A smart privacy preserving framework for industrial IoT using hybrid meta-heuristic algorithm - Scientific Reports

• Read original article here

A smart privacy preserving framework for industrial IoT using hybrid meta-heuristic algorithm
Your article has downloaded
Similar articles being viewed by others
Slider with three articles shown per slide. Use the Previous and Next buttons to navigate the slides or the slide controller buttons at the end to navigate through each slide.
16 November 2022
Mohammed Abutaha, Basil Atawneh, … Georges Kaddoum
10 June 2022
Amir Masoud Rahmani, Saqib Ali, … Mehdi Hosseinzadeh
03 September 2022
Zhu Rongrong
Scientific Reports volume 13, Article number: 5372 (2023) Cite this article
1 Altmetric
Information technology
Abstract
Industrial Internet of Things (IIoT) seeks more attention in attaining enormous opportunities in the field of Industry 4.0. But there exist severe challenges related to data privacy and security when processing the automatic and practical data collection and monitoring over industrial applications in IIoT. Traditional user authentication strategies in IIoT are affected by single factor authentication, which leads to poor adaptability along with the increasing users count and different user categories. For addressing such issue, this paper aims to implement the privacy preservation model in IIoT using the advancements of artificial intelligent techniques. The two major stages of the designed system are the sanitization and restoration of IIoT data. Data sanitization hides the sensitive information in IIoT for preventing it from leakage of information. Moreover, the designed sanitization procedure performs the optimal key generation by a new Grasshopper–Black Hole Optimization (G–BHO) algorithm. A multi-objective function involving the parameters like degree of modification, hiding rate, correlation coefficient between the actual data and restored data, and information preservation rate was derived and utilized for generating optimal key. The simulation result establishes the dominance of the proposed model over other state-of the-art models in terms of various performance metrics. In respect of privacy preservation, the proposed G–BHO algorithm has achieved 1%, 15.2%, 12.6%, and 1% enhanced result than JA, GWO, GOA, and BHO, respectively.
Introduction
Over the past few years, the industrial infrastructures and standards are gradually developed owing to the combination of industrial equipment and IoT in the industrial applications, which is termed to be IIoT 1 , 2 , 3 . Recently, due to the most significant application of the IoT, IIoT has a great opportunity and also plays an important part in the further improvement of the Industry 4.0 4 , 5 , 6 . IIoT is an integrated technology, which includes big data analysis, cloud computing, artificial intelligence, mobile communications and IoT for performing all the industrial production process 7 , 8 . On evaluating the data collection acquired from the industrial equipment and further processing towards the predictive maintenance for optimizing the production processes, IIoT enhances the product qualities and efficiency of manufacturing along with reducing the resource computation and product cost, which simultaneously improves the level of the traditional industry 9 , 10 . As it is being an openly available and scalable information communication medium, IIoT allows the exchanging of diverse data over the industrial devices that are used for industrial operations in both the local and wider areas. When generating an enormous volume of data through the connected IIoT devices, it creates more requirements regarding the accuracy and efficiency at the time of practical data collection, monitoring, and processing. As there are many challenges in the data privacy and security, it seeks more attention of the researchers to develop a new model towards securing the IIoT data 11 , 12 .
