System and Method for Providing Smart Health Analysis

Abstract

The present invention provides a smart health analysis system and method that performs big data processing and communication load balancing using iterative hashing and regression analysis in an IoT communication network, and applies machine learning techniques for IoT networks. The present invention has the effect of improving the performance of network throughput, latency, and packet drop rate according to changes in node speed and data detection speed. The present invention can provide a minimum cost solution for e-health services and transmission using regression analysis. The present invention has the effect of enabling IoT systems to more easily achieve confidentiality, authentication, and communication integrity even when there are potential threats.

Description

Technology Field

[0001] The present invention relates to a smart health analysis system and method, and more specifically, to a smart health analysis system and method that performs big data processing and communication load balancing using iterative hashing and regression analysis in an IoT communication network, and applies machine learning techniques for IoT networks. Background Technology

[0002] In today's intelligent world, all intelligent devices or sensors can collect and process data from the environment and then send it to other connected networking systems.

[0003] However, since IoT devices have limited resources, the network's ability to maintain data stability is degraded, and routing holes become more common.

[0004] In addition, there is a problem where the node is damaged during the process of sending large amounts of data to the sink node, making it impossible for application users to accurately and effectively collect the data they need.

[0005] In addition to the operational objectives of IoT-based sensors that detect, process, and transmit network data, one of the core and critical performance factors is data latency. This is significantly affected by constraint devices, and as a result, communication networks experience failures in routing large amounts of data.

[0006] The use of optimization technology for the development of IoT-based green computing systems is rapidly increasing. Unlike other communication systems, sensor nodes have limited resources, which affects network resources for big data management. Prior art literature

[0007] Korean Registration No. 10-2109771 The problem to be solved

[0008] To solve these problems, the present invention aims to provide a smart health analysis system and method that performs big data processing and communication load balancing using iterative hashing and regression analysis in an IoT communication network, and applies machine learning techniques for IoT networks. means of solving the problem

[0009] A smart health analysis system according to the features of the present invention for achieving the above objective is,

[0010] A data collection unit that collects health data from one or more sensors and registers the sensors to a node adjacent to the sensors;

[0011] A router selection unit that randomly selects a router, which is a node to relay health data received from the above sensor, and the cost of each selected router ( A data relay unit comprising a cost calculation unit that calculates ) by a cost function, and a traffic prediction unit that calculates a weighted score by accumulating and summing the costs of each router calculated above, and predicts traffic according to the calculated weighted score to select the address of the next hop router; and

[0012] It includes a data security unit that encrypts health data to be transmitted to the address of the selected router using a blockchain.

[0013] A smart health analysis method according to the features of the present invention is,

[0014] A smart health analysis method using a smart health analysis system including a data collection unit, a data relay unit, and a data security unit,

[0015] The above data collection unit collects health data from one or more sensors and registers the sensors to a node adjacent to the sensors;

[0016] The above data relay unit randomly selects a router, which is a node to relay health data received from the sensor, and the cost of each selected router ( Step of calculating ) by the cost function;

[0017] The data relay unit calculates a weighted score by accumulating and summing the costs of each router calculated above, and predicts traffic based on the calculated weighted score to select the address of the next hop router; and

[0018] The above data security unit includes the step of encrypting health data to be transmitted to the address of the selected router into a blockchain. Effects of the invention

[0019] With the aforementioned configuration, the present invention has the effect of improving the performance of network throughput, latency, and packet drop rate according to changes in node speed and data detection speed.

[0020] The present invention can provide a minimum cost solution for e-health services and transmission using regression analysis.

[0021] The present invention has the effect of enabling IoT systems to more easily achieve confidentiality, authentication, and communication integrity even when there are potential threats. Brief explanation of the drawing

[0022] FIG. 1 is a diagram showing the configuration of a smart health analysis system using regression analysis with iterative hashing in an IoT communication network according to an embodiment of the present invention. FIG. 2 is a diagram illustrating a method for predicting the optimal router to transmit health data using machine learning according to an embodiment of the present invention. FIG. 3 is a diagram showing an algorithm for performing data security when transmitting health data according to an embodiment of the present invention. Specific details for implementing the invention

[0023] Throughout the specification, when a part is described as "including" a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0024] FIG. 1 is a diagram showing the configuration of a smart health analysis system using regression analysis with iterative hashing in an IoT communication network according to an embodiment of the present invention, and FIG. 2 is a diagram showing a method for predicting the optimal router to transmit health data using machine learning according to an embodiment of the present invention.

[0025] The smart health analysis system (100) provides the main functions of network initialization, health data collection, data relay and data security.

[0026] Through network initialization, all biosensors (111) are implanted in the patient's body and share their location information with the nearest node. Additionally, each sensor must be registered with the nearest relay node to verify its authenticity.

[0027] The biosensor (111) can also be described as a medical sensor.

