In step with rapid advancements in computer vision, vehicle classification demonstrates a considerable potential to reshape intelligent transportation systems. In the last couple of decades, image processing and pattern recognition‐based vehicle classification systems have been used to improve the effectiveness of automated highway toll collection and traffic monitoring systems. However, these methods are trained on limited handcrafted features extracted from small datasets, which do not cater the real‐time road traffic conditions. Deep learning‐based classification systems have been proposed to incorporate the above‐mentioned issues in traditional methods. However, convolutional neural networks require piles of data including noise, weather, and illumination factors to ensure robustness in real‐time applications. Moreover, there is no generalized dataset available to validate the efficacy of vehicle classification systems. To overcome these issues, we propose a convolutional neural network‐based vehicle classification system to improve robustness of vehicle classification in real‐time applications. We present a vehicle dataset comprising of 10,000 images categorized into six‐common vehicle classes considering adverse illuminous conditions to achieve robustness in real‐time vehicle classification systems. Initially, pretrained AlexNet, GoogleNet, Inception‐v3, VGG, and ResNet are fine‐tuned on self‐constructed vehicle dataset to evaluate their performance in terms of accuracy and convergence. Based on better performance, ResNet architecture is further improved by adding a new classification block in the network. To ensure generalization, we fine‐tuned the network on the public VeRi dataset containing 50,000 images, which have been categorized into six vehicle classes. Finally, a comparison study has been carried out between the proposed and existing vehicle classification methods to evaluate the effectiveness of the proposed vehicle classification system. Consequently, our proposed system achieved 99.68%, 99.65%, and 99.56% accuracy, precision, and F1‐score on our self‐constructed dataset.

Cloud Computing (CC) is a promising technology due to its pervasive features, such as online storage, high scalability, and seamless accessibility, in that it plays an important role in reduction of the capital cost and workforce, which attracts organizations to conduct their businesses and financial activities over the cloud. Even though CC is a great innovation in the aspect of computing with ease of access, it also has some drawbacks. With the increase of cloud usage, security issues are proportional to the increase. To address these, there has been much work done in this domain, whereas research work considering the growing constrained applications provided by the Internet of Things (IoT) and smart city networks are still lacking. In this survey, we provide a comprehensive security analysis of CC-enabled IoT and present state-of-the-art in the research area. Finally, future research work and possible areas of implementation and consideration are given to discuss open issues.

In modern network infrastructure, Distributed Denial of Service (DDoS) attacks are considered as severe network security threats. For conventional network security tools it is extremely difficult to distinguish between the higher traffic volume of a DDoS attack and large number of legitimate users accessing a targeted network service or a resource. Although these attacks have been widely studied, there are few works which collect and analyse truly representative characteristics of DDoS traffic. The current research mostly focuses on DDoS detection and mitigation with predefined DDoS data-sets which are often hard to generalise for various network services and legitimate users’ traffic patterns. In order to deal with considerably large DDoS traffic flow in a Software Defined Networking (SDN), in this work we proposed a fast and an effective entropy-based DDoS detection. We deployed generalised entropy calculation by combining Shannon and Renyi entropy to identify distributed features of DDoS traffic—it also helped SDN controller to effectively deal with heavy malicious traffic. To lower down the network traffic overhead, we collected data-plane traffic with signature-based Snort detection. We then analysed the collected traffic for entropy-based features to improve the detection accuracy of deep learning models: Stacked Auto Encoder (SAE) and Convolutional Neural Network (CNN). This work also investigated the trade-off between SAE and CNN classifiers by using accuracy and false-positive results. Quantitative results demonstrated SAE achieved relatively higher detection accuracy of 94% with only 6% of false-positive alerts, whereas the CNN classifier achieved an average accuracy of 93%.

The health industry is one of the most auspicious domains for the application of Internet of Things (IoT) based technologies. Lots of studies have been carried out in the health industry field to minimize the use of resources and increase the efficiency. The use of IoT combined with other technologies has brought quality advancement in the health sector at minimum expense. One such technology is the use of wireless body area networks (WBANs), which will help patients incredibly in the future and will make them more productive because there will be no need for staying at home or a hospital for a long time. WBANs and IoT have an integrated future as WBANs, like any IoT application, are a collection of heterogeneous sensor-based devices. For the better amalgamation of the IoT and WBANs, several hindrances blocking their integration need to be addressed. One such problem is the efficient routing of data in limited resource sensor nodes (SNs) in WBANs. To solve this and other problems, such as transmission of duplicate sensed data, limited network lifetime, etc., energy harvested and cooperative-enabled efficient routing protocol (EHCRP) for IoT-WBANs is proposed. The proposed protocol considers multiple parameters of WBANs for efficient routing such as residual energy of SNs, number of hops towards the sink, node congestion levels, signal-to-noise ratio (SNR) and available network bandwidth. A path cost estimation function is calculated to select forwarder node using these parameters. Due to the efficient use of the path-cost estimation process, the proposed mechanism achieves efficient and effective multi-hop routing of data and improves the reliability and efficiency of data transmission over the network. After extensive simulations, the achieved results of the proposed protocol are compared with state-of-the-art techniques, i.e., E-HARP, EB-MADM, PCRP and EERP. The results show significant improvement in network lifetime, network throughout, and end-to-end delay.

Due to unavoidable environmental factors, wireless sensor networks are facing numerous tribulations regarding network coverage. These arose due to the uncouth deployment of the sensor nodes in the wireless coverage area that ultimately degrades the performance and confines the coverage range. In order to enhance the network coverage range, an instance (node) redeployment-based Bodacious-instance Coverage Mechanism (BiCM) is proposed. The proposed mechanism creates new instance positions in the coverage area. It operates in two stages; in the first stage, it locates the intended instance position through the Dissimilitude Enhancement Scheme (DES) and moves the instance to a new position, while the second stage is called the depuration, when the moving distance between the initial and intended instance positions is sagaciously reduced. Further, the variations of various parameters of BiCM such as loudness, pulse emission rate, maximum frequency, grid points, and sensing radius have been explored, and the optimized parameters are identified. The performance metric has been meticulously analyzed through simulation results and is compared with the state-of-the-art Fruit Fly Optimization Algorithm (FOA) and, one step above, the tuned BiCM algorithm in terms of mean coverage rate, computation time, and standard deviation. The coverage range curve for various numbers of iterations and sensor nodes is also presented for the tuned Bodacious-instance Coverage Mechanism (tuned BiCM), BiCM, and FOA. The performance metrics generated by the simulation have vouched for the effectiveness of tuned BiCM as it achieved more coverage range than BiCM and FOA.

Mining social network data and developing user profile from unstructured and informal data are a challenging task. The proposed research builds user profile using Twitter data which is later helpful to provide the user with personalized recommendations. Publicly available tweets are fetched and classified and sentiments expressed in tweets are extracted and normalized. This research uses domain-specific seed list to classify tweets. Semantic and syntactic analysis on tweets is performed to minimize information loss during the process of tweets classification. After precise classification and sentiment analysis, the system builds user interest-based profile by analyzing user’s post on Twitter to know about user interests. The proposed system was tested on a dataset of almost 1 million tweets and was able to classify up to 96% tweets accurately.