Search

The unprecedented ballooning of network traffic flow, specifically, Internet of Things (IoT) network traffic, has big stressed of congestion on todays Internet. Non-recurring network traffic flow may be caused by temporary disruptions, such as packet drop, poor quality of services, delay, etc. Hence, the network traffic flow estimation is important in IoT networks to predict congestion. As the data in IoT networks is collected from a large number of diversified devices which have unlike format of data and also manifest complex correlations, so the generated data is heterogeneous and nonlinear in nature. Conventional machine learning approaches unable to deal with nonlinear datasets and suffer from misclassification of real network traffic due to overfitting. Therefore, it also becomes really hard for conventional machine learning tools like shallow neural networks to predict the congestion accurately. Accuracy of congestion prediction algorithms play an important role to control the congestion by regulating the send rate of the source. Various deeplearning methods (LSTM, CNN, GRU, etc.) are considered in designing network traffic flow predictors, which have shown promising results. In this work, we propose a novel congestion predictor for IoT, that uses Temporal Convolutional Network (TCN). Furthermore, we use Taguchi method to optimize the TCN model that reduces the number of runs of the experiments. We compare TCN with other four deep learning-based models concerning Mean Absolute Error (MAE) and Mean Relative Error (MRE). The experimental results show that TCN based deep learning framework achieves improved performance with 95.52% accuracy in predicting network congestion. Further, we design the Home IoT network testbed to capture the real network traffic flows as no standard dataset is available.

Online event detection (OED) has seen a rise in the research community as it can provide quick identification of possible events happening at times in the world. Through these systems, potential events can be indicated well before they are reported by the news media, by grouping similar documents shared over social media by users. Most OED systems use textual similarities for this purpose. Similar documents, that may indicate a potential event, are further strengthened by the replies made by other users, thereby improving the potentiality of the group. However, these documents are at times unusable as independent documents, as they may replace previously appeared noun phrases with pronouns, leading OED systems to fail while grouping these replies to their suitable clusters. In this paper, a pronoun resolution system that tries to replace pronouns with relevant nouns over social media data is proposed. Results show significant improvement in performance using the proposed system.

Constrained Application Protocol (CoAP) is a communication protocol at application layer in the Internet of things (IoT). It employs NSTART parameter to maintain outstanding parallel connections toward individual destinations for data transfer. In the existing implementation of CoAP, the value for NSTART is defined statically. Where the higher values of this parameter up to a threshold give better throughput in moderate traffic, the lower values for the same control the congestion in overload situations. In overload situations, higher values of NSTART result in increased overhead and thereby perform poorly in terms of throughput. The static nature of the parameter and its relevant performance issues demand to manage NSTART dynamically in CoAP. In this work, we propose an algorithm to adjust the value of NSTART dynamically based on the congestion in the network. The proposed approach is an attempt to make the client free from managing NSTART dynamically. For the proof of concept, we implement and run simulations on Cooja mote at IoT/FIT-Lab testbed and Californium CoAP client. The results show 15% improvement in throughput, 15% reduction in average exchange duration, and 48% reduction in average number of retries (average taken of three scenarios) by proposed approach in comparison with the best case of CoCoA (NSTART = 4).

A common practice in the Internet of Things (IoT) is to integrate the Constrained Application Protocol (CoAP), a lightweight application layer protocol based on UDP (User Datagram Protocol), in its architecture. While UDP does not have any provision to control congestion at the transport layer, CoAP and its other variants, such as CoCoA and CoCoA+, get around this issue by employing their own congestion control mechanism, a send-rate-based congestion control technique, at the application layer. These techniques estimate the send-rate through Round Trip Time (RTT), however, fail to map acknowledgments and re-transmitted messages, leading to inaccurate RTT in case of bursty traffic and thus performs poorly. To tackle this problem, the article proposes novel a technique to regulate the send-rate considering the current and previous congestion ratio, congestion factor, and throughput. The proposed method aims to regulate the send-rate according to congestion level in the network. The performance measurements of the proposed approach show significant increments in goodput and fairness as compared to the techniques used in CoCoA+ and CoAP in burst traffic scenarios.

Hand gesture recognition is a vital aspect of robotic vision models. This paper presents a fusion based approach for hand gesture recognition. In this approach, we first extract the Gaussian scale space of an image and compute features on different scales. Kirsch’s convolution mask is then applied on the feature map. The aim of the proposed approach is to remove unwanted information extract scale, rotation, and illumination invariant patterns from hand gestures. The final feature vector is aggregated through the concatenation of multiscale histograms. The Support Vector Machine classifier is demonstrated using extracted features. Moreover, we calculate the progress efficiency of proposed methods on three distinct databases by conducting experiments viz, Thomson, Bochum, and HGRI. The proposed method achieves classification accuracies of 94.25%, 92.77%, and 95.78% respectively on the investigated databases that outperform the existing approaches for hand gesture recognition

