Internet of Things (IoT) is bringing an increasing number of connected devices that have a di- rect impact on the growth of data and energy-hungry services. These services are relying on Cloud infrastructures for storage and computing capabilities, transforming their architecture into more a distributed one based on edge facilities provided by Internet Service Providers (ISP). Yet, between the IoT device, communication network and Cloud infrastructure, it is unclear which part is the largest in terms of energy consumption. In this paper, we provide end-to-endenergymodels for EdgeCloud-basedIoTplatforms. These models are applied to a concrete scenario: datastreamanalysis produced by cameras embedded on vehicles. The validation combines mea- surements on real test-beds running the targeted application and simulations on well-known sim- ulators for studying the scaling-up with an increasing number of IoT devices. Our results show that, for our scenario, the edgeCloud part embedding the computing resources consumes 3 times more than the IoT part comprising the IoT devices and the wireless access point.
The device layer is utilized, and it monitors the different open infrastructure at the edge. It sends the separated data that demands services used up locally by the fog layer. In the fog layer, the device layer spreads the filtered raw data to the fog layer, where the high- performance distributed SDN controller is located. Every fog node covers a small area and provides dataanalysis and service delivery in a timely manner. The fog layer communicates with the cloud layer to inform it with the result of the processed data, and sometimes informs the device layer about the output reports. The fog layer introduces localization services, whereas the cloud layer provides central monitoring and controlling. Fog layer provides a wide range of events detection, behavioural analysis, and semi-permanent pattern recognition by introducing distributed computing and storage. Cloud layer proposes a distributed cloudbased on the block chain technique that has secure, low-cost, and on-demand access to the foremost competitive computing infrastructures. Clients can use all the computing resources like servers, data storage, and applications that they need. For the fog layer, the use of a block chain-based, distributed computing is to secure SDN controller network architecture in the fog node. All the SDN controllers are communicated in a distributed pattern using the block-chain technique . Every SDN controller is approved by an analysis function of the stream rule and a packet migration function to secure the network from overload attacks. At the edge of the network, the multi-interfaced Base Stations (BSs) are used
Internet-of-Things (IoT) systems are becoming increasingly com- plex, heterogeneous and pervasive, integrating a variety of physical devices and virtual services that are spread across architecture lay- ers (cloud, fog, edge) using different connection types. As such, research and design of such systems have proven to be challeng- ing. Despite the influx in IoT research and the significant benefits of simulation-based approaches in supporting research, there is a general lack of appropriate modelling and simulation platforms to create a detailed representation of end-to-endIoT services, i.e. from the underlying IoT nodes to the application layer in the cloud along with the underlying networking infrastructure. To aid re- searchers and practitioners in overcoming these challenges, we propose IoTNetSim, a novel self-contained extendable platform for modelling and simulation of end-to-endIoT services. The platform supports modelling heterogeneous IoT nodes (sensors, actuators, gateways, etc.) with their fine-grained details (mobility, energy profile, etc.), as well as different models of application logic and net- work connectivity. The proposed work is distinct from the current literature, being an all-in-one tool for end-to-endIoT services with a multi-layered architecture that allows modelling IoT systems with different structures. We experimentally validate and evaluate our IoTNetSim implementation using two very large-scale real-world cases from the natural environment and disaster monitoring IoT domains.
Sensors of dierent kinds connect to the IoT network and generate a large number of data streams. We explore the possibility of performing stream processing at the network edge and an architecture to do so. This thesis work is based on a prototype solution developed by Nokia. The system operates close to the data sources and retrieves the databased on requests made by applications through the system. Processing the data close to the place where it is generated can save bandwidth and assist in decision making. This work proposes a processing component operating at the far edge. The applicability of the prototype solution given the proposed processing component was illustrated in three use cases. Those use cases involve analysis performed on values of Key Performance Indicators, data streams generated by air quality sensors called Sensordrones, and recognizing car license plates by an application of deep learning.
