Amiza Amir

Internet of Things (IoT): Taxonomy of security attacks

The Internet of Things (IoT) comprises a complex network of smart devices, which frequently exchange data through the Internet. Given the significant growth of IoT as a new technological paradigm, which may involve safety-critical operations and sensitive data to be put online, its security aspect is vital. This paper studies the network security matters in the smart home, health care and transportation domains. It is possible that the interruption might occur in IoT devices during operation causing them to be in the shutdown mode. Taxonomy of security attacks within IoT networks is constructed to assist IoT developers for better awareness of the risk of security flaws so that better protections shall be incorporated.

On the effectiveness of congestion control mechanisms for remote healthcare monitoring system in IoT environment — A review

A progressive advancement in biosensors and wireless technology are the major contributors to the realization of continuous remote health monitoring system (RHMS). Wireless Body Area Network (WBAN) is part of this technology due to the deployment of multiple sensors such as Electrocardiogram (ECG) to collect vital body signals for processing and diagnosis. Among the benefits offered by this technology include remote monitoring of patients health status and early detection of abnormalities in the collected signals. Once detected, several preventive measurements can be taken. However, this system needs to encounter some challenges in the wireless network such as delay, packet loss and throughput due to network congestion when transmitting and receiving a bulk of multiple data. Generally, the presence of these problems in transmitting vital body signals may result in incorrect medical diagnosing which can increase mortality rate and cause severe impact to the overall systems performance. Thus, a suitable design of congestion control mechanism is urgently needed in designing a reliable and efficient remote health monitoring system.

A Communication-Efficient Distributed Algorithm for Large-scale Classification within P2P Networks

This paper proposes a supervised and fully-distributed intelligent classification algorithm that is accurate and scalable for large networks. In addition, the resulting algorithm has the following interesting features: fully-distributed, asynchronous, light-weight, online learning, and fast responses. These characteristics make it scalable for large networks. A major distinction of our method compared to the other approaches is that it forms a single global classifier, instead of building many local classifiers (one at every site). Fine-granularity components of the classifier are distributed across the network by using Distributed Hash Table (DHT) --- which provides efficient linking to these components and ensures the system remains fully-distributed. Our simulation results also show that the proposed method is more communication-efficient than several other distributed algorithms. The results also show that the distributed algorithm is able to produce accurate results that are comparable to the available state-of-the-art machine learning techniques.

Developing Machine Intelligence within P2P Networks Using a Distributed Associative Memory

In this paper, we discuss machine intelligence for conducting routine tasks within the Internet. We demonstrate a technique, called the Distributed Associative Memory Tree (DASMET), to deal with multi-feature recognition in a peer-to-peer (P2P)-based system. Shared content in P2P-based system is predominantly multimedia files. Multi-feature is an appealing way to tackle pattern recognition in this domain. In our scheme, the information held at individual peers is integrated into a common knowledge base in a logical tree like structure and relies on the robustness of a well-designed structured P2P overlay to cope with dynamic networks. Additionally, we also incorporate a consistent and secure backup scheme to ensure its reliability. We compare our scheme to the Backpropagation network and the Radial Basis Function (RBF) network on two standard datasets, for comparative accuracy. We also show that our scheme is scalable as increasing the number of stored patterns does not significantly affect the processing time.

Distributed classification for image spam detection

Spam appears in various forms and the current trend in spamming is moving towards multimedia spam objects. Image spam is a new type of spam attacks which attempts to bypass the spam filters that mostly text-based. Spamming attacks the users in many ways and these are usually countered by having a server to filter the spammers. This paper provides a fully-distributed pattern recognition system within P2P networks using the distributed associative memory tree (DASMET) algorithm to detect spam which is cost-efficient and not prone to a single point of failure, unlike the server-based systems. This algorithm is scalable for large and frequently updated data sets, and specifically designed for data sets that consist of similar occurring patterns.We have evaluated our system against centralised state-of-the-art algorithms (NN, k-NN, naive Bayes, BPNN and RBFN) and distributed P2P-based algorithms (Ivote-DPV, ensemble k-NN, ensemble naive Bayes, and P2P-GN). The experimental results show that our method is highly accurate with a 98 to 99% accuracy rate, and incurs a small number of messages—in the best-case, it requires only two messages per recall test. In summary, our experimental results show that the DAS-MET performs best with a relatively small amount of resources for the spam detection compared to other distributed methods.