In this article, the authors have discussed about the connection between Internet of Things and growth of big data. They have also discussed short reference on the evolution, features, lifecycle, and implementation of Big Data from IoT over the cloud. Internet of Things represents a platform or environment that consists of enormous number of sensors and mediators interconnecting heterogeneous physical devices over the internet. IoT applications are available in many real-world areas such as smart city, smart workplace, smart home, smart transportation and various other ubiquitous computing areas. Using IoT applications generates tremendous amount of data for storage and management in the internet. With the time and research evolution integration of the IoT platforms and cloud comes in the market and IoT platforms data storage and management started shifting to the cloud from the internet connected physical systems for many real-world application areas. Meanwhile when this data becomes huge termed as Big Data. Handling of Big Data over the cloud develops many new areas of research and attention.

Trust and Reputation Systems (TRS) have been witnessing the most promising solutions to ensure the security of a system in the presence of insider adversaries. The methodology behind TRS is to analyze the behavior of entities in the system to compute trustworthiness on them by the mean of direct (using own experience) and indirect (using recommendations from the others) trust computation. For the Internet of Things (IoT), where devices collaborate with each other to offer multiple services, TRS solve the issues of access control, decision making, reliable service delivery etc. The focus of this paper is towards robust indirect trust computation in a service-oriented, dynamic IoT environment. We propose a robust recommendation aggregation scheme which alleviates the effect of false or dishonest recommendations in indirect trust computation by performing the objective and subjective evaluation. In the objective evaluation, the value of a received recommendation is weighted by computing deviation from the average value of all the recommendation received. Subjective evaluation is performed to weight the recommender based on their age of interaction. The scheme deploys a retraining module which retrains the credibility of a recommender based on the dissimilarity experienced with the outcome of recommendation. The effectiveness of the proposed scheme has been demonstrated by the results of the simulation.

Proliferation of ubiquitous communication gave rise to new paradigm, Internet of Things (IoT), in which various heterogeneous physical objects can communicate with each other seamlessly. One major concern in deployment of IoT is to get reliable services and maintain trustworthy environment of device. Wireless channel, resource constraint environment, limited bandwidth, internet connectivity, and others make IoT vulnerable toward various kinds of security attacks. In an uncertain IoT environment, devices can behave maliciously to interrupt the services. Managing trust between the devices is an important aspect of security. A trusted environment, where devices can share, process, and access information without bothering about security and privacy, is needed to effectively utilize IoT services. This chapter is intended to introduce the readers about importance of trust management, its major components, computation models; attacks against trust management and various research challenges exist in deploying trust management system in IoT.

Advancement in digitization gave rise to a new paradigm of communication, called Internet of Things (IoT), where various heterogeneous physical objects (IoT devices) with embedded intelligence can provide or receive services from each other seamlessly. One concerned issue is to maintain reliability in the system with these heterogeneous devices for effective utilization of services. Trust and Reputation management schemes solve this problem by analyzing the behavior of entities experienced in the previous instances. However, existing strategies are not well suitable to meet IoT requirement where network consist of heterogeneous devices having multi-service characteristics. In this paper, we propose a trust management scheme for IoT environment where erratic behavior may be observed for some of the IoT devices. To evaluate trust on a device, our scheme considers following three factors: Success Rate which measures the availability of a device for the previous service requests, Satisfaction Level that gives the degree of satisfaction on the received responses from the device, and Credibility that assesses device susceptibility towards being reliable in the system. Simulation results validate the effectiveness of our proposed scheme by obviating devices' capricious behavior with the different values of trust.

Link failure due to node mobility is routine in mobile ad hoc networks (MANETs). A standard approach to deal with such failures is to select stable routes instead of the shortest route. However, protocols that select stable route do not attempt to optimize the number of hops and hence do not guarantee improvement in performance. Another approach to handle mobility induced route failure is multipath reactive protocols. These protocols start forwarding the packet on the first available route and switches path when a route failure is detected. Although these multipath protocols save on route rediscovery overhead, we establish that they still incur a substantial overhead called route handover cost, while switching among the alternate paths. In this work, we propose a light weight metric link lifetime, which helps to choose a stable path in multipath protocols. We incorporate link lifetime in standard multipath protocol and show that it reduces routing overhead and consequently improves throughput.

Event Detection has gained a lot of attention as a potential research area in the last decade due to emergence and huge adoption of online social media platforms. Though the problem is examined extensively by many researchers, it is the first and foremost important to understand events in detail and then apply detection techniques to identify them. In this paper, we will first study the properties of events and classify them based on these properties. Then, we propose a generic framework for event detection as a possible guideline to develop such systems.