E out (t i , t j ) = min(µ(t j − t i ), ηC − C available )
−E self-discharge (t j − t i ) (4)
V. E XPERIMENTATION
The first half of our experiment is to measure the power consumption and performance degradation with different resolutions on Grid’5000, a French platform for experi- menting distributed system . The used servers are Dell PowerEdge R720 from the Taurus cluster at Grid’5000 Lyon site. Each server is composed of two Intel Xeon E5-2630 processors (2.3GHz) each with 6 cores, 32 GB of RAM and 600 GB of disk space. The processors support hyper- threading technology thus the total of 12 physical cores servers can provide 24 virtual CPUs. KVM is the virtual- ization solution along with Linux on x86-based servers. The experiment results are used for building power and perfor- mance models. The network energy consumption model is defined in a similar way in  and based on bit. These models were integrated into the simulator we developed in . In order to extrapolate to large-scale, the second half of our experiments are held using this simulator. A. Setup
The integration of a 5G/LTE PHY-layer provided by either the GEDOMIS testbed or CASTLE testbed with the LENA LTE-EPC emulated protocol stack  running over the EXTREME Testbed will allow full-stack experimentation of multiple 5G use cases that exploit the flexibility of SDN/NFV when applied to mobile networks (e.g., virtual base stations or self-organized networking). This may consider, for example, the virtual base station use case defined by the ETSI NFV group, or experiments on coverage and capacity optimization involving the whole radio protocol stack and its interactions with core network elements and real applications. The availability of full-stack mobile network testbed allowing experimentation from PHY up to applications and services is currently rare. In fact, experiments on the above topics are often limited to either 1) PHY layer platforms with minimal MAC layer support, due to the cost of commercial protocol stacks, and the limitations of open source ones, or 2) IP-level testbeds with limited access to low-level PHY configuration. In contrast, our integrated testbed will allow evaluating full-stack NFV solutions in a real wireless propagation environment with GEDOMIS testbed, or in a real-time emulation environment with CASTLE emulator, with the possibility of combining both emulated and real (e.g., fiber) backhaul/aggregation network links and applications, and of including additional emulated cells in order to achieve a larger experiment scale.
Internet of Things (IoT) is the trending Internet based revolution, which has come to limelight in the recent years. IOT consists of multiple physical objects that can be easily accessed with the help of internet. It provides the flexibility of accessing billions of devices that are connected with each other and can communicate in order to share information. IOT can improve the quality of our daily life by making our daily tasks easy. On the other hand, Cloud Computing is a paradigm that provides on-demand delivery of IT resources over the internet. Both cloud computing and IOT together can be beneficial for us. The IoT has issues such as limited capabilities in terms of processing power, storage, performance, security and reliability. The integration of the IoT with Cloud Computing is the best way to overcome most of these issues. In addition, the Cloud can be benefited from the IoT by breaking the stereotype and dealing with real world objects in a more dynamic and distributed way, and providing new services for billions of devices in different conditions. However, this integration can resolve many issues but there are many issues in the integration of Cloud and IoT. This paper provides an overview of the integration of the Cloud Computing with IoT by highlighting the benefits of integration and implementation challenges that may occur. In this paper we will also discuss the architecture of the new integrated Cloud-basedIoT paradigm and the areas of applications.
We propose a principled approach to designing and deploying end-to- end secure, distributed software by means of thorough, relentless tagging of the security meaning of data, analogous to what is already done for data types. The aim is to guarantee that—above a small trusted code base—data cannot be leaked by buggy or malicious software compo- nents. This is crucial for cloud infrastructures, in which the stored data and hosted services all have diﬀerent owners whose interests are not aligned (and may even be in competition). We have developed data tag- ging schemes and enforcement techniques that can help form the afore- mentioned trusted code base. Our big idea—cloud-hosted services that have end-to-end information ﬂow control—preempts worries about secu- rity and privacy violations retarding the evolution of large-scale cloud computing.
