Integrating AI into blockchain consensus, such as Proof-of-Learning and Proof of Useful Work, necessitates AI enablers. However, current consensus protocols cannot ensure AI enabler quality, crucial for AI-powered distributed blockchain and federated learning. Traditional consensus middleware between network and application layers proves inadequate for AI-focused blockchain and federated learning. Thus, an AI-driven application-level consensus with quality-assured enablers is imperative. We propose Proof-of-INtelligence (PIN), an application-level consensus for AI-based blockchain and federated learning, ensuring AI enabler quality. To the best of our knowledge, PIN pioneers the first AI-centric application-level consensus for distributed environments. Employing enablers like accuracy and training quality, PIN is showcased in the federated learning setup “PIN in BlOckchAin-based fedeRateD learning (PIN-BOARD),” the first AI-specific consensus application in blockchain-based federated learning. Both PIN and PIN-BOARD are the highlights of our contributions to the presented work and emphasize the novelty. PIN is the first AI-centric application-level consensus for blockchain and pioneers decentralized AI assurance; PIN addresses the limitations of existing consensus protocols and advances blockchain-based federated learning through the novel framework called PIN-BOARD. Experimental evaluation involves PIN’s accuracy, confirmation time, and a new AI-assurance factor metric. PIN-BOARD’s assessment includes testing accuracy and reward accuracy. A thorough security analysis ensures the strength of PIN and PIN-BOARD. The comparative evaluation highlights PIN’s 20% throughput enhancement and efficient artificial index. PIN-BOARD reduces epochs by 28.5% for peak federated learning accuracy as compared to existing federated models. Thus, PIN emerges as an efficient AI-driven application-level consensus with AI assurance.

The integration of Electric Vehicles (EVs) into the energynet, the network from power generation to EV charging station, presents a symbiotic relationship with potential benefits for sustainable and efficient transportation. However, the existing research has revealed challenges in maintaining an equilibrium between energy supply and demand, often resulting in underutilization or overutilization of energy networks. Blockchain technology has emerged as a promising solution to enhance transparency and secure decentralized energy distribution; however fails to connect the equilibrium in the presence of uncertainty of demand-supply and/or handling information cascading. In this paper, we introduce TAURITE (sTAckelberg eqUilibRium in blockchaIned energyneT with Evs), a novel blockchain-based energynet framework that explicitly leverages the Stackelberg model for energy flow equilibrium within EV interfaces. TAURITE employs Subgame Perfect Nash Equilibrium (SPNE) to address demand uncertainty in dynamic vehicular environments. It also tackles information cascades’ impact on energy distribution, demonstrating its ability to maintain equilibrium even in such scenarios. TAURITE introduces a multi-variate polynomial-based key generation process through the smart contractAVTALand incorporatesProof-of-Energy-Equilibrium (PoEE)as an energy sector consensus mechanism. Experimental results show that TAURITE significantly improves throughput, latency, and energy efficiency, with an average$30\%$enhancement in these metrics. Notably, TAURITE ensures$100\%$allocation stability, even in the presence of information cascades, marking a substantial advancement in sustainable and efficient energy management within the evolving energynet-EV ecosystem.

De-anonymization attacks in blockchains are significant concerns as they compromise the privacy of users on a public ledger. Such attacks, in the form of network analysis and transaction patterns, aim to link a blockchain address to the identity of its owner, potentially revealing sensitive information. Though researchers introduce various solutions using Tor, VPN, and i2P to protect against de-anonymization in blockchains, they have certain limitations: i) non-verification of the private transactions, ii) reveal of the transaction graph, and iii) requirement of a trusted setup that is itself vulnerable to the adversary. All these lead to the revocation of de-anonymization problems. In this paper, we show a novel privacy assurance framework for blockchains. The proposed framework is called De-Anonymization Withstanding Solution (DAWS). DAWS is the first privacy-preserved blockchain framework against de-anonymization attacks. DAWS uses privacy-classifying smart contract execution and a novel consensus called Proof-of-Privacy (PoPri). A set of experiments is executed on PoPri as well as DAWS. The blockchain transactions are modified by including user-defined privacy labels. DAWS can handle attacker advantage ≥0.008 with a privacy breach probability < 0.01% under our threat model. Besides, an improvement in the throughput of DAWS is noticed as compared to Ethereum (almost 80 times) with the Hyperledger configuration for consensus. The gas consumption improvement is 20%. All the listed features enhance the appeal of the proposed DAWS as a robust privacy-preserving solution against blockchain de-anonymization attacks.

