Game Theoretic Analysis of Dynamic Spectrum Leasing in Large Scale Wireless Networks

In this thesis, the concept of dynamic spectrum leasing has been suggested as a solution to improve the spectral efficiency (SE) of wireless networks while reducing the energy expense. Wireless communication is facing the ever growing challenge of improving information transfer rate. This grand task is inherently coupled with managing the energy costs required to meet the capacity enhancement. To this end, the idea of cooperation among various networks which is inherently supported by the wireless medium, is utilized in this thesis. An intelligent division/sharing of resources between wireless terminals situated in close proximity of each other are the natural mechanisms, which when accurately modelled, can lead to commendable gains in the spectral utility of a wireless network. Cooperative use of existing resources in the network such that they are scheduled/divided in a way to fulfil the demands of each type of network can reap enhanced data transmission without additional energy sources. With this ideological foundation, the sharing of spectral resources between a legacy and a secondary network is suggested such that the quality of service of both types of networks is maintained/improved while improving the energy utilization of the network. Either access to the legacy spectrum is granted to the entire secondary network for a limited duration of time or a fixed term spectral access license is issued to a limited number of secondary users as a reward for their help in cooperating with the legacy network for its communication. An intelligent determination of this division of time/number of licenses, based on the geographical locations of all users is casted under the framework of dynamic spectrum leasing (DSL). This unified approach suggesting dynamic spectrum leasing as an alternative to both uni-directional and bi-directional communication in legacy networks is the main contribution and novelty of this thesis. It is shown that DSL improves the energy efficiency (EE) of uni-directional communication by 10 times where the secondary network cooperates to relay the data while maintaining the desired quality of service of the legacy network. In doing so it improves the spectral efficiency of the secondary network and reserves up to 40% time for their activity. Otherwise, the secondary network does not have any right to carry out its own transmission in the frequencies owned by the primary user. The thesis investigates how the techniques of network coding, beamforming and space time coding can help to harness better spectral and energy efficiency in DSL. It is shown that DSL with packet level network coding can help in bi-directional communication between a pair of primary nodes by improving the number of bits exchanged between them by 54%. The total energy cost in doing so is 10 times lower than direct bi-directional communication. DSL when applied with physical layer network coding and beamforming further enhances the SE gain by 17x or more. The thesis also studies how spectrum leasing can be applied to infrastructured cellular wireless networks. It is shown that if the cellular network of Leeds is considered, 12x improvement in the SE and 16x in the EE can be observed. Hence DSL can serve as a spectral and energy efficient alternative for wireless communication.