Reducing Power Consumption In Embedding Virtual Infrastructure

Communication technology is a substance to the digitalization of society,and a significant contributor to Gross Domestic Product(GDP)growth across the world.The birth of the concept of(IOT)turns the future in a new direction.The Internet of Things(IOT)is the extension of Internet connectivity into physical devices and everyday objects in the design of embedded systems.Moreover,that system has to produce the desired outputs for the given inputs and there are number of other requirements which must be satisfied.New technological breakthroughs and massive production provide cheap and easy-to use products that are more accessible to the average person,which leads to worldwide usage of emerging technologies.One of the main facilitator of technological progress(modern civilization more generally)is the energy that run these systems.However,due to its global usage,technological tools create an increasing demand for more energy.Therefore,it is a big trial to achieve the accurate requirements by reducing the power consumption.Internet Service Providers(ISPs)are facing the challenge of minimizing the power/energy consumption.Recent work in the research community and the industry has been devoted to reduce the energy cost at both of the server and network levels.Network virtualization is considered to be not only an organizer to overcome the inflexibility of the current internet infrastructure but also an enabler to achieve an energy-efficient future of the Internet.Virtual Network Embedding(VNE)is a critical issue in network virtualization technology.Because,it can save energy to turn powering / switching / off unused devices in the internet environment which are committed to research the power aware VNE problem under different power consumption models embedded in VN.Aiming at low energy consumption,we have discussed the joint power-aware node and link resource allocation method to solve the VNE problems in this dissertation.Firstly,we proposed a generalized power consumption model(GPC)of embedding a VN.In addition,the GPC model can get any combination of fixed and variable power consumption which is linked with physical nodes and network devices.Formerly,the problem is transformed into a mixed integer program and two embedding algorithms are proposed for better coordination between power-aware node and link mapping Stage.The simulation results show that the power performance of the algorithm is better than the existing algorithm in the case of an overprovisioned scenario.However,we actually matched the performance of six algorithms,namely ENM-SNP,ENM-SPN,RW-SNP,RW-SPN,R-Vi NE-SNP,and R-Vi NE-SPN.Whereas “RW” denotes using the energy-aware node mapping algorithm proposed to map virtual nodes.Furthermore,we made an effort to investigate the power-aware VNE problem with different power consumption models of embedding a VN.The rest of our work contribution is to formulate the GPC model and GPC-PE-VNE problem with the aspect of reducing energy cost.After that,we also provide the linear programming(LP)formulation as well as the embedding algorithm.Although,we evaluate the performance of our proposed algorithm and compare them with existing algorithm in term of energy consumption.Results of the analysis show that our proposed algorithm has better outcomes than the prevailing algorithms.Ultimately,we suggest a possible way to improve and extend this work in the future.