selection S_{I+1, j}；If | A_{I, j}| >=1 and i=L_max- 1, then randomly choose candidate's child node and be used as S_{I+1, j}；If | A_{I, j}|=0 or i= L_max, then just stop choosing S_{I, j}Child node.

Assuming that L_max=5 and C_maxWhen=4, Fig. 2 is shown for node S_uThe exemplary diagnostic test structure set up, It will be appreciated by those skilled in the art that the diagnostic test structure is an exemplary diagnostic test structure, rather than uniquely Diagnostic test structure.

According to described in above-mentioned steps on treating diagnosis node S_uDiagnostic test structure, can further perform diagnosis Operation, to obtain S_uFailure whether state.The present invention is carried out using the temporal correlation of wireless sensor network sensitive information Diagnostic test.So-called temporal correlation, refers to what is gathered in physical space apart from two close nodes close at the time of Information can be same or like.C (x, y) is made to represent that node x carries out sensitive information result of the comparison to node y, if sensitive information phase Closely, then C (x, y)=0 is made；If conversely, sensitive information difference is larger, making C (x, y)=1.When node x wants to carry out with node y Sensitive information compares, then node x is sent to node y compares request and sensing moment t, and then x and y senses ring in moment t simultaneously Environment information is to obtain the sensing value v of oneself_xWith v_y(such as sensing temperature, humidity), and y sends v_yTo x.When x receives y measurement After value, by v_xWith v_yCompare, if | v_x-v_y|≤θ, wherein θ are the threshold values being previously set, then illustrate x and y measured value Difference be that is, sensitive information is close in the difference range of permission, then C (x, y)=0；If | v_x-v_y| ＞ θ, i.e. sensitive information Difference is larger, then C (x, y)=1.

Make l_jRepresent S_uDeagnostic structure in jth arrange the actual maximum number of plies.The process of diagnostic test is detailed below, The process one is divided into four-wheel, sequentially carries out.All diagnostic tests in each round can be carried out parallel simultaneously.

First round diagnostic test：To 1 all≤j≤C_maxAndBy S_{3r-1, j}To S_{3r-2, j}Sensed Information compares.That is, carrying out sensing letter to 1+3i node layers to the 2+3i node layers of all diagnostic test structures Breath compares, i=0, and 1,2 ....

Second wheel diagnostic test：To 1 all≤j≤C_maxAndBy S_{3r, j}To S_{3r-1, j}Carry out sensing letter Breath compares.That is, carrying out sensitive information to 2+3i node layers to the 3+3i node layers of all diagnostic test structures Compare, i=0,1,2 ....

Third round diagnostic test：To 1 all≤j≤C_maxAndBy S_{3r+1, j}To S_{3r, j}Carry out sensing letter Breath compares.That is, carrying out sensitive information to 3+3i node layers to the 4+3i node layers of all diagnostic test structures Compare, i=0,1,2 ....

Fourth round diagnostic test：To 1 all≤j≤C_max, S first_{1, j}Collect the sensing that three-wheel is completed before this is listed in Information comparative resultThen the comparative result is sent to S_u。S_uThe first node layer of that most row of continuous 0 number since the first in sensitive information comparative result is selected to S_uCarry out Sensitive information compares.If sensitive information result of the comparison is 0, S is judged_uIt is malfunctioning node；Otherwise, it is determined that S_uIt is fault-free section Point.

As shown in figure 4, Bold arrows represent the diagnostic operation of first round execution.Because the maximum number of plies of each row is divided herein Not Wei 5,3,4,5, so in the first round simultaneously use node S_2,1Test node S_1,1, use node S_2,2Test node S_1,2, use node S_2,3Test node S_1,3, use node S_2,4Test node S_{Isosorbide-5-Nitrae}, use node S_5,1Test node S_4,1, use node S_5,4Test node S_4,4。

As shown in figure 5, Bold arrows represent the diagnostic operation that the second wheel is performed.Node S is used in second wheel simultaneously_3,1Test Node S_2,1, use node S_3,2Test node S_2,2, use node S_3,3Test node S_2,3, use node S_3,4Test node S_2,4。