[0028] A smart health analysis system (100) according to an embodiment of the present invention may include a data collection unit (110), a data relay unit (120), and a data security unit (130). In addition, it may further include an external medical server.

[0029] The data collection unit (110) includes one or more biosensors (111), a gateway node (112), and a sink node (113).

[0030] Each biosensor (111) can share local information with a nearby node, sense health data, register with the nearest relay node, and consider duplicate data sensing within a preset radius (S100).

[0031] The data collection unit (110) can collect and store health data from one or more biosensors (111).

[0032] The biosensor (111) includes a biosensor (111) having static sink nodes (113) that interact with each other using an undirected graph G(N, E) of edge E. Each node has health data Assume there is a unique identity (ID) that aims to collect.

[0033] The gateway node (112) is searched for additional transmission to the medical server for data collection and analysis.

[0034] The medical server receives health data from the smart health analysis system (100) and interacts with the cloud service for data storage and high processing resources.

[0035] The gateway node (112) contains enough memory to store neighbor data.

[0036] The biosensor (111) also considers redundant data sensing within a preset radius. When two nodes are in the same radius, the collected health data D i is considered duplicated.

[0037] A list of symbols and abbreviations used in the present invention is shown in Table 1 and Table 2.

[0038]

[0039]

[0040] The key components of the present invention are network infrastructure, statistical regression analysis, and smart security. Each component plays a crucial role in the development and improvement of medical systems by supporting data privacy and reliability.

[0041] The following provides details of an IoT-based smart health analysis system that supports the machine learning techniques listed in Algorithm 1.

[0042] A data relay unit (120) according to an embodiment of the present invention includes a router selection unit (121), a cost calculation unit (122), and a traffic prediction unit (123).

[0043] The router selection unit (121) calculates a cost value based on a cost function, determines that it is optimal if the cost value is smaller than a threshold value, and otherwise selects an alternative router.

[0044] The router selection unit (121) randomly selects an arbitrary router for communication link setup (S101). An arbitrary router can be represented as shown in the following mathematical formula 1.

[0045]

[0046] Here, can represent a random router.

[0047] The cost calculation unit (122) uses the following mathematical formula 2 to calculate each router The cost of ( ) can be calculated by the cost function (S101).

[0048] Each router has a cost based on the initially calculated distance. Possesses . Subsequent costs The value can be recalculated based on the following mathematical formula 2 using packet reception and link interference parameters.

[0049]

[0050] Here, is the router cost value, pks is the number of transmitted packets, and trans is the link at active time t. Indicates the number of retransmissions through. If the router containing is greater than the preset threshold T, the specific router is marked as untrusted as shown in Equation 3 below.

[0051]

[0052] Here, is a random router, T is a preset threshold, represents the cost value of the router.

[0053] The cost calculation unit (122) can determine the optimal cost value by mathematical formula 2 and mathematical formula 3 (S102).

[0054] The traffic prediction unit (123) can calculate a weighted score by accumulating the costs of each router using the following mathematical formula 4 (S103).

[0055] Weighted Score of each router It integrates the cost values ​​calculated for a specific communication link as follows.

[0056]

[0057] Here, is the weighted score of each router, and n is the total number of routers in a specific communication link, is the cost value of the router, and i represents the sequence number of the router in a specific communication link.

[0058] The traffic prediction unit (123) calculated The value is stored in the routing table and calculated The value can be used as a parameter in statistical regression analysis.

[0059] The traffic prediction unit (123) can calculate the traffic delay rate by performing statistical regression analysis of each router using the following mathematical formula 5 (S104).

[0060]

[0061] Here, is the traffic latency, and is y Intercept, and is the regression coefficient, and is the weighted score of all router devices, and represents an error variable.

[0062] The primary purpose of predicting values ​​is to determine high packet forwarding and minimal link interference for each router (node).

[0063] The traffic prediction unit (123) determines whether each router has the lowest traffic delay rate (S105), and selects the transmission router with the lowest calculated traffic delay rate as the address of the next hop router (S106).

[0064] The traffic prediction unit (123) can transmit the address of the selected router and encrypted health data to the sink node (113) through the gateway node (112).

[0065] The gateway node (112) must register the nearest biosensor (111) to ensure data reliability. Then, the data relay unit (120) considers various quality recognition factors (link failure and packet reception information) to calculate a cost function.

[0066] The data relay unit (120) determines the cumulative cost from the source node to the destination node and, in the case of optimal traffic prediction, can select the next hop for health data transmission.

[0067] The following Algorithm 1 shows the pseudocode for next-hop prediction through statistical regression analysis.

[0068]

[0069] FIG. 3 is a diagram showing an algorithm for performing data security when transmitting health data according to an embodiment of the present invention.

[0070] The data security unit (130) according to an embodiment of the present invention may further include a secret key encryption unit (131).