Smart contracts are integral to decentralized systems like blockchains and enable the automation of processes through programmable conditions. However, their immutability, once deployed, poses challenges when addressing errors or bugs. Existing solutions, such as proxy contracts, facilitate upgrades while preserving application integrity. Yet, proxy contracts bring issues such as storage constraints and proxy selector clashes - along with complex inheritance management. This paper introduces a novel upgradeable smart contract framework with version control, named ”decentraLized vErsion control and updAte manaGement in upgrAdeable smart coNtracts (LEAGAN).” LEAGAN is the first decentralized updatable smart contract framework that employs data separation with Incremental Hash (IH) and Revision Control System (RCS). It updates multiple contract versions without starting anew for each update, and reduces time complexity, and where RCS optimizes space utilization through differentiated version control. LEAGAN also introduces the first status contract in upgradeable smart contracts, and which reduces overhead while maintaining immutability. In Ethereum Virtual Machine (EVM) experiments, LEAGAN shows 40% better space utilization, 30% improved time complexity, and 25% lower gas consumption compared to state-of-the-art models. It thus stands as a promising solution for enhancing blockchain system efficiency.

The Internet of Things (IoT) is a network of interconnected objects, which congregate and exchange gigantic amounts of data. Usually, pre-deployed embedded sensors sense this massive data. Soon, several applications of IoT are anticipated to exploit emerging 6G technology. Healthcare is one of them, where the 6G-inspired paradigm may facilitate the users to exchange information through hundreds of sensors under the assumption of Artificial Intelligence of Things (AIoT). Integration of medical sensors with AIoT is known as Artificial Intelligence of Medical Things (AIoMT). The secure and seamless interactions among 6G-enabled AIoMT users should be the primary challenge. Furthermore, resource-constrained wearable sensing devices, with their inability to execute complex security solutions, provide an ideal attraction for malicious entities to launch diverse attacks. These challenges have motivated us to design a cost-effective authenticated solution (CAS) for 6G-enabled AIoMT healthcare applications. Our CAS protocol not only prevents cyber threats like impersonation session key secrecy, but it can also prevent physical threats like hardware tampering. We observe formal and informal security validations to endorse its robustness and effectiveness. Performance comparison reveals that CAS protocol offers maximum security enrichment. Moreover, CAS is cost-effective as it has achieved 33% and 60% reduction in computation and communication overheads, respectively, compared to contemporary competing related protocols.

Access control is an important security parameter in industrial networks; a mismanaged access control system leads to security breaches. The existing security solutions significantly consider the access control methods in the Industrial Internet of Things (IIoT); however, falsified identity can bypass the secure access control system. Thus, a centralized access control method leads to risks for data security. We are the first to address the risk factors of granted access in an industrial environment and present a risk-adaptive access control framework for IIoTs. Our proposed solution framework uses blockchain to provide secure decentralized access control in the industrial environment with privacy-preserved multi-party data sharing. We name our framework “Adaptive Access controL for Multi-party data cOmputation in iNdustrial Decentralization (AALMOND)”. AALMOND uses lightweight cryptographic operations to reduce the complexity of the execution and loosen up the tight bounds on resource-constrained industrial devices. Further, the risk-adaptive access control in AALMOND provides a better security analysis of the multi-party sharing data. Our framework uses role-based, attribute-based, and organization-based access controls to map the assets for risk calculation. We put all the required policies in a smart contract for the ease of multi-party data sharing to obtain a transparent access control execution more suitably. We also pioneer in the calculation of the risk adaptivity of AALMOND considering the NIST recommendations of operation risk, security risk, and heuristic risk. We measure the performance of AALMOND with state-of-the-art frameworks based on throughput, latency, complexity analysis, and risk adaptivity factors. We find that AALMOND is efficient for IIoTs as it shows 24% reduced latency and 20% better throughput as compared to the other existing models.

