Reliability comprehensive evaluation method for sensing nodes of mobile crowd sensing network

Abstract

The invention discloses a method for comprehensively evaluating the reliability of sensing nodes of a mobile crowd sensing network, and belongs to the technical field of mobile crowd sensing networks. Comprising a sensing node sensing task type classification method, a sensing node transmission rate scoring method, a sensing node stability scoring method, a sensing node sensing task completion efficiency scoring method and a sensing node reliability comprehensive scoring method. According to the method, multiple dimensions such as the sensing task type, the transmission rate, the stability, the sensing task completion efficiency and the reliability of the sensing nodes are considered, the sensing nodes in the mobile crowd sensing network are comprehensively, accurately and visually evaluated, and a scientific basis is provided for planning and optimization of the mobile crowd sensing network.

Description

Technical Field

[0001] The present invention belongs to the technical field of mobile crowd-sensing networks, and in particular relates to a method for comprehensively evaluating the reliability of perception nodes in a mobile crowd-sensing network. Background Art

[0002] Mobile crowd-intelligence perception network is a new paradigm for acquiring perception data in the Internet of Things. It uses perception nodes carrying mobile smart terminal devices as data sources. With the improvement of sensor capabilities and the popularization of mobile smart terminal devices, the application prospects of mobile crowd-intelligence perception network are broader. A typical mobile crowd-intelligence perception network consists of three parts: perception nodes, perception platforms, and service requesters. Service requesters initiate perception task requests to the perception platform. The perception platform issues perception tasks according to task requirements and selects appropriate perception nodes to participate. Perception nodes use mobile smart terminal devices to acquire perception data and upload perception data to the perception platform. Finally, the perception platform analyzes and aggregates the uploaded perception data and returns the analysis results to the service requester. At present, the following problems still exist in the reliability assessment of perception nodes in mobile crowd-intelligence perception networks:

[0003] (1) Existing evaluation methods lack consideration of indicators such as the transmission rate, stability, and efficiency of the sensing nodes in completing sensing tasks. Different sensing tasks have different emphases on different indicators, so it is necessary to formulate evaluation standards for relevant indicators.

[0004] (2) Existing evaluation methods usually only evaluate the reliability of perception nodes based on a single indicator and cannot conduct targeted evaluation of different types of perception tasks based on multiple indicators. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a comprehensive reliability evaluation method for sensing nodes in a mobile crowd sensing network, which can make a comprehensive, accurate and intuitive evaluation of the sensing nodes in the mobile crowd sensing network from multiple dimensions.

[0006] A method for comprehensively evaluating the reliability of sensing nodes in a mobile crowd sensing network comprises the following steps:

[0007] Step 1: Classify the sensing tasks of the sensing nodes from three perspectives, including classification from the perspective of the duration of the sensing tasks, classification from the perspective of the time sensitivity of the sensing tasks to the data requirements, and classification from the perspective of the type of sensing task data;

[0008] Step 2: Calculate the transmission rate score of the sensing node, the stability score of the sensing node, and the efficiency score of the sensing node in completing the sensing task;

[0009] Step 3: Use the scores calculated in step 2 to perform reliability scores on the three-angle classification perception node perception tasks in step 1.

[0010] The classification from the perspective of the duration of the perception task in step one includes continuous perception tasks and non-continuous perception tasks; the classification from the perspective of the time sensitivity of the perception task to the data requirements includes time-sensitive perception tasks and non-time-sensitive perception tasks; the classification from the perspective of the data type of the perception task includes binary perception tasks and numerical perception tasks.