Most significant challenges of the IoT security are caused due to the large scale and heterogeneity of the objects 13 , 14 . The challenges are related on ensuring the integrity of the involved records in the naming architecture while identifying the object 14 , 15 , 16 . At the same time, the Domain Name System (DNS) gives the services on translating the name for the internet users, which can be represented to be insecure naming system. It may be sensitive to diverse attacks like DNS cache poisoning attack. These attacks insert the fake DNS records into the cache of the users and directly affect the resolution mapping among the addressing architecture and naming architecture 17 , 18 , 19 . Thus, the entire naming architecture gets insecure owing to the lack of integrity protection of the user records. The security extension belongs to the DNS is used for ensuring the authenticity and integrity of the user’s resource record. It is also used as the tool for distributing the cryptographic public keys, which acts as the solution for the naming service. However, the challenging part is to deploy the service extensions of DNS in IoT 18 , 19 , 20 , 21 , 22 , which suffers from communication overhead and high computation and is not acceptable for IoT devices 23 , 24 . Privacy is known to be more complicated when compared to security due to its requirements in Cloud Service Providers (CSPs) 25 , 26 . The involvement of trusted CSP makes handling and transferring sensitive data simpler. Yet, diverse issues are occurring in the cloud 27 , 28 . Further, the possibility to authorize the user’s data, diverse public CSPs avails their services without any costs. Recently, various models are generated for handling the existing challenges, re-establishing the user’s control, and also for providing data protection towards the cloud 29 , 30 . But all existing models need a masking strategy for the sensitive data, where the masked values are stored in the cloud. These masked data are only accessed by the user, who controls the data that are obtained from the cloud 31 . Still, it is difficult to manage both the cloud storage and computational power for the users because of the data protection, which is highly suitable on the masked data of the cloud platforms. Thus, it is necessary to design a novel privacy preservation model in IIoT using hybrid optimization algorithm. Privacy preserving in IIoT requires amalgamation of of policies and technical measures for ensuring collection, storing and sharing of data can be done in secure manner. The use of multi-objective optimization in privacy preservation can act as an influential approach as it gives significant flexibility, transparency, efficiency and personalization. In the context of privacy preserving the kay generation plays a very important role for felicitating secure communication. The heuristic algorithms are often utilized in the generation of keys. Combining different heuristic algorithms is advantageous over using single heuristic algorithm. Some of the important advantages of using combination of heuristics are increased robustness, reduced bias, enhanced security and improved privacy. Altogether, the combination of heuristic algorithms provides robust and powerful approach for key generation in privacy preserving background. The paper is contributed towards the IIoT data privacy that is mentioned as follows.
To investigate an IIoT-based privacy protection model with the generation of optimal key by utilizing the implemented hybrid heuristic approach to assure the security among the shared information and the privacy across the IIoT data.
To construct the privacy protection mechanism by involving the data restoration followed by sanitization tasks with IIoT data with the aid of implemented G–BHO-based optimal key to secure the data transmission in IIoT network.
To implement the hybridized form of heuristic algorithm termed G–BHO to pick out the optimal key for performing the data restoration followed by sanitization tasks.
To estimate the potency of the developed privacy protection model based on proposed G–BHO by comparing it with existing techniques over various performance metrics.
The further sections in the proposed model are simplified below. The earlier developments in the IIoT data privacy protection model is discussed in Part 2. The implemented privacy protection model considering the IIoT data is described in Part 3. The multi-objective strategy involved in the developed model is depicted in Part 4. The IIoT data-based sanitization and restoration tasks with developed G–BHO algorithm is given in Part 5. The analysis and the observed results are explained in Part 6. The Conclusion and Future scope are summarized in Part 7.
Literature review
Related work
Ref. 32 have designed an exhaustic model for helping the energy researchers and medical practitioners by performing the optimization of energy resource through enhancing the privacy and also better perceptive of industry 4.0 infrastructure based on 5G. The suggested framework was estimated with diverse case studies and also with the mathematical modeling. Ref. 33 have proposed the model in the cloud scenario for privacy preservation based on the artificial intelligence. The suggested sanitization process mainly relies on the performance of generating optimal key that was done through the hybrid optimization algorithm. At last, the efficacy of the proposed model has showcased through the evaluation of the traditional models by improving the cloud security.
Ref. 34 have developed a highly effective technique for performing the privacy preservation through monitoring the correlation among the multivariate streams obtained from the network of IIoT devices. Here, the “data covariance matrix” was utilized for adding the noises, which would not be removed using the filtering to prevent the unauthenticated access of the user data. For enhancing the communication efficiency among the connected IoT devices, the suggested model has established the inherent properties belongs to the correlation matrix and has monitored the significant coefficients of a minimum subset of correlation values. The analysis was performed for validating the robust and effective performance of the developed approach.
Ref. 11 have investigated a privacy preservation model with the help of multi-keyword ranked searching algorithm. The simulation analysis establishes the supremacy of the proposed scheme in terms of verification time, storage, and computation by comparing with the existing searching encryption approaches. Ref. 35 have developed a security mechanism and trust management for preserving the communications in the IoT networks. The artificial intelligence-based approach was implemented for solving the problems on computing and communicating over the 5G-incorporated IoT networks that have been unidentified in the existing models.