As shown in fig. 6, Bold arrows represent the diagnostic operation that third round is performed.Node S is used in third round simultaneously_4,1Test Node S_3,1, use node S_4,3Test node S_3,3, use node S_4,4Test node S_3,4。

As shown in fig. 7, the test result sequence of four row nodes is respectively：＜ 0,0,0,0 ＞, ＜ 0,0 ＞, ＜ 0,1,0 ＞, The ＞ of ＜ 0,0,1,0.Obviously in first row, four continuous 0 and continuous 0 number are begun with this four sequences most from first place It is many, therefore selection node S_1,1Diagnosis node S is treated to test_u.It must be noted that, it is assumed that it is the ＞ of ＜ 1,1,0,0,1 to have a sequence, Because the first place numeral of sequence is 1, the sequence continuous 0 number since the first is 0.

Present invention also provides a kind of parallel diagnosis test system, the system includes：

Determining module, for determining to treat diagnosis node；

Judge module, diagnostic test structure is set up for judging whether to need for the band diagnosis node, if it is, logical Know that diagnostic test structure sets up module to set up based on the diagnostic test structure for treating diagnosis node, if it is not, then notifying diagnosis to survey Die trial block performs diagnostic test；

Diagnostic test structure sets up module, for setting up based on the diagnostic test structure for treating diagnosis node；

Diagnostic test module, for performing diagnostic test according to diagnostic test structure；

Diagnostic result judge module, for being judged to treat whether diagnosis node is malfunctioning node according to diagnostic test results.

Diagnostic test of the diagnostic test mode of the present invention for treating diagnosis node only needs to collect the node periphery merogenesis The test result of point, largely reduces unnecessary test, and the diagnostic test method is surveyed using the diagnosis designed by the present invention Structure is tried, the problem of occurring signal conflict when signal can be avoided to transmit during diagnostic test.

In addition, the time of parallel processing, effectively reduction diagnostic test is capable of in the diagnostic test of the diagnostic test method.Due to Diagnostic test needs to collect based on the diagnostic test results in deagnostic structure, while road wheel is entered in the diagnostic test in a parallel fashion Test, so the time of diagnostic test be approximately equal to signal transmission used in the time, that is, with treating in diagnostic test structure The distance of the farthest node of diagnostic test nodal distance, is then tested the spent time plus single diagnosis again, and diagnosis is surveyed Total cost time of examination is shorter.

Above example is merely to illustrate the present invention, and not limitation of the present invention, about the common skill of technical field Art personnel, without departing from the spirit and scope of the present invention, can also make a variety of changes and modification, therefore all etc. Same technical scheme falls within scope of the invention, and scope of patent protection of the invention should be defined by the claims.

Claims (9)

1. a kind of sensor network fault nodal parallel diagnostic test method, this method comprises the following steps：

(1) determine to treat diagnosis node；

(2) judge whether to need to set up diagnostic test structure for the band diagnosis node, if it is, step (3) is performed, if It is no, directly perform step (4)；

(3) set up based on the diagnostic test structure for treating diagnosis node；

(4) diagnostic test is performed according to diagnostic test structure；

(5) judged to treat whether diagnosis node is malfunctioning node according to diagnostic test results.

2. parallel diagnosis method of testing as claimed in claim 1, it is characterised in that：

The foundation specifically includes following steps based on the diagnostic test structure for treating diagnosis node：

By treating diagnosis node S_uOr the maximum number of plies L of diagnostic test structure described in default_maxWith maximum child node number C_max, Wherein 1≤C_max≤|N(S_u) |, N (S_u) it is to treat diagnosis node S_uNeighbor node set, by S_uNeighbor node concentrate select C_maxIndividual nodeAs the first layer set of node of deagnostic structure；

For i-th layer of node S_{I, j}, wherein 2≤i≤L_maxAnd 1≤j≤C_max, determine node S_{I, j}Candidate's child node collection A_{I, j}；

If | A_{I, j}| >=1 and i ＜ L_max- 1, then in A_{I, j}In select that the candidate's child node conduct for possessing most neighbor nodes S_{I+1, j}Or one candidate's child node of random selection is used as S_{I+1, j}；If | A_{I, j}| >=1 and i=L_max- 1, then randomly choose one Candidate's child node is used as S_{I+1, j}；If | A_{I, j}|=0 or i=L_max, then just stop choosing S_{I, j}Child node.