[0011] The method for scoring the transmission rate of the sensing node in step 2 is:

[0012] There are m perception tasks in total, and the perception task set T = {t1, t2, L, t m}, there are n sensing nodes in total, and the set of sensing nodes U = {u1,u2,L,u n}, the relationship between the transmission rate of the sensing node and the network environment and transmission mode of the sensing node is expressed as:

[0013] The i-th sensor node u i Transmission rate V i The calculation method is:

[0014]

[0015] Where V i The unit is Mbps, t 1i The sensing platform receives the sensing node u i The time to upload the perception data, in seconds, t 0i is the sensing node u i The time to start uploading sensor data, in seconds, c i is the sensing node u i The size of the uploaded sensing data, in Mb, i=1,2,L,n;

[0016] The i-th sensor node u i Transfer Rate Ratings 1i The calculation method is:

[0017]

[0018] The perception node stability scoring method is:

[0019] The i-th sensor node u i The number of perception tasks completed each day within three days is recorded as C 0i , C 1i , C 2i , get the sensing node u i The average value within three days Ci , and the stability variance X of the sensing node completing the sensing task every day within three days i

[0020]

[0021] The i-th sensor node u i Stability Ratings 2i The calculation method is:

[0022]

[0023] The scoring method for the efficiency of the sensing node in completing the sensing task is:

[0024] The i-th sensor node u i The time it takes to complete all perception tasks is o i , the average time for all n sensing nodes in the sensing node set U to complete all sensing tasks is:

[0025]

[0026] Among them, the time unit is second;

[0027] The i-th sensor node u i Complete the perceived task efficiency score 3i The calculation method is:

[0028]

[0029] The reliability scoring method for the three-angle classification perception node perception tasks in step 3 is:

[0030] Classification from the perspective of perceived task duration:

[0031] The sensing tasks of the sensing nodes are divided into continuous sensing tasks and non-continuous sensing tasks. m The jth perception task t in j , perception task type D j The value method of:

[0032]

[0033] For the i-th sensor node u in the sensor node set U i , assuming u i The remaining power is E i , in milliampere-hours, sensing node u i The rated capacity is The unit is milliampere-hour. For the sensing task t j Sensing node ui The reliability score SD of the sensing node for continuous sensing tasks and non-continuous sensing tasks ij The calculation method is:

[0034]

[0035] Where ω1 is u i The remaining power weight, ω2 is the weight of the transmission rate score of the sensing node, ω3 is the weight of the stability score of the sensing node, and ω4 is the weight of the efficiency score of the sensing node to complete the sensing task. The value range of each weight is (0,1);

[0036] SD ij The range is 0≤SD ij ≤10, when 0≤SD ij <6, perception task t j ,u i The reliability is low; when 6≤SD ij ≤7, the perception task t j ,u i The reliability is qualified; when 7<SD ij <9: Perception task t j ,u i The reliability is good; when 9≤SD ij ≤10, the perception task t j ,u i Excellent reliability;

[0037] Classification from the perspective of the time sensitivity of perception tasks to data requirements:

[0038] When the sensing tasks of the sensing nodes include time-sensitive sensing tasks and non-time-sensitive sensing tasks, the sensing task set T = {t1, t2, L, t m The jth perception task t in j , perception task type F j The value of is:

[0039]

[0040] The perception platform releases the perception task t j The time is St j , in seconds, the sensing platform receives the sensing node u i Time Dt for uploading sensory data ij , in seconds, for the perception task t j Sensing node u i The reliability score SF of the sensing node for time-sensitive sensing tasks and non-time-sensitive sensing tasks ij The calculation method is:

[0041]

[0042] Where ω5 is the time sensitivity weight of the perception task to the data demand, and its value range is (0,1); SF ij The range is 0≤SF ij ≤10, when 0≤SF ij <6, perception task t j ,u i The reliability is low; when 6≤SF ij ≤7, the perception task t j ,u i The reliability is qualified; when 7<SF ij <9: Perception task t j ,u i The reliability is good; when 9≤SF ij ≤10, the perception task t j ,u i The reliability is excellent.

[0043] From the perspective of perception task data type, it can be classified into:

[0044] The sensing tasks of the sensing nodes are divided into binary sensing tasks and numerical sensing tasks. m The jth perception task t in j , perception task type Q j The value of is:

[0045]

[0046] If Q j =1, the i-th sensor node u in the sensor node set U i The number of binary perception tasks completed within three days that match the perception task results is χ0, in units, and the perception node u i The total number of binary perception tasks completed within three days is χ, in units, and the perception node u i The percentage of binary perception task results completed within three days that match θ i for:

[0047]

