Search

Modeling individual’s variation in speech pattern can be challenging in Automatic Speech Recognition (ASR). In Classical Arabic (CA) language, 20 Quranic accents are permitted for Quranic recitation. An ASR system for CA with accent detection requires a modeling method that can capture speech pattern changes. Here, we study the accentual influences on Malay speakers’ pronunciation and its prosodic impacts towards ASR system for CA language with seven Quranic accents identification. The proposed ASR system was developed over three stages. First, a dataset of Surah Al-Fatihah recitation was recorded from 14 Malay speakers in seven Quranic accents, forming a total of 5,684 words. Second, various spectral and prosodic features are extracted from the dataset for further classification process. The final stage includes training and testing the classification model. The existing ASR systems are often enabled by Gaussian Mixture Models (GMM) because of its capability to represent a wide range of sample distributions. However, GMM is susceptible to overfitting when the model complexity is high, due to the presence of singularities. To support identification of seven Quranic accents, Universal Background Model (UBM) is adapted to GMM using Maximum A Posteriori (MAP) estimation method. The UBM models were trained over each of Quranic accents, and combined to establish final UBM with 512 mixture components. The proposed ASR system utilizing the GMM-UBM outperformed k-NN, GMM, and GMM-iVector in identifying Al-Fatihah recitation to the corresponding Quranic accents. The GMM-UBM yields a testing accuracy of 86.148%, which is an increment of 4.435% from utilizing GMM alone.

Security and privacy are among the most critical challenges on the internet of drones (IoD) network. The communication entities of the IoD network can communicate securely with the use of authenticated key agreement (AKA) based techniques. However, the design of such techniques must balance the tradeoff between security and lightweight features. Recently, Chen et al. proposed an authentication and key sharing scheme for IoD deployment. It is, however, realized after scrutiny that the proposed technique is vulnerable to security attacks under the well-accepted Canetti-Krawczyk (CK) adversary model. Moreover, the scheme applies to the IoD network with only one drones' flying zone. To solve these challenges, this paper proposed a secure lightweight proven authenticated key agreement (SLPAKA) technique for IoD deployment. The technique is free from all the problems identified in the scheme of Chen et al. To ensure the reliability of the SLPAKA, the security of the technique has been assessed from a theoretical method and formal way using the ProVerif cryptographic protocol verification tool. Apart from comparing the performance of SLPAKA with the benchmarking schemes in terms of security, computational cost, and communication cost, the SLPAKA and the technique proposed by Chen et al. are implemented using a python programming language to evaluate and compare their performance in terms of energy consumption and computational time metrics. The results show that the SLPAKA outperforms the technique of Chen et al. and all the other benchmarking techniques in terms of security and lightweight features.

Prior 6LoWPAN intrusion detection system (IDS) utilized several features to detect various malicious activities. However, these IDS methods only detect specific attack but fails when the attacks are combined. In this paper, we propose an IDS known as compression header analyzer intrusion detection system (CHA-IDS) that analyzes 6LoWPAN compression header data to mitigate the individual and combination routing attacks. CHA-IDS is a multi-agent system framework that capture and manage raw data for data collection, analysis, and system actions. The proposed CHA-IDS utilize best first and greedy stepwise with correlation-based feature selection to determine only significant features needed for the intrusion detection. These features are then tested using six machine learning algorithms to find the best classification method that able to distinguish between an attack and non-attack and then from the best classification method, we devise a rule to be implemented in Tmote Sky. To ensure the reliability of our proposed method, we evaluate the CHA-IDS with three types of combination attacks known as hello flood, sinkhole, and wormhole. We also compare our results in term of accuracy of detection, energy overhead, and memory consumption with the prior 6LoWPAN-IDS implementation such as SVELTE and Pongle’s IDS. The results show that CHA-IDS performs better than the aforementioned methods with 99% true positive rate and consumed low energy overhead and memory that fit in constrained device such Tmote Sky.