The proliferation of new technologies and enhanced communication protocols rising expectations of internet users towards ubiquitous connectivity. Internet of Things (IoT) is an evolving technology that facilitates ubiquitous communication between uniquely identifiable heterogeneous physical objects seamlessly through the internet. Although, lots of efforts have been made to adopt IoT globally, its deployment suffers various research challenges. A critical issue that adversely affects the performance of IoT is Security. Inherent characteristics of IoT environment like limited resources, wireless channel, heterogeneous communication protocols, battery operated devices etc. restrict the use of traditional security solutions for IoT environment. Trust Management is an important aspect of security which ensures the reliability of the system by analyzing the behaviour of devices over the time to capture their malicious activities. This paper is intended to introduce about the importance of Trust Management over Security and Privacy in uncertain environments like IoT. In this paper, we present a generic framework to manage Trust considering all the qualitative and quantitative parameters. The paper also highlights the major requirement that needs to be considered while embedding Trust Management with IoT.

Rapid advances in digitalization, Internet and world wide web enable the communication between two computing hosts across any location in the globe. Pervasive or ubiquitous computing enhanced this to a level that any two intelligent physical objects can communicate with each other, anytime and anywhere. Internet of Things (IoT) is a novel paradigm emerging from ubiquitous computing that facilitates communication between many real world objects by collaboration of various technologies. Integrating IoT with Cloud Computing derives manifold advantages to store and process enormous data generated by heterogeneous devices. Though both the technologies together have profound effect in many application areas like smart home, smart agriculture, healthcare etc., their integration involves many challenges. Security is one of the most important issues confronted by IoT-Cloud architecture. In this paper, we propose a Secure, Hybrid, Cloud Enabled architecture for IoT (SHCEI), which uses hybrid cloud (both public and private cloud). This architecture ensures security of intradomain data and will also address issues of scalability and interoperability. Further, we highlight some of the research challenges in implementing the combined architecture of Cloud and IoT.

In mobile ad hoc networks, routes are mostly multi-hop and nodes communicate via packet radios. With increasing number of mobile devices, the network is subjected to frequent topology changes. In such networks, a major challenge is to provide seamless connectivity without incurring excessive routing overhead. Routing protocols for ad hoc networks generally select the best possible path based on the current topology. However, these protocols cannot handle mobility induced route failures. A few ad hoc routing protocols have been proposed which select multiple, disjoint path to the destination. These multipath protocols start forwarding packet on the first available route and in case the route fails, it switches over to the next available path. In this way, they save on the route rediscovery overhead, but there is still a substantial overhead in switching among the alternate paths. In this paper, we propose to use link availability estimation technique to select the most stable route from among the alternate paths. The proposed route selection technique was tested by extending AOMDV, a well-known ad hoc, on-demand, multipath routing protocol. Empirical results indicate that selecting stable route results in higher throughput in dynamic network topology scenarios.

Since the inception of opportunistic routing in ad hoc networks, a number of OR protocols have been proposed. The performance measure of these protocols is predominantly based on simulation. Although simulation provides a simple and economical mechanism to test complex routing algorithms, they have been reported to produce inconsistent results. Simulation-based evaluation of ad hoc routing protocols should, therefore, be complemented with mathematical modeling and verification. OR protocols try to restrict the number of transmissions, therefore, the most important performance evaluation metric of OR protocols is packet transmission. In this chapter we examine the various mathematical evaluation models available in literature. During our study we found that these models do not consider retransmitted packets. We, therefore, propose an analytical model which account for packet retransmissions. The mathematical models are validated using prominent OR protocols. The performance of OR protocols depend not only on design issues like candidate selection and forwarder prioritization but also on network parameters like node topology, node density, etc. While the user has no control on the network parameters the performance of the protocol can be improved by carefully tuning the design issues. As a proof of concept we propose an OR protocol that minimizes packet retransmissions.

Opportunistic Routing (OR) is a class of routing protocol that exploits the broadcast nature of wireless network to improve routing efficiency. The primary step in OR is to select a group of nodes as forwarders and prioritize them. The current approach of selecting the forwarders is through simulation which does not always give the best results, select forwarders on diverse paths and can be compute intensive for large networks. In this paper we formulate the problem using OR algebra. We propose a forwarder selection scheme primary-brook forwarder selection (PBFS) that computes the forwarder list such that a particular network objective is maximized. The main criteria while selecting the forwarders is to ensure that adjacent forwarders can listen to each other. Analytical as well as empirical results show that our scheme requires less number of retransmission as compared to some of the popular OR protocols.

Although opportunistic routing OR for ad hoc networks have been shown to improve network throughput, energy audit of these protocols have not been done. In this paper, an analytical model to characterise the energy consumption of OR protocols is presented. Total energy consumption is computed taking into account the energy consumed in exchanging control packets, data packet transmission including retransmission and reception. The model considers packet retransmissions that can occur due to network conditions and protocol inaccuracies. The proposed analytical model is used to compute the energy consumption of some well known OR protocols available in literature. The result of the mathematical model is compared with simulation results. The theoretical and experimental results are found to be in conformance.