MRLC filter banks for blocker detection in LTE systems

Within the last three decades, the mobile communication sector evolved significantly. Meanwhile, the fourth generation (4G) system is in operation. While in recent years the focus was set on voice calls, when using a mobile phone, today we check emails everywhere we go, watch videos from the internet on our phones, organize video conferences while being on the road, etc. The enabling technology is a combination of more powerful devices (e.g. Tablet PC, Smart Phone) and more powerful networks (e.g. long term evolution (LTE), LTE-Advanced). However, while devices shrink in size, the usage of the various applications mentioned before does not support minimizing the energy consumption. As the evolution of batteries went slower than the users' demands, possibilities of energy saving have to be investigated elsewhere. While by Smart Phone usually a user equipment (UE) supporting many applications is meant, the term could be used ulteriorly. In this thesis a device being smart concerning the energy consumption is considered to be a smart device. An introduction to a concept of intelligent analogue (AFE) and digital frontends (DFE), based on interference detection, will be given. The main idea is simple: old fashioned UEs are developed to fulfill the worst case scenarios described in the according 3rd generation partnership project (3GPP) standard. However, at most of the times the transmission scenario is much more relaxed compared to the worst case. Knowledge about this relaxed scenario allows to reconfigure the AFE and DFE to the actual needs. The scope of this thesis is the detection of interferences, also called blockers, around a specific UE at very low computational cost and classify the instantaneous spectral environment of the device. The information gained is aimed to be used for adopting the filters, amplifiers and analog-to-digital converters in the AFE and DFE accordingly to save energy. The presented implementation is based on a multi-rate low-cost (MRLC) filter approach, building the blocker detection unit. This unit delivers information about the spectral environment of the UE and enables a smart power on demand receiver architecture, adopting itself to the actual needs. In this thesis the developed filter chains for an LTE-specific approach and also a general spectral sensing approach are presented. A comparison, not limited to implementation cost, to well known spectral sensing techniques is given. Finally, an FPGA based prototype implementing the structures, and an estimate on the overall complexity for manufacturing the structures on silicon using the 65 nm- CMOS process is given.