Mobile Cloud Resource Management

Introduction

Mobile devices such as tablets, phablets, smartphones etc. are increasingly becoming an essential part of our daily life as the most effective and convenient tool regarding communication, work, general knowledge access, maps, diary, health monitor, etc. The explosion in numbers of such devices increased mobile broadband coverage and as a result the need for always-on collaborative services. Microsoft CEO Satya Nadella stated in his keynote speech at Worldwide Partner Conference 2014 in Washington DC on the 16^th of July 2014 that there is a “tremendous opportunity” with mobile and cloud solutions and that Microsoft is focused on defining the “next-generation of productivity” (Nadella, 2014). He also stated that there are going to be more than 3 billion people with connected devices on top of 200 billion sensors all of which will generate tons and tons of zettabytes of data and they are going to consume and reason over the large data. It is true that there has been significant advances in cloud computing and mobile technologies both changing our lives, the way we do business and the way we communicate and this is primarily because of the rapid scalability of cloud services, the ubiquitous network access, on-demand self-service etc. Mobile users experience on a daily basis various mobile cloud services, which either run on their devices and / or remote servers with both data storage and data processing happening outside their mobile device. However, this experience is highly dependent on how cloud-resources are managed.

A cloud-computing infrastructure as described by Pallis (2010), is a complex system that is constantly evolving, consisted of numerous devices (infrastructure, platforms and software) with a great deal of requirements, aiming towards the evolution and convergence of several independent computing trends such as internet delivery, distributed computing, storage, content outsourcing, security etc. Cloud computing is based on a large number of shared resources that are subject to spatial and temporal based varying resource requests. The nature of resource requests varies thus making resource management a complex task. Resource management can play a decisive role in optimizing the performance of such a complex system by efficiently and fairly managing performance, functionality and cost. Inefficient resource management can result in negatively affecting the performance and cost while it may also impair system functionality. Due to its complexity and inhomogeneity in the type of services it offers, cloud resource management requires complex policies and decisions to fulfil a multi-objective optimization. This is extremely challenging because of the vast scale of the system, which makes it impossible to have accurate global state information that are not subjected to unpredictable interactions. According to Marinescu (2013), the main strategies for cloud resource management are associated with the three types of cloud delivery models. These are: Infrastructure as a Service (IaaS), Platform as a Service (PaaS) and Software as a Service (SaaS). In all cases, the cloud services providers are faced with large, fluctuating loads that sometimes are unable to serve.