[0071] The data security department (130) provides a security algorithm for analyzing potential threats.

[0072] The sink node (113) processes the key sharing process for the biosensor (111) and gateway node (112) before transmitting health data to the medical server.

[0073] The gateway node (112) transmits health data received from the biosensor (111) to the data relay unit (120).

[0074] The data relay unit (120) transmits health data to the data security unit (130), and the data security unit (130) encrypts the health data using a blockchain and transmits it to the data relay unit (120).

[0075] The data relay unit (120) transmits encrypted health data and the address information of the selected router to the sink node (113) via the gateway node (112).

[0076] Additionally, the gateway node (112) can register the nearest sensor to ensure data reliability, and receive information about the next hop to transmit health data through the data relay unit (120), and receive health data encrypted with a blockchain.

[0077] The sink node (113) transmits the encrypted health data received from the gateway to an external medical server.

[0078] The data security unit (130) can receive health data through the data relay unit (120) (S110).

[0079] In addition, key information of each node (gateway node (112), sink node (113), etc.) can be stored in a routing table. For example, the secret key encryption unit (131) is a node A secret key Generate, select the address of the next hop's router, and then the same secret key is the next hop np's public key It is encrypted as.

[0080]

[0081] Here, is a node, is encryption using the public key, np is the next hop, and can represent the secret key, and ID can represent the unique ID (Identity) of the source node.

[0082] When health data is received, np searches for the private key, obtains the node's ID and key information, and decrypts the secret data.

[0083] The secret key generation of the secret key encryption unit (131) can control the sink node (113) and later perform encryption functions to achieve data integrity using iterative hashing on the IoT sensor and gateway node (112).

[0084] The secret key encryption unit (131) maintains data privacy and security using a hash technique. The secret key encryption unit (131) has been verified and tested to safely transmit health data in block form using a security algorithm, and performs hashing technology for each iteration.

[0085] As a result, the secret key encryption unit (131) is provided with a high level of data integrity through a trust chain. The secret key encryption unit (131) can perform information processing such as the following mathematical formula 7 by executing Cipher Block Chaining (CBC). The secret key encryption unit (131) can generate a block of health data using mathematical formula 6 (S111).

[0086]

[0087] Here, is a Cipher Block, and E is a secret key It is an encryption process based on, is an encryption with a secret key, and is the previous hash, and It can represent health data.

[0088] Subsequently, the proposed security algorithm is the next level secret key You can use it and calculate the MAC for one block.

[0089]

[0090] Accordingly, health data is received from the sink node (113) via a digital hash and provides a confidence level for each block.

[0091] The secret key encryption unit (131) can generate an encrypted block of health data by performing iterative level hashing of the health data according to mathematical formulas 7 and 8 (S112).

[0092] The secret key encryption unit (131) transmits the encrypted health data to the sink node (113) via the data relay unit (120) and the gateway node (112).

[0093] At this time, the secret key encryption unit (131) transmits an identity request signal from the gateway node (112) to verify the identity (S113).

[0094] The secret key encryption unit (131) receives an identity response signal corresponding to an identity request signal from the gateway node (112) and determines whether the identity has been verified (S114).

[0095] When the identity is verified, the secret key encryption unit (131) transmits all health data to the sink node (113) (S115).

[0096] The sink node (113) can transmit encrypted health data to an external medical server (S116).

[0097] Encryption technology obtains a calculated hashing block to retrieve the collected data.

[0098] The decryption technique for a hashing block can be performed using the following mathematical formula 9.

[0099]

[0100] Here, D can represent the decoded data.

[0101] It consists of two main procedures: cryptographic block generation and data verification. In block generation, all health data is divided into fixed-size chunks, and each chunk operates separately for data transmission through a routing path.

[0102] Each individual data block is a digital hash using the XoR operation with the security key, and once completed, the hash for the next iteration is calculated.

[0103] In this way, health data is transmitted to the sink node (113) through the discovery of the gateway node (112). The health data is stored on a cloud server where privacy and integrity are guaranteed.

[0104] Algorithm 2 describes a security hashing technique using a gateway.

[0105]

[0106] Accordingly, the present invention provides a smart health analysis system utilizing machine learning techniques for IoT networks that optimize the learning process through link and traffic prediction support. The system of the present invention utilizes regression analysis to provide a minimum cost solution for e-health services and transmission.