A secure Multi-Party Computation (MPC) is one of the distributed computational methods, where it computes a function over the inputs given by more than one party jointly and keeps those inputs private from the parties involved in the process. Randomization in secret sharing leading to MPC is a requirement for privacy enhancements; however, most of the available MPC models use the trust assumptions of sharing and combining values. Thus, randomization in secret sharing and MPC modules is neglected. As a result, the available MPC models are prone to information leakage problems, where the models can reveal the partial values of the sharing secrets. In this paper, we propose the first model of utilizing a random function generator as an MPC primitive. More specifically, we analyze our previous development of the Symmetric Random Function Generator (SRFG) for information-theoretic security, where the system is considered to have unconditional security if it is secure against adversaries with unlimited computing resources and time. Further, we apply SRFG to eradicate the problem of information leakage in the general MPC model. Through a set of experiments, we show that SRFG is a function generator that can generate the combined functions (combination of logic GATEs) with$n/2$-private to$n$-private norms. As the main goal of MPC is privacy preservation of the inputs, we analyze the applicability of SRFG properties in secret sharing and MPC and observe that SRFG is eligible to be a cryptographic primitive in MPC developments. We also measure the performance of our proposed SRFG-based MPC framework with the other randomness generation-based MPC frameworks and analyze the comparative attributes with the state-of-the-art models. We observe that our posed SRFG-based MPC is$\approx$30% better in terms of throughput and also shows 100% privacy attainment.

Abstract In this pandemic, providing a quality environment is considered one of the essential objectives of smart wellbeing observation. Therefore, the prediction of irregular events has become a fundamental requirement of assistive or clinical consideration. By concentrating on this need, a dew computation-inspired irregular physical event determination solution is presented to determine the symptoms of COVID-19 by analyzing the physical activities of the individuals at their initial stage. The benefit of the proposed solution is enhanced by forwarding the predicted outcomes and their occurrence within a speculated time on a private cloud database to decide the health seriousness. Furthermore, a dynamic decisive-module is introduced to notify medical specialists about the current wellbeing status of the individuals under monitoring. The real-time prediction efficiency of the proposed solution is determined by implementing and calculating the outcomes on both the dew and cloud platforms. The calculated outcomes exhibit the improved viability of the dew platform over the cloud platform by increasing the prediction speed of 46.27% for 40 and 45.54% for 30 frames per second. Moreover, the event prediction performance is justified over the state-of-the-art monitoring arrangements by achieving accuracy (92.88%), specificity (90.87%), sensitivity (88.26%) and F1-measure (89.53%) with the least decision-making delay.

The heterogeneous nature of the Industrial Internet of Thing (IIoT) has a considerable impact on the development of an effective Intrusion Detection System (IDS). The proliferation of linked devices results in multiple inputs from industrial sensors. IDS faces challenges in analyzing the features of the traffic and identifying anonymous behavior. Due to the unavailability of a comprehensive feature mapping method, the present IDS solutions are non-usable to identify zero-day vulnerabilities. In this paper, we introduce the first comprehensive IDS framework that combines an efficient feature-mapping technique and cascading model to solve the above-mentioned problems. We call our proposed solution deeP learnIG model intrusioN detection in indUStrial internet-of things (PIGNUS). PIGNUS integrates Auto Encoders (AE) to select optimal features and Cascade Forward Back Propagation Neural Network (CFBPNN) for classification and attack detection. The cascading model uses interconnected links from the initial layer to the output layer and determines the normal and abnormal behavior patterns and produces a perfect classification. We execute a set of experiments on five popular IIoT datasets: gas pipeline, water storage tank, NSLKDD+, UNSW-NB15, and X-IIoTID. We compare PIGNUS to the state-of-the-art models in terms of accuracy, False Positive Ratio (FPR), precision, and recall. The results show that PIGNUS provides more than 95% accuracy, which is 25% better on average than the existing models. In the other parameters, PIGNUS shows 20% improved FPR, 10% better recall, and 10% better in precision. Overall, PIGNUS proves its efficiency as an IDS solution for IIoTs. Thus, PIGNUS is an efficient solution for IIoTs.

