Abstract
In recent years, there has been a tremendous increase in the number of handsets, in particular smartphones, supporting a wide range of applications. Furthermore, the amount of mobile data traffic is expected to increase dramatically in the coming years. If the current traffic demand growth rate were maintained, current cellular system capacity would not be able to cope with it. Therefore future cellular systems have to be designed to contain the expected traffic growth. On the other hand, energy efficiency of cellular systems has emanated as one of the important performance indicators coupled with the current international focus on climate-change issues and increasing energy prices. There is always a trade-off between the coverage, quality of service, power consumption and capacity issues when considering the recent forecasts of expected traffic growth. The capability of optimizing the energy consumption of a cellular system without reducing from the user experience (i. e. compromising operational parameters) by means of exploiting the variation in traffic in both time and space domains is investigated. A bottom-up approach is used to model the power consumption of several types of LTE base stations. An energy-efficient algorithm based on a dynamic multi architecture deployment strategy is proposed demonstrating the ability to consume less energy by capitalizing on traffic diversity in the spatial and time domains. There is a need to design an energy-efficient framework for decision-making for future cellular systems. In particular, a novel scheme is proposed, the fuzzy-logic architecture selection (FLAS). We show how using multiple network parameters in architecture decision reduces energy consumption. Network densification is envisioned as a key enabler for 2020 vision that requires cellular systems to grow in capacity by hundreds of times. However, increased energy consumption and complex mobility management associated with network densification remains as the two main challenges to be addressed before further network densification can be exploited on a wide scale. In the wake of these challenges, a novel dense network deployment strategy for increasing the capacity of future cellular systems without sacrificing energy efficiency and compromising mobility performance is proposed and evaluated. The proposed deployment architecture consists of smart small cells, called cloud nodes, which provide data coverage to individual users on a demand basis while taking into account the spatiotemporal dynamics of user mobility and traffic. Furthermore, the comparison of the performance of the proposed architectures with both conventional macro deployment as well as pure micro cell based dense deployment in terms of number of KPIs is conducted and discuss and quantify the trade-off therein.