The biggest project in this topic is the website IoT ONE  which is - according to its own statement - “mapping the Global Ecosystem of the Industrial Internet of Things”. Case studies, use cases, suppliers, software and hardware are listed on its website amongst a lot of other sections. IoTplatforms as discussed in this thesis are part of the software section. The website counts 158 IoT applications, 76 middleware softwares and 54 Application Enablement Platforms (AEPs). The benefit of this work is of course the huge source of platforms to be selected for the analysis in this thesis. One drawback of IoT ONE is that the available data is superficial. A software category, deployment type, first launch, current version and a short description are available for most platforms, for some there is an overview of support offering or pricing methods. For a few platforms, there are also case studies linked. However, a lot of the given information on IoT software is not up to date. For example, Google CloudIoT Core is still listed as private beta despite of the official launch for general availability in February 2018  and Macchina.io has not listed any other version since its first release in January 2014 which is not even listed on Macchina’s own release notes and despite having had nine releases since then . Other platforms are listed as AEP despite being just a full-built application with no possibility to create an application on top and vice-versa. The best thing that could have been used from IoT ONE would be its pricing rating which rated software in four steps: “price leader”, ”below market average”, “market average” and “above market average”. However, this rating was deactivated in April 2019. Overall, IoT ONE gives an overview of IoT software without technical details. In comparison, the goal of this thesis is to offer a comparison on a more in-depth and more technical level.
Although the IoT design paradigm does not restrict the type of communication technology to be used for connect- ing the objects, there are many smart city use-cases in which wireless is the only feasible option. In many applications, long range transmissions are required in addition to low energy consumption and the battery-powered nodes are expected to function for at least 10 years without battery recharge or replacement. However, many of the conven- tional long-range wireless technologies consume too much energy (e.g. WiMAX, 2G, 3G and 4G technologies) and are not optimised for smart city scenarios. Low-power wide area network (LP-WAN) technologies provide features that can be explored to achieve the delicate balance between these seemingly conflicting requirements through some trade- offs between performance, complexity and cost, e.g. data rate, data size vs energy consumption. The aim of this paper is therefore to investigate the effects of the traffic characteristics of smart city applications on the battery life of the end devices and explore the trade-offs using the features available in each LP-WAN technology.
The discussed surveys , , ,  note the potential of edge computing in data analytics and point out the impor- tance of edge computing in IoT for handling the rapid increase of the number of connected devices. Our study contributes to employing edge computing for data analytics by combining edge and cloud computing for the delivery of ML applications. Smart cities are one of the commonly discussed use cases and applications of edge computing. Mohammad et al.  ex- amined possibilities of service-oriented middleware for cloud and fog enabled smart city services. They did not discuss specific smart city services but focused on the middleware. Their experiments demonstrated the benefits of edge com- puting in terms of response time. Tang et al.  presented a hierarchical fog computing architecture for the support of connected devices in smart cities. In addition to the hierarchy of fog nodes, the proposed model includes the cloud as the top layer. The evaluation was performed on an event detection task in a smart pipeline monitoring system: the preliminary results demonstrated the feasibility of the proposed architec- ture. Similar to Mohammad et al.  and Tang et al. , our study also employs both edge/fog and cloud, but differs from theirs in that it also includes an extensive evaluation of the presented edge-cloud architecture.
o gateway e a TTP de autorização ao próprio utilizador. Apenas este poderá autorizar a autenticação verificando e comparando dois simples códigos.
