Energy Efficiency in Hybrid Mobile and Wireless Networks
Doctoral thesis, Peer reviewed
MetadataVis full innførsel
Wireless Internet access is almost pervasive nowadays, and many types of wireless networks can be used to access the Internet. However, along with this growth, there is an even greater concern about the energy consumption and efficiency of mobile devices as well as of the supporting networks, triggering the appearance of the concept of green communication. While some efforts have been made towards this direction, challenges still exist and need to be tackled from diverse perspectives. Cellular networks, WLANs, and ad hoc networks in the form of wireless mesh networks are the most popular technologies for wireless Internet access. The availability of such a variety of access networks has also paved the way to explore synergistic approaches for Internet access, leading to the concept of hybrid networks and relay communications. In addition, many mobile devices are being equipped with multiple interfaces, enabling them to operate in hybrid networks. In contrast, the improvements in the battery technology itself have not matched the pace of the emerging mobile applications. The situation becomes more sophisticated when a mobile device functions also as a relay node to forward other station’s data. In the literature, energy efficiency of mobile devices has been addressed from various perspectives such as protocol-level efforts, battery management efforts, etc. However, there is little work on energy efficiency in hybrid mobile and wireless networks and devices with heterogeneous connections. For example, when there are multiple networks available to a mobile device, how to achieve optimum long-term energy consumption of such a device is an open question. Furthermore, in today’s cellular networks, micro-, pico-, and femto-cells are the most popular network topologies in order to support high data rate services and high user density. With the growth of such small-cell solutions, the energy consumption of these networks is also becoming an important concern for operators. Towards this direction, various solutions have been proposed, ranging from deployment strategies for base stations to cooperative techniques etc. However, as base stations have the largest share in a network’s energy consumption, methods that allow lightly-loaded base stations sleep or be switched off are possible means as a feasible step towards green communications. In this dissertation, we tackle the above mentioned problems from two perspectives, i.e., mobile station’s and operator’s perspectives. More specifically, by taking into account the amount of transferred data in uplinks and downlinks individually for various components in a hybrid network, strategies are proposed to reduce mobile station’s battery energy consumption. For this purpose, other parameters such as link distance and remaining battery energy can also be considered for handover decision making, in order to maximize energy efficiency of the mobile station. To optimize long-term energy consumption of the mobile stations operated in such scenarios, a Markov decision process-based methodology is proposed as our contribution to this topic. Moreover, from operator’s perspective, a network energy conservation scheme which may switch off a base station is proposed for micro- or pico-cells scenarios. Both deterministic and probabilistic schemes are proposed for network energy conservation. The problems considered and the solutions proposed in this dissertation advance the frontiers of the research work within the theme of energy efficiency for mobile devices as well as hybrid mobile and wireless networks.
Doktorgradsavhandling i informasjons- og kommunikasjonsteknologi, Universitetet i Agder, Grimstad, 2012