[0048] If Q j = 0, the sensing node set U has completed the sensing task t j The i-th sensor node u i , let the set U contain u i A subset of γ sensor nodes Completed the numerical perception task t j, for the set U J The xth sensor node in The perception data uploaded to the perception platform is q x ,x=1,2,L,γ,numerical perception task t j The actual result data is q0, q x The absolute value of the difference with q0 is k x for:

[0049] k x =|q x -q0|

[0050] The average difference of γ sensor nodes for:

[0051]

[0052] For the perception task t j Sensing node u i The comprehensive score SQ of the reliability of the sensing node that combines the correctness of the sensing node in completing the sensing task ij The calculation method is:

[0053]

[0054] Where ω6 is the correctness weight of the sensing node completing the sensing task, and its value range is (0,1); SQ ij The range is 0≤SQ ij ≤10, when 0≤SQ ij <6, perception task t j ,u i The reliability is low; when 6≤SQ ij ≤7, the perception task t j ,u i The reliability is qualified; when 7<SQ ij <9: Perception task t j ,u i The reliability is good; when 9≤SQ ij ≤10, the perception task t j ,u i The reliability is excellent.

[0055] Through the above design scheme, the present invention can bring the following beneficial effects: a comprehensive evaluation method for the reliability of perception nodes in a mobile crowd-intelligence perception network, which takes into account multiple dimensions such as the perception task type, transmission rate, stability, efficiency of completing perception tasks, and reliability of the perception nodes, and makes a comprehensive, accurate, and intuitive evaluation of the perception nodes in the mobile crowd-intelligence perception network, providing a scientific basis for the planning and optimization of the mobile crowd-intelligence perception network. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] The present invention is further described below with reference to the accompanying drawings and specific embodiments:

[0057] Figure 1 The present invention is a flowchart of a method for comprehensive reliability evaluation of sensing nodes in a mobile crowd sensing network. DETAILED DESCRIPTION

[0058] A comprehensive evaluation method for the reliability of sensing nodes in a mobile crowdsensing network, such as Figure 1 As shown, it includes a classification method for the sensing task type of the sensing node, a scoring method for the transmission rate of the sensing node, a scoring method for the stability of the sensing node, a scoring method for the efficiency of the sensing node in completing the sensing task, and a comprehensive scoring method for the reliability of the sensing node. The specific contents are as follows:

[0059] (1) Classification method of sensing node sensing task types;

[0060] The perception tasks of perception nodes can be classified from different perspectives. From the perspective of the duration of the perception tasks, the perception tasks can be divided into continuous perception tasks and non-continuous perception tasks. Continuous perception tasks refer to perception tasks in which the perception nodes continuously upload perception data within a period of 60 minutes or more, otherwise they are non-continuous perception tasks; from the perspective of the timeliness sensitivity of the perception tasks to data requirements, they can be divided into timeliness-sensitive perception tasks and non-timeliness-sensitive perception tasks. Timeliness-sensitive perception tasks refer to perception tasks that require the perception nodes to be completed within 5 minutes, otherwise they are non-timeliness-sensitive perception tasks; from the perspective of the data type of the perception tasks, they can be divided into binary perception tasks and numerical perception tasks. Binary perception tasks refer to perception data uploaded by the perception nodes to the perception platform as "yes" or "no", "1" or "0" and "true" or "false", and numerical perception tasks refer to perception data uploaded by the perception nodes to the perception platform as a specific value;

[0061] (2) A method for scoring the transmission rate of sensing nodes;

[0062] Assume that there are m perception tasks, and the perception task set T = {t1, t2, L, t m}, there are n sensing nodes in total, and the set of sensing nodes U = {u1,u2,L,u n}, the transmission rate of the sensing node is mainly related to the network environment and transmission mode of the sensing node. i Transmission rate V i The calculation method is:

[0063]

[0064] Among them, V i The unit is Mbps, t 1i The sensing platform receives the sensing node ui The time to upload the perception data, in seconds, t 0i is the sensing node u i The time to start uploading sensor data, in seconds, c i is the sensing node u i The size of the uploaded sensing data, in Mb, i=1,2,L,n;

[0065] The i-th sensor node u i Transfer Rate Ratings 1i The calculation method is:

[0066]

[0067] (3) Perception node stability scoring method;

[0068] Assume that the i-th sensor node u i The number of perception tasks completed each day within three days is C 0i , C 1i , C 2i , the unit is: , then the sensing node u i The average number of perception tasks completed each day within three days C i The calculation method is:

[0069]

[0070] The stability variance of the sensing node completing the sensing task every day within three days is X i It can be expressed as:

[0071]

[0072] The i-th sensor node u i Stability Ratings 2i The calculation method is:

[0073]

[0074] (4) Method for scoring the efficiency of sensing nodes in completing sensing tasks;

[0075] Assume that the i-th sensor node u i The time it takes to complete all perception tasks is o i , in seconds, then the average time for all n sensing nodes in the sensing node set U to complete all sensing tasks is:

[0076]

[0077] Among them, the unit of time is seconds;

[0078] The i-th sensor node ui Complete the perceived task efficiency score 3i The calculation method is:

[0079]

[0080] (5) Comprehensive scoring method for perception node reliability

[0081] When the sensing tasks of the sensing nodes are divided into continuous sensing tasks and non-continuous sensing tasks, for the sensing task set T = {t1, t2, L, t m The jth perception task t in j , perception task type D j The value of is as follows:

[0082]

[0083] For the i-th sensor node u in the sensor node set U i , assuming u i The remaining power is E i , in milliampere-hours, sensing node u i The rated capacity is The unit is milliampere-hour. For the sensing task t j Sensing node u i The reliability score SD of the sensing node for continuous sensing tasks and non-continuous sensing tasks ij The calculation method is:

[0084]

[0085] Among them, ω1 is u i The remaining power weight, ω2 is the weight of the perception node transmission rate score, ω3 is the weight of the perception node stability score, and ω4 is the weight of the perception node to complete the perception task efficiency score. Each weight can be adjusted according to the specific situation. The value range is (0,1). The default value is SD ij The range is 0≤SD ij ≤10, when 0≤SD ij <6, it is considered as the perception task t j ,u i The reliability is low, when 6≤SD ij ≤7, it is considered as the perception task t j ,u i The reliability is qualified when 7<SD ij When <9, it is considered as the perceptual task t j ,u i The reliability is good, when 9≤SD ij ≤10, it is considered as the perception task tj ,u i The reliability is excellent.

[0086] When the sensing tasks of the sensing nodes are divided into time-sensitive sensing tasks and non-time-sensitive sensing tasks, for the sensing task set T = {t1, t2, L, t m The jth perception task t in j , perception task type F j The value of is as follows:

[0087]

[0088] Assume that the perception platform releases the perception task t j The time is St j , in seconds, the sensing platform receives the sensing node u i Time Dt for uploading sensory data ij , in seconds, for the perception task t j Sensing node u i The reliability score SF of the sensing node for time-sensitive sensing tasks and non-time-sensitive sensing tasks ij The calculation method is:

[0089]

[0090] Among them, ω5 is the time sensitivity weight of the perception task to the data demand. Each weight can be adjusted according to the specific situation. The value range is (0,1). The default value is SF ij The range is 0≤SF ij ≤10, when 0≤SF ij <6, it is considered as the perception task t j ,u i The reliability is low, when 6≤SF ij ≤7, it is considered as the perception task t j ,u i The reliability is qualified when 7<SF ij <9, it is considered as the perception task t j ,u i The reliability is good, when 9≤SF ij ≤10, it is considered as the perception task t j ,u i The reliability is excellent.