Ref. 36 have suggested a novel authentication strategy using the transfer learning utilizing blockchain technology. Here, the blockchain technology has involved for achieving the superior performance in privacy preservation regarding the industrial applications. Also, the transfer learning was used for authentication strategy for constructing the trustworthy blockchains with the enhanced privacy preservation in the industrial applications. The experimental results have been carried out for ensuring the accurate authentications along with the low latency and high throughput. Ref. 37 have implemented an enhanced clustering structure to preserve the data privacy using the optimal clustering protocol into the model. This protocol was used for improving the energy efficient and data privacy routing over the heterogeneous network that has utilized the multi-hop communication and clustering for minimizing the energy consumption among the sensor nodes and also for extending the lifetime of the network. The simulation results have shown that the enhanced performance on data security was observed through the proposed approach when considering the network lifetime and computational time.
Ref. 38 utilized a novel hybrid optimization algorithm to develop a strategy for privacy preservation using the business data in the cloud environment. The hybrid optimization algorithm has achieved the high convergence, and control parameters used in this model have been reduced when solution generation. Finally, various analyses were conducted for estimating the supremacy of the proposed algorithm. The evaluation of the suggested model among the existing models was done for showing the effective performance of the proposed model.
Ref. 39 had designed a secure Fog-based architecture for IIoT. In order to reduce computational overheads few jobs were offloaded to Fog nodes. The authors used existing security schemes and made suitable changes in them to make the architecture robust. However, the disadvantage was distribution of same data to several users required encryption for every user.
Ref. 40 had introduced GMGW for performing process of sanitization. For improving restoration accuracy, a hybrid algorithm GMGW was proposed in this work. This work also possessed some limitations like falling into local minimum particularly in case of complex problems.
Ref. 41 had designed PSV-GWO for finding the optimal key. It contained less parameters and it didn’t fall into local optimum easily. But it had major flaws like poor local search ability and low precision solving.
Ref. 42 designed OI-CSA for finding optimal key. They used modified version of cuckoo search algorithm. The results obtained very encouraging however they didn’t focus on combining it with web mining.
Ref. 43 introduced a software-defined IIoT for making network more flexible. However, it had its disadvantage as use of SDN is still in its infant stage and using it can also result in higher latency in data forwarding. Ref. 44 developed an IIoT by focusing on use of fog as middleware. Nonetheless, security issues involved were never discussed in the proposed architecture.
Ref. 45 had proposed (BS-WOA) for the identification of secret key. In order to preserve privacy, the database was modified using optimal secret key. However, the data pool consisted of large number of users and hence maintaining privacy of every database was a severe challenge. Ref. 6 designed a novel privacy model based on decision tree. The main feature of this model was entire independence from any kind of back ground knowledge. However, this model didn’t provide accurate access to loss in privacy.
Problem statement
Numerous privacy preservation is reviewed in Table 1 . Privacy protection and energy resource optimization framework minimizes the energy consumption in 5G network and provides better runtime and scalability. However, it lacks in collecting some real-world statistics from Industry 4.0 for analyzing the solution. Artificial Intelligence utilizes the cloud data to evaluate the practical challenges and to attain the desired security requirements. But, it fails to manage the optimal privacy while handling the sensitive data. The accuracy of restoration is very poor. Fast adaptive correlation matrix Completion method minimizes the risks related to operations and also in the security and privacy problems. But, it fails to satisfy the practical needs of the IoT services owing to its high network latency. PPP-MKRS scheme is applicable for e-health system. It is highly complex to provide the optimal security and privacy while managing the remote data services. Artificial Intelligence 46 reduce the threat of privacy leakage efficiently. However, due to high mobility, the long-term operations cannot work effectively. Transfer learning ensures the appropriate authentications in the applications of IIoT. It also attains the superior performance with regard to throughput and latency in different IIoT schemes. But, the conditions related to privacy of the practical IoT data are not much improved. Multihop Dynamic Clustering Routing Protocol helps to maximize the lifetime of WSN. It provides a better privacy for solving the problems of data security attacks. However, in some scenarios, the information obtained from the mobile according to the system requirements needs the data identity authenticity. Red deer-bird swarm algorithm ensures sufficient solutions to secure the privacy of the data and also provides higher convergence. However, the performance of the model is needs to be improved. Therefore, a new privacy preservation model for IIoT is required to be developed considering these abovementioned drawbacks.
Table 1 Benefits and issues of privacy preservation using industrial IoT.