Green IT Engineering: Concepts, Models, Complex Systems Architectures

Among the challenges faced by modern civilization, are the key challenges associated with the shortage of energy resources, environmental issues and safety. Information technologies are an important factor of energy consumption, safety and green culture. The paper describes concepts and taxonomy of green IT engineering. Main principles of development and implementation, indicators and values of green computing are analyzed. Description of EU project GreenCo is provided.

It is given the basic principles and directions of Green IT Engineering development, as well as the principles of ecological research in the software industry. It is performed the generalized procedure for foresight-project implementation to determine scientific-technical development Green IT Engineering perspective directions using the proposed information technology. It is proposed the approach to construction of foresight technology system model in the form of the two-level hierarchical system consisting of the functional and methodical levels. As a formal basis, methods of bibliometrics and scientometrics (calculating the number of publications, analysis of citing), multi-criterion decision-making problems (t-ordering, Pareto optimality) and patent analysis (trace analysis of the dynamics of inventive activity) are chosen. The result of the foresight-research implementation will be a number of priorities for Green IT Engineering development. In future members of foresight-projects will implement the theoretical development in the form of interactive decision-making support system.

State-of-Art and development of Green IT engineering in the view of resource-based approach which analyzes integration of the artificial world created by human into the natural world (NW) are considered. The challenges solved in integration process include problems of the green technologies. The solution is based on achievement of a certain throughput, trustworthiness and investment of resources: models, methods and means. Throughput is aimed at receiving the maximum return from resources, and trustworthiness—at harmonization with the NW. Resources structured under parallelism and fuzziness of the NW show in development three levels: replication, diversification and autonomy. They are serviced by methods improving throughput, trustworthiness and access to resources, respectively. Green technologies as model occupy autonomy level. The used models and methods belong to the level of diversification. Means are built with replication of operational elements. The methods of multiply effect and its focusing on parameters important for green technology are offered. These methods are based on increasing a level of resource development.

The green computing systems should be not only energy-aware or energy-efficient but also naturally reliable computing systems. Combination of these approaches gives many advantages for applied computing systems. However, estimations for them should be done very carefully. Delay insensitive circuits occupy important role in designing of green hardware, but reliability improvement techniques of delay-insensitive circuits should be further researched. In this chapter, the new indexes of estimating efficiency of computing systems are proposed. These indexes help developers in complex estimation of computing systems operating in wide supply voltage range and temperatures. The semi-modularity for fault-tolerant circuits is analyzed. The synthesis technique of fault-tolerant delay-insensitive circuits is proposed. The technique helps to researchers and developers of Green Hardware in optimization energy consumption, complexity, reliability and performance.

Many of today’s portable computers are not inferior to the computing power of a desktop computer. The main problem of portable computers is short duration of activity of the device in standalone mode, while ensuring the quality of service, i.e., subjective user satisfaction. There is no single way of measuring subjective user satisfaction as the current quality of the service requirements conflicts with the requirements to ensure uptime portable computers. Power consumption model of the portable computer is presented as 3-level power graph. The developed model of the multitasking system’s scheduling in a portable computer battery life is based on the assumption that the problem under consideration belongs to the soft real-time subject. Synthesis of schedules multitasking system in a portable computer battery life is carried out by solving the bi-criteria optimization problem with the release of the Pareto-optimal solutions. These criteria includes minimum penalty for the decline in the quality of service and maximum battery life with the current profile of power consumption. For experimental investigation of the power consumption PCMark-7 and Microsoft Joulemeter were used. As a result of experiments it is found that during operation on portable computers minimum and maximum energy levels differ by more than 2 times. A time transition to a low or high voltage processor, respectively, can lead to a twofold increase in battery life. The latter is, of course, the most optimistic estimate, since an increase in battery life portable computers only 10–20 % will give the user quite noticeable and necessary advantages.