O Mediador é um componente novo introduzido nesta arquitetura comparativamente a uma arquitetura para uma plataforma IoT convencional e não segura. Este componente não possui papéis muito complexos, tratando-se, fundamentalmente, de um proxy. Sempre que uma aplicação consumidora pretenda aceder aos dados armazenados pela plataforma, ao invés de ser feito um acesso direto ao componente de armazenamento com o intuito de recuperar dados, os pedidos passam primeiro por este componente chamado de Mediador. Ao receber um pedido de consulta de dados, este componente recorre à TTP de autorização de forma a obter a autorização expressa do utilizador através da receção de uma asserção SAML (Secção 5.2). Caso a aplicação em causa não possua a autorização necessária para aceder a estes dados devido ao pedido de autorização efetuado pelo Mediador à TTP, o acesso aos dados termina aqui. Para além da consulta efetuada à TTP para efeitos de autorização. Para que a apresentação dos dados seja feita sem que a privacidade dos donos dos dados seja comprometida, é implementado nesta arquitetura um mecanismo de generalização, transformação e anonimato de dados conseguido através da execução de código certificado. A execução é feita pelas TTPs de processamento. No entanto, o código compilado e meta-dados são armazenados no Mediador mediante um registo de código compilado prévio entre a aplicação consumidora e o Mediador (pertencente à plataforma). Após o Mediador armazenar o código compilado e os respetivos meta-dados, este retorna um identificador para o consumidor. Os meta-dados referidos incluem informação sobre chaves públicas, assinaturas e ouras informações úteis relacionadas com o código certificado a ser executado e aprovados pela TTP de certificação de código. Estes são incluídos num manifesto presente num ficheiro.
Ways to mitigate IoT security challenges pertaining to device cloning and sensitive data exposure was addressed in (Naik & Maral 2017). Security issues like data tampering, denial of service, unauthorized device access and control were also discussed. Oliot-Discovery Service (Kwon et al, 2016) deals with performance as well as security issues by focusing on intra discovery service aspect. The authors discussed that data storage is a critical problem to solve at scale and they utilized the proposed service to securely identify data storage. Researchers have also highlighted the urgent need to constrain system component to a reasonably secure and private behavior (Han et al, 2015). The authors recommended following security best practices in home automation. Moreover, a way of providing privacy in a multi-trust-domain environment was presented in (Banerjee et al, 2014). The authors introduced a methodology for privacy-aware communication by utilizing a novel zero-exposure slot allocation scheme and anonymous communication between devices. (Caiming et al, 2013) introduces the novel idea of mitigating IoT exploits by using an analogue to the human immune system. The author’s rightfully focus on the increasing way IoT devices are exploited and consider treating the IoT environment as a pathogen ridden system. Their idea uses sound defense in depth practices and combines security mechanisms into an adaptive approach. It focuses on shifting mitigating features as attacks change over time. Another work (De Rubertis et al, 2013) focuses on two end-to-end encryption protocols and proposes a comparison between the two, as well as a means to assist network designers in choosing the appropriate protocol, for a given set of hardware. This paper identifies a singular point within the IOT security landscape and proposes the comparison of two well-known solutions.
Increasingly complex IoT solutions require more advanced communication platforms and middleware that facilitate seamless integration of devices, networks and applications. There is a wide range of software platforms developed for the purpose of supporting and enabling IoT solutions. The intention is to enable rapid development and lower costs by offering standardised components that can be shared across multiple solutions in many industry verticals. Third party IoTplatforms are relatively new in the market and display a great diversity in terms of functionality and application areas. Broadly speaking, most IoTplatforms fall into one of the following three categories: connectivity management platforms, device management platforms and application enablement platforms. Berg Insight estimates that total revenues for third party IoTplatforms will grow at a compound annual growth rate (CAGR) of 32.2 percent from € 450 million in 2014 to € 2.4 billion in 2020.