Image Classification for Snake Species Using Machine Learning Techniques

This paper investigates the accuracy of five state-of-the-art machine learning techniques — decision tree J48, nearest neighbors, k-nearest neighbors (k-NN), backpropagation neural network, and naive Bayes — for image-based snake species identification problem. Conventionally, snake species identification is conducted manually based on the observation of the characteristics such head shape, body pattern, body color, and eyes shape. Images of 22 species of snakes that can be found in Malaysia were collected into a database, namely the Snakes of Perlis Corpus. Then, an intelligent approach is proposed to automatically identify a snake species based on an image which is useful for content retrieval purpose where a snake species can be predicted whenever a snake image is given as input. Our experiment shows that backpropagation neural network and nearest neighbour are highly accurate with greater than 87 % accuracy on CEDD descriptor in this problem.

Performances of Machine Learning Algorithms for Binary Classification of Network Anomaly Detection System

The rapid growth of technologies might endanger them to various network attacks due to the nature of data which are frequently exchange their data through Internet and large-scale data that need to be handle. Moreover, network anomaly detection using machine learning faced difficulty when dealing the involvement of dataset where the number of labelled network dataset is very few in public and this caused many researchers keep used the most commonly network dataset (KDDCup99) which is not relevant to employ the machine learning (ML) algorithms for a classification. Several issues regarding these available labelled network datasets are discussed in this paper. The aim of this paper to build a network anomaly detection system using machine learning algorithms that are efficient, effective and fast processing. The finding showed that AODE algorithm is performed well in term of accuracy and processing time for binary classification towards UNSW-NB15 dataset.

Distributed pattern recognition within DHT-based networks for imbalanced datasets

This paper studies the accuracy of a fully-distributed pattern recognition algorithm, namely the P2P-GN, with imbalanced datasets problem which is often neglected by most of the available distributed algorithms. A major distinction of the P2P-GN compared to the other approaches is that it forms a single global classifier, instead of building many local classifiers (one at every site). Fine-granularity components of the classifier are distributed across the network by using Distributed Hash Table (DHT) - which provides efficient linking to these components and ensures the system remains fully-distributed. Our experimental results also show that the P2P-GN can produce highly accurate results despite imbalanced data distribution compared to other distributed algorithms (Ivote-DPV, ensemble ID3, ensemble k-NN, and ensemble naive Bayes).

A Lightweight Graph-Based Pattern Recognition Scheme in Mobile Ad Hoc Networks

A lightweight, low-computation, distributed intrusion detection scheme termed the distributed hierarchical graph neuron (DHGN) was proposed to be incorporated into a cooperative intrusion detection system (IDS) in mobile ad hoc networks (MANETs). Its onecycle learning and divide-and-distribute recognition task approach allows DHGN to detect similar patterns in short of time. An IDS of such properties is essential in the resource constrained MANETs environment. MANETs are distributed and self-configuring networks, with limited resources and dynamic nodes.

A Wheel Graph Structured Associative Memory for Single-Cycle Pattern Recognition within P2P Networks

A novel and efficient associative-memory-based pattern recognition scheme within P2P networks is proposed and implemented. The proposed scheme, known as the multi-wheel Graph Neuron, is adapted from Graph Neuron-based algorithms which are single-cycle, light-weight, and scalable associative-memory-based pattern recognition algorithms for wireless sensor networks, and has been implemented over a structured P2P Chord overlay network. The proposed approach promotes collaboration among peers during the detection process within the P2P networks. Since the scheme only required single cycle learning, the communication cost amongst peers is minimized. The preliminary results show that the proposed single-cycle recognition scheme guarantees high detection accuracy.