The energy conservative extension of the Internet of Things (IoT), green IoT, is a revolutionary approach in connecting people, processes, and things in an energy-efficient way. The existing research works in the domain of green IoT forbid the use of decentralized management (e.g., blockchain) of data due to its intrinsic disadvantages of block mining, transaction incentives, and less throughput. However, the advantages of blockchains urge the development of new decentralized strategies utilizing the architecture of green IoT. The infancy stage of the conjunction between green IoT and blockchains, and the need of decentralization in energy management in green IoT motivate us for the present research. In this article, we address the problems of decentralization, energy conservation, and privacy simultaneously. We introduce the first blockchain-based privacy-preserving framework for green IoT. We name this framework “Blockchain-based energy-efficient and privacy-preserving data management scheme for GREEN-iot (BENIGREEN)” for smart cities. BENIGREEN uses weight metrics for energy-efficient cluster heads (CHs) selection. The use of weight metrics is a novel contribution in the field of green IoT. Furthermore, we integrate a decentralized blockchain framework with an authentication scheme for secure transmission among base station (BS) and sensor nodes by employing registration, certification, and revocation phases. Consequently, BS allocates the collected information from CHs to decentralized blockchain and cloud storage. The BS eliminates all malicious nodes from the network by employing a certificate revocation process. We execute thorough experiments in terms of the operation time, throughput, average energy consumption, and computational latency. The comparative analysis with the state-of-the-art schemes show that BENIGREEN is efficient for IoT paradigm.

Internet-of-Thing (IoT) faces various security attacks. Different solutions exist to mitigate the intrusion problems. However, the existing solutions lack behind in dealing with heterogeneity of attack sources and features. The future anticipated demand of devices’ connections also urge the need of new solutions addressing the concerns of time consumption and complexity. In this article, we show a novel solution for the intrusion detection in IoT framework. We configure the intrusion detection in the edge computing layer so that the effect of the attack is not propagated to the clouds. Our solution uses cerebellar model articulation controller with kernel map. This combination is very new in the direction of intrusion detection; hence, it emphasizes the novelty of our proposed intrusion detection solution. We name our solution asCerebellar Model Articulation Controller based Intrusion Detection System (CMACIDS). Additionally, we use spline fitting to the kernel mapping for the model fit; this adds on another novel contribution to CMACIDS. The results obtained with our detection system are compared with the state-of-the-art solutions in terms of complexity, false alarms, and precision of detection. The analysis of the comparative study proves the efficiency of the solution and makes CMACIDS suitable for IoT paradigm.

The decentralization and transparency have provided wide acceptance of blockchain technology in various sectors through numerous applications. The claimed security services by blockchain have been proved using various cryptographic techniques, mainly public key infrastructure and digital signatures. However, the use of generic cryptographic primitives using large prime numbers or elliptic curves with logarithms is going to be an issue with quantum computers as those techniques are vulnerable in post-quantum era. Therefore, the paradigm shift from pre-quantum to the post-quantum era has necessitated new cryptographic developments which are robust against quantum attacks and applicable in blockchain for post-quantum decentralization. Therefore, we have presented a solution for post-quantum decentralization in the blockchain. It uses lattices with polynomials for identity-based encryption (IBE) and aggregate signatures for the consensus to ensure efficiency and suitability in post-quantum blockchain applications. We experiment the proposed approach based on delay, throughput, energy consumption and complexity. The comparative results prove that the presented work is efficient.

Internet of Things (IoT) composed of large number of sensing devices with a variety of features applicable for various applications. In such scenarios, due to low data handling capabilities, limited storage, and security aspects, it is quite challenging to protect networks against illegal information access and utilizes storage efficiently. Though researchers provide various solutions for security and data storage, but a few solutions are appropriate for wireless sensor networks (WSNs)-enabled IoTs. Therefore, a blockchain-based decentralized framework integrated with authentication and privacy-preserving schemes is developed for the secure communication in WSNs-enabled IoTs. Registration, certification, and revocation process are employed for the communication with sensor nodes and base station (BS) in a cloud computing environment. In this scheme cluster heads forward the collected information to the BS. Consequently, BS records all the key parameters on the distributed blockchain and large data is forwarded to clouds for the storage. The revoked certificates of all malicious nodes are eliminated from blockchain by BS. The performance of the proposed scheme is scrutinized in terms of detection accuracy, certification delay, computational, and communicational overheads. The simulated results, comparative analysis, and security validation support the superiority of the proposed solution over the existing approaches.