[0107] Although embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements by those skilled in the art using the basic concept of the present invention as defined in the following claims also fall within the scope of the present invention. Explanation of the symbols

[0108] 100: Smart Health Analysis System 110: Data Collection Unit 111: Biosensor 112: Gateway Node 113: Sync Node 120: Data Relay Department 121: Router selection 122: Cost Calculation Department 123: Traffic Prediction Unit 130: Data Security Department 131: Secret key encryption section

Claims

Claim 1 A data collection unit that collects health data from one or more sensors and registers the sensors to the node closest to the sensors; a router selection unit that randomly selects a router, which is a node to relay health data received from the sensors, and the cost of each selected router ( A data relay unit comprising a cost calculation unit that calculates ) by a cost function, and a traffic prediction unit that calculates a weighted score by accumulating and summing the costs of each router calculated above, and predicts traffic according to the calculated weighted score to select the address of the next hop router; and a data security unit that encrypts health data to be transmitted to the address of the selected router using a blockchain, wherein the data collection unit further includes: a Gateway Node that registers the nearest sensor to ensure data reliability and receives information of the next hop to which the health data is to be transmitted through the data relay unit, and receives the health data encrypted using a blockchain; and a Sink Node that transmits the encrypted health data received from the Gateway to an external medical server, wherein the cost calculation unit holds the calculated cost from each router, recalculates the cost according to the following mathematical formula 1, and if the recalculated cost according to the following mathematical formula 2 is greater than a preset threshold T, determines that the router corresponding to the recalculated cost cannot be trusted, [Mathematical Formula 1] Here, is the router cost value, pks is the number of transmitted packets, and trans is the link at active time t. This is the number of retransmissions through.[Equation 2] Here, is a random router, T is a preset threshold, is the cost value of the router. The above traffic prediction unit calculates a weighted score by accumulating and summing the costs of each router calculated above using the following mathematical formula 3, stores the calculated weighted score in a routing table, and [Mathematical Formula 3] Here, is the weighted score of each router, and n is the total number of routers in a specific communication link, is the cost value of the router, and i is the sequence number of the router in a specific communication link. The traffic prediction unit calculates the traffic delay rate by performing statistical regression analysis of each router using the following mathematical formula 4 and the calculated weighted score, and selects the transmission router with the smallest calculated traffic delay rate as the address of the next hop router, [Mathematical Formula 4] Here, is the traffic latency, and is y Intercept, and is the regression coefficient, and is the weighted score of all router devices, and is an error variable. Claim 2 delete Claim 3 delete Claim 4 delete Claim 5 delete Claim 6 In claim 1, the data security unit receives health data received from the data relay unit from the node A secret key Generate, select the address of the router of the next hop, and then the same secret key is the public key of the next hop np A smart health analysis system further comprising a secret key cryptographic unit encrypted by Claim 7 In claim 6, the secret key encryption unit is a smart health analysis system in which Cipher Block Chaining (CBC) is executed to perform an information processing process as shown in the following Equation 5. [Equation 5] Here, is a Cipher Block, and E is a secret key It is an encryption process based on, is an encryption with a secret key, and is the previous hash, and represents health data. Claim 8 A smart health analysis system according to claim 7, wherein the secret key encryption unit generates the health data into the encryption block, transmits the generated encryption block to the sink node, and the sink node verifies the identity request of the gateway node and then transmits the health data of the encryption block to an external medical server. Claim 9 A smart health analysis method using a smart health analysis system comprising a data collection unit, a data relay unit, and a data security unit, wherein the data collection unit collects health data from one or more sensors and registers the sensors to the node closest to the sensors; the data relay unit randomly selects a router, which is a node to relay the health data received from the sensors, and the cost of each selected router ( A smart health analysis method comprising: a step of calculating ) by a cost function; a step in which the data relay unit calculates a weighted score by accumulating and summing the costs of each router calculated above, and predicts traffic according to the calculated weighted score to select the address of the next hop router; and a step in which the data security unit encrypts health data to be transmitted to the address of the selected router using a blockchain, wherein the data relay unit retains the calculated costs from each router and recalculates the costs according to the following mathematical formula 1; and a step in which the data relay unit determines that the router corresponding to the recalculated cost is untrustworthy if the recalculated cost according to the following mathematical formula 2 is greater than a preset threshold T. [Mathematical Formula 1] Here, is the router cost value, pks is the number of transmitted packets, and trans is the link at active time t. This is the number of retransmissions through.[Equation 2] Here, is a random router, T is a preset threshold, is the cost value of the router. The above data relay unit further includes the step of calculating a weighted score by accumulating and summing the costs of each router calculated above using the following mathematical formula 3, and [Mathematical Formula 3] Here, is the weighted score of each router, and n is the total number of routers in a specific communication link, is the cost value of a router, and i is the sequence number of a router in a specific communication link. The data relay unit calculates the traffic delay rate by performing statistical regression analysis of each router using the following mathematical formula 4 and the calculated weighted score; the smart health analysis method further comprises the step of selecting the transmission router having the lowest value of the calculated traffic delay rate as the address of the next hop router. [Mathematical Formula 4] Here, is the traffic latency, and is y Intercept, and is the regression coefficient, and is the weighted score of all router devices, and is an error variable. Claim 10 delete Claim 11 delete Claim 12 delete

Images