3. parallel diagnosis method of testing as claimed in claim 2, it is characterised in that：

Determine node S_{I, j}Candidate's child node collection A_{I, j}Specially directly use S_{I, j}Neighborhood generation A_{I, j}。

4. parallel diagnosis method of testing as claimed in claim 2, it is characterised in that：

Determine node S_{I, j}Candidate's child node collection A_{I, j}Specific method is：

To all and node S_{I, j}Adjacent node y, node S is put into if node y meets following two condition by node y_{I, j}'s Candidate's child node collection A_{I, j}In：

(1) node y and node S_uDistance be equal to node S_{I, j}With node S_uDistance add 1；

(2) node y and node S_uI-th layer of deagnostic structure in except node S_{I, j}Outside all nodes it is all non-conterminous.

5. any parallel diagnosis method of testing as described in claim 2-4, it is characterised in that：

It is described to be specially according to the execution diagnostic test of diagnostic test structure：

First round diagnostic test：To 1 all≤j≤C_maxAndBy S_{3r-1, j}To S_{3r-2, j}Carry out sensitive information ratio Compared with, wherein, l_jRepresent S_uDeagnostic structure in jth arrange the actual maximum number of plies；

Second wheel diagnostic test：To 1 all≤j≤C_maxAndBy S_{3r, j}To S_{3r-1, j}Carry out sensitive information ratio Compared with；

Third round diagnostic test：To 1 all≤j≤C_maxAndBy S3_{R+1, j}To S_{3r, j}Carry out sensitive information ratio Compared with.

6. parallel diagnosis method of testing as claimed in claim 5, it is characterised in that：

The sensitive information is more specific to be：

Node x is sent to node y compares request and sensing moment t；

Node x and y senses environmental information to obtain the sensing value v of oneself in moment t simultaneously_xWith v_y；

Node y is by v_yIt is sent to node x；

Node x is received after y measured value, by v_xWith v_yCompare, if | v_x-v_y|≤θ, wherein θ are the threshold values being previously set, then Sensitive information comparative result C (x, y)=0；If | v_x-v_y| ＞ θ, then sensitive information comparative result C (x, y)=1.

7. parallel diagnosis method of testing as claimed in claim 6, it is characterised in that：

It is described to be judged to treat whether diagnosis node is that malfunctioning node is specially according to diagnostic test results：

To 1 all≤j≤C_max, by S_{1, j}Collect the sensitive information comparative result that three-wheel is completed before jth is listed in Then the comparative result is sent to S_u；S_uSelect sensing Since the first continuous 0 most that first node layer arranged of number of starting to S in information comparative result_uCarry out sensitive information ratio Compared with；If sensitive information result of the comparison is 0, S is judged_uIt is malfunctioning node；Otherwise, it is determined that S_uIt is fault-free node.

8. parallel diagnosis method of testing as claimed in claim 6, it is characterised in that：

The environmental information includes temperature, humidity.

9. a kind of sensor network fault nodal parallel diagnostic test system that 1-8 either method is required for perform claim, should System includes：

Determining module, for determining to treat diagnosis node；

Judge module, diagnostic test structure is set up for judging whether to need for the band diagnosis node, if it is, notifying to examine Disconnected test structure sets up module to set up based on the diagnostic test structure for treating diagnosis node, if it is not, then notifying diagnostic test mould Block performs diagnostic test；

Diagnostic test structure sets up module, for setting up based on the diagnostic test structure for treating diagnosis node；

Diagnostic test module, for performing diagnostic test according to diagnostic test structure；

Diagnostic result judge module, for being judged to treat whether diagnosis node is malfunctioning node according to diagnostic test results.