A method of absolute value estimation of the computer energy consumption in performing the programs is proposed in the chapter. The evaluation is made on the basis of the program source code and can help to choose the most optimal solution from the viewpoint of energy saving at the Software development stage. The method is based on the indication of energy consumption by the computer Random Access Memory (RAM) depending upon how intensive the RAM is used by Software. Selection of RAM due to the fact that it is a necessary equipment, and any computing device cannot function without it. Evaluation of energy consumed by RAM is made in using the two proposed mathematical models. The method for computer’s estimating power consumption based on assembler source code was made on the basis of the proposed models. This allows you to create a green software with the control of “green” degree on all stages of its development.

The article presents a new model of the effect of the Internet resources structure impact on the energy efficiency of information search. Existing studies have revealed that the number of Internet queries is growing exponentially. The execution of each one consumes energy. The article also present the state of the art of search engines energy consumption, characteristics of hypertext systems and search engines. Besides, a model of the relationship between the hypertext characteristics and the number of information search steps is developed. For this purpose, the impact of hypertext structure on the steps number on searching for relevant and pertinent information; and the impact of the steps number on energy consumption were studied. As a result, approaches for optimization of hypertext structure in the conditions of uncertainty were formulated. A simulation model allowed testing the adequacy of the developed model of the effect of the structure of distributed hypertext systems on the energy efficiency of information search. To this end, a hypertext model was generated as a random hypergraph. The results can be used to create automated systems for hypertext systems optimization.

Simple Network Management Protocol (SNMP) is an application-layer protocol that is used to monitor IP based devices. Also it can monitor and manage IP based devices without impacting their performance. SNMP is also used in a cloud computing environment to monitor and control virtual machines. However, currently, there is limited published research work on the deployment of SNMP for monitoring and controlling virtual machines in such type of environment. This would allow load balancing in cloud environment during peak hours leading to reduced power consumption helping in Green IT cause. This paper discusses the deployment SNMP for monitoring and controlling Type 1 hypervisor (in a cloud environment). This is followed by the customization of MIB and net-snmp with Agent X to provide more SNMP management features. The completed research work will provide a rigorous physical experimentation involving SNMP monitoring and management for Type I hypervisor Xen.

The rapid advances of communication technologies that aim to increase the information transmission speeds, aggravate problems of reliable data exchanges. Especially the expansion of the use of wireless telecommunications technologies is accompanied by a noticeable increase of the intensity of the electromagnetic field and consequently by an increase in the number of errors caused by external interference. The importance of the classical criteria, such as the number of control bits, is reduced and more attention is paid to other parameters, such as the computational and temporal complexity of the procedures for correcting errors, as well as transmission energy requirements. The above factors dictate the necessity for sufficient developments of the means for ensuring the reliability of communication systems, including methods for data transmission error correction. This chapter proposes a collection of techniques for correcting transmission burst errors in data transmitted over signal channels suffering from strong electromagnetic interference, such as encountered in distributed and embedded systems. Efficiency is achieved by separating the error detection from the correction process and using different codes for each case. The proposed error control techniques are based on simple mathematical operations and are suitable for implementation in FPGA devices. It hence becomes possible to replace energy demanding retransmission operations, including the overheads they entail with energy efficient local error correction calculations. The techniques employed are shown to be more efficient than existing ones, according to criteria that are relevant to current applications. These techniques reduce the need for error recovery by retransmission and hence the environmental effect of data transmission in terms of energy consumption and electromagnetic emissions.

Nowadays, there is a great attention towards cautious usage of energy sources to be employed in disparate application domains, including critical infrastructures, to save both in financial terms and in environmental impact. This chapter focuses on stochastic model-based as a support to the analysis of energy saving systems, in combination with other non functional properties, such as reliability, safety and availability. We discuss general guidelines to build a model-based framework to analyse critical cyber-physical systems, where effective energy consumption is required, while assuring imposed levels of resilience. Also, an overview of the most commonly employed methodologies and tools for model-based analysis is provided, and extensive literature is indicated as pointers to relevant research activities performed on this attractive topic over the last decades. Finally, in order to corroborate the proposed framework, a case study in the railway domain is proposed. By adopting the Stochastic Activity Networks formalism, the framework is instantiated to analyse effective trade-offs between energy consumption and satisfaction of other dependability related requirements.