[0091] When the sensing tasks of the sensing nodes are divided into binary sensing tasks and numerical sensing tasks, for the sensing task set T = {t1, t2, L, t m The jth perception task t in j , perception task type Qj The value of is as follows:

[0092]

[0093] If Q j =1, for the i-th sensor node u in the sensor node set U i The number of binary perception tasks completed within three days that match the perception task results is χ0, in units, and the perception node u i The total number of binary perception tasks completed within three days is χ, in units, and the perception node u i The percentage of binary perception task results completed within three days that match θ i for:

[0094]

[0095] If Q j = 0, for the sensing node set U that has completed the sensing task t j The i-th sensor node u i , let the set U contain u i A subset of γ sensor nodes Completed the numerical perception task t j , for the set U J The xth sensor node in The perception data uploaded to the perception platform is q x ,x=1,2,L,γ,numerical perception task t j The actual result data is q0, q x The absolute value of the difference with q0 is k x for:

[0096]

[0097] The average difference of γ sensor nodes for:

[0098]

[0099] For the perception task t j Sensing node u i The comprehensive score SQ of the reliability of the sensing node that combines the correctness of the sensing node in completing the sensing task ij The calculation method is:

[0100]

[0101] Among them, ω6 is the correctness weight of the sensing node completing the sensing task. Each weight can be adjusted according to the specific situation. The value range is (0,1). The default value is SQ ij The range is 0≤SQ ij ≤10, when 0≤SQ ij <6, it is considered as the perception task t j ,u i The reliability is low, when 6≤SQ ij ≤7, it is considered as the perception task t j ,u i The reliability is qualified, when 7<SQ ij When <9, it is considered as the perceptual task t j ,u i The reliability is good, when 9≤SQ ij ≤10, it is considered as the perception task t j ,u i The reliability is excellent.

[0102] The software program involved in the present invention is compiled based on mobile communication, network and computer processing technologies, which are technologies familiar to those skilled in the art.

[0103] The specific embodiments of the present invention are merely examples for clear explanation and are not limitations on the implementation methods. For ordinary technicians in the field, other different forms of changes or modifications can be made based on the above description. There is no need to list all the implementation methods here, and the obvious changes or modifications derived therefrom are still within the scope of protection of the present invention.

Claims

1. A method for comprehensive reliability evaluation of sensing nodes in a mobile crowd-sensing network, characterized by: The following steps are included: Step 1: Classify the sensing tasks of the sensing nodes from three perspectives, including classification from the perspective of the duration of the sensing tasks, classification from the perspective of the time sensitivity of the sensing tasks to the data requirements, and classification from the perspective of the type of sensing task data; Step 2: Calculate the transmission rate score of the sensing node, the stability score of the sensing node, and the efficiency score of the sensing node in completing the sensing task; Step 3: Use the scores calculated in step 2 to perform reliability scores on the three-angle classification perception node perception tasks in step 1.

2. According to claim 1, a method for comprehensive evaluation of the reliability of sensing nodes in a mobile crowd-sensing network is characterized by:

1. The classification from the perspective of the duration of the perception task includes continuous perception tasks and non-continuous perception tasks; the classification from the perspective of the time sensitivity of the perception task to the data demand includes time-sensitive perception tasks and non-time-sensitive perception tasks; the classification from the perspective of the data type of the perception task includes binary perception tasks and numerical perception tasks.

3. According to claim 1, a method for comprehensive reliability evaluation of sensing nodes in a mobile crowd-sensing network is characterized by: The method for scoring the transmission rate of the sensing node in step 2 is: There are m perception tasks in total, and the perception task set T = {t1, t2, L, t m }, there are n sensing nodes in total, and the set of sensing nodes U = {u1,u2,L,u n }, the relationship between the transmission rate of the sensing node and the network environment and transmission mode of the sensing node is expressed as: The i-th sensor node u i Transmission rate V i The calculation method is: Where V i The unit is Mbps, t 1i The sensing platform receives the sensing node u i The time to upload the perception data, in seconds, t 0i is the sensing node u i The time to start uploading sensor data, in seconds, c i is the sensing node u i The size of the uploaded sensing data, in Mb, i=1,2,L,n; The i-th sensor node u i Transfer Rate Ratings 1i The calculation method is: The perception node stability scoring method is: The i-th sensor node u i The number of perception tasks completed each day within three days is recorded as C 0i , C 1i , C 2i , get the sensing node u i The average value within three days C i , and the stability variance X of the sensing node completing the sensing task every day within three days i The i-th sensor node u i Stability Ratings 2i The calculation method is: The scoring method for the efficiency of the sensing node in completing the sensing task is: The i-th sensor node u i The time it takes to complete all perception tasks is o i , the average time for all n sensing nodes in the sensing node set U to complete all sensing tasks is: Among them, the time unit is second; The i-th sensor node u i Complete the perceived task efficiency score 3i The calculation method is:

4. According to claim 1, a method for comprehensive reliability evaluation of sensing nodes in a mobile crowd-sensing network is characterized by: The reliability scoring method for the three-angle classification perception node perception tasks in step 3 is: Classification from the perspective of perceived task duration: The sensing tasks of the sensing nodes are divided into continuous sensing tasks and non-continuous sensing tasks. m The jth perception task t in j , perception task type D j The value method of: For the i-th sensor node u in the sensor node set U i , assuming u i The remaining power is E i , in milliampere-hours, sensing node u i The rated capacity is The unit is milliampere-hour. For the sensing task t j Sensing node u i The reliability score SD of the sensing node for continuous sensing tasks and non-continuous sensing tasks ij The calculation method is: Where ω1 is u i The remaining power weight, ω2 is the weight of the transmission rate score of the sensing node, ω3 is the weight of the stability score of the sensing node, and ω4 is the weight of the efficiency score of the sensing node to complete the sensing task. The value range of each weight is (0,1); SD ij The range is 0≤SD ij ≤10, when 0≤SD ij <6, perception task t j ,u i The reliability is low; when 6≤SD ij ≤7, the perception task t j ,u i The reliability is qualified; when 7<SD ij <9: Perception task t j ,u i The reliability is good; when 9≤SD ij ≤10, the perception task t j ,u i Excellent reliability; Classification from the perspective of the time sensitivity of perception tasks to data requirements: When the sensing tasks of the sensing nodes include time-sensitive sensing tasks and non-time-sensitive sensing tasks, the sensing task set T = {t1, t2, L, t m The jth perception task t in j , perception task type F j The value of is: The perception platform releases the perception task t j The time is St j , in seconds, the sensing platform receives the sensing node u i Time Dt for uploading sensory data ij , in seconds, for the perception task t j Sensing node u i The reliability score SF of the sensing node for time-sensitive sensing tasks and non-time-sensitive sensing tasks ij The calculation method is: Where ω5 is the time sensitivity weight of the perception task to the data demand, and its value range is (0,1); SF ij The range is 0≤SF ij ≤10, when 0≤SF ij <6, perception task t j ,u i The reliability is low; when 6≤SF ij ≤7, the perception task t j ,u i The reliability is qualified; when 7<SF ij <9: Perception task t j ,u i The reliability is good; when 9≤SF ij ≤10, the perception task t j ,u i The reliability is excellent. From the perspective of perception task data type, it can be classified into: The sensing tasks of the sensing nodes are divided into binary sensing tasks and numerical sensing tasks. m The jth perception task t in j , perception task type Q j The value of is: If Q j =1, the i-th sensor node u in the sensor node set U i The number of binary perception tasks completed within three days that match the perception task results is χ0, in units, and the perception node u i The total number of binary perception tasks completed within three days is χ, in units, and the perception node u i The percentage of binary perception task results completed within three days that match θ i for: If Q j = 0, the sensing node set U has completed the sensing task t j The i-th sensor node u i , let the set U contain u i A subset of γ sensor nodes Completed the numerical perception task t j , for the set U J The xth sensor node in The perception data uploaded to the perception platform is q x ,x=1,2,L,γ,numerical perception task t j The actual result data is q0, q x The absolute value of the difference with q0 is k x for: k x =|q x -q0| The average difference of γ sensor nodes for: For the perception task t j Sensing node u i The comprehensive score SQ of the reliability of the sensing node that combines the correctness of the sensing node in completing the sensing task ij The calculation method is: Where ω6 is the correctness weight of the sensing node completing the sensing task, and its value range is (0,1); SQ ij The range is 0≤SQ ij ≤10, when 0≤SQ ij <6, perception task t j ,u i The reliability is low; when 6≤SQ ij ≤7, the perception task t j ,u i The reliability is qualified; when 7<SQ ij <9: Perception task t j ,u i The reliability is good; when 9≤SQ ij ≤10, the perception task t j ,u i The reliability is excellent.

Images