ABSTRACT Recently, big data analytics has received important attention in a variety of application domains including business, finance, space science, healthcare, telecommunication and Internet of Things (IoT). Among these areas, IoT is considered as an important platform in bringing people, processes, data and things/objects together in order to enhance the quality of our everyday lives. However, the key challenges are how to effectively extract useful features from the massive amount of heterogeneous data generated by resource-constrained IoT devices in order to provide real-time information and feedback to the end- users, and how to utilize this data-aware intelligence in enhancing the performance of wireless IoT networks. Although there are parallel advances in cloud computing and edge computing for addressing some issues in data analytics, they have their own benefits and limitations. The convergence of these two computing paradigms, i.e., massive virtually shared pool of computing and storage resources from the cloud and real- time data processing by edge computing, could effectively enable live data analytics in wireless IoT networks. In this regard, we propose a novel framework for coordinated processing between edge and cloud comput- ing/processing by integrating advantages from both the platforms. The proposed framework can exploit the network-wide knowledge and historical information available at the cloud center to guide edge computing units towards satisfying various performance requirements of heterogeneous wireless IoT networks. Starting with the main features, key enablers and the challenges of big data analytics, we provide various synergies and distinctions between cloud and edge processing. More importantly, we identify and describe the potential key enablers for the proposed edge-cloud collaborative framework, the associated key challenges and some interesting future research directions.
Control section: The energy meter, loads, digital display and power supply constitute the control section. The theft circuit and serial communication end are arranged at the home end. However the main home section is assigned the task of counting the energy meter reading. The digital energy meter shows the blinking and records the reading. Usually the representative from the KSEB visits houses regularly and notes the reading. The current reading and previous reading are compared and the difference is noted as the usage for that specific period. Cost of energy or the electricity bill is calculated based on the tariff.
The work in this paper proposes an architecture based on Fog and Cloud Computing to solve the performance degradation in the Cloud response time when large volume of Things traffic towards that Cloud. Introducing Fog layer nearby the Things layer contributes to network performance enhancement in terms of response time and packet loss. Two IoT architectures are proposed with LB (formed by two types of load balancer, namely: HAProxy and SLB router) and proposed DM technique using MQTT protocol with QoS level 0 and 1 in the Fog layer. The LB distribute the received messages from Things across a specified number of Fog servers according to LeastConn technique and results in reduction of packet loss, delay, RTT, and response time to half that without LB in the carried tests and achieves the highest throughput. Up to the author's knowledge, both LBs have not been evaluated on Fog layer by researchers previously. Finally, the results obtained have been conducted through practical implementation and can be considered as a possible solution to IoTbased Fog/Cloud performance enhancement.
AT mega 328 has 1KB Electrically Erasable Programmable Read Only Memory (EEPROM). This property shows if the electric supply supplied to the micro-controller is removed, even then it can store the data and can provide results after providing it with the electric supply. Moreover, ATmega-328 has 2KB Static Random Access Memory. AT mega 328 have several different features which make it the most popular device in today‟s market. These features consist of advanced RISC architecture, good performance, low power consumption, real timer counter having separate oscillator, 6 PWM pins, programmable Serial USART, programming lock for software security, throughput up to 20 MIPS etc.ATmega328 is an 8- bit and 28 Pins AVR Microcontroller, manufactured by Microchip, follows RISC
remaining resource type, in a descending order. For example, if a pCPE node has 90% of vCPU left but only 20% of memory left, then the remaining memory will be used for sorting. 3.2. Cost Estimation. With the list of eligible candidate places for a VNF instance 𝑓, we can further estimate the cost of 𝑓 deployed at each place. Algorithms 2 and 3 provide imple- mentation of the cost model from Section 2. Algorithm 3 defines the function to choose the place for VNF instance 𝑓 at the lowest cost, namely, ChoosePlace(𝑓). The function first calls GetCandidates(𝑓) in Algorithm 1 to get the places eligible for deploying 𝑓. Then, for each eligible place, the cost is checked based on the type of the place based on Algorithm 2. If the place is the cloud, CloudCost(𝑓) is invoked for cost; if the place is a pCPE node, the function BnbCost(𝑓, V) is called instead. After iterating all eligible places, the place with the lowest cost is selected and returned. 3.3. Time Complexity. Algorithm 1 has the time complexity of 𝑂(𝑛 log(𝑛)) because of sorting the pCPE nodes by remaining