Digital substations (DS) of smart grid are complex multi-component maintained systems, consisting of a lot of hardware and software components. Failures of the components cause functional and parametrical degradation of the substations. According with systems (DS) structure the reliability-block diagram (RBD), the structure function and the structure function and the Direct Partial Logical Derivatives (DPLDs) for RMSS “electronic transformers—merging unit” are considered. The mathematical tool of logical differential calculus and DPLD in particular are used in many application problems. One of them is reliability engineering. The principal condition of the DPLD application in reliability analysis is the representation of system under investigation by the structure function. We consider the calculation some of these measures as structural, Birnbaum’s, and criticality for the analysis of the electronic transformers—merging unit. The structure function of this unit based on the operation conditions of this system (unit). The construction of this function allows estimating the most important components of this system in stationary state through structural, Birnbaum’s, and criticality importance measures. According with developed technique an example of DS availability model is presented and discussed.

In this book chapter, we have considered a topical problem of intelligent energy management, which arises in the context of an intensively developed science direction—Green IT. The hybrid neuro-neo-fuzzy system and its high-speed learning algorithm are proposed. This system can be used for on-line prediction of essentially non-stationary nonlinear chaotic and stochastic time series, which describe electrical load producing and consuming processes. The considered hybrid adaptive system of computational intelligence has some advantages over the conventional artificial neural networks and neuro-fuzzy systems. The proposed hybrid neuro-neo-fuzzy prediction system provides a high quality load prediction that is very important for power systems.

This paper presents the development of PLC-based systems for data acquisition and supervisory control of environment-friendly energy-saving complex high-tech technologies. The functional structure and main components of PLC-based SCADA-systems for environment-friendly energy-saving technological processes are given. The examples of SCADA applications in design of PLC-based systems for monitoring and automatic control of (a) ecopyrogenesis (EPG) and (b) thermoacoustic technological processes are presented. Paper considers the criteria of energy and economic efficiency of the EPG technological process. The functional structures, software and hardware implementation as well as multi-level human-machine interfaces of the developed PLC-based systems for data acquisition and supervisory control are given. Considerable attention is given to particular qualities of computing of ecopyrogenesis and thermoacoustic processes technological parameters by the proposed SCADA-systems. The developed PLC-based SCADA-systems provide: significant increasing of energy and economic efficiency criteria of the EPG and TAD complexes, high precision control of both technological processes, monitoring of current technological parameters, using the indirect methods for parameters measuring and identifying, and automatic control with high quality indicators and optimal parameters.

The paper describes an approach to measure energy consumption of programmable components (MCU, FPGA etc.) which are a core of Programmable Logic Controller (PLC). Investigation of energy consumption dependence from a type of programmable components as well as from features of software gives a road map for implementation of energy efficiency measures to make PLCs “greener”. Special toolset named GreenCo Controller has been designed, implemented and tested for the above research. A designed architecture includes the following components: a commercially available motherboard (Arduino UNO) programmed with software to perform data processing in respect to energy consumption measurement of the target board-shield (GreenCo board); GreenCo board on the base of Microchip dsPIC30F3011 MCU programmed with different type of application software as a subject of energy consumption investigation; a software monitor for PC desktop using named GreenCo Controller GUI. “Green” features of PLC are analyzed after that formal optimization problems in accordance with criterion “energy consumption level/safety level” have been stated. Such problems can be solved with dynamic programming method. Concept and design of GreenCo Controller are described. Areas of future researches with GreenCo Controller are proposed as a conclusion.

This chapter presents an approach to industrial control systems design, which is based on application of “green” microcontrollers (with low power consumption) used in complex real-time control systems. A wide range of capabilities of modern microcontroller peripheral units allows us to implement a system with wide functionality based on a single microcontroller chip. We describe the order of connection of microcontroller pins, which reduces noise influence on the results of signal processing. As an example, the proposed approach is considered in the context of development of the power supply control system for magnetic elements of linear electron accelerators.