The pandemic/epidemic of COVID-19 has affected people worldwide. A huge number of lives succumbed to death due to the sudden outbreak of this corona virus infection. The specified symptoms of COVID-19 detection are very common like normal flu; asymptomatic version of COVID-19 has become a critical issue. Therefore, as a precautionary measurement, the oxygen level needs to be monitored by every individual if no other critical condition is found. It is not the only parameter for COVID-19 detection but, as per the suggestions by different medical organizations such as the World Health Organization, it is better to use oximeter to monitor the oxygen level in probable patients as a precaution. People are using the oximeters personally; however, not having any clue or guidance regarding the measurements obtained. Therefore, in this article, we have shown a framework of oxygen level monitoring and severity calculation and probabilistic decision of being a COVID-19 patient. This framework is also able to maintain the privacy of patient information and uses probabilistic classification to measure the severity. Results are measured based on latency of blockchain creation and overall response, throughput, detection, and severity accuracy. The analysis finds the solution efficient and significant in the Internet of Things framework for the present health hazard in our world.

The integration between blockchains, Internet-of-Thing (IoT), and smart contracts is an emerging and promising technology. The advantages of this technology have raised the importance of Industrial Internet-of-Thing (IIoT) and have paved the pathway for “Industry 4.0.” Surprisingly, access control has received less attention in IIoTs. Though there are some solutions coming forward to use blockchains for IIoT to enable secure and resilient access control management, the challenge is to satisfy the low-latency requirements of IIoTs for validating and adding the blocks to the chain. Besides, role-based and rule-based access controls in the existing systems can be forged without organizational access controls and compliance. Therefore, we address these problems in this article. In the present work, we proposeDHACS, aDecentralized Hybrid Access Control for Smart contract, for IIoTs. DHACS aims to provide transparency, reliability, and robustness to the existing access control mechanism in IIoTs. The framework is based on blockchain feasibilities that contribute to an interconnected hybrid access control through smart contract provision. It is a novel idea in the domain of IIoTs. We use three access control strategies, role-based, rule-based, and organization-based, to develop a hybrid approach for smart contract in DHACS. The operational transactions along with their access controls are accounted and blocks are made by the transaction pooler and block creator. We use a private blockchain environment; however, it can be extended to a public blockchain or consortium blockchain for geographical distributed dependency. We compare DHACS with three existing approaches in recent time. We measure the performance in terms of computational costs, storage complexity, and energy consumption. DHACS outperforms the others approaches and is considered to be efficient for IIoT applications with more than 30% better efficiency in access control management. To the best of our knowledge, DHACS is the first attempt to use decentralized blockchains with smart contract for hybrid access control in IIoTs.

In this paper, we address the highly challenging problem of object goal navigation. The agent, in an unseen environment, has to perceive its surroundings to identify and navigate towards potential regions where the specified goal category can occur. Rather than developing goal driven exploration policies, we aim to adapt the existing exploration policies that maximize scene coverage to be goal-conditioned. Thus, we propose a standalone scene understanding module to identify potential regions where the goal occurs. We also propose Goal-Conditioned Exploration (GCExp), an algorithm that entails the integration of our novel scene understanding module with any existing exploration policy. We test our solution in photo-realistic simulation environments using state-of-the-art exploration policy, Active Neural Slam [1], and show improved performance over the same on every evaluation metric.

Digital forensic in Internet-of-Thing (IoT) paradigm is critical due to its heterogeneity and lack of transparency of evidence processing. Moreover, cross-border legalization makes a hindrance in such process pertaining to the cloud forensic issues. This urges a forensic framework for IoT which provides distributed computing, decentralization, and transparency of forensic investigation of digital evidences in cross-border perspectives. To this end, we propose a framework for IoT forensics that addresses the above mentioned issues. The proposed solution called Internet-of-Forensics (IoF) considers a blockchain tailored IoT framework for digital forensics. It provides a transparent view of the investigation process that involves all the stakeholders (e.g., heterogeneous devices, and cloud service providers) in a single framework. It uses blockchain-based case chain to deal with the investigation process including chain-of-custody and evidence chain. Consensus is used for consortium to solve the problems of cross-border legalization. This is also beneficial for a transparent and ease of forensic reference. The programmable lattice-based cryptographic primitives produce reduced complexities. It shows benefits for power-aware devices and puts an add-on to the novelty of the presented idea. IoF is generic; hence, it can be used by autonomous security operation centers, cyber-forensic investigators and manually initiated evidences under chain-of-custody for man-made crimes. Security services are assured as required by the framework. IoF is experimented and compared with the other state-of-the-art frameworks. The outcomes and analysis prove the efficiency of IoF concerning complexity, time consumption, memory and CPU utilization, gas consumption, and energy analysis.

Interconnection and intercommunication between people, processes, and things have been enhanced with the development of the Internet of Things (IoT) and extended with fog computing for better efficiency. Fog computing improves the network services and circumvents the problem of escalated data management as an interface between cloud and terminal with the requirement of secure data transmission. In this article, a novel security framework is designed for fog computing intended for improving the security of IoT. This framework uses single point of aggregation from fog interface and lattice cryptographic approach to establish the security for the services. Azure cloud platform and Sage lattice cryptographic base have been employed to evaluate the performance of the framework. It is noticed that the Level 1-Level 2 (L1-L2) cache implementation in the system significantly improves the efficiency of the framework. The results show that the proposed framework is robust in preventing attacks on fog nodes and additionally provides an advantage for low communication overhead.

In this paper, we discussed an efficient statistical method with proof-of-work consensus approach for cloud and fog computing. With this method, solution with precise probability in minimal time is realized. We have used the expectation maximization algorithm and polynomial matrix factorization. The advantages of this statistical method are the less iteration to converge to the consensus solution and easiness to configure the complete mathematical model as per the requirement. Moreover, the energy and memory consumption are also less which make this approach appealing for cloud and fog computing. The experimental results also show that the proposed approach is significantly efficient in terms of time and memory consumption. This novel approach seems beneficial for Internet-of-Things (IoT), one of the most fast-growing technologies in network computing.

Wireless LAN networks are considered to be widely used and efficient infrastructure used in different domains of communication. In this paper, we worked on Network Intrusion Detection System (NIDS) to prevent intruder's activities by using snooping agents and honeypot on the network. The idea behind using snooping agents and honeypot is to provide network management in term of monitoring. Honey pot is placed just after the Firewall and intrusion system have strongly coupled synchronize with snooping agents Monitoring is considered at packet level and pattern level of the traffic. Simulation filtered and monitor traffic for highlight the intrusion in the network. Further attack sequence has been created and have shown the effects of attack sequence on scenario which have both honey pot and snoop agent with different network performance parameters like throughput, network load, queuing delay, retransmission attempt and packet. The simulation scenario shows the impact of attack on the network performance.

Advanced Encryption Standard (AES) is a standard algorithm for block ciphers for providing security services. A number of variations of this algorithm are available in network security domain. In spite of the strong security features, this algorithm has been recently broken down by the cryptanalysis processes. Therefore, it is required to improve the security strength of this algorithm as AES is popular in commercial use. In this paper, we have shown the reasons of the loopholes in AES and also have provided a solution by using our Symmetric Random Function Generator (SRFG). The use of randomness in the key generation process in block cipher is novel in this domain. We have also compared our results with the original AES based upon some parameters such as nonlinearity, resiliency, balancedness, propagation characteristics, and immunity. The results show that our proposed version of AES is better in withstanding attacks.

An advance multiresolution wavelet based approach for wideband spectrum sensing for cognitive radio system is proposed in this paper. Prime focus is made on the coarse detection part for interweaved system, in which unoccupied spectrum can be used efficiently by the cognitive users. Quick and immediate shifting over the sensed vacant channel is extremely vital and is a challenging task. To overcome this issue, fast and efficient spectrum sensing technique is proposed for cognitive radios by improvising the Discrete Wavelet Packet Transform (DWPT) for multiresolution interweaved systems. This proposed scheme not only increases the system speed but also reduces complexity. Simulation results are used to analyse the system performance and numerical analysis for computing system complexity.

Abstract Feature selection methods play a significant role during classification of data having high dimensions of features. The methods select most relevant subset of features that describe data appropriately. Mutual information (MI) based upon information theory is one of the metrics used for measuring relevance of features. This paper analyses various feature selection methods for (1) reduction in number of features; (2) performance of Naïve Bayes classification model trained on reduced set of features. Research gaps identified are: (1) computation of MI from the whole sample space instead of unclassified sample subspace; (2) consideration of relevance of features only or tradeoff between relevance and redundancy, but class conditional interaction of features is ignored. In this paper, we propose a general evaluation function using MI for feature selection. The proposed evaluation function is implemented which use dynamically computed MI values from unclassified instances. Effectiveness of the proposed feature selection method is done empirically by comparing classification results using KDD 1999 benchmarked dataset of intrusion detection. The results indicate practicability and effectiveness of the proposed method for applications concerned with high accuracy and stability of predictions. Copyright © 2011 John Wiley